diff --git a/modules/webalert/lib/alert_engine.py b/modules/webalert/lib/alert_engine.py
index 27b00ca8e..2b3e759ea 100644
--- a/modules/webalert/lib/alert_engine.py
+++ b/modules/webalert/lib/alert_engine.py
@@ -1,348 +1,487 @@
# -*- coding: utf-8 -*-
##
## This file is part of CDS Invenio.
## Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 CERN.
##
## CDS Invenio is free software; you can redistribute it and/or
## modify it under the terms of the GNU General Public License as
## published by the Free Software Foundation; either version 2 of the
## License, or (at your option) any later version.
##
## CDS Invenio is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with CDS Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
"""Alert engine implementation."""
## rest of the Python code goes below
__revision__ = "$Id$"
from cgi import parse_qs
from re import search, sub
from time import strftime
import datetime
from invenio.config import \
CFG_LOGDIR, \
CFG_SITE_SUPPORT_EMAIL, \
CFG_SITE_URL, \
CFG_WEBALERT_SEND_EMAIL_NUMBER_OF_TRIES, \
- CFG_WEBALERT_SEND_EMAIL_SLEEPTIME_BETWEEN_TRIES
+ CFG_WEBALERT_SEND_EMAIL_SLEEPTIME_BETWEEN_TRIES, \
+ CFG_SITE_NAME
from invenio.webbasket_dblayer import get_basket_owner_id, add_to_basket
-from invenio.search_engine import perform_request_search
+from invenio.search_engine import perform_request_search, wash_colls, get_coll_sons, is_hosted_collection
from invenio.webinterface_handler import wash_urlargd
from invenio.dbquery import run_sql
from invenio.webuser import get_email
from invenio.mailutils import send_email
from invenio.errorlib import register_exception
from invenio.alert_engine_config import CFG_WEBALERT_DEBUG_LEVEL
+from invenio.websearch_external_collections_config import CFG_EXTERNAL_COLLECTION_TIMEOUT
+from invenio.websearch_external_collections_config import CFG_EXTERNAL_COLLECTION_MAXRESULTS_ALERTS
+from invenio.websearch_external_collections_getter import HTTPAsyncPageGetter, async_download
+from invenio.websearch_external_collections_utils import get_collection_id
+
import invenio.template
websearch_templates = invenio.template.load('websearch')
webalert_templates = invenio.template.load('webalert')
def update_date_lastrun(alert):
- return run_sql('update user_query_basket set date_lastrun=%s where id_user=%s and id_query=%s and id_basket=%s;', (strftime("%Y-%m-%d"), alert[0], alert[1], alert[2],))
+ """Update the last time this alert was ran in the database."""
+ return run_sql('update user_query_basket set date_lastrun=%s where id_user=%s and id_query=%s and id_basket=%s;', (strftime("%Y-%m-%d"), alert[0], alert[1], alert[2],))
def get_alert_queries(frequency):
+ """Return all the queries for the given frequency."""
+
return run_sql('select distinct id, urlargs from query q, user_query_basket uqb where q.id=uqb.id_query and uqb.frequency=%s and uqb.date_lastrun <= now();', (frequency,))
def get_alert_queries_for_user(uid):
+ """Returns all the queries for the given user id."""
+
return run_sql('select distinct id, urlargs, uqb.frequency from query q, user_query_basket uqb where q.id=uqb.id_query and uqb.id_user=%s and uqb.date_lastrun <= now();', (uid,))
def get_alerts(query, frequency):
+ """Returns a dictionary of all the records found for a specific query and frequency along with other informationm"""
+
r = run_sql('select id_user, id_query, id_basket, frequency, date_lastrun, alert_name, notification from user_query_basket where id_query=%s and frequency=%s;', (query['id_query'], frequency,))
return {'alerts': r, 'records': query['records'], 'argstr': query['argstr'], 'date_from': query['date_from'], 'date_until': query['date_until']}
-# Optimized version:
-def add_records_to_basket(record_ids, basket_id):
- nrec = len(record_ids)
+def add_records_to_basket(records, basket_id):
+ """Add the given records to the given baskets"""
+
+ nrec = len(records[0])
if nrec > 0:
if CFG_WEBALERT_DEBUG_LEVEL > 0:
- print "-> adding %s records into basket %s: %s" % (nrec, basket_id, record_ids)
+ print "-> adding %s records into basket %s: %s" % (nrec, basket_id, records[0])
try:
if CFG_WEBALERT_DEBUG_LEVEL < 4:
owner_uid = get_basket_owner_id(basket_id)
- add_to_basket(owner_uid, record_ids, [basket_id])
+ add_to_basket(owner_uid, records[0], 0, [basket_id])
else:
print ' NOT ADDED, DEBUG LEVEL == 4'
except Exception:
register_exception()
-
+ for external_collection_results in records[1][0]:
+ nrec = len(external_collection_results[1][0])
+ if nrec > 0:
+ if CFG_WEBALERT_DEBUG_LEVEL > 0:
+ print "-> adding %s external records (collection \"%s\") into basket %s: %s" % (nrec, external_collection_results[0], basket_id, external_collection_results[1][0])
+ try:
+ if CFG_WEBALERT_DEBUG_LEVEL < 4:
+ owner_uid = get_basket_owner_id(basket_id)
+ collection_id = get_collection_id(external_collection_results[0])
+ add_to_basket(owner_uid, external_collection_results[1][0], collection_id, [basket_id])
+ # TBD: maybe cache here the html brief format of the external records for the baskets later?
+ else:
+ print ' NOT ADDED, DEBUG LEVEL == 4'
+ except Exception:
+ register_exception()
def get_query(alert_id):
+ """Returns the query for that corresponds to this alert id."""
+
r = run_sql('select urlargs from query where id=%s', (alert_id,))
return r[0][0]
def email_notify(alert, records, argstr):
+ """Send the notification e-mail for a specific alert."""
- if len(records) == 0:
- return
+ if len(records[0]) == 0:
+ total_n_external_records = 0
+ for external_collection_results in records[1][0]:
+ total_n_external_records += len(external_collection_results[1][0])
+ if total_n_external_records == 0:
+ return
msg = ""
if CFG_WEBALERT_DEBUG_LEVEL > 0:
msg = "*** THIS MESSAGE WAS SENT IN DEBUG MODE ***\n\n"
url = CFG_SITE_URL + "/search?" + argstr
- # Extract the pattern and catalogue list from the formatted query
+ # Extract the pattern, the collection list, the current collection
+ # and the sc (split collection) from the formatted query
query = parse_qs(argstr)
pattern = query.get('p', [''])[0]
- catalogues = query.get('c', [])
+ collection_list = query.get('c', [])
+ current_collection = query.get('cc', [''])
+ sc = query.get('sc', ['1'])
+ collections = calculate_desired_collection_list(collection_list, current_collection, int(sc[0]))
frequency = alert[3]
msg += webalert_templates.tmpl_alert_email_body(
- alert[5], url, records, pattern, catalogues, frequency)
+ alert[5], url, records, pattern, collections, frequency)
email = get_email(alert[0])
if email == 'guest':
print "********************************************************************************"
print "The following alert was not send, because cannot detect user email address:"
print " " + repr(argstr)
print "********************************************************************************"
return
if CFG_WEBALERT_DEBUG_LEVEL > 0:
print "********************************************************************************"
print msg
print "********************************************************************************"
if CFG_WEBALERT_DEBUG_LEVEL < 2:
send_email(fromaddr=webalert_templates.tmpl_alert_email_from(),
toaddr=email,
subject=webalert_templates.tmpl_alert_email_title(alert[5]),
content=msg,
header='',
footer='',
attempt_times=CFG_WEBALERT_SEND_EMAIL_NUMBER_OF_TRIES,
attempt_sleeptime=CFG_WEBALERT_SEND_EMAIL_SLEEPTIME_BETWEEN_TRIES)
if CFG_WEBALERT_DEBUG_LEVEL == 4:
send_email(fromaddr=webalert_templates.tmpl_alert_email_from(),
toaddr=CFG_SITE_SUPPORT_EMAIL,
subject=webalert_templates.tmpl_alert_email_title(alert[5]),
content=msg,
header='',
footer='',
attempt_times=CFG_WEBALERT_SEND_EMAIL_NUMBER_OF_TRIES,
attempt_sleeptime=CFG_WEBALERT_SEND_EMAIL_SLEEPTIME_BETWEEN_TRIES)
-def get_argument(args, argname):
- if args.has_key(argname):
- return args[argname]
- else:
- return []
-
def _date_to_tuple(date):
+ """Private function. Converts a date as a tuple of string into a list of integers."""
+
return [int(part) for part in (date.year, date.month, date.day)]
def get_record_ids(argstr, date_from, date_until):
+ """Returns the local and external records found for a specific query and timeframe."""
+
argd = wash_urlargd(parse_qs(argstr), websearch_templates.search_results_default_urlargd)
- p = get_argument(argd, 'p')
- c = get_argument(argd, 'c')
- cc = get_argument(argd, 'cc')
- aas = get_argument(argd, 'aas')
- f = get_argument(argd, 'f')
- so = get_argument(argd, 'so')
- sp = get_argument(argd, 'sp')
- ot = get_argument(argd, 'ot')
- p1 = get_argument(argd, 'p1')
- f1 = get_argument(argd, 'f1')
- m1 = get_argument(argd, 'm1')
- op1 = get_argument(argd, 'op1')
- p2 = get_argument(argd, 'p2')
- f2 = get_argument(argd, 'f2')
- m2 = get_argument(argd, 'm2')
- op2 = get_argument(argd, 'op2')
- p3 = get_argument(argd, 'p3')
- f3 = get_argument(argd, 'f3')
- m3 = get_argument(argd, 'm3')
- sc = get_argument(argd, 'sc')
+ p = argd.get('p', [])
+ c = argd.get('c', [])
+ cc = argd.get('cc', [])
+ aas = argd.get('aas', [])
+ f = argd.get('f', [])
+ so = argd.get('so', [])
+ sp = argd.get('sp', [])
+ ot = argd.get('ot', [])
+ p1 = argd.get('p1', [])
+ f1 = argd.get('f1', [])
+ m1 = argd.get('m1', [])
+ op1 = argd.get('op1', [])
+ p2 = argd.get('p2', [])
+ f2 = argd.get('f2', [])
+ m2 = argd.get('m2', [])
+ op2 = argd.get('op3', [])
+ p3 = argd.get('p3', [])
+ f3 = argd.get('f3', [])
+ m3 = argd.get('m3', [])
+ sc = argd.get('sc', [])
d1y, d1m, d1d = _date_to_tuple(date_from)
d2y, d2m, d2d = _date_to_tuple(date_until)
- return perform_request_search(of='id', p=p, c=c, cc=cc, f=f, so=so, sp=sp, ot=ot,
+ washed_colls = wash_colls(cc, c, sc, 0)
+ hosted_colls = washed_colls[3]
+ if hosted_colls:
+ req_args = "p=%s&d1d=%s&d1m=%s&d1y=%s&d2d=%s&d2m=%s&d2y=%s&ap=%i" % (p, d1d, d1m, d1y, d2d, d2m, d2y, 0)
+ external_records = calculate_external_records(req_args, [p, p1, p2, p3], f, hosted_colls, CFG_EXTERNAL_COLLECTION_TIMEOUT, CFG_EXTERNAL_COLLECTION_MAXRESULTS_ALERTS)
+ else:
+ external_records = ([],[])
+
+ recids = perform_request_search(of='id', p=p, c=c, cc=cc, f=f, so=so, sp=sp, ot=ot,
aas=aas, p1=p1, f1=f1, m1=m1, op1=op1, p2=p2, f2=f2,
m2=m2, op2=op2, p3=p3, f3=f3, m3=m3, sc=sc, d1y=d1y,
d1m=d1m, d1d=d1d, d2y=d2y, d2m=d2m, d2d=d2d)
-
-def get_argument_as_string(argstr, argname):
- args = parse_qs(argstr)
- a = get_argument(args, argname)
- r = ''
- if len(a):
- r = a[0]
- for i in a[1:len(a)]:
- r += ", %s" % i
- return r
-
-def get_pattern(argstr):
- return get_argument_as_string(argstr, 'p')
-
-def get_catalogue(argstr):
- return get_argument_as_string(argstr, 'c')
-
-def get_catalogue_num(argstr):
- args = parse_qs(argstr)
- a = get_argument(args, 'c')
- return len(a)
-
+ return (recids, external_records)
def run_query(query, frequency, date_until):
"""Return a dictionary containing the information of the performed query.
The information contains the id of the query, the arguments as a
string, and the list of found records."""
if frequency == 'day':
date_from = date_until - datetime.timedelta(days=1)
elif frequency == 'week':
date_from = date_until - datetime.timedelta(weeks=1)
else:
# Months are not an explicit notion of timedelta (it's the
# most ambiguous too). So we explicitely take the same day of
# the previous month.
d, m, y = (date_until.day, date_until.month, date_until.year)
m = m - 1
if m == 0:
m = 12
y = y - 1
date_from = datetime.date(year=y, month=m, day=d)
recs = get_record_ids(query[1], date_from, date_until)
- n = len(recs)
+ n = len(recs[0])
if n:
- log('query %08s produced %08s records' % (query[0], len(recs)))
+ log('query %08s produced %08s records for all the local collections' % (query[0], n))
+
+ for external_collection_results in recs[1][0]:
+ n = len(external_collection_results[1][0])
+ if n:
+ log('query %08s produced %08s records for external collection \"%s\"' % (query[0], n, external_collection_results[0]))
if CFG_WEBALERT_DEBUG_LEVEL > 2:
print "[%s] run query: %s with dates: from=%s, until=%s\n found rec ids: %s" % (
strftime("%c"), query, date_from, date_until, recs)
return {'id_query': query[0], 'argstr': query[1],
'records': recs, 'date_from': date_from, 'date_until': date_until}
-
def process_alert_queries(frequency, date):
"""Run the alerts according to the frequency.
Retrieves the queries for which an alert exists, performs it, and
processes the corresponding alerts."""
alert_queries = get_alert_queries(frequency)
for aq in alert_queries:
q = run_query(aq, frequency, date)
alerts = get_alerts(q, frequency)
process_alerts(alerts)
+def process_alert_queries_for_user(uid, date):
+ """Process the alerts for the given user id.
+
+ All alerts are with reference date set as the current local time."""
+
+ alert_queries = get_alert_queries_for_user(uid)
+
+ for aq in alert_queries:
+ frequency = aq[2]
+ q = run_query(aq, frequency, date)
+ alerts = get_alerts(q, frequency)
+ process_alerts(alerts)
def replace_argument(argstr, argname, argval):
"""Replace the given date argument value with the new one.
If the argument is missing, it is added."""
if search('%s=\d+' % argname, argstr):
r = sub('%s=\d+' % argname, '%s=%s' % (argname, argval), argstr)
else:
r = argstr + '&%s=%s' % (argname, argval)
return r
def update_arguments(argstr, date_from, date_until):
"""Replace date arguments in argstr with the ones specified by date_from and date_until.
Absent arguments are added."""
d1y, d1m, d1d = _date_to_tuple(date_from)
d2y, d2m, d2d = _date_to_tuple(date_until)
r = replace_argument(argstr, 'd1y', d1y)
r = replace_argument(r, 'd1m', d1m)
r = replace_argument(r, 'd1d', d1d)
r = replace_argument(r, 'd2y', d2y)
r = replace_argument(r, 'd2m', d2m)
r = replace_argument(r, 'd2d', d2d)
return r
def log(msg):
+ """Logs the given message in the alert engine log."""
+
try:
logfile = open(CFG_LOGDIR + '/alertengine.log', 'a')
logfile.write(strftime('%Y%m%d%H%M%S#'))
logfile.write(msg + '\n')
logfile.close()
except Exception:
register_exception()
def process_alerts(alerts):
+ """Process the given alerts and store the records found to the user defined baskets
+ and/or notify them by e-mail"""
+
# TBD: do not generate the email each time, forge it once and then
# send it to all appropriate people
for a in alerts['alerts']:
if alert_use_basket_p(a):
add_records_to_basket(alerts['records'], a[2])
if alert_use_notification_p(a):
argstr = update_arguments(alerts['argstr'], alerts['date_from'], alerts['date_until'])
try:
email_notify(a, alerts['records'], argstr)
except Exception:
# There were troubles sending this alert, so register
# this exception and continue with other alerts:
register_exception(alert_admin=True,
prefix="Error when sending alert %s, %s\n." % \
(repr(a), repr(argstr)))
+ # Inform the admin when external collections time out
+ if len(alerts['records'][1][1]) > 0:
+ register_exception(alert_admin=True,
+ prefix="External collections %s timed out when sending alert %s, %s\n." % \
+ (", ".join(alerts['records'][1][1]), repr(a), repr(argstr)))
update_date_lastrun(a)
-
def alert_use_basket_p(alert):
- return alert[2] != 0
+ """Boolean. Should this alert store the records found in a basket?"""
+ return alert[2] != 0
def alert_use_notification_p(alert):
- return alert[6] == 'y'
+ """Boolean. Should this alert send a notification e-mail about the records found?"""
+ return alert[6] == 'y'
def run_alerts(date):
"""Run the alerts.
First decide which alerts to run according to the current local
time, and runs them."""
if date.day == 1:
process_alert_queries('month', date)
if date.isoweekday() == 1: # first day of the week
process_alert_queries('week', date)
process_alert_queries('day', date)
-def process_alert_queries_for_user(uid, date):
- """Process the alerts for the given user id.
-
- All alerts are with reference date set as the current local time."""
-
- alert_queries = get_alert_queries_for_user(uid)
- print alert_queries
-
- for aq in alert_queries:
- frequency = aq[2]
- q = run_query(aq, frequency, date)
- alerts = get_alerts(q, frequency)
- process_alerts(alerts)
+# External records related functions
+def calculate_external_records(req_args, pattern_list, field, hosted_colls, timeout=CFG_EXTERNAL_COLLECTION_TIMEOUT, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS_ALERTS):
+ """Function that returns the external records found and the potential time outs
+ given a search pattern and a list of hosted collections."""
+
+ (external_search_engines, basic_search_units) = calculate_external_search_params(req_args, pattern_list, field, hosted_colls)
+
+ return do_calculate_external_records(req_args, basic_search_units, external_search_engines, timeout, limit)
+
+def calculate_external_search_params(req_args, pattern_list, field, hosted_colls):
+ """Function that calculates the basic search units given the search pattern.
+ Also returns a set of hosted collections engines."""
+
+ from invenio.search_engine import create_basic_search_units
+ from invenio.websearch_external_collections import bind_patterns
+ from invenio.websearch_external_collections import select_hosted_search_engines as select_external_search_engines
+
+ pattern = bind_patterns(pattern_list)
+ basic_search_units = create_basic_search_units(None, pattern, field)
+
+ external_search_engines = select_external_search_engines(hosted_colls)
+
+ return (external_search_engines, basic_search_units)
+
+def do_calculate_external_records(req_args, basic_search_units, external_search_engines, timeout=CFG_EXTERNAL_COLLECTION_TIMEOUT, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS_ALERTS):
+ """Function that returns the external records found and the potential time outs
+ given the basic search units or the req arguments and a set of hosted collections engines."""
+
+ # list to hold the hosted search engines and their respective search urls
+ engines_list = []
+ # list to hold the non timed out results
+ results_list = []
+ # list to hold all the results
+ full_results_list = []
+ # list to hold all the timeouts
+ timeout_list = []
+
+ for engine in external_search_engines:
+ url = engine.build_search_url(basic_search_units, req_args, limit=limit)
+ if url:
+ engines_list.append([url, engine])
+ # we end up with a [[search url], [engine]] kind of list
+
+ # create the list of search urls to be handed to the asynchronous getter
+ pagegetters_list = [HTTPAsyncPageGetter(engine[0]) for engine in engines_list]
+
+ # function to be run on every result
+ def finished(pagegetter, data, current_time):
+ """Function called, each time the download of a web page finish.
+ Will parse and print the results of this page."""
+ # each pagegetter that didn't timeout is added to this list
+ results_list.append((pagegetter, data))
+
+ # run the asynchronous getter
+ finished_list = async_download(pagegetters_list, finished, engines_list, timeout)
+
+ # create the complete list of tuples, one for each hosted collection, with the results and other information,
+ # including those that timed out
+ for (finished, engine) in zip(finished_list, engines_list): #finished_and_engines_list:
+ if finished:
+ for result in results_list:
+ if result[1] == engine:
+ engine[1].parser.parse_and_get_results(result[0].data, feedonly=True)
+ full_results_list.append((engine[1].name, engine[1].parser.parse_and_extract_records(of="xm")))
+ break
+ elif not finished:
+ timeout_list.append(engine[1].name)
+
+ return (full_results_list, timeout_list)
+
+def calculate_desired_collection_list(c, cc, sc):
+ """Function that calculates the user desired collection list when sending a webalert e-mail"""
+
+ if not cc[0]:
+ cc = [CFG_SITE_NAME]
+
+ # quickly create the reverse function of is_hosted_collection
+ is_not_hosted_collection = lambda coll: not is_hosted_collection(coll)
+
+ # calculate the list of non hosted, non restricted, regular sons of cc
+ washed_cc_sons = filter(is_not_hosted_collection, get_coll_sons(cc[0]))
+ # clean up c removing hosted collections
+ washed_c = filter(is_not_hosted_collection, c)
+
+ # try to simulate the wash_colls function behavior when calculating the collections to return
+ if washed_cc_sons == washed_c:
+ if sc == 0:
+ return cc
+ elif sc == 1:
+ return washed_c
+ else:
+ if sc == 0:
+ return washed_c
+ elif sc == 1:
+ washed_c_sons = []
+ for coll in washed_c:
+ if coll in washed_cc_sons:
+ washed_c_sons.extend(get_coll_sons(coll))
+ else:
+ washed_c_sons.append(coll)
+ return washed_c_sons
diff --git a/modules/webalert/lib/htmlparser.py b/modules/webalert/lib/htmlparser.py
index 98a4363ea..5f95fe834 100644
--- a/modules/webalert/lib/htmlparser.py
+++ b/modules/webalert/lib/htmlparser.py
@@ -1,187 +1,189 @@
## This file is part of CDS Invenio.
## Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 CERN.
##
## CDS Invenio is free software; you can redistribute it and/or
## modify it under the terms of the GNU General Public License as
## published by the Free Software Foundation; either version 2 of the
## License, or (at your option) any later version.
##
## CDS Invenio is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with CDS Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
"""HTML parser for records."""
__revision__ = "$Id$"
import re
from HTMLParser import HTMLParser
import textwrap
import htmlentitydefs
from invenio.config import \
CFG_WEBALERT_MAX_NUM_OF_CHARS_PER_LINE_IN_ALERT_EMAIL, \
CFG_SITE_LANG
from invenio.bibformat import format_record
from invenio.htmlutils import remove_html_markup
from invenio.messages import gettext_set_language
whitespaces_pattern = re.compile(r'[ \t]+')
def wrap(text):
"""Limits the number of characters per line in given text.
The function does not preserve new lines.
"""
lines = textwrap.wrap(text, CFG_WEBALERT_MAX_NUM_OF_CHARS_PER_LINE_IN_ALERT_EMAIL)
r = ''
for l in lines:
r += l + '\n'
return r
def wrap_records(text):
"""Limits the number of characters per line in given text.
The function preserves new lines.
"""
lines = text.split('\n')
result_lines = []
for l in lines:
newlines = textwrap.wrap(l, CFG_WEBALERT_MAX_NUM_OF_CHARS_PER_LINE_IN_ALERT_EMAIL)
for ll in newlines:
result_lines.append(ll)
return '\n'.join(result_lines)
class RecordHTMLParser(HTMLParser):
"""A parser for the HTML returned by invenio.search_engine.print_record.
The parser provides methods to transform the HTML returned by
invenio.search_engine.print_record into plain text, with some
minor formatting.
"""
silent = False
new_line = True # Are we at the beginning of a new line? (after
# ,
or at the beginning of the text)
def __init__(self):
HTMLParser.__init__(self)
self.result = ''
def handle_starttag(self, tag, attrs):
if tag == 'strong':
# self.result += '*'
pass
elif tag == 'a':
self.printURL = 0
self.unclosedBracket = 0
for f in attrs:
#if f[1] == 'note':
# self.result += 'Fulltext : <'
# self.unclosedBracket = 1
if f[1] == 'moreinfo':
self.result += 'Detailed record : '
self.printURL = 1
if (self.printURL == 1) and (f[0] == 'href'):
self.result += '<' + f[1] + '>'
elif tag == 'br':
self.result += '\n'
elif tag == 'style' or tag == 'script':
self.silent = True
elif tag == 'p':
if not self.new_line:
self.result += '\n'
self.new_line = True
def handle_startendtag(self, tag, attrs):
if tag == 'br':
self.result += '\n'
self.new_line = True
def handle_endtag(self, tag):
if tag == 'strong':
# self.result += '\n'
pass
elif tag == 'a':
if self.unclosedBracket == 1:
self.result += '>'
self.unclosedBracket = 0
elif tag == 'style' or tag == 'script':
self.silent = False
elif tag == 'p':
self.result += '\n'
self.new_line = True
def handle_data(self, data):
if data.lower() in ['detailed record', 'similar record', 'cited by']:
pass
elif self.silent == False:
if self.new_line:
# Remove unnecessary trailing whitespace at the
# beginning of a line
data = data.lstrip()
# Merge all consecutive whitespaces into a single one
self.result += data
self.new_line = False
def handle_comment(self, data):
if 'START_NOT_FOR_TEXT' == data.upper().strip():
self.silent = True
elif 'END_NOT_FOR_TEXT' == data.upper().strip():
self.silent = False
def handle_charref(self, name):
"""Process character references of the form "ref;". Transform to text whenever possible."""
try:
self.result += unichr(int(name)).encode("utf-8")
except:
return
def handle_entityref(self, name):
"""Process a general entity reference of the form "&name;".
Transform to text whenever possible."""
if name == 'nbsp':
# Keep them for the moment. Will be processed at the end.
# It is needed to consider them separately, as all
# consecutive whitespaces will be merged into a single one
# at the end. It should not be the case of non breakable
# space. Note that because of this trick, input
# will be considered as a
self.result += ' '
return
char_code = htmlentitydefs.name2codepoint.get(name, None)
if char_code is not None:
try:
self.result += unichr(char_code).encode("utf-8")
except:
return
-def get_as_text(record_id, ln=CFG_SITE_LANG):
+def get_as_text(record_id=0, xml_record=None, ln=CFG_SITE_LANG):
"""Return the record in a textual format"""
_ = gettext_set_language(ln)
out = ""
- rec_in_hb = format_record(record_id, of="hb")
+ if record_id != 0:
+ rec_in_hb = format_record(record_id, of="hb")
+ elif xml_record:
+ rec_in_hb = format_record(0, of="hb", xml_record=xml_record)
rec_in_hb = rec_in_hb.replace('\n', ' ')
htparser = RecordHTMLParser()
try:
htparser.feed(rec_in_hb)
htparser.close()
out = htparser.result
except:
out = remove_html_markup(rec_in_hb)
# Remove trailing whitespace and linefeeds
out = out.strip('\n').strip()
# Merge consecutive whitespaces. Must be done here, once all HTML
# tags have been removed
out = whitespaces_pattern.sub(' ', out)
# Now consider non-breakable spaces
out = out.replace(' ', ' ')
out = re.sub(r"[\-:]?\s*%s\s*[\-:]?" % _("Detailed record"), "", out)
out = re.sub(r"[\-:]?\s*%s\s*[\-:]?" % _("Similar records"), "", out)
out = re.sub(r"[\-:]?\s*%s\s*[\-:]?" % _("Cited by"), "", out)
return out.strip()
-
diff --git a/modules/webalert/lib/webalert_templates.py b/modules/webalert/lib/webalert_templates.py
index 4e50a2650..568dd90f0 100644
--- a/modules/webalert/lib/webalert_templates.py
+++ b/modules/webalert/lib/webalert_templates.py
@@ -1,576 +1,650 @@
## This file is part of CDS Invenio.
## Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 CERN.
##
## CDS Invenio is free software; you can redistribute it and/or
## modify it under the terms of the GNU General Public License as
## published by the Free Software Foundation; either version 2 of the
## License, or (at your option) any later version.
##
## CDS Invenio is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with CDS Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
__revision__ = "$Id$"
import cgi
import time
from invenio.config import \
CFG_WEBALERT_ALERT_ENGINE_EMAIL, \
CFG_SITE_NAME, \
CFG_SITE_SUPPORT_EMAIL, \
CFG_SITE_URL, \
CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL
from invenio.messages import gettext_set_language
from invenio.htmlparser import get_as_text, wrap, wrap_records
from invenio.urlutils import create_html_link
+from invenio.search_engine import guess_primary_collection_of_a_record, get_coll_ancestors
+
class Template:
def tmpl_errorMsg(self, ln, error_msg, rest = ""):
"""
Adds an error message to the output
Parameters:
- 'ln' *string* - The language to display the interface in
- 'error_msg' *string* - The error message
- 'rest' *string* - The rest of the page
"""
# load the right message language
_ = gettext_set_language(ln)
out = """
%(error)s
%(rest)s""" % {
'error' : error_msg,
'rest' : rest
}
return out
def tmpl_textual_query_info_from_urlargs(self, ln, args):
"""
Displays a human inteligible textual representation of a query
Parameters:
- 'ln' *string* - The language to display the interface in
- 'args' *array* - The URL arguments array (parsed)
"""
# load the right message language
_ = gettext_set_language(ln)
out = ""
if args.has_key('p'):
out += "" + _("Pattern") + ": " + "; ".join(args['p']) + " "
if args.has_key('f'):
out += "" + _("Field") + ": " + "; ".join(args['f']) + " "
if args.has_key('p1'):
out += "" + _("Pattern 1") + ": " + "; ".join(args['p1']) + " "
if args.has_key('f1'):
out += "" + _("Field 1") + ": " + "; ".join(args['f1']) + " "
if args.has_key('p2'):
out += "" + _("Pattern 2") + ": " + "; ".join(args['p2']) + " "
if args.has_key('f2'):
out += "" + _("Field 2") + ": " + "; ".join(args['f2']) + " "
if args.has_key('p3'):
out += "" + _("Pattern 3") + ": " + "; ".join(args['p3']) + " "
if args.has_key('f3'):
out += "" + _("Field 3") + ": " + "; ".join(args['f3']) + " "
if args.has_key('c'):
out += "" + _("Collections") + ": " + "; ".join(args['c']) + " "
elif args.has_key('cc'):
out += "" + _("Collection") + ": " + "; ".join(args['cc']) + " "
return out
def tmpl_account_list_alerts(self, ln, alerts):
"""
Displays all the alerts in the main "Your account" page
Parameters:
- 'ln' *string* - The language to display the interface in
- 'alerts' *array* - The existing alerts IDs ('id' + 'name' pairs)
"""
# load the right message language
_ = gettext_set_language(ln)
out = """""" % {
'show' : _("SHOW"),
}
return out
def tmpl_input_alert(self, ln, query, alert_name, action, frequency, notification,
baskets, old_id_basket, id_basket, id_query,
guest, guesttxt):
"""
Displays an alert adding form.
Parameters:
- 'ln' *string* - The language to display the interface in
- 'query' *string* - The HTML code of the textual representation of the query (as returned ultimately by tmpl_textual_query_info_from_urlargs...)
- 'alert_name' *string* - The alert name
- 'action' *string* - The action to complete ('update' or 'add')
- 'frequency' *string* - The frequency of alert running ('day', 'week', 'month')
- 'notification' *string* - If notification should be sent by email ('y', 'n')
- 'baskets' *array* - The existing baskets ('id' + 'name' pairs)
- 'old_id_basket' *string* - The id of the previous basket of this alert
- 'id_basket' *string* - The id of the basket of this alert
- 'id_query' *string* - The id of the query associated to this alert
- 'guest' *bool* - If the user is a guest user
- 'guesttxt' *string* - The HTML content of the warning box for guest users (produced by webaccount.tmpl_warning_guest_user)
"""
# load the right message language
_ = gettext_set_language(ln)
out = ""
out += """
%(notify_cond)s
%(query_text)s:
%(query)s
""" % {
'notify_cond' : _("This alert will notify you each time/only if a new item satisfies the following query:"),
'query_text' : _("QUERY"),
'query' : query,
}
out += """"
if guest:
out += guesttxt
return out
def tmpl_list_alerts(self, ln, alerts, guest, guesttxt):
"""
Displays the list of alerts
Parameters:
- 'ln' *string* - The language to display the interface in
- 'alerts' *array* - The existing alerts:
- 'queryid' *string* - The id of the associated query
- 'queryargs' *string* - The query string
- 'textargs' *string* - The textual description of the query string
- 'userid' *string* - The user id
- 'basketid' *string* - The basket id
- 'basketname' *string* - The basket name
- 'alertname' *string* - The alert name
- 'frequency' *string* - The frequency of alert running ('day', 'week', 'month')
- 'notification' *string* - If notification should be sent by email ('y', 'n')
- 'created' *string* - The date of alert creation
- 'lastrun' *string* - The last running date
- 'guest' *bool* - If the user is a guest user
- 'guesttxt' *string* - The HTML content of the warning box for guest users (produced by webaccount.tmpl_warning_guest_user)
"""
# load the right message language
_ = gettext_set_language(ln)
out = '
' + _("Set a new alert from %(x_url1_open)syour searches%(x_url1_close)s, the %(x_url2_open)spopular searches%(x_url2_close)s, or the input form.") + '
""" % {
'no' : _("No"),
'name' : _("Name"),
'search_freq' : _("Search checking frequency"),
'notification' : _("Notification by email"),
'result_basket' : _("Result in basket"),
'date_run' : _("Date last run"),
'date_created' : _("Creation date"),
'query' : _("Query"),
'action' : _("Action"),
}
i = 0
for alert in alerts:
i += 1
if alert['frequency'] == "day":
frequency = _("daily")
else:
if alert['frequency'] == "week":
frequency = _("weekly")
else:
frequency = _("monthly")
if alert['notification'] == "y":
notification = _("yes")
else:
notification = _("no")
# we clean up the HH:MM part of lastrun, since it is always 00:00
lastrun = alert['lastrun'].split(',')[0]
created = alert['created'].split(',')[0]
out += """
' + (_("You have defined %s alerts.") % ('' + str(len(alerts)) + '' )) + '
'
if guest:
out += guesttxt
return out
def tmpl_display_alerts(self, ln, permanent, nb_queries_total, nb_queries_distinct, queries, guest, guesttxt):
"""
Displays the list of alerts
Parameters:
- 'ln' *string* - The language to display the interface in
- 'permanent' *string* - If displaying most popular searches ('y') or only personal searches ('n')
- 'nb_queries_total' *string* - The number of personal queries in the last period
- 'nb_queries_distinct' *string* - The number of distinct queries in the last period
- 'queries' *array* - The existing queries:
- 'id' *string* - The id of the associated query
- 'args' *string* - The query string
- 'textargs' *string* - The textual description of the query string
- 'lastrun' *string* - The last running date (only for personal queries)
- 'guest' *bool* - If the user is a guest user
- 'guesttxt' *string* - The HTML content of the warning box for guest users (produced by webaccount.tmpl_warning_guest_user)
"""
# load the right message language
_ = gettext_set_language(ln)
if len(queries) == 0:
out = _("You have not executed any search yet. Please go to the %(x_url_open)ssearch interface%(x_url_close)s first.") % \
{'x_url_open': '',
'x_url_close': ''}
return out
out = ''
# display message: number of items in the list
if permanent == "n":
msg = _("You have performed %(x_nb1)s searches (%(x_nb2)s different questions) during the last 30 days or so.") % {'x_nb1': nb_queries_total,
'x_nb2': nb_queries_distinct}
out += '
' + msg + '
'
else:
# permanent="y"
msg = _("Here are the %s most popular searches.")
msg %= ('' + str(len(queries)) + '')
out += '
' + msg + '
'
# display the list of searches
out += """
%(no)s
%(question)s
%(action)s
""" % {
'no' : "#",
'question' : _("Question"),
'action' : _("Action")
}
if permanent == "n":
out += '
%s
' % _("Last Run")
out += "
\n"
i = 0
for query in queries :
i += 1
# id, pattern, base, search url and search set alert, date
out += """
""" % {
'index' : i,
'textargs' : query['textargs'],
'siteurl' : CFG_SITE_URL,
'args' : cgi.escape(query['args']),
'id' : query['id'],
'ln': ln,
'execute_query' : _("Execute search"),
'set_alert' : _("Set new alert")
}
if permanent == "n":
out += '
%s
' % query['lastrun']
out += """
\n"""
out += "
\n"
if guest :
out += guesttxt
return out
def tmpl_alert_email_title(self, name):
return 'Alert %s run on %s' % (
name, time.strftime("%Y-%m-%d"))
def tmpl_alert_email_from(self):
return '%s Alert Engine <%s>' % (CFG_SITE_NAME, CFG_WEBALERT_ALERT_ENGINE_EMAIL)
def tmpl_alert_email_body(self, name, url, records, pattern,
- catalogues, frequency):
-
- l = len(catalogues)
+ collection_list, frequency):
+
+ recids_by_collection = {}
+ for recid in records[0]:
+ primary_collection = guess_primary_collection_of_a_record(recid)
+ if primary_collection in collection_list:
+ if not recids_by_collection.has_key(primary_collection):
+ recids_by_collection[primary_collection] = []
+ recids_by_collection[primary_collection].append(recid)
+ else:
+ ancestors = get_coll_ancestors(primary_collection)
+ ancestors.reverse()
+ nancestors = 0
+ for ancestor in ancestors:
+ nancestors += 1
+ if ancestor in collection_list:
+ if not recids_by_collection.has_key(ancestor):
+ recids_by_collection[ancestor] = []
+ recids_by_collection[ancestor].append(recid)
+ break
+ elif len(ancestors) == nancestors:
+ if not recids_by_collection.has_key('None of the above'):
+ recids_by_collection['None of the above'] = []
+ recids_by_collection['None of the above'].append(recid)
+
+ collection_list = [coll for coll in recids_by_collection.keys() if coll != 'None of the above']
+ for external_collection_results in records[1][0]:
+ if external_collection_results[1][0]:
+ collection_list.append(external_collection_results[0])
+
+ l = len(collection_list)
if l == 0:
collections = ''
elif l == 1:
- collections = "collection: %s\n" % catalogues[0]
+ collections = "collection: %s\n" % collection_list[0]
+ else:
+ collections = "collections: %s\n" % wrap(', '.join(collection_list))
+
+ l = len(records[0])
+ for external_collection_results in records[1][0]:
+ l += len(external_collection_results[1][0])
+ if l == 1:
+ total = '1 record'
else:
- collections = "collections: %s\n" % wrap(', '.join(catalogues))
+ total = '%d records' % l
if pattern:
pattern = 'pattern: %s\n' % pattern
frequency = {'day': 'daily',
'week': 'weekly',
'month': 'monthly'}[frequency]
- l = len(records)
- if l == 1:
- total = '1 record'
- else:
- total = '%d records' % l
-
-
body = """\
Hello:
Below are the results of the email notification alert that
you set up with the %(sitename)s.
This is an automatic message, please don't reply to it.
For any question, please use <%(sitesupportemail)s> instead.
alert name: %(name)s
%(pattern)s%(collections)sfrequency: %(frequency)s
run time: %(runtime)s
found: %(total)s
url: <%(url)s>
""" % {'sitesupportemail': CFG_SITE_SUPPORT_EMAIL,
'name': name,
'sitename': CFG_SITE_NAME,
'pattern': pattern,
'collections': collections,
'frequency': frequency,
'runtime': time.strftime("%a %Y-%m-%d %H:%M:%S"),
'total': total,
'url': url}
-
- for index, recid in enumerate(records[:CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL]):
- body += "\n%i) " % (index + 1)
- body += self.tmpl_alert_email_record(recid)
- body += "\n"
-
- if len(records) > CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ index = 0
+
+ for collection_recids in recids_by_collection.items():
+ if collection_recids[0] != 'None of the above':
+ body += "\nCollection: %s\n" % collection_recids[0]
+ for recid in collection_recids[1]:
+ index += 1
+ body += "\n%i) " % (index)
+ body += self.tmpl_alert_email_record(recid=recid)
+ body += "\n"
+ if index == CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ break
+ if index == CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ break
+
+ if index < CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ if recids_by_collection.has_key('None of the above'):
+ if len(recids_by_collection.keys()) > 1:
+ body += "\nNone of the above collections:\n"
+ else:
+ # if the uncategorized collection is the only collection then present
+ # all the records as belonging to CFG_SITE_NAME
+ body += "\nCollection: %s\n" % CFG_SITE_NAME
+ for recid in recids_by_collection['None of the above']:
+ index += 1
+ body += "\n%i) " % (index)
+ body += self.tmpl_alert_email_record(recid=recid)
+ body += "\n"
+ if index == CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ break
+
+ if index < CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ for external_collection_results in records[1][0]:
+ body += "\nCollection: %s\n" % external_collection_results[0]
+ for recid in external_collection_results[1][0]:
+ index += 1
+ body += "\n%i) " % (index)
+ # TODO: extend function to accept xml_record!
+ body += self.tmpl_alert_email_record(xml_record=external_collection_results[1][1][recid])
+ body += "\n"
+ if index == CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ break
+ if index == CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
+ break
+
+ if l > CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL:
body += '''
Only the first %s records were displayed. Please consult the search
URL given at the top of this email to see all the results.
''' % CFG_WEBALERT_MAX_NUM_OF_RECORDS_IN_ALERT_EMAIL
body += '''
--
%s Alert Service <%s>
Unsubscribe? See <%s>
Need human intervention? Contact <%s>
''' % (CFG_SITE_NAME, CFG_SITE_URL, CFG_SITE_URL + '/youralerts/list', CFG_SITE_SUPPORT_EMAIL)
return body
- def tmpl_alert_email_record(self, recid):
+ def tmpl_alert_email_record(self, recid=0, xml_record=None):
""" Format a single record."""
- out = wrap_records(get_as_text(recid))
- out += "\nDetailed record: <%s/record/%s>" % (CFG_SITE_URL, recid)
+ if recid != 0:
+ out = wrap_records(get_as_text(record_id=recid))
+ out += "\nDetailed record: <%s/record/%s>" % (CFG_SITE_URL, recid)
+ elif xml_record:
+ out = wrap_records(get_as_text(xml_record=xml_record))
+ # TODO: add Detailed record url for external records?
return out
diff --git a/modules/webbasket/lib/webbasket.py b/modules/webbasket/lib/webbasket.py
index 6d1a85eb7..0b518e982 100644
--- a/modules/webbasket/lib/webbasket.py
+++ b/modules/webbasket/lib/webbasket.py
@@ -1,1019 +1,1018 @@
## This file is part of CDS Invenio.
## Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 CERN.
##
## CDS Invenio is free software; you can redistribute it and/or
## modify it under the terms of the GNU General Public License as
## published by the Free Software Foundation; either version 2 of the
## License, or (at your option) any later version.
##
## CDS Invenio is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with CDS Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
"""Web Baskets features."""
__revision__ = "$Id$"
import cgi
from zlib import decompress
from invenio.config import CFG_SITE_LANG, CFG_SITE_URL, \
CFG_WEBBASKET_MAX_NUMBER_OF_DISPLAYED_BASKETS
from invenio.messages import gettext_set_language
from invenio.dateutils import convert_datetext_to_dategui, \
convert_datetext_to_datestruct,\
convert_datestruct_to_dategui
from invenio.bibformat import format_record
from invenio.webbasket_config import CFG_WEBBASKET_SHARE_LEVELS, \
CFG_WEBBASKET_SHARE_LEVELS_ORDERED, \
CFG_WEBBASKET_CATEGORIES, \
CFG_WEBBASKET_WARNING_MESSAGES
from invenio.webuser import isGuestUser, collect_user_info
from invenio.search_engine import \
record_exists, \
check_user_can_view_record
from invenio.webcomment import check_user_can_attach_file_to_comments
import invenio.webbasket_dblayer as db
try:
import invenio.template
webbasket_templates = invenio.template.load('webbasket')
except ImportError:
pass
from invenio.websearch_external_collections_utils import get_collection_name_by_id
from invenio.websearch_external_collections import select_hosted_search_engines
from invenio.websearch_external_collections_config import CFG_EXTERNAL_COLLECTION_TIMEOUT
-from invenio.websearch_external_collections_searcher import external_collections_dictionary
from invenio.websearch_external_collections_getter import HTTPAsyncPageGetter, async_download
def perform_request_display(uid,
category=CFG_WEBBASKET_CATEGORIES['PRIVATE'],
selected_topic=0,
selected_group_id=0,
ln=CFG_SITE_LANG):
"""Display all the baskets of given category, topic or group.
@param uid: user id
@param category: selected category (see webbasket_config.py)
@param selected_topic: # of selected topic to display baskets
@param selected_group_id: id of group to display baskets
@param ln: language"""
warnings = []
errors = []
baskets_html = []
baskets = []
_ = gettext_set_language(ln)
nb_groups = db.count_groups_user_member_of(uid)
nb_external_baskets = db.count_external_baskets(uid)
selectionbox = ''
infobox = ''
if category == CFG_WEBBASKET_CATEGORIES['EXTERNAL']:
baskets = db.get_external_baskets_infos(uid)
if len(baskets):
map(list, baskets)
else:
category = CFG_WEBBASKET_CATEGORIES['PRIVATE']
if category == CFG_WEBBASKET_CATEGORIES['GROUP']:
groups = db.get_group_infos(uid)
if len(groups):
if selected_group_id == 0 and len(groups):
selected_group_id = groups[0][0]
selectionbox = webbasket_templates.tmpl_group_selection(groups,
selected_group_id,
ln)
baskets = db.get_group_baskets_infos(selected_group_id)
def adapt_group_rights(item):
"""Suppress unused element in tuple."""
out = list(item)
if out[-1] == uid:
out[-2] = CFG_WEBBASKET_SHARE_LEVELS['MANAGE']
return out[:-1]
baskets = map(adapt_group_rights, baskets)
else:
category = CFG_WEBBASKET_CATEGORIES['PRIVATE']
if category == CFG_WEBBASKET_CATEGORIES['PRIVATE']:
topics_list = db.get_personal_topics_infos(uid)
if not selected_topic and len(topics_list):
selected_topic = 0
selectionbox = webbasket_templates.tmpl_topic_selection(topics_list,
selected_topic,
ln)
if selected_topic >= len(topics_list):
selected_topic = len(topics_list) - 1
if len(topics_list) > 0:
baskets = db.get_personal_baskets_infos(uid, topics_list[selected_topic][0])
else:
baskets = []
def add_manage_rights(item):
""" Convert a tuple to a list and add rights"""
out = list(item)
out.append(CFG_WEBBASKET_SHARE_LEVELS['MANAGE'])
return out
baskets = map(add_manage_rights, baskets)
bskids = []
for basket in baskets:
bskids.append(basket[0])
levels = dict(db.is_shared_to(bskids))
create_link = ''
if category == CFG_WEBBASKET_CATEGORIES['PRIVATE']:
create_link = webbasket_templates.tmpl_create_basket_link(selected_topic, ln)
infobox = webbasket_templates.tmpl_baskets_infobox(map(lambda x: (x[0], x[1], x[2]),
baskets),
create_link,
ln)
for (bskid, name, date_modification,
nb_views, nb_items, last_added, share_level) in baskets:
(bsk_html, bsk_e, bsk_w) = __display_basket(bskid,
name,
date_modification,
nb_views,
nb_items,
last_added,
share_level,
levels[bskid],
category,
selected_topic,
selected_group_id,
ln)
baskets_html.append(bsk_html)
errors.extend(bsk_e)
warnings.extend(bsk_w)
body = webbasket_templates.tmpl_display(selectionbox,
infobox,
baskets_html,
category,
nb_groups,
nb_external_baskets,
ln)
return (body, errors, warnings)
def __display_basket(bskid, name, date_modification, nb_views,
nb_items, last_added,
share_level, group_sharing_level,
category=CFG_WEBBASKET_CATEGORIES['PRIVATE'],
selected_topic=0, selected_group_id=0,
ln=CFG_SITE_LANG):
"""Private function. Display a basket giving its category and topic or group.
@param share_level: rights user has on basket
@param group_sharing_level: None if basket is not shared,
0 if public basket,
> 0 if shared to usergroups but not public.
@param category: selected category (see webbasket_config.py)
@param selected_topic: # of selected topic to display baskets
@param selected_group_id: id of group to display baskets
@param ln: language"""
_ = gettext_set_language(ln)
errors = []
warnings = []
nb_bsk_cmts = 0
last_cmt = _("N/A")
records = []
cmt_dates = []
date_modification = convert_datetext_to_dategui(date_modification, ln)
items = db.get_basket_content(bskid, 'hb')
external_recids = []
for (recid, nb_cmt, last_cmt, ext_val, int_val, score) in items:
cmt_dates.append(convert_datetext_to_datestruct(last_cmt))
last_cmt = convert_datetext_to_dategui(last_cmt, ln)
val = ''
nb_bsk_cmts += nb_cmt
if recid < 0:
if ext_val:
val = decompress(ext_val)
else:
external_recids.append(recid)
else:
if int_val:
val = decompress(int_val)
else:
val = format_record(recid, 'hb', on_the_fly=True)
records.append((recid, nb_cmt, last_cmt, val, score))
if external_recids:
external_records = format_external_records(external_recids, 'hb')
for external_record in external_records:
for record in records:
if record[0] == -external_record[0]:
idx = records.index(record)
tuple_to_list = list(records.pop(idx))
tuple_to_list[3] = external_record[1]
records.insert(idx,tuple(tuple_to_list))
break
if len(cmt_dates) > 0:
last_cmt = convert_datestruct_to_dategui(max(cmt_dates), ln)
body = webbasket_templates.tmpl_basket(bskid,
name,
date_modification,
nb_views,
nb_items, last_added,
(check_sufficient_rights(share_level, CFG_WEBBASKET_SHARE_LEVELS['READITM']),
check_sufficient_rights(share_level, CFG_WEBBASKET_SHARE_LEVELS['MANAGE']),
check_sufficient_rights(share_level, CFG_WEBBASKET_SHARE_LEVELS['READCMT']),
check_sufficient_rights(share_level, CFG_WEBBASKET_SHARE_LEVELS['ADDITM']),
check_sufficient_rights(share_level, CFG_WEBBASKET_SHARE_LEVELS['DELITM'])),
nb_bsk_cmts, last_cmt,
group_sharing_level,
category, selected_topic, selected_group_id,
records,
ln)
return (body, errors, warnings)
def perform_request_display_item(uid, bskid, recid, format='hb',
category=CFG_WEBBASKET_CATEGORIES['PRIVATE'],
topic=0, group_id=0, ln=CFG_SITE_LANG):
"""Display an item of a basket of given category, topic or group.
@param uid: user id
@param bskid: basket_id
@param recid: record id
@param format: format of the record (hb, hd, etc.)
@param category: selected category (see webbasket_config.py)
@param topic: # of selected topic to display baskets
@param group_id: id of group to display baskets
@param ln: language"""
body = ''
errors = []
warnings = []
rights = db.get_max_user_rights_on_basket(uid, bskid)
if not(check_sufficient_rights(rights, CFG_WEBBASKET_SHARE_LEVELS['READITM'])):
errors.append('ERR_WEBBASKET_NO_RIGHTS')
return (body, errors, warnings)
if category == CFG_WEBBASKET_CATEGORIES['PRIVATE']:
topics_list = db.get_personal_topics_infos(uid)
if not topic and len(topics_list):
topic = 0
topicsbox = webbasket_templates.tmpl_topic_selection(topics_list, topic, ln)
elif category == CFG_WEBBASKET_CATEGORIES['GROUP']:
groups = db.get_group_infos(uid)
if group_id == 0 and len(groups):
group_id = groups[0][0]
topicsbox = webbasket_templates.tmpl_group_selection(groups, group_id, ln)
else:
topicsbox = ''
record = db.get_basket_record(bskid, recid, format)
comments = db.get_comments(bskid, recid)
group_sharing_level = None
levels = db.is_shared_to(bskid)
if len(levels):
group_sharing_level = levels[0][1]
basket = db.get_basket_general_infos(bskid)
if not(len(basket)):
errors.append('ERR_WEBBASKET_DB_ERROR')
return (body, errors, warnings)
item_html = webbasket_templates.tmpl_item(basket,
recid, record, comments,
group_sharing_level,
(check_sufficient_rights(rights, CFG_WEBBASKET_SHARE_LEVELS['READCMT']),
check_sufficient_rights(rights, CFG_WEBBASKET_SHARE_LEVELS['ADDCMT']),
check_sufficient_rights(rights, CFG_WEBBASKET_SHARE_LEVELS['DELCMT'])),
selected_category=category, selected_topic=topic, selected_group_id=group_id,
ln=ln)
body = webbasket_templates.tmpl_display(topicsbox=topicsbox, baskets=[item_html],
selected_category=category,
nb_groups=db.count_groups_user_member_of(uid),
nb_external_baskets=db.count_external_baskets(uid),
ln=ln)
return (body, errors, warnings)
def perform_request_write_comment(uid, bskid, recid, cmtid=0,
category=CFG_WEBBASKET_CATEGORIES['PRIVATE'],
topic=0, group_id=0,
ln=CFG_SITE_LANG):
"""Display a comment writing form
@param uid: user id
@param bskid: basket id
@param recid: record id (comments are on a specific record in a specific basket)
@param cmtid: if provided this comment is a reply to comment cmtid.
@param category: selected category
@param topic: selected topic
@param group_id: selected group id
@param ln: language
"""
body = ''
warnings = []
errors = []
textual_msg = '' # initial value in replies
html_msg = '' # initial value in replies (if FCKeditor)
title = '' # initial title in replies
if not check_user_can_comment(uid, bskid):
errors.append(('ERR_WEBBASKET_CANNOT_COMMENT'))
return (body, errors, warnings)
if cmtid:
# this is a reply to another comment
comment = db.get_comment(cmtid)
if comment:
# Title
if comment[2]:
title = 'Re: ' + comment[2]
# Build two msg: one mostly textual, the other one with HTML markup, for the FCKeditor.
textual_msg = webbasket_templates.tmpl_quote_comment_textual(comment[2], # title
uid,
comment[0], # nickname
comment[4], # date
comment[3],
ln)
html_msg = webbasket_templates.tmpl_quote_comment_html(comment[2], # title
uid,
comment[0], # nickname
comment[4], # date
comment[3],
ln)
else:
warning = (CFG_WEBBASKET_WARNING_MESSAGES['ERR_WEBBASKET_cmtid_INVALID'], cmtid)
warnings.append(warning)
record = db.get_basket_record(bskid, recid, 'hb')
# Check that user can attach file. To simplify we use the same
# checking as in WebComment, though it is not completely adequate.
user_info = collect_user_info(uid)
can_attach_files = check_user_can_attach_file_to_comments(user_info, recid)
body = webbasket_templates.tmpl_write_comment(bskid=bskid,
recid=recid,
cmt_title=title,
cmt_body_textual=textual_msg,
cmt_body_html=html_msg,
record = record,
selected_category=category,
selected_topic=topic,
selected_group_id=group_id,
warnings=warnings,
can_attach_files=can_attach_files)
if category == CFG_WEBBASKET_CATEGORIES['PRIVATE']:
topics_list = db.get_personal_topics_infos(uid)
if not topic and len(topics_list):
topic = 0
topicsbox = webbasket_templates.tmpl_topic_selection(topics_list, topic, ln)
elif category == CFG_WEBBASKET_CATEGORIES['GROUP']:
groups = db.get_group_infos(uid)
if group_id == 0 and len(groups):
group_id = groups[0][0]
topicsbox = webbasket_templates.tmpl_group_selection(groups, group_id, ln)
else:
topicsbox = ''
body = webbasket_templates.tmpl_display(topicsbox, '', [ body ], category, ln)
return (body, errors, warnings)
def perform_request_save_comment(uid, bskid, recid, title='', text='',
ln=CFG_SITE_LANG,
editor_type='textarea'):
""" Save a given comment if able to.
@param uid: user id (int)
@param bskid: basket id (int)
@param recid: record id (int)
@param title: title of comment (string)
@param text: comment's body (string)
@param ln: language (string)
@param editor_type: the kind of editor/input used for the comment: 'textarea', 'fckeditor'
@return: (errors, infos) where errors: list of errors while saving
infos: list of informations to display"""
_ = gettext_set_language(ln)
errors = []
infos = []
if not check_user_can_comment(uid, bskid):
errors.append(('ERR_WEBBASKET_CANNOT_COMMENT'))
return (errors, infos)
if editor_type == 'fckeditor':
# Here we remove the line feeds introduced by FCKeditor (they
# have no meaning for the user) and replace the HTML line
# breaks by linefeeds, so that we are close to an input that
# would be done without the FCKeditor. That's much better if a
# reply to a comment is made with a browser that does not
# support FCKeditor.
text = text.replace('\n', '').replace('\r', '').replace(' ', '\n')
if not(db.save_comment(uid, bskid, recid, title, text)):
errors.append(('ERR_WEBBASKET_DB_ERROR'))
else:
infos.append(_('Your comment has been successfully posted'))
return (errors, infos)
def perform_request_delete_comment(uid, bskid, recid, cmtid):
"""Delete comment cmtid on record recid for basket bskid."""
errors = []
if __check_user_can_perform_action(uid, bskid, CFG_WEBBASKET_SHARE_LEVELS['DELCMT']):
db.delete_comment(bskid, recid, cmtid)
else:
errors.append('ERR_WEBBASKET_NO_RIGHTS')
return errors
def perform_request_add(uid, recids=[], colid=0, bskids=[], referer='',
new_basket_name='', new_topic_name='', create_in_topic='',
ln=CFG_SITE_LANG):
"""Add records to baskets
@param uid: user id
@param recids: list of records to add
@param colid: in case of external collections, the id of the collection the records belong to
@param bskids: list of baskets to add records to. if not provided, will return a
page where user can select baskets
@param referer: URL of the referring page
@param new_basket_name: add record to new basket
@param new_topic_name: new basket goes into new topic
@param create_in_topic: # of topic to put basket into
@param ln: language
@return: (body, errors, warnings) tuple
"""
body = ''
errors = []
warnings = []
if not(type(recids) == list):
recids = [recids]
validated_recids = []
if colid == 0:
# local records
for recid in recids:
recid = int(recid)
if record_exists(recid) == 1:
validated_recids.append(recid)
user_info = collect_user_info(uid)
for recid in validated_recids:
(auth_code, auth_msg) = check_user_can_view_record(user_info, recid)
if auth_code:
# User not authorized to view record
validated_recids.remove(recid)
warnings.append(('WRN_WEBBASKET_NO_RIGHTS_TO_ADD_THIS_RECORD', recid))
elif colid > 0:
# external records, no need to validate
validated_recids = recids
elif colid == -1:
# external url / resource.
# TODO: how to handle?
pass
if not(len(validated_recids)):
warnings.append('WRN_WEBBASKET_NO_RECORD')
body += webbasket_templates.tmpl_warnings(warnings, ln)
if referer and not(referer.find(CFG_SITE_URL) == -1):
body += webbasket_templates.tmpl_back_link(referer, ln)
return (body, errors, warnings)
if new_basket_name != '':
new_topic_name = new_topic_name.strip()
if new_topic_name:
topic = new_topic_name
elif create_in_topic != -1:
topics = map(lambda x: x[0], db.get_personal_topics_infos(uid))
try:
topic = topics[create_in_topic]
except IndexError:
topic = ''
else:
topic = ''
warnings.append('WRN_WEBBASKET_NO_GIVEN_TOPIC')
body += webbasket_templates.tmpl_warnings(warnings, ln)
bskids = []
if topic:
id_bsk = db.create_basket(uid, new_basket_name, topic)
bskids.append(id_bsk)
if bskids:
bskids = [int(bskid) for bskid in bskids if int(bskid) > 0]
if bskids:
for bskid in bskids:
if not(__check_user_can_perform_action(uid,
bskid,
CFG_WEBBASKET_SHARE_LEVELS['ADDITM'])):
errors.append('ERR_WEBBASKET_NO_RIGHTS')
return (body, errors, warnings)
nb_modified_baskets = db.add_to_basket(uid, validated_recids, bskids)
body = webbasket_templates.tmpl_added_to_basket(nb_modified_baskets, ln)
body_tmp, warnings_temp, errors_tmp = perform_request_display(uid,
category=CFG_WEBBASKET_CATEGORIES['PRIVATE'],
selected_topic=create_in_topic != -1 and create_in_topic or 0,
selected_group_id=0,
ln=CFG_SITE_LANG)
body += body_tmp
warnings += warnings_temp
errors += errors_tmp
body += webbasket_templates.tmpl_back_link(referer, ln)
return (body, warnings, errors)
else:
warnings.append('WRN_WEBBASKET_NO_BASKET_SELECTED')
body += webbasket_templates.tmpl_warnings(warnings, ln)
# Display basket_selection
personal_baskets = db.get_all_personal_baskets_names(uid)
group_baskets = db.get_all_group_baskets_names(uid)
external_baskets = db.get_all_external_baskets_names(uid)
topics = map(lambda x: x[0], db.get_personal_topics_infos(uid))
body += webbasket_templates.tmpl_add(recids=validated_recids,
colid=colid,
personal_baskets=personal_baskets,
group_baskets=group_baskets,
external_baskets=external_baskets,
topics=topics,
referer=referer,
ln=ln)
body += webbasket_templates.tmpl_back_link(referer, ln)
return (body, errors, warnings)
def perform_request_delete(uid, bskid, confirmed=0,
category=CFG_WEBBASKET_CATEGORIES['PRIVATE'],
selected_topic=0, selected_group_id=0,
ln=CFG_SITE_LANG):
"""Delete a given basket.
@param uid: user id (user has to be owner of this basket)
@param bskid: basket id
@param confirmed: if 0 will return a confirmation page; if 1 will delete basket.
@param category: category currently displayed
@param selected_topic: topic currently displayed
@param selected_group_id: if category is group, id of the group currently displayed
@param ln: language"""
body = ''
errors = []
warnings = []
if not(db.check_user_owns_baskets(uid, [bskid])):
errors.append(('ERR_WEBBASKET_NO_RIGHTS',))
return (body, errors, warnings)
if confirmed:
success = db.delete_basket(bskid)
if not success:
errors.append(('ERR_WEBBASKET_DB_ERROR',))
else:
body = webbasket_templates.tmpl_confirm_delete(bskid,
db.count_subscribers(uid, bskid),
category,
selected_topic, selected_group_id,
ln)
return (body, errors, warnings)
def delete_record(uid, bskid, recid):
"""Delete a given record in a given basket.
@param uid: user id (user has to have sufficient rights on basket
@param bskid: basket id
@param recid: record id
"""
if __check_user_can_perform_action(uid,
bskid,
CFG_WEBBASKET_SHARE_LEVELS['DELITM']):
db.delete_item(bskid, recid)
def move_record(uid, bskid, recid, direction):
"""Move a record up or down in a basket (change score).
@param uid: user id (user has to have manage rights over this basket)
@param bskid: basket id
@param recid: record we want to move
@param direction: CFG_WEBBASKET_ACTIONS['UP'] or CFG_WEBBASKET_ACTIONS['DOWN']
"""
if __check_user_can_perform_action(uid,
bskid,
CFG_WEBBASKET_SHARE_LEVELS['MANAGE']):
db.move_item(bskid, recid, direction)
def perform_request_edit(uid, bskid, topic=0, new_name='',
new_topic = '', new_topic_name='',
groups=[], external='',
ln=CFG_SITE_LANG):
"""Interface for management of basket. If names, groups or external is
provided, will save new rights into database, else will provide interface.
@param uid: user id (user has to have sufficient rights on this basket
@param bskid: basket id to change rights on
@param topic: topic currently used (int)
@param new_name: new name of basket
@param new_topic: topic in which to move basket (int),
new_topic_name must be left blank
@param new_topic_name: new topic in which to move basket
(will overwrite param new_topic)
@param groups: list of strings formed in this way: group_id + '_' + rights
@param external: rights for everybody (can be 'NO')
@param ln: language
"""
body = ''
errors = []
warnings = []
rights = db.get_max_user_rights_on_basket(uid, bskid)
if rights != CFG_WEBBASKET_SHARE_LEVELS['MANAGE']:
errors.append(('ERR_WEBBASKET_NO_RIGHTS',))
return (body, errors, warnings)
bsk_name = db.get_basket_name(bskid)
if not(groups) and not(external) and not(new_name) and not(new_topic) and not(new_topic_name):
# display interface
topics = map(lambda x: x[0], db.get_personal_topics_infos(uid))
groups_rights = db.get_groups_subscribing_to_basket(bskid)
external_rights = ''
if groups_rights and groups_rights[0][0] == 0:
external_rights = groups_rights[0][2]
groups_rights = groups_rights[1:]
display_delete = db.check_user_owns_baskets(uid, bskid)
display_general = display_delete
if isGuestUser(uid):
display_sharing = 0
else:
display_sharing = 1
body = webbasket_templates.tmpl_edit(bskid=bskid, bsk_name=bsk_name,
display_general=display_general,
topics=topics, topic=topic,
display_delete=display_delete,
display_sharing=display_sharing,
groups_rights=groups_rights,
external_rights=external_rights,
ln=ln)
else:
out_groups = {}
if len(groups):
for group in groups:
(group_id, group_rights) = group.split('_')
out_groups[group_id] = group_rights
out_groups['0'] = external
if not(isGuestUser(uid)):
db.update_rights(bskid, out_groups)
if new_name != bsk_name:
db.rename_basket(bskid, new_name)
if new_topic_name:
db.move_baskets_to_topic(uid, bskid, new_topic_name)
elif new_topic != -1:
if db.check_user_owns_baskets(uid, bskid):
topics = map(lambda x: x[0], db.get_personal_topics_infos(uid))
try:
new_topic_name = topics[new_topic]
db.move_baskets_to_topic(uid, bskid, new_topic_name)
except:
errors.append(('ERR_WEBBASKET_DB_ERROR'))
else:
topic = 0
errors.append(('ERR_WEBBASKET_NOT_OWNER'))
return (body, errors, warnings)
def perform_request_add_group(uid, bskid, topic=0, group_id=0, ln=CFG_SITE_LANG):
"""If group id is specified, share basket bskid to this group with
READITM rights;
else return a page for selection of a group.
@param uid: user id (selection only of groups user is member of)
@param bskid: basket id
@param topic: topic currently displayed
@param group_id: id of group to share basket to
@param ln: language
"""
if group_id:
db.share_basket_with_group(bskid,
group_id,
CFG_WEBBASKET_SHARE_LEVELS['READITM'])
else:
groups = db.get_groups_user_member_of(uid)
body = webbasket_templates.tmpl_add_group(bskid, topic, groups, ln)
return body
def perform_request_create_basket(uid,
new_basket_name='',
new_topic_name='', create_in_topic=-1,
topic_number=-1,
ln=CFG_SITE_LANG):
"""if new_basket_name and topic infos are given create a basket and return topic number,
else return (body, errors, warnings) tuple of basket creation form.
@param uid: user id (int)
@param new_basket_name: name of the basket to create (str)
@param new_topic_name: name of new topic to create new basket in (str)
@param create_in_topic: identification number of topic to create new basket in (int)
@param topic_number: number of topic to preselect on the creation form.
@pram ln: language
"""
if new_basket_name and (new_topic_name or create_in_topic != -1):
topics_infos = map(lambda x: x[0], db.get_personal_topics_infos(uid))
new_topic_name = new_topic_name.strip()
if new_topic_name:
topic = new_topic_name
else:
try:
topic = topics_infos[create_in_topic]
except IndexError:
return 0
db.create_basket(uid, new_basket_name, topic)
topics = map(lambda x: x[0], topics_infos)
try:
return topics.index(topic)
except ValueError:
return 0
else:
topics = map(lambda x: x[0], db.get_personal_topics_infos(uid))
if topic_number in range (0, len(topics)):
create_in_topic = topics[topic_number]
body = webbasket_templates.tmpl_create_basket(new_basket_name,
new_topic_name,
create_in_topic,
topics,
ln)
return (body, [], [])
def perform_request_display_public(bskid=0, of='hb', ln=CFG_SITE_LANG):
"""return html representation of a public basket
@param bskid: basket id
@param of: format
@param ln: language"""
_ = gettext_set_language(ln)
body = ''
errors = []
warnings = []
basket = db.get_public_basket_infos(bskid)
if of[0] == 'x':
items = []
if len(basket) == 7:
content = db.get_basket_content(bskid)
for item in content:
# TODO: return xml for external records too?
items.append(format_record(item[0], of))
return webbasket_templates.tmpl_xml_basket(items)
if len(basket) == 7:
items = db.get_basket_content(bskid)
external_recids = []
last_cmt = _("N/A")
records = []
cmt_dates = []
for (recid, nb_cmt, last_cmt, ext_val, int_val, score) in items:
cmt_dates.append(convert_datetext_to_datestruct(last_cmt))
last_cmt = convert_datetext_to_dategui(last_cmt, ln)
val = ''
if recid < 0:
if ext_val:
val = decompress(ext_val)
else:
external_recids.append(recid)
else:
if int_val:
val = format_record(recid, 'hb')
records.append((recid, nb_cmt, last_cmt, val, score))
if external_recids:
external_records = format_external_records(external_recids, 'hb')
for external_record in external_records:
for record in records:
if record[0] == -external_record[0]:
idx = records.index(record)
tuple_to_list = list(records.pop(idx))
tuple_to_list[3] = external_record[1]
records.insert(idx,tuple(tuple_to_list))
break
body = webbasket_templates.tmpl_display_public(basket, records, ln)
else:
errors.append('ERR_WEBBASKET_RESTRICTED_ACCESS')
return (body, errors, warnings)
def perform_request_list_public_baskets(inf_limit=0, order=1, asc=1, ln=CFG_SITE_LANG):
"""Display list of public baskets.
@param inf_limit: display baskets from inf_limit
@param order: 1: order by name of basket, 2: number of views, 3: owner
@param asc: ascending order if 1, descending if 0
@param ln: language
"""
errors = []
warnings = []
total_baskets = db.count_public_baskets()
baskets = db.get_public_baskets_list(inf_limit, CFG_WEBBASKET_MAX_NUMBER_OF_DISPLAYED_BASKETS, order, asc)
body = webbasket_templates.tmpl_display_list_public_baskets(baskets, inf_limit, total_baskets, order, asc, ln)
return (body, errors, warnings)
def perform_request_subscribe(uid, bskid):
"""subscribe to external basket bskid"""
errors = []
if db.is_public(bskid):
db.subscribe(uid, bskid)
else:
errors.append('ERR_WEBBASKET_RESTRICTED_ACCESS')
return errors
def perform_request_unsubscribe(uid, bskid):
"""unsubscribe from external basket bskid"""
db.unsubscribe(uid, bskid)
def check_user_can_comment(uid, bskid):
""" Private function. check if a user can comment """
min_right = CFG_WEBBASKET_SHARE_LEVELS['ADDCMT']
rights = db.get_max_user_rights_on_basket(uid, bskid)
if rights:
if CFG_WEBBASKET_SHARE_LEVELS_ORDERED.index(rights) >= CFG_WEBBASKET_SHARE_LEVELS_ORDERED.index(min_right):
return 1
return 0
def __check_user_can_perform_action(uid, bskid, rights):
""" Private function, check if a user has sufficient rights"""
min_right = rights
rights = db.get_max_user_rights_on_basket(uid, bskid)
if rights:
if CFG_WEBBASKET_SHARE_LEVELS_ORDERED.index(rights) >= CFG_WEBBASKET_SHARE_LEVELS_ORDERED.index(min_right):
return 1
return 0
def check_sufficient_rights(rights_user_has, rights_needed):
"""Private function, check if the rights are sufficient."""
try:
out = CFG_WEBBASKET_SHARE_LEVELS_ORDERED.index(rights_user_has) >= \
CFG_WEBBASKET_SHARE_LEVELS_ORDERED.index(rights_needed)
except ValueError:
out = 0
return out
def create_guest_warning_box(ln=CFG_SITE_LANG):
"""return a warning message about logging into system"""
return webbasket_templates.tmpl_create_guest_warning_box(ln)
def create_personal_baskets_selection_box(uid,
html_select_box_name='baskets',
selected_bskid=None,
ln=CFG_SITE_LANG):
"""Return HTML box for basket selection. Only for personal baskets.
@param uid: user id
@param html_select_box_name: name used in html form
@param selected_bskid: basket currently selected
@param ln: language
"""
baskets = db.get_all_personal_baskets_names(uid)
return webbasket_templates.tmpl_personal_baskets_selection_box(
baskets,
html_select_box_name,
selected_bskid,
ln)
def create_basket_navtrail(uid,
category=CFG_WEBBASKET_CATEGORIES['PRIVATE'],
topic=0, group=0,
bskid=0, ln=CFG_SITE_LANG):
"""display navtrail for basket navigation.
@param uid: user id (int)
@param category: selected category (see CFG_WEBBASKET_CATEGORIES)
@param topic: selected topic # if personal baskets
@param group: selected group id for displaying (int)
@param bskid: basket id (int)
@param ln: language"""
_ = gettext_set_language(ln)
out = ''
if category == CFG_WEBBASKET_CATEGORIES['PRIVATE']:
out += ' > '\
'%s'
out %= (CFG_SITE_URL,
'category=' + category + '&ln=' + ln,
_("Personal baskets"))
topics = map(lambda x: x[0], db.get_personal_topics_infos(uid))
if topic in range(0, len(topics)):
out += ' > '
out += ''\
'%s'
out %= (CFG_SITE_URL,
'category=' + category + '&topic=' + \
str(topic) + '&ln=' + ln,
cgi.escape(topics[topic]))
if bskid:
basket = db.get_public_basket_infos(bskid)
if basket:
out += ' > '
out += '%s'
out %= (CFG_SITE_URL,
'category=' + category + '&topic=' + \
str(topic) + '&ln=' + ln + '#bsk' + str(bskid),
cgi.escape(basket[1]))
elif category == CFG_WEBBASKET_CATEGORIES['GROUP']:
out += ' > '\
'%s'
out %= (CFG_SITE_URL, 'category=' + category + '&ln=' + ln, _("Group baskets"))
groups = db.get_group_infos(uid)
if group:
groups = filter(lambda x: x[0] == group, groups)
if len(groups):
out += ' > '
out += '%s'
out %= (CFG_SITE_URL,
'category=' + category + '&group=' + \
str(group) + '&ln=' + ln,
cgi.escape(groups[0][1]))
if bskid:
basket = db.get_public_basket_infos(bskid)
if basket:
out += ' > '
out += '%s'
out %= (CFG_SITE_URL,
'category=' + category + '&group=' + \
str(group) + '&ln=' + ln + '#bsk' + str(bskid),
cgi.escape(basket[1]))
elif category == CFG_WEBBASKET_CATEGORIES['EXTERNAL']:
out += ' > '\
'%s'
out %= (CFG_SITE_URL,
'category=' + category + '&ln=' + ln,
_("Others' baskets"))
if bskid:
basket = db.get_public_basket_infos(bskid)
if basket:
out += ' > '
out += ''\
'%s'
out %= (CFG_SITE_URL,
'category=' + category + '&ln=' + ln + \
'#bsk' + str(bskid),
cgi.escape(basket[1]))
return out
def create_infobox(infos=[]):
"""Create an infos box. infos param should be a list of strings.
Return formatted infos"""
return webbasket_templates.tmpl_create_infobox(infos)
def account_list_baskets(uid, ln=CFG_SITE_LANG):
"""Display baskets informations on account page"""
_ = gettext_set_language(ln)
(personal, group, external) = db.count_baskets(uid)
link = '%s'
base_url = CFG_SITE_URL + '/yourbaskets/display?category=%s&ln=' + ln
personal_text = personal
if personal:
url = base_url % CFG_WEBBASKET_CATEGORIES['PRIVATE']
personal_text = link % (url, personal_text)
group_text = group
if group:
url = base_url % CFG_WEBBASKET_CATEGORIES['GROUP']
group_text = link % (url, group_text)
external_text = external
if external:
url = base_url % CFG_WEBBASKET_CATEGORIES['EXTERNAL']
else:
url = CFG_SITE_URL + '/yourbaskets/list_public_baskets?ln=' + ln
external_text = link % (url, external_text)
out = _("You have %(x_nb_perso)s personal baskets and are subscribed to %(x_nb_group)s group baskets and %(x_nb_public)s other users public baskets.") %\
{'x_nb_perso': personal_text,
'x_nb_group': group_text,
'x_nb_public': external_text}
return out
## External records functions
def format_external_records(recids, of='hb'):
"""Given a list of external records' recids, this function returns a list of tuples
with each recid and the actual formatted record using the selected output format.
It also stores the formatted record in the database for future use."""
if type(recids) is not list:
recids = [recids]
records_grouped_by_collection = db.get_external_records_by_collection(recids)
formatted_records = []
for records in records_grouped_by_collection:
external_records = fetch_and_store_external_records(records, of)
formatted_records.extend(external_records)
return tuple(formatted_records)
def fetch_and_store_external_records(records, of="hb"):
"""Function that fetches the formatted records for one collection and stores them
into the database. It also calculates and stores the original external url for each
record."""
results = []
parsed_results_list = []
parsed_results_dict = {}
formatted_records = []
ids = records[0].split(",")
external_ids = records[1].split(",")
collection_name = get_collection_name_by_id(records[2])
collection_engine_set = select_hosted_search_engines(collection_name)
collection_engine = collection_engine_set.pop()
external_ids_urls = collection_engine.build_record_urls(external_ids)
external_urls = [external_id_url[1] for external_id_url in external_ids_urls]
external_urls_dict = {}
for (local_id, url) in zip(ids,external_urls):
external_urls_dict[local_id] = url
db.store_external_urls(external_urls_dict)
url = collection_engine.build_search_url(None, req_args=external_ids)
pagegetters = [HTTPAsyncPageGetter(url)]
def finished(pagegetter, data, time):
"""Function to be called when a page has been downloaded."""
results.append(pagegetter)
finished_list = async_download(pagegetters, finish_function=finished, timeout=CFG_EXTERNAL_COLLECTION_TIMEOUT)
if finished_list[0]:
collection_engine.parser.parse_and_get_results(results[0].data, feedonly=True)
(parsed_results_list, parsed_results_dict) = collection_engine.parser.parse_and_extract_records(of=of)
for (id,external_id) in zip(ids,external_ids):
formatted_records.append((int(id), parsed_results_dict[external_id]))
db.store_external_records(formatted_records, of)
else:
for (id,external_id) in zip(ids,external_ids):
formatted_records.append((int(id), "There was a timeout when fetching the record."))
return formatted_records
diff --git a/modules/websearch/lib/search_engine.py b/modules/websearch/lib/search_engine.py
index 1bf517221..0e5adb194 100644
--- a/modules/websearch/lib/search_engine.py
+++ b/modules/websearch/lib/search_engine.py
@@ -1,4884 +1,4890 @@
# -*- coding: utf-8 -*-
## This file is part of CDS Invenio.
## Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 CERN.
##
## CDS Invenio is free software; you can redistribute it and/or
## modify it under the terms of the GNU General Public License as
## published by the Free Software Foundation; either version 2 of the
## License, or (at your option) any later version.
##
## CDS Invenio is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with CDS Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
# pylint: disable-msg=C0301
"""CDS Invenio Search Engine in mod_python."""
__lastupdated__ = """$Date$"""
__revision__ = "$Id$"
## import general modules:
import cgi
import cStringIO
import copy
import string
import os
import re
import time
import urllib
import urlparse
import zlib
import sys
if sys.hexversion < 0x2040000:
# pylint: disable-msg=W0622
from sets import Set as set
# pylint: enable-msg=W0622
## import CDS Invenio stuff:
from invenio.config import \
CFG_CERN_SITE, \
CFG_OAI_ID_FIELD, \
CFG_WEBCOMMENT_ALLOW_REVIEWS, \
CFG_WEBSEARCH_CALL_BIBFORMAT, \
CFG_WEBSEARCH_CREATE_SIMILARLY_NAMED_AUTHORS_LINK_BOX, \
CFG_WEBSEARCH_FIELDS_CONVERT, \
CFG_WEBSEARCH_NB_RECORDS_TO_SORT, \
CFG_WEBSEARCH_SEARCH_CACHE_SIZE, \
CFG_WEBSEARCH_USE_JSMATH_FOR_FORMATS, \
CFG_WEBSEARCH_USE_ALEPH_SYSNOS, \
CFG_BIBUPLOAD_SERIALIZE_RECORD_STRUCTURE, \
CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG, \
CFG_BIBRANK_SHOW_DOWNLOAD_GRAPHS, \
CFG_SITE_LANG, \
CFG_SITE_NAME, \
CFG_LOGDIR, \
CFG_SITE_URL
from invenio.search_engine_config import InvenioWebSearchUnknownCollectionError
from invenio.bibrecord import create_record
from invenio.bibrank_record_sorter import get_bibrank_methods, rank_records, is_method_valid
from invenio.bibrank_downloads_similarity import register_page_view_event, calculate_reading_similarity_list
from invenio.bibindex_engine_stemmer import stem
from invenio.bibformat import format_record, format_records, get_output_format_content_type, create_excel
from invenio.bibformat_config import CFG_BIBFORMAT_USE_OLD_BIBFORMAT
from invenio.bibrank_downloads_grapher import create_download_history_graph_and_box
from invenio.data_cacher import DataCacher
from invenio.websearch_external_collections import print_external_results_overview, perform_external_collection_search
from invenio.access_control_admin import acc_get_action_id
from invenio.access_control_config import VIEWRESTRCOLL, \
CFG_ACC_GRANT_AUTHOR_RIGHTS_TO_EMAILS_IN_TAGS
from invenio.websearchadminlib import get_detailed_page_tabs
from invenio.intbitset import intbitset as HitSet
from invenio.dbquery import DatabaseError
from invenio.access_control_engine import acc_authorize_action
from invenio.errorlib import register_exception
from invenio.textutils import encode_for_xml
import invenio.template
webstyle_templates = invenio.template.load('webstyle')
webcomment_templates = invenio.template.load('webcomment')
from invenio.bibrank_citation_searcher import calculate_cited_by_list, \
calculate_co_cited_with_list, get_records_with_num_cites, get_self_cited_by
from invenio.bibrank_citation_grapher import create_citation_history_graph_and_box
from invenio.dbquery import run_sql, run_sql_cached, get_table_update_time, Error
from invenio.webuser import getUid, collect_user_info
from invenio.webpage import pageheaderonly, pagefooteronly, create_error_box
from invenio.messages import gettext_set_language
from invenio.search_engine_query_parser import SearchQueryParenthesisedParser, \
InvenioWebSearchQueryParserException, SpiresToInvenioSyntaxConverter
from invenio import webinterface_handler_wsgi_utils as apache
try:
import invenio.template
websearch_templates = invenio.template.load('websearch')
except:
pass
from invenio.websearch_external_collections import calculate_hosted_collections_results, do_calculate_hosted_collections_results
from invenio.websearch_external_collections_config import CFG_HOSTED_COLLECTION_TIMEOUT_ANTE_SEARCH
from invenio.websearch_external_collections_config import CFG_HOSTED_COLLECTION_TIMEOUT_POST_SEARCH
+from invenio.websearch_external_collections_config import CFG_EXTERNAL_COLLECTION_MAXRESULTS
## global vars:
cfg_nb_browse_seen_records = 100 # limit of the number of records to check when browsing certain collection
cfg_nicely_ordered_collection_list = 0 # do we propose collection list nicely ordered or alphabetical?
## precompile some often-used regexp for speed reasons:
re_word = re.compile('[\s]')
re_quotes = re.compile('[\'\"]')
re_doublequote = re.compile('\"')
re_equal = re.compile('\=')
re_logical_and = re.compile('\sand\s', re.I)
re_logical_or = re.compile('\sor\s', re.I)
re_logical_not = re.compile('\snot\s', re.I)
re_operators = re.compile(r'\s([\+\-\|])\s')
re_pattern_wildcards_at_beginning = re.compile(r'(\s)[\*\%]+')
re_pattern_single_quotes = re.compile("'(.*?)'")
re_pattern_double_quotes = re.compile("\"(.*?)\"")
re_pattern_regexp_quotes = re.compile("\/(.*?)\/")
re_pattern_short_words = re.compile(r'([\s\"]\w{1,3})[\*\%]+')
re_pattern_space = re.compile("__SPACE__")
re_pattern_today = re.compile("\$TODAY\$")
re_pattern_parens = re.compile(r'\([^\)]+\s+[^\)]+\)')
re_unicode_lowercase_a = re.compile(unicode(r"(?u)[áàäâãå]", "utf-8"))
re_unicode_lowercase_ae = re.compile(unicode(r"(?u)[æ]", "utf-8"))
re_unicode_lowercase_e = re.compile(unicode(r"(?u)[éèëê]", "utf-8"))
re_unicode_lowercase_i = re.compile(unicode(r"(?u)[íìïî]", "utf-8"))
re_unicode_lowercase_o = re.compile(unicode(r"(?u)[óòöôõø]", "utf-8"))
re_unicode_lowercase_u = re.compile(unicode(r"(?u)[úùüû]", "utf-8"))
re_unicode_lowercase_y = re.compile(unicode(r"(?u)[ýÿ]", "utf-8"))
re_unicode_lowercase_c = re.compile(unicode(r"(?u)[çć]", "utf-8"))
re_unicode_lowercase_n = re.compile(unicode(r"(?u)[ñ]", "utf-8"))
re_unicode_uppercase_a = re.compile(unicode(r"(?u)[ÁÀÄÂÃÅ]", "utf-8"))
re_unicode_uppercase_ae = re.compile(unicode(r"(?u)[Æ]", "utf-8"))
re_unicode_uppercase_e = re.compile(unicode(r"(?u)[ÉÈËÊ]", "utf-8"))
re_unicode_uppercase_i = re.compile(unicode(r"(?u)[ÍÌÏÎ]", "utf-8"))
re_unicode_uppercase_o = re.compile(unicode(r"(?u)[ÓÒÖÔÕØ]", "utf-8"))
re_unicode_uppercase_u = re.compile(unicode(r"(?u)[ÚÙÜÛ]", "utf-8"))
re_unicode_uppercase_y = re.compile(unicode(r"(?u)[Ý]", "utf-8"))
re_unicode_uppercase_c = re.compile(unicode(r"(?u)[ÇĆ]", "utf-8"))
re_unicode_uppercase_n = re.compile(unicode(r"(?u)[Ñ]", "utf-8"))
re_latex_lowercase_a = re.compile("\\\\[\"H'`~^vu=k]\{?a\}?")
re_latex_lowercase_ae = re.compile("\\\\ae\\{\\}?")
re_latex_lowercase_e = re.compile("\\\\[\"H'`~^vu=k]\\{?e\\}?")
re_latex_lowercase_i = re.compile("\\\\[\"H'`~^vu=k]\\{?i\\}?")
re_latex_lowercase_o = re.compile("\\\\[\"H'`~^vu=k]\\{?o\\}?")
re_latex_lowercase_u = re.compile("\\\\[\"H'`~^vu=k]\\{?u\\}?")
re_latex_lowercase_y = re.compile("\\\\[\"']\\{?y\\}?")
re_latex_lowercase_c = re.compile("\\\\['uc]\\{?c\\}?")
re_latex_lowercase_n = re.compile("\\\\[c'~^vu]\\{?n\\}?")
re_latex_uppercase_a = re.compile("\\\\[\"H'`~^vu=k]\\{?A\\}?")
re_latex_uppercase_ae = re.compile("\\\\AE\\{?\\}?")
re_latex_uppercase_e = re.compile("\\\\[\"H'`~^vu=k]\\{?E\\}?")
re_latex_uppercase_i = re.compile("\\\\[\"H'`~^vu=k]\\{?I\\}?")
re_latex_uppercase_o = re.compile("\\\\[\"H'`~^vu=k]\\{?O\\}?")
re_latex_uppercase_u = re.compile("\\\\[\"H'`~^vu=k]\\{?U\\}?")
re_latex_uppercase_y = re.compile("\\\\[\"']\\{?Y\\}?")
re_latex_uppercase_c = re.compile("\\\\['uc]\\{?C\\}?")
re_latex_uppercase_n = re.compile("\\\\[c'~^vu]\\{?N\\}?")
class RestrictedCollectionDataCacher(DataCacher):
def __init__(self):
def cache_filler():
ret = []
try:
viewcollid = acc_get_action_id(VIEWRESTRCOLL)
res = run_sql("""SELECT DISTINCT ar.value
FROM accROLE_accACTION_accARGUMENT raa JOIN accARGUMENT ar ON raa.id_accARGUMENT = ar.id
WHERE ar.keyword = 'collection' AND raa.id_accACTION = %s""", (viewcollid,))
except Exception:
# database problems, return empty cache
return []
for coll in res:
ret.append(coll[0])
return ret
def timestamp_verifier():
return max(get_table_update_time('accROLE_accACTION_accARGUMENT'), get_table_update_time('accARGUMENT'))
DataCacher.__init__(self, cache_filler, timestamp_verifier)
def collection_restricted_p(collection):
restricted_collection_cache.recreate_cache_if_needed()
return collection in restricted_collection_cache.cache
try:
restricted_collection_cache.is_ok_p
except Exception:
restricted_collection_cache = RestrictedCollectionDataCacher()
def get_permitted_restricted_collections(user_info):
"""Return a list of collection that are restricted but for which the user
is authorized."""
restricted_collection_cache.recreate_cache_if_needed()
ret = []
for collection in restricted_collection_cache.cache:
if acc_authorize_action(user_info, 'viewrestrcoll', collection=collection)[0] == 0:
ret.append(collection)
return ret
def is_user_owner_of_record(user_info, recid):
"""
Check if the user is owner of the record, i.e. he is the submitter
and/or belongs to a owner-like group authorized to 'see' the record.
@param user_info: the user_info dictionary that describe the user.
@type user_info: user_info dictionary
@param recid: the record identifier.
@type recid: positive integer
@return: True if the user is 'owner' of the record; False otherwise
@rtype: bool
"""
authorized_emails_or_group = []
for tag in CFG_ACC_GRANT_AUTHOR_RIGHTS_TO_EMAILS_IN_TAGS:
authorized_emails_or_group.extend(get_fieldvalues(recid, tag))
for email_or_group in authorized_emails_or_group:
if email_or_group in user_info['group']:
return True
email = email_or_group.strip().lower()
if user_info['email'].strip().lower() == email:
return True
return False
def check_user_can_view_record(user_info, recid):
"""
Check if the user is authorized to view the given recid. The function
grants access in two cases: either user has author rights on this
record, or he has view rights to the primary collection this record
belongs to.
@param user_info: the user_info dictionary that describe the user.
@type user_info: user_info dictionary
@param recid: the record identifier.
@type recid: positive integer
@return: (0, ''), when authorization is granted, (>0, 'message') when
authorization is not granted
@rtype: (int, string)
"""
record_primary_collection = guess_primary_collection_of_a_record(recid)
if collection_restricted_p(record_primary_collection):
(auth_code, auth_msg) = acc_authorize_action(user_info, VIEWRESTRCOLL, collection=record_primary_collection)
if auth_code == 0 or is_user_owner_of_record(user_info, recid):
return (0, '')
else:
return (auth_code, auth_msg)
else:
return (0, '')
class IndexStemmingDataCacher(DataCacher):
"""
Provides cache for stemming information for word/phrase indexes.
This class is not to be used directly; use function
get_index_stemming_language() instead.
"""
def __init__(self):
def cache_filler():
try:
res = run_sql("""SELECT id, stemming_language FROM idxINDEX""")
except DatabaseError:
# database problems, return empty cache
return {}
return dict(res)
def timestamp_verifier():
return get_table_update_time('idxINDEX')
DataCacher.__init__(self, cache_filler, timestamp_verifier)
try:
index_stemming_cache.is_ok_p
except Exception:
index_stemming_cache = IndexStemmingDataCacher()
def get_index_stemming_language(index_id):
"""Return stemming langugage for given index."""
index_stemming_cache.recreate_cache_if_needed()
return index_stemming_cache.cache[index_id]
class CollectionRecListDataCacher(DataCacher):
"""
Provides cache for collection reclist hitsets. This class is not
to be used directly; use function get_collection_reclist() instead.
"""
def __init__(self):
def cache_filler():
ret = {}
try:
res = run_sql("SELECT name,reclist FROM collection")
except Exception:
# database problems, return empty cache
return {}
for name, reclist in res:
ret[name] = None # this will be filled later during runtime by calling get_collection_reclist(coll)
return ret
def timestamp_verifier():
return get_table_update_time('collection')
DataCacher.__init__(self, cache_filler, timestamp_verifier)
try:
if not collection_reclist_cache.is_ok_p:
raise Exception
except Exception:
collection_reclist_cache = CollectionRecListDataCacher()
def get_collection_reclist(coll):
"""Return hitset of recIDs that belong to the collection 'coll'."""
collection_reclist_cache.recreate_cache_if_needed()
if not collection_reclist_cache.cache[coll]:
# not yet it the cache, so calculate it and fill the cache:
set = HitSet()
query = "SELECT nbrecs,reclist FROM collection WHERE name=%s"
res = run_sql(query, (coll, ), 1)
if res:
try:
set = HitSet(res[0][1])
except:
pass
collection_reclist_cache.cache[coll] = set
# finally, return reclist:
return collection_reclist_cache.cache[coll]
class SearchResultsCache(DataCacher):
"""
Provides temporary lazy cache for Search Results.
Useful when users click on `next page'.
"""
def __init__(self):
def cache_filler():
return {}
def timestamp_verifier():
return '1970-01-01 00:00:00' # lazy cache is always okay;
# its filling is governed by
# CFG_WEBSEARCH_SEARCH_CACHE_SIZE
DataCacher.__init__(self, cache_filler, timestamp_verifier)
try:
if not search_results_cache.is_ok_p:
raise Exception
except Exception:
search_results_cache = SearchResultsCache()
class CollectionI18nNameDataCacher(DataCacher):
"""
Provides cache for I18N collection names. This class is not to be
used directly; use function get_coll_i18nname() instead.
"""
def __init__(self):
def cache_filler():
ret = {}
try:
res = run_sql("SELECT c.name,cn.ln,cn.value FROM collectionname AS cn, collection AS c WHERE cn.id_collection=c.id AND cn.type='ln'") # ln=long name
except Exception:
# database problems
return {}
for c, ln, i18nname in res:
if i18nname:
if not ret.has_key(c):
ret[c] = {}
ret[c][ln] = i18nname
return ret
def timestamp_verifier():
return get_table_update_time('collectionname')
DataCacher.__init__(self, cache_filler, timestamp_verifier)
try:
if not collection_i18nname_cache.is_ok_p:
raise Exception
except Exception:
collection_i18nname_cache = CollectionI18nNameDataCacher()
def get_coll_i18nname(c, ln=CFG_SITE_LANG, verify_cache_timestamp=True):
"""
Return nicely formatted collection name (of the name type `ln'
(=long name)) for collection C in language LN.
This function uses collection_i18nname_cache, but it verifies
whether the cache is up-to-date first by default. This
verification step is performed by checking the DB table update
time. So, if you call this function 1000 times, it can get very
slow because it will do 1000 table update time verifications, even
though collection names change not that often.
Hence the parameter VERIFY_CACHE_TIMESTAMP which, when set to
False, will assume the cache is already up-to-date. This is
useful namely in the generation of collection lists for the search
results page.
"""
if verify_cache_timestamp:
collection_i18nname_cache.recreate_cache_if_needed()
out = c
try:
out = collection_i18nname_cache.cache[c][ln]
except KeyError:
pass # translation in LN does not exist
return out
class FieldI18nNameDataCacher(DataCacher):
"""
Provides cache for I18N field names. This class is not to be used
directly; use function get_field_i18nname() instead.
"""
def __init__(self):
def cache_filler():
ret = {}
try:
res = run_sql("SELECT f.name,fn.ln,fn.value FROM fieldname AS fn, field AS f WHERE fn.id_field=f.id AND fn.type='ln'") # ln=long name
except Exception:
# database problems, return empty cache
return {}
for f, ln, i18nname in res:
if i18nname:
if not ret.has_key(f):
ret[f] = {}
ret[f][ln] = i18nname
return ret
def timestamp_verifier():
return get_table_update_time('fieldname')
DataCacher.__init__(self, cache_filler, timestamp_verifier)
try:
if not field_i18nname_cache.is_ok_p:
raise Exception
except Exception:
field_i18nname_cache = FieldI18nNameDataCacher()
def get_field_i18nname(f, ln=CFG_SITE_LANG, verify_cache_timestamp=True):
"""
Return nicely formatted field name (of type 'ln', 'long name') for
field F in language LN.
If VERIFY_CACHE_TIMESTAMP is set to True, then verify DB timestamp
and field I18N name cache timestamp and refresh cache from the DB
if needed. Otherwise don't bother checking DB timestamp and
return the cached value. (This is useful when get_field_i18nname
is called inside a loop.)
"""
if verify_cache_timestamp:
field_i18nname_cache.recreate_cache_if_needed()
out = f
try:
out = field_i18nname_cache.cache[f][ln]
except KeyError:
pass # translation in LN does not exist
return out
def get_alphabetically_ordered_collection_list(level=0, ln=CFG_SITE_LANG):
"""Returns nicely ordered (score respected) list of collections, more exactly list of tuples
(collection name, printable collection name).
Suitable for create_search_box()."""
out = []
res = run_sql_cached("SELECT id,name FROM collection ORDER BY name ASC",
affected_tables=['collection',])
for c_id, c_name in res:
# make a nice printable name (e.g. truncate c_printable for
# long collection names in given language):
c_printable_fullname = get_coll_i18nname(c_name, ln, False)
c_printable = wash_index_term(c_printable_fullname, 30, False)
if c_printable != c_printable_fullname:
c_printable = c_printable + "..."
if level:
c_printable = " " + level * '-' + " " + c_printable
out.append([c_name, c_printable])
return out
def get_nicely_ordered_collection_list(collid=1, level=0, ln=CFG_SITE_LANG):
"""Returns nicely ordered (score respected) list of collections, more exactly list of tuples
(collection name, printable collection name).
Suitable for create_search_box()."""
colls_nicely_ordered = []
res = run_sql("""SELECT c.name,cc.id_son FROM collection_collection AS cc, collection AS c
WHERE c.id=cc.id_son AND cc.id_dad=%s ORDER BY score DESC""", (collid, ))
for c, cid in res:
# make a nice printable name (e.g. truncate c_printable for
# long collection names in given language):
c_printable_fullname = get_coll_i18nname(c, ln, False)
c_printable = wash_index_term(c_printable_fullname, 30, False)
if c_printable != c_printable_fullname:
c_printable = c_printable + "..."
if level:
c_printable = " " + level * '-' + " " + c_printable
colls_nicely_ordered.append([c, c_printable])
colls_nicely_ordered = colls_nicely_ordered + get_nicely_ordered_collection_list(cid, level+1, ln=ln)
return colls_nicely_ordered
def get_index_id_from_field(field):
"""
Return index id with name corresponding to FIELD, or the first
index id where the logical field code named FIELD is indexed.
Return zero in case there is no index defined for this field.
Example: field='author', output=4.
"""
out = 0
if field == '':
field = 'global' # empty string field means 'global' index (field 'anyfield')
# first look in the index table:
res = run_sql("""SELECT id FROM idxINDEX WHERE name=%s""", (field,))
if res:
out = res[0][0]
return out
# not found in the index table, now look in the field table:
res = run_sql("""SELECT w.id FROM idxINDEX AS w, idxINDEX_field AS wf, field AS f
WHERE f.code=%s AND wf.id_field=f.id AND w.id=wf.id_idxINDEX
LIMIT 1""", (field,))
if res:
out = res[0][0]
return out
def get_words_from_pattern(pattern):
"Returns list of whitespace-separated words from pattern."
words = {}
for word in string.split(pattern):
if not words.has_key(word):
words[word] = 1;
return words.keys()
def create_basic_search_units(req, p, f, m=None, of='hb'):
"""Splits search pattern and search field into a list of independently searchable units.
- A search unit consists of '(operator, pattern, field, type, hitset)' tuples where
'operator' is set union (|), set intersection (+) or set exclusion (-);
'pattern' is either a word (e.g. muon*) or a phrase (e.g. 'nuclear physics');
'field' is either a code like 'title' or MARC tag like '100__a';
'type' is the search type ('w' for word file search, 'a' for access file search).
- Optionally, the function accepts the match type argument 'm'.
If it is set (e.g. from advanced search interface), then it
performs this kind of matching. If it is not set, then a guess is made.
'm' can have values: 'a'='all of the words', 'o'='any of the words',
'p'='phrase/substring', 'r'='regular expression',
'e'='exact value'.
- Warnings are printed on req (when not None) in case of HTML output formats."""
opfts = [] # will hold (o,p,f,t,h) units
# FIXME: quick hack for the journal index
if f == 'journal':
opfts.append(['+', p, f, 'w'])
return opfts
## check arguments: is desired matching type set?
if m:
## A - matching type is known; good!
if m == 'e':
# A1 - exact value:
opfts.append(['+', p, f, 'a']) # '+' since we have only one unit
elif m == 'p':
# A2 - phrase/substring:
opfts.append(['+', "%" + p + "%", f, 'a']) # '+' since we have only one unit
elif m == 'r':
# A3 - regular expression:
opfts.append(['+', p, f, 'r']) # '+' since we have only one unit
elif m == 'a' or m == 'w':
# A4 - all of the words:
p = strip_accents(p) # strip accents for 'w' mode, FIXME: delete when not needed
for word in get_words_from_pattern(p):
opfts.append(['+', word, f, 'w']) # '+' in all units
elif m == 'o':
# A5 - any of the words:
p = strip_accents(p) # strip accents for 'w' mode, FIXME: delete when not needed
for word in get_words_from_pattern(p):
if len(opfts)==0:
opfts.append(['+', word, f, 'w']) # '+' in the first unit
else:
opfts.append(['|', word, f, 'w']) # '|' in further units
else:
if of.startswith("h"):
print_warning(req, "Matching type '%s' is not implemented yet." % cgi.escape(m), "Warning")
opfts.append(['+', "%" + p + "%", f, 'w'])
else:
## B - matching type is not known: let us try to determine it by some heuristics
if f and p[0] == '"' and p[-1] == '"':
## B0 - does 'p' start and end by double quote, and is 'f' defined? => doing ACC search
opfts.append(['+', p[1:-1], f, 'a'])
elif f and p[0] == "'" and p[-1] == "'":
## B0bis - does 'p' start and end by single quote, and is 'f' defined? => doing ACC search
opfts.append(['+', '%' + p[1:-1] + '%', f, 'a'])
elif f and p[0] == "/" and p[-1] == "/":
## B0ter - does 'p' start and end by a slash, and is 'f' defined? => doing regexp search
opfts.append(['+', p[1:-1], f, 'r'])
elif f and string.find(p, ',') >= 0:
## B1 - does 'p' contain comma, and is 'f' defined? => doing ACC search
opfts.append(['+', p, f, 'a'])
elif f and str(f[0:2]).isdigit():
## B2 - does 'f' exist and starts by two digits? => doing ACC search
opfts.append(['+', p, f, 'a'])
else:
## B3 - doing WRD search, but maybe ACC too
# search units are separated by spaces unless the space is within single or double quotes
# so, let us replace temporarily any space within quotes by '__SPACE__'
p = re_pattern_single_quotes.sub(lambda x: "'"+string.replace(x.group(1), ' ', '__SPACE__')+"'", p)
p = re_pattern_double_quotes.sub(lambda x: "\""+string.replace(x.group(1), ' ', '__SPACE__')+"\"", p)
p = re_pattern_regexp_quotes.sub(lambda x: "/"+string.replace(x.group(1), ' ', '__SPACE__')+"/", p)
# wash argument:
p = re_equal.sub(":", p)
p = re_logical_and.sub(" ", p)
p = re_logical_or.sub(" |", p)
p = re_logical_not.sub(" -", p)
p = re_operators.sub(r' \1', p)
for pi in string.split(p): # iterate through separated units (or items, as "pi" stands for "p item")
pi = re_pattern_space.sub(" ", pi) # replace back '__SPACE__' by ' '
# firstly, determine set operator
if pi[0] == '+' or pi[0] == '-' or pi[0] == '|':
oi = pi[0]
pi = pi[1:]
else:
# okay, there is no operator, so let us decide what to do by default
oi = '+' # by default we are doing set intersection...
# secondly, determine search pattern and field:
if string.find(pi, ":") > 0:
fi, pi = string.split(pi, ":", 1)
# test whether fi is a real index code or a MARC-tag defined code:
if fi in get_fieldcodes() or '00' <= fi[:2] <= '99':
pass
else:
# it is not, so join it back:
fi, pi = f, fi + ":" + pi
else:
fi, pi = f, pi
# look also for old ALEPH field names:
if fi and CFG_WEBSEARCH_FIELDS_CONVERT.has_key(string.lower(fi)):
fi = CFG_WEBSEARCH_FIELDS_CONVERT[string.lower(fi)]
# wash 'pi' argument:
if re_quotes.match(pi):
# B3a - quotes are found => do ACC search (phrase search)
if pi[0] == '"' and pi[-1] == '"':
pi = string.replace(pi, '"', '') # remove quote signs
opfts.append([oi, pi, fi, 'a'])
elif pi[0] == "'" and pi[-1] == "'":
pi = string.replace(pi, "'", "") # remove quote signs
opfts.append([oi, "%" + pi + "%", fi, 'a'])
else: # unbalanced quotes, so fall back to WRD query:
opfts.append([oi, pi, fi, 'w'])
elif fi and str(fi[0]).isdigit() and str(fi[0]).isdigit():
# B3b - fi exists and starts by two digits => do ACC search
opfts.append([oi, pi, fi, 'a'])
elif fi and not get_index_id_from_field(fi) and get_field_name(fi):
# B3c - logical field fi exists but there is no WRD index for fi => try ACC search
opfts.append([oi, pi, fi, 'a'])
elif pi.startswith('/') and pi.endswith('/'):
# B3d - pi has slashes around => do regexp search
opfts.append([oi, pi[1:-1], fi, 'r'])
else:
# B3e - general case => do WRD search
pi = strip_accents(pi) # strip accents for 'w' mode, FIXME: delete when not needed
for pii in get_words_from_pattern(pi):
opfts.append([oi, pii, fi, 'w'])
## sanity check:
for i in range(0, len(opfts)):
try:
pi = opfts[i][1]
if pi == '*':
if of.startswith("h"):
print_warning(req, "Ignoring standalone wildcard word.", "Warning")
del opfts[i]
if pi == '' or pi == ' ':
fi = opfts[i][2]
if fi:
if of.startswith("h"):
print_warning(req, "Ignoring empty %s search term." % fi, "Warning")
del opfts[i]
except:
pass
## return search units:
return opfts
def page_start(req, of, cc, aas, ln, uid, title_message=None,
description='', keywords='', recID=-1, tab='', p=''):
"Start page according to given output format."
_ = gettext_set_language(ln)
if not req or isinstance(req, cStringIO.OutputType):
return # we were called from CLI
if not title_message:
title_message = _("Search Results")
content_type = get_output_format_content_type(of)
if of.startswith('x'):
if of == 'xr':
# we are doing RSS output
req.content_type = "application/rss+xml"
req.send_http_header()
req.write("""\n""")
else:
# we are doing XML output:
req.content_type = "text/xml"
req.send_http_header()
req.write("""\n""")
elif of.startswith('t') or str(of[0:3]).isdigit():
# we are doing plain text output:
req.content_type = "text/plain"
req.send_http_header()
elif of == "id":
pass # nothing to do, we shall only return list of recIDs
elif content_type == 'text/html':
# we are doing HTML output:
req.content_type = "text/html"
req.send_http_header()
if not description:
description = "%s %s." % (cc, _("Search Results"))
if not keywords:
keywords = "%s, WebSearch, %s" % (get_coll_i18nname(CFG_SITE_NAME, ln, False), get_coll_i18nname(cc, ln, False))
## generate RSS URL:
argd = {}
if req.args:
argd = cgi.parse_qs(req.args)
rssurl = websearch_templates.build_rss_url(argd)
## add jsmath if displaying single records (FIXME: find
## eventual better place to this code)
if of.lower() in CFG_WEBSEARCH_USE_JSMATH_FOR_FORMATS:
metaheaderadd = """
"""
else:
metaheaderadd = ''
## generate navtrail:
navtrail = create_navtrail_links(cc, aas, ln)
if navtrail != '':
navtrail += ' > '
if (tab != '' or ((of != '' or of.lower() != 'hd') and of != 'hb')) and \
recID != -1:
# If we are not in information tab in HD format, customize
# the nav. trail to have a link back to main record. (Due
# to the way perform_request_search() works, hb
# (lowercase) is equal to hd)
navtrail += ' %s' % \
(CFG_SITE_URL, recID, title_message)
if (of != '' or of.lower() != 'hd') and of != 'hb':
# Export
format_name = of
query = "SELECT name FROM format WHERE code=%s"
res = run_sql(query, (of,))
if res:
format_name = res[0][0]
navtrail += ' > ' + format_name
else:
# Discussion, citations, etc. tabs
tab_label = get_detailed_page_tabs(cc, ln=ln)[tab]['label']
navtrail += ' > ' + _(tab_label)
else:
navtrail += title_message
if p:
# we are serving search/browse results pages, so insert pattern:
navtrail += ": " + cgi.escape(p)
title_message = cgi.escape(p) + " - " + title_message
## finally, print page header:
req.write(pageheaderonly(req=req, title=title_message,
navtrail=navtrail,
description=description,
keywords=keywords,
metaheaderadd=metaheaderadd,
uid=uid,
language=ln,
navmenuid='search',
navtrail_append_title_p=0,
rssurl=rssurl))
req.write(websearch_templates.tmpl_search_pagestart(ln=ln))
#else:
# req.send_http_header()
def page_end(req, of="hb", ln=CFG_SITE_LANG):
"End page according to given output format: e.g. close XML tags, add HTML footer, etc."
if of == "id":
return [] # empty recID list
if not req:
return # we were called from CLI
if of.startswith('h'):
req.write(websearch_templates.tmpl_search_pageend(ln = ln)) # pagebody end
req.write(pagefooteronly(lastupdated=__lastupdated__, language=ln, req=req))
return
def create_page_title_search_pattern_info(p, p1, p2, p3):
"""Create the search pattern bit for the page web page
HTML header. Basically combine p and (p1,p2,p3) together so that
the page header may be filled whether we are in the Simple Search
or Advanced Search interface contexts."""
out = ""
if p:
out = p
else:
out = p1
if p2:
out += ' ' + p2
if p3:
out += ' ' + p3
return out
def create_inputdate_box(name="d1", selected_year=0, selected_month=0, selected_day=0, ln=CFG_SITE_LANG):
"Produces 'From Date', 'Until Date' kind of selection box. Suitable for search options."
_ = gettext_set_language(ln)
box = ""
# day
box += """"""
# month
box += """"""
# year
box += """"""
return box
def create_search_box(cc, colls, p, f, rg, sf, so, sp, rm, of, ot, aas,
ln, p1, f1, m1, op1, p2, f2, m2, op2, p3, f3,
m3, sc, pl, d1y, d1m, d1d, d2y, d2m, d2d, dt, jrec, ec,
action=""):
"""Create search box for 'search again in the results page' functionality."""
# load the right message language
_ = gettext_set_language(ln)
# some computations
cc_intl = get_coll_i18nname(cc, ln, False)
cc_colID = get_colID(cc)
colls_nicely_ordered = []
if cfg_nicely_ordered_collection_list:
colls_nicely_ordered = get_nicely_ordered_collection_list(ln=ln)
else:
colls_nicely_ordered = get_alphabetically_ordered_collection_list(ln=ln)
colls_nice = []
for (cx, cx_printable) in colls_nicely_ordered:
if not cx.startswith("Unnamed collection"):
colls_nice.append({ 'value' : cx,
'text' : cx_printable
})
coll_selects = []
if colls and colls[0] != CFG_SITE_NAME:
# some collections are defined, so print these first, and only then print 'add another collection' heading:
for c in colls:
if c:
temp = []
temp.append({ 'value' : CFG_SITE_NAME,
'text' : '*** %s ***' % _("any public collection")
})
# this field is used to remove the current collection from the ones to be searched.
temp.append({ 'value' : '',
'text' : '*** %s ***' % _("remove this collection")
})
for val in colls_nice:
# print collection:
if not cx.startswith("Unnamed collection"):
temp.append({ 'value' : val['value'],
'text' : val['text'],
'selected' : (c == re.sub("^[\s\-]*","", val['value']))
})
coll_selects.append(temp)
coll_selects.append([{ 'value' : '',
'text' : '*** %s ***' % _("add another collection")
}] + colls_nice)
else: # we searched in CFG_SITE_NAME, so print 'any public collection' heading
coll_selects.append([{ 'value' : CFG_SITE_NAME,
'text' : '*** %s ***' % _("any public collection")
}] + colls_nice)
## ranking methods
ranks = [{
'value' : '',
'text' : "- %s %s -" % (_("OR").lower (), _("rank by")),
}]
for (code, name) in get_bibrank_methods(cc_colID, ln):
# propose found rank methods:
ranks.append({
'value' : code,
'text' : name,
})
formats = []
query = """SELECT code,name FROM format WHERE visibility='1' ORDER BY name ASC"""
res = run_sql(query)
if res:
# propose found formats:
for code, name in res:
formats.append({ 'value' : code,
'text' : name
})
else:
formats.append({'value' : 'hb',
'text' : _("HTML brief")
})
# show collections in the search box? (not if there is only one
# collection defined, and not if we are in light search)
show_colls = True
show_title = True
if len(collection_reclist_cache.cache.keys()) == 1 or \
aas == -1:
show_colls = False
show_title = False
if cc == CFG_SITE_NAME:
show_title = False
return websearch_templates.tmpl_search_box(
ln = ln,
aas = aas,
cc_intl = cc_intl,
cc = cc,
ot = ot,
sp = sp,
action = action,
fieldslist = get_searchwithin_fields(ln=ln, colID=cc_colID),
f1 = f1,
f2 = f2,
f3 = f3,
m1 = m1,
m2 = m2,
m3 = m3,
p1 = p1,
p2 = p2,
p3 = p3,
op1 = op1,
op2 = op2,
rm = rm,
p = p,
f = f,
coll_selects = coll_selects,
d1y = d1y, d2y = d2y, d1m = d1m, d2m = d2m, d1d = d1d, d2d = d2d,
dt = dt,
sort_fields = get_sortby_fields(ln=ln, colID=cc_colID),
sf = sf,
so = so,
ranks = ranks,
sc = sc,
rg = rg,
formats = formats,
of = of,
pl = pl,
jrec = jrec,
ec = ec,
show_colls = show_colls,
show_title = show_title,
)
def create_navtrail_links(cc=CFG_SITE_NAME, aas=0, ln=CFG_SITE_LANG, self_p=1, tab=''):
"""Creates navigation trail links, i.e. links to collection
ancestors (except Home collection). If aas==1, then links to
Advanced Search interfaces; otherwise Simple Search.
"""
dads = []
for dad in get_coll_ancestors(cc):
if dad != CFG_SITE_NAME: # exclude Home collection
dads.append ((dad, get_coll_i18nname(dad, ln, False)))
if self_p and cc != CFG_SITE_NAME:
dads.append((cc, get_coll_i18nname(cc, ln, False)))
return websearch_templates.tmpl_navtrail_links(
aas=aas, ln=ln, dads=dads)
def get_searchwithin_fields(ln='en', colID=None):
"""Retrieves the fields name used in the 'search within' selection box for the collection ID colID."""
res = None
if colID:
res = run_sql_cached("""SELECT f.code,f.name FROM field AS f, collection_field_fieldvalue AS cff
WHERE cff.type='sew' AND cff.id_collection=%s AND cff.id_field=f.id
ORDER BY cff.score DESC, f.name ASC""", (colID,),
affected_tables=['field', 'collection_field_fieldvalue'])
if not res:
res = run_sql_cached("SELECT code,name FROM field ORDER BY name ASC",
affected_tables=['field',])
fields = [{
'value' : '',
'text' : get_field_i18nname("any field", ln, False)
}]
for field_code, field_name in res:
if field_code and field_code != "anyfield":
fields.append({ 'value' : field_code,
'text' : get_field_i18nname(field_name, ln, False)
})
return fields
def get_sortby_fields(ln='en', colID=None):
"""Retrieves the fields name used in the 'sort by' selection box for the collection ID colID."""
_ = gettext_set_language(ln)
res = None
if colID:
res = run_sql_cached("""SELECT DISTINCT(f.code),f.name FROM field AS f, collection_field_fieldvalue AS cff
WHERE cff.type='soo' AND cff.id_collection=%s AND cff.id_field=f.id
ORDER BY cff.score DESC, f.name ASC""", (colID,),
affected_tables=['field', 'collection_field_fieldvalue'])
if not res:
# no sort fields defined for this colID, try to take Home collection:
res = run_sql_cached("""SELECT DISTINCT(f.code),f.name FROM field AS f, collection_field_fieldvalue AS cff
WHERE cff.type='soo' AND cff.id_collection=%s AND cff.id_field=f.id
ORDER BY cff.score DESC, f.name ASC""", (1,),
affected_tables=['field', 'collection_field_fieldvalue'])
if not res:
# no sort fields defined for the Home collection, take all sort fields defined wherever they are:
res = run_sql_cached("""SELECT DISTINCT(f.code),f.name FROM field AS f, collection_field_fieldvalue AS cff
WHERE cff.type='soo' AND cff.id_field=f.id
ORDER BY cff.score DESC, f.name ASC""",
affected_tables=['field', 'collection_field_fieldvalue'])
fields = [{
'value' : '',
'text' : _("latest first")
}]
for field_code, field_name in res:
if field_code and field_code != "anyfield":
fields.append({ 'value' : field_code,
'text' : get_field_i18nname(field_name, ln, False)
})
return fields
def create_andornot_box(name='op', value='', ln='en'):
"Returns HTML code for the AND/OR/NOT selection box."
_ = gettext_set_language(ln)
out = """
""" % (name,
is_selected('a', value), _("AND"),
is_selected('o', value), _("OR"),
is_selected('n', value), _("AND NOT"))
return out
def create_matchtype_box(name='m', value='', ln='en'):
"Returns HTML code for the 'match type' selection box."
_ = gettext_set_language(ln)
out = """
""" % (name,
is_selected('a', value), _("All of the words:"),
is_selected('o', value), _("Any of the words:"),
is_selected('e', value), _("Exact phrase:"),
is_selected('p', value), _("Partial phrase:"),
is_selected('r', value), _("Regular expression:"))
return out
def is_selected(var, fld):
"Checks if the two are equal, and if yes, returns ' selected'. Useful for select boxes."
if type(var) is int and type(fld) is int:
if var == fld:
return " selected"
elif str(var) == str(fld):
return " selected"
elif fld and len(fld)==3 and fld[0] == "w" and var == fld[1:]:
return " selected"
return ""
def wash_colls(cc, c, split_colls=0, verbose=0):
"""Wash collection list by checking whether user has deselected
anything under 'Narrow search'. Checks also if cc is a list or not.
Return list of cc, colls_to_display, colls_to_search since the list
of collections to display is different from that to search in.
This is because users might have chosen 'split by collection'
functionality.
The behaviour of "collections to display" depends solely whether
user has deselected a particular collection: e.g. if it started
from 'Articles and Preprints' page, and deselected 'Preprints',
then collection to display is 'Articles'. If he did not deselect
anything, then collection to display is 'Articles & Preprints'.
The behaviour of "collections to search in" depends on the
'split_colls' parameter:
* if is equal to 1, then we can wash the colls list down
and search solely in the collection the user started from;
* if is equal to 0, then we are splitting to the first level
of collections, i.e. collections as they appear on the page
we started to search from;
The function raises exception
InvenioWebSearchUnknownCollectionError
if cc or one of c collections is not known.
"""
colls_out = []
colls_out_for_display = []
# list to hold the hosted collections to be searched and displayed
hosted_colls_out = []
debug = ""
if verbose:
debug += " "
debug += " 1) --- initial parameters ---"
debug += " cc : %s" % cc
debug += " c : %s" % c
debug += " "
# check what type is 'cc':
if type(cc) is list:
for ci in cc:
if collection_reclist_cache.cache.has_key(ci):
# yes this collection is real, so use it:
cc = ci
break
else:
# check once if cc is real:
if not collection_reclist_cache.cache.has_key(cc):
if cc:
raise InvenioWebSearchUnknownCollectionError(cc)
else:
cc = CFG_SITE_NAME # cc is not set, so replace it with Home collection
# check type of 'c' argument:
if type(c) is list:
colls = c
else:
colls = [c]
if verbose:
debug += " 2) --- after check for the integrity of cc and the being or not c a list ---"
debug += " cc : %s" % cc
debug += " c : %s" % c
debug += " "
# remove all 'unreal' collections:
colls_real = []
for coll in colls:
if collection_reclist_cache.cache.has_key(coll):
colls_real.append(coll)
else:
if coll:
raise InvenioWebSearchUnknownCollectionError(coll)
colls = colls_real
if verbose:
debug += " 3) --- keeping only the real colls of c ---"
debug += " colls : %s" % colls
debug += " "
# check if some real collections remain:
if len(colls)==0:
colls = [cc]
if verbose:
debug += " 4) --- in case no colls were left we use cc directly ---"
debug += " colls : %s" % colls
debug += " "
# then let us check the list of non-restricted "real" sons of 'cc' and compare it to 'coll':
res = run_sql("""SELECT c.name FROM collection AS c,
collection_collection AS cc,
collection AS ccc
WHERE c.id=cc.id_son AND cc.id_dad=ccc.id
AND ccc.name=%s AND cc.type='r'""", (cc,))
# list that holds all the non restricted sons of cc that are also not hosted collections
l_cc_nonrestricted_sons_and_nonhosted_colls = []
res_hosted = run_sql("""SELECT c.name FROM collection AS c,
collection_collection AS cc,
collection AS ccc
WHERE c.id=cc.id_son AND cc.id_dad=ccc.id
AND ccc.name=%s AND cc.type='r'
AND (c.dbquery NOT LIKE 'hostedcollection:%%' OR c.dbquery IS NULL)""", (cc,))
for row_hosted in res_hosted:
l_cc_nonrestricted_sons_and_nonhosted_colls.append(row_hosted[0])
l_cc_nonrestricted_sons_and_nonhosted_colls.sort()
l_cc_nonrestricted_sons = []
l_c = colls
for row in res:
if not collection_restricted_p(row[0]):
l_cc_nonrestricted_sons.append(row[0])
l_c.sort()
l_cc_nonrestricted_sons.sort()
if l_cc_nonrestricted_sons == l_c:
colls_out_for_display = [cc] # yep, washing permitted, it is sufficient to display 'cc'
# the following elif is a hack that preserves the above funcionality when we start searching from
# the frontpage with some hosted collections deselected (either by default or manually)
elif set(l_cc_nonrestricted_sons_and_nonhosted_colls).issubset(set(l_c)):
colls_out_for_display = colls
split_colls = 0
else:
colls_out_for_display = colls # nope, we need to display all 'colls' successively
# remove duplicates:
#colls_out_for_display_nondups=filter(lambda x, colls_out_for_display=colls_out_for_display: colls_out_for_display[x-1] not in colls_out_for_display[x:], range(1, len(colls_out_for_display)+1))
#colls_out_for_display = map(lambda x, colls_out_for_display=colls_out_for_display:colls_out_for_display[x-1], colls_out_for_display_nondups)
colls_out_for_display = list(set(colls_out_for_display))
if verbose:
debug += " 5) --- decide whether colls_out_for_diplay should be colls or is it sufficient for it to be cc; remove duplicates ---"
debug += " colls_out_for_display : %s" % colls_out_for_display
debug += " "
# the following piece of code takes care of removing collections whose ancestors are going to be searched anyway
# list to hold the collections to be removed
colls_to_be_removed = []
# first calculate the collections that can safely be removed
for coll in colls_out_for_display:
for ancestor in get_coll_ancestors(coll):
#if ancestor in colls_out_for_display: colls_to_be_removed.append(coll)
if ancestor in colls_out_for_display and not is_hosted_collection(coll): colls_to_be_removed.append(coll)
# secondly remove the collections
for coll in colls_to_be_removed:
colls_out_for_display.remove(coll)
if verbose:
debug += " 6) --- remove collections that have ancestors about to be search, unless they are hosted ---"
debug += " colls_out_for_display : %s" % colls_out_for_display
debug += " "
# calculate the hosted collections to be searched.
if colls_out_for_display == [cc]:
if is_hosted_collection(cc):
hosted_colls_out.append(cc)
else:
for coll in get_coll_sons(cc):
if is_hosted_collection(coll):
hosted_colls_out.append(coll)
else:
for coll in colls_out_for_display:
if is_hosted_collection(coll):
hosted_colls_out.append(coll)
if verbose:
debug += " 7) --- calculate the hosted_colls_out ---"
debug += " hosted_colls_out : %s" % hosted_colls_out
debug += " "
# second, let us decide on collection splitting:
if split_colls == 0:
# type A - no sons are wanted
colls_out = colls_out_for_display
else:
# type B - sons (first-level descendants) are wanted
for coll in colls_out_for_display:
coll_sons = get_coll_sons(coll)
if coll_sons == []:
colls_out.append(coll)
else:
for coll_son in coll_sons:
if not is_hosted_collection(coll_son):
colls_out.append(coll_son)
#else:
# colls_out = colls_out + coll_sons
# remove duplicates:
#colls_out_nondups=filter(lambda x, colls_out=colls_out: colls_out[x-1] not in colls_out[x:], range(1, len(colls_out)+1))
#colls_out = map(lambda x, colls_out=colls_out:colls_out[x-1], colls_out_nondups)
colls_out = list(set(colls_out))
if verbose:
debug += " 8) --- calculate the colls_out; remove duplicates ---"
debug += " colls_out : %s" % colls_out
debug += " "
# remove the hosted collections from the collections to be searched
if hosted_colls_out:
for coll in hosted_colls_out:
try:
colls_out.remove(coll)
except ValueError:
# in case coll was not found in colls_out
pass
if verbose:
debug += " 9) --- remove the hosted_colls from the colls_out ---"
debug += " colls_out : %s" % colls_out
return (cc, colls_out_for_display, colls_out, hosted_colls_out, debug)
def strip_accents(x):
"""Strip accents in the input phrase X (assumed in UTF-8) by replacing
accented characters with their unaccented cousins (e.g. é by e).
Return such a stripped X."""
x = re_latex_lowercase_a.sub("a", x)
x = re_latex_lowercase_ae.sub("ae", x)
x = re_latex_lowercase_e.sub("e", x)
x = re_latex_lowercase_i.sub("i", x)
x = re_latex_lowercase_o.sub("o", x)
x = re_latex_lowercase_u.sub("u", x)
x = re_latex_lowercase_y.sub("x", x)
x = re_latex_lowercase_c.sub("c", x)
x = re_latex_lowercase_n.sub("n", x)
x = re_latex_uppercase_a.sub("A", x)
x = re_latex_uppercase_ae.sub("AE", x)
x = re_latex_uppercase_e.sub("E", x)
x = re_latex_uppercase_i.sub("I", x)
x = re_latex_uppercase_o.sub("O", x)
x = re_latex_uppercase_u.sub("U", x)
x = re_latex_uppercase_y.sub("Y", x)
x = re_latex_uppercase_c.sub("C", x)
x = re_latex_uppercase_n.sub("N", x)
# convert input into Unicode string:
try:
y = unicode(x, "utf-8")
except:
return x # something went wrong, probably the input wasn't UTF-8
# asciify Latin-1 lowercase characters:
y = re_unicode_lowercase_a.sub("a", y)
y = re_unicode_lowercase_ae.sub("ae", y)
y = re_unicode_lowercase_e.sub("e", y)
y = re_unicode_lowercase_i.sub("i", y)
y = re_unicode_lowercase_o.sub("o", y)
y = re_unicode_lowercase_u.sub("u", y)
y = re_unicode_lowercase_y.sub("y", y)
y = re_unicode_lowercase_c.sub("c", y)
y = re_unicode_lowercase_n.sub("n", y)
# asciify Latin-1 uppercase characters:
y = re_unicode_uppercase_a.sub("A", y)
y = re_unicode_uppercase_ae.sub("AE", y)
y = re_unicode_uppercase_e.sub("E", y)
y = re_unicode_uppercase_i.sub("I", y)
y = re_unicode_uppercase_o.sub("O", y)
y = re_unicode_uppercase_u.sub("U", y)
y = re_unicode_uppercase_y.sub("Y", y)
y = re_unicode_uppercase_c.sub("C", y)
y = re_unicode_uppercase_n.sub("N", y)
# return UTF-8 representation of the Unicode string:
return y.encode("utf-8")
def wash_index_term(term, max_char_length=50, lower_term=True):
"""
Return washed form of the index term TERM that would be suitable
for storing into idxWORD* tables. I.e., lower the TERM if
LOWER_TERM is True, and truncate it safely to MAX_CHAR_LENGTH
UTF-8 characters (meaning, in principle, 4*MAX_CHAR_LENGTH bytes).
The function works by an internal conversion of TERM, when needed,
from its input Python UTF-8 binary string format into Python
Unicode format, and then truncating it safely to the given number
of UTF-8 characters, without possible mis-truncation in the middle
of a multi-byte UTF-8 character that could otherwise happen if we
would have been working with UTF-8 binary representation directly.
Note that MAX_CHAR_LENGTH corresponds to the length of the term
column in idxINDEX* tables.
"""
if lower_term:
washed_term = unicode(term, 'utf-8').lower()
else:
washed_term = unicode(term, 'utf-8')
if len(washed_term) <= max_char_length:
# no need to truncate the term, because it will fit
# nicely even if it uses four-byte UTF-8 characters
return washed_term.encode('utf-8')
else:
# truncate the term in a safe position:
return washed_term[:max_char_length].encode('utf-8')
def lower_index_term(term):
"""
Return safely lowered index term TERM. This is done by converting
to UTF-8 first, because standard Python lower() function is not
UTF-8 safe. To be called by both the search engine and the
indexer when appropriate (e.g. before stemming).
In case of problems with UTF-8 compliance, this function raises
UnicodeDecodeError, so the client code may want to catch it.
"""
return unicode(term, 'utf-8').lower().encode('utf-8')
def wash_output_format(format):
"""Wash output format FORMAT. Currently only prevents input like
'of=9' for backwards-compatible format that prints certain fields
only. (for this task, 'of=tm' is preferred)"""
if str(format[0:3]).isdigit() and len(format) != 6:
# asked to print MARC tags, but not enough digits,
# so let's switch back to HTML brief default
return 'hb'
else:
return format
def wash_pattern(p):
"""Wash pattern passed by URL. Check for sanity of the wildcard by
removing wildcards if they are appended to extremely short words
(1-3 letters). TODO: instead of this approximative treatment, it
will be much better to introduce a temporal limit, e.g. to kill a
query if it does not finish in 10 seconds."""
# strip accents:
# p = strip_accents(p) # FIXME: when available, strip accents all the time
# add leading/trailing whitespace for the two following wildcard-sanity checking regexps:
p = " " + p + " "
# get rid of wildcards at the beginning of words:
p = re_pattern_wildcards_at_beginning.sub("\\1", p)
# replace spaces within quotes by __SPACE__ temporarily:
p = re_pattern_single_quotes.sub(lambda x: "'"+string.replace(x.group(1), ' ', '__SPACE__')+"'", p)
p = re_pattern_double_quotes.sub(lambda x: "\""+string.replace(x.group(1), ' ', '__SPACE__')+"\"", p)
p = re_pattern_regexp_quotes.sub(lambda x: "/"+string.replace(x.group(1), ' ', '__SPACE__')+"/", p)
# get rid of extremely short words (1-3 letters with wildcards):
p = re_pattern_short_words.sub("\\1", p)
# replace back __SPACE__ by spaces:
p = re_pattern_space.sub(" ", p)
# replace special terms:
p = re_pattern_today.sub(time.strftime("%Y-%m-%d", time.localtime()), p)
# remove unnecessary whitespace:
p = string.strip(p)
return p
def wash_field(f):
"""Wash field passed by URL."""
# get rid of unnecessary whitespace:
f = string.strip(f)
# wash old-style CDS Invenio/ALEPH 'f' field argument, e.g. replaces 'wau' and 'au' by 'author'
if CFG_WEBSEARCH_FIELDS_CONVERT.has_key(string.lower(f)):
f = CFG_WEBSEARCH_FIELDS_CONVERT[f]
return f
def wash_dates(d1="", d1y=0, d1m=0, d1d=0, d2="", d2y=0, d2m=0, d2d=0):
"""
Take user-submitted date arguments D1 (full datetime string) or
(D1Y, D1M, D1Y) year, month, day tuple and D2 or (D2Y, D2M, D2Y)
and return (YYY1-M1-D2 H1:M1:S2, YYY2-M2-D2 H2:M2:S2) datetime
strings in the YYYY-MM-DD HH:MM:SS format suitable for time
restricted searching.
Note that when both D1 and (D1Y, D1M, D1D) parameters are present,
the precedence goes to D1. Ditto for D2*.
Note that when (D1Y, D1M, D1D) are taken into account, some values
may be missing and are completed e.g. to 01 or 12 according to
whether it is the starting or the ending date.
"""
datetext1, datetext2 = "", ""
# sanity checking:
if d1 == "" and d1y == 0 and d1m == 0 and d1d == 0 and d2 == "" and d2y == 0 and d2m == 0 and d2d == 0:
return ("", "") # nothing selected, so return empty values
# wash first (starting) date:
if d1:
# full datetime string takes precedence:
datetext1 = d1
else:
# okay, first date passed as (year,month,day):
if d1y:
datetext1 += "%04d" % d1y
else:
datetext1 += "0000"
if d1m:
datetext1 += "-%02d" % d1m
else:
datetext1 += "-01"
if d1d:
datetext1 += "-%02d" % d1d
else:
datetext1 += "-01"
datetext1 += " 00:00:00"
# wash second (ending) date:
if d2:
# full datetime string takes precedence:
datetext2 = d2
else:
# okay, second date passed as (year,month,day):
if d2y:
datetext2 += "%04d" % d2y
else:
datetext2 += "9999"
if d2m:
datetext2 += "-%02d" % d2m
else:
datetext2 += "-12"
if d2d:
datetext2 += "-%02d" % d2d
else:
datetext2 += "-31" # NOTE: perhaps we should add max(datenumber) in
# given month, but for our quering it's not
# needed, 31 will always do
datetext2 += " 00:00:00"
# okay, return constructed YYYY-MM-DD HH:MM:SS datetexts:
return (datetext1, datetext2)
def is_hosted_collection(coll):
"""Check if the given collection is a hosted one; i.e. its dbquery starts with hostedcollection:
Returns True if it is, False if it's not or if the result is empty or if the query failed"""
res = run_sql("SELECT dbquery FROM collection WHERE name=%s", (coll, ))
try:
return res[0][0].startswith("hostedcollection:")
except:
return False
def get_colID(c):
"Return collection ID for collection name C. Return None if no match found."
colID = None
res = run_sql("SELECT id FROM collection WHERE name=%s", (c,), 1)
if res:
colID = res[0][0]
return colID
def get_coll_ancestors(coll):
"Returns a list of ancestors for collection 'coll'."
coll_ancestors = []
coll_ancestor = coll
while 1:
res = run_sql("""SELECT c.name FROM collection AS c
LEFT JOIN collection_collection AS cc ON c.id=cc.id_dad
LEFT JOIN collection AS ccc ON ccc.id=cc.id_son
WHERE ccc.name=%s ORDER BY cc.id_dad ASC LIMIT 1""",
(coll_ancestor,))
if res:
coll_name = res[0][0]
coll_ancestors.append(coll_name)
coll_ancestor = coll_name
else:
break
# ancestors found, return reversed list:
coll_ancestors.reverse()
return coll_ancestors
def get_coll_sons(coll, type='r', public_only=1):
"""Return a list of sons (first-level descendants) of type 'type' for collection 'coll'.
If public_only, then return only non-restricted son collections.
"""
coll_sons = []
query = "SELECT c.name FROM collection AS c "\
"LEFT JOIN collection_collection AS cc ON c.id=cc.id_son "\
"LEFT JOIN collection AS ccc ON ccc.id=cc.id_dad "\
"WHERE cc.type=%s AND ccc.name=%s"
query += " ORDER BY cc.score DESC"
res = run_sql(query, (type, coll))
for name in res:
if not public_only or not collection_restricted_p(name[0]):
coll_sons.append(name[0])
return coll_sons
def get_coll_real_descendants(coll, type='_', get_hosted_colls=True):
"""Return a list of all descendants of collection 'coll' that are defined by a 'dbquery'.
IOW, we need to decompose compound collections like "A & B" into "A" and "B" provided
that "A & B" has no associated database query defined.
"""
coll_sons = []
res = run_sql("""SELECT c.name,c.dbquery FROM collection AS c
LEFT JOIN collection_collection AS cc ON c.id=cc.id_son
LEFT JOIN collection AS ccc ON ccc.id=cc.id_dad
WHERE ccc.name=%s AND cc.type LIKE %s ORDER BY cc.score DESC""",
(coll, type,))
for name, dbquery in res:
if dbquery: # this is 'real' collection, so return it:
if get_hosted_colls:
coll_sons.append(name)
else:
if not dbquery.startswith("hostedcollection:"):
coll_sons.append(name)
else: # this is 'composed' collection, so recurse:
coll_sons.extend(get_coll_real_descendants(name))
return coll_sons
def browse_pattern(req, colls, p, f, rg, ln=CFG_SITE_LANG):
"""Browse either biliographic phrases or words indexes, and display it."""
# load the right message language
_ = gettext_set_language(ln)
## is p enclosed in quotes? (coming from exact search)
if p.startswith('"') and p.endswith('"'):
p = p[1:-1]
p_orig = p
## okay, "real browse" follows:
## FIXME: the maths in the get_nearest_terms_in_bibxxx is just a test
if not f and string.find(p, ":") > 0: # does 'p' contain ':'?
f, p = string.split(p, ":", 1)
## do we search in words indexes?
if not f:
return browse_in_bibwords(req, p, f)
index_id = get_index_id_from_field(f)
if index_id != 0:
coll = HitSet()
for coll_name in colls:
coll |= get_collection_reclist(coll_name)
browsed_phrases_in_colls = get_nearest_terms_in_idxphrase_with_collection(p, index_id, rg/2, rg/2, coll)
else:
browsed_phrases = get_nearest_terms_in_bibxxx(p, f, (rg+1)/2+1, (rg-1)/2+1)
while not browsed_phrases:
# try again and again with shorter and shorter pattern:
try:
p = p[:-1]
browsed_phrases = get_nearest_terms_in_bibxxx(p, f, (rg+1)/2+1, (rg-1)/2+1)
except:
# probably there are no hits at all:
req.write(_("No values found."))
return
## try to check hits in these particular collection selection:
browsed_phrases_in_colls = []
if 0:
for phrase in browsed_phrases:
phrase_hitset = HitSet()
phrase_hitsets = search_pattern("", phrase, f, 'e')
for coll in colls:
phrase_hitset.union_update(phrase_hitsets[coll])
if len(phrase_hitset) > 0:
# okay, this phrase has some hits in colls, so add it:
browsed_phrases_in_colls.append([phrase, len(phrase_hitset)])
## were there hits in collections?
if browsed_phrases_in_colls == []:
if browsed_phrases != []:
#print_warning(req, """
No match close to %s found in given collections.
#Please try different term.
Displaying matches in any collection...""" % p_orig)
## try to get nbhits for these phrases in any collection:
for phrase in browsed_phrases:
browsed_phrases_in_colls.append([phrase, get_nbhits_in_bibxxx(phrase, f)])
## display results now:
out = websearch_templates.tmpl_browse_pattern(
f=f,
fn=get_field_i18nname(get_field_name(f) or f, ln, False),
ln=ln,
browsed_phrases_in_colls=browsed_phrases_in_colls,
colls=colls,
rg=rg,
)
req.write(out)
return
def browse_in_bibwords(req, p, f, ln=CFG_SITE_LANG):
"""Browse inside words indexes."""
if not p:
return
_ = gettext_set_language(ln)
urlargd = {}
urlargd.update(req.argd)
urlargd['action'] = 'search'
nearest_box = create_nearest_terms_box(urlargd, p, f, 'w', ln=ln, intro_text_p=0)
req.write(websearch_templates.tmpl_search_in_bibwords(
p = p,
f = f,
ln = ln,
nearest_box = nearest_box
))
return
def search_pattern(req=None, p=None, f=None, m=None, ap=0, of="id", verbose=0, ln=CFG_SITE_LANG, display_nearest_terms_box=True):
"""Search for complex pattern 'p' within field 'f' according to
matching type 'm'. Return hitset of recIDs.
The function uses multi-stage searching algorithm in case of no
exact match found. See the Search Internals document for
detailed description.
The 'ap' argument governs whether an alternative patterns are to
be used in case there is no direct hit for (p,f,m). For
example, whether to replace non-alphanumeric characters by
spaces if it would give some hits. See the Search Internals
document for detailed description. (ap=0 forbits the
alternative pattern usage, ap=1 permits it.)
The 'of' argument governs whether to print or not some
information to the user in case of no match found. (Usually it
prints the information in case of HTML formats, otherwise it's
silent).
The 'verbose' argument controls the level of debugging information
to be printed (0=least, 9=most).
All the parameters are assumed to have been previously washed.
This function is suitable as a mid-level API.
"""
_ = gettext_set_language(ln)
hitset_empty = HitSet()
# sanity check:
if not p:
hitset_full = HitSet(trailing_bits=1)
hitset_full.discard(0)
# no pattern, so return all universe
return hitset_full
# search stage 1: break up arguments into basic search units:
if verbose and of.startswith("h"):
t1 = os.times()[4]
basic_search_units = create_basic_search_units(req, p, f, m, of)
if verbose and of.startswith("h"):
t2 = os.times()[4]
print_warning(req, "Search stage 1: basic search units are: %s" % cgi.escape(repr(basic_search_units)))
print_warning(req, "Search stage 1: execution took %.2f seconds." % (t2 - t1))
# search stage 2: do search for each search unit and verify hit presence:
if verbose and of.startswith("h"):
t1 = os.times()[4]
basic_search_units_hitsets = []
for idx_unit in xrange(len(basic_search_units)):
bsu_o, bsu_p, bsu_f, bsu_m = basic_search_units[idx_unit]
basic_search_unit_hitset = search_unit(bsu_p, bsu_f, bsu_m)
if verbose >= 9 and of.startswith("h"):
print_warning(req, "Search stage 1: pattern %s gave hitlist %s" % (cgi.escape(bsu_p), basic_search_unit_hitset))
if len(basic_search_unit_hitset) > 0 or \
ap==0 or \
bsu_o=="|" or \
((idx_unit+1) 0:
# we retain the new unit instead
if of.startswith('h'):
print_warning(req, _("No exact match found for %(x_query1)s, using %(x_query2)s instead...") % \
{'x_query1': "" + cgi.escape(bsu_p) + "",
'x_query2': "" + cgi.escape(bsu_pn) + ""})
basic_search_units[idx_unit][1] = bsu_pn
basic_search_units_hitsets.append(basic_search_unit_hitset)
else:
# stage 2-3: no hits found either, propose nearest indexed terms:
if of.startswith('h') and display_nearest_terms_box:
if req:
if bsu_f == "recid":
print_warning(req, "Requested record does not seem to exist.")
else:
print_warning(req, create_nearest_terms_box(req.argd, bsu_p, bsu_f, bsu_m, ln=ln))
return hitset_empty
else:
# stage 2-3: no hits found either, propose nearest indexed terms:
if of.startswith('h') and display_nearest_terms_box:
if req:
if bsu_f == "recid":
print_warning(req, "Requested record does not seem to exist.")
else:
print_warning(req, create_nearest_terms_box(req.argd, bsu_p, bsu_f, bsu_m, ln=ln))
return hitset_empty
if verbose and of.startswith("h"):
t2 = os.times()[4]
for idx_unit in range(0, len(basic_search_units)):
print_warning(req, "Search stage 2: basic search unit %s gave %d hits." %
(basic_search_units[idx_unit][1:], len(basic_search_units_hitsets[idx_unit])))
print_warning(req, "Search stage 2: execution took %.2f seconds." % (t2 - t1))
# search stage 3: apply boolean query for each search unit:
if verbose and of.startswith("h"):
t1 = os.times()[4]
# let the initial set be the complete universe:
hitset_in_any_collection = HitSet(trailing_bits=1)
hitset_in_any_collection.discard(0)
for idx_unit in xrange(len(basic_search_units)):
this_unit_operation = basic_search_units[idx_unit][0]
this_unit_hitset = basic_search_units_hitsets[idx_unit]
if this_unit_operation == '+':
hitset_in_any_collection.intersection_update(this_unit_hitset)
elif this_unit_operation == '-':
hitset_in_any_collection.difference_update(this_unit_hitset)
elif this_unit_operation == '|':
hitset_in_any_collection.union_update(this_unit_hitset)
else:
if of.startswith("h"):
print_warning(req, "Invalid set operation %s." % cgi.escape(this_unit_operation), "Error")
if len(hitset_in_any_collection) == 0:
# no hits found, propose alternative boolean query:
if of.startswith('h') and display_nearest_terms_box:
nearestterms = []
for idx_unit in range(0, len(basic_search_units)):
bsu_o, bsu_p, bsu_f, bsu_m = basic_search_units[idx_unit]
if bsu_p.startswith("%") and bsu_p.endswith("%"):
bsu_p = "'" + bsu_p[1:-1] + "'"
bsu_nbhits = len(basic_search_units_hitsets[idx_unit])
# create a similar query, but with the basic search unit only
argd = {}
argd.update(req.argd)
argd['p'] = bsu_p
argd['f'] = bsu_f
nearestterms.append((bsu_p, bsu_nbhits, argd))
text = websearch_templates.tmpl_search_no_boolean_hits(
ln=ln, nearestterms=nearestterms)
print_warning(req, text)
if verbose and of.startswith("h"):
t2 = os.times()[4]
print_warning(req, "Search stage 3: boolean query gave %d hits." % len(hitset_in_any_collection))
print_warning(req, "Search stage 3: execution took %.2f seconds." % (t2 - t1))
return hitset_in_any_collection
def search_pattern_parenthesised(req=None, p=None, f=None, m=None, ap=0, of="id", verbose=0, ln=CFG_SITE_LANG, display_nearest_terms_box=True):
"""Search for complex pattern 'p' containing parenthesis within field 'f' according to
matching type 'm'. Return hitset of recIDs.
For more details on the parameters see 'search_pattern'
"""
_ = gettext_set_language(ln)
# if the pattern uses SPIRES search syntax, convert it to Invenio syntax
spires_syntax_converter = SpiresToInvenioSyntaxConverter()
p = spires_syntax_converter.convert_query(p)
# sanity check: do not call parenthesised parser for search terms
# like U(1):
if not re_pattern_parens.search(p):
return search_pattern(req, p, f, m, ap, of, verbose, ln, display_nearest_terms_box=display_nearest_terms_box)
# Try searching with parentheses
try:
parser = SearchQueryParenthesisedParser()
# get a hitset with all recids
result_hitset = HitSet(trailing_bits=1)
# parse the query. The result is list of [op1, expr1, op2, expr2, ..., opN, exprN]
parsing_result = parser.parse_query(p)
if verbose and of.startswith("h"):
print_warning(req, "Search stage 1: search_pattern_parenthesised() returned %s." % repr(parsing_result))
# go through every pattern
# calculate hitset for it
# combine pattern's hitset with the result using the corresponding operator
for index in xrange(0, len(parsing_result)-1, 2 ):
current_operator = parsing_result[index]
current_pattern = parsing_result[index+1]
# obtain a hitset for the current pattern
current_hitset = search_pattern(req, current_pattern, f, m, ap, of, verbose, ln, display_nearest_terms_box=display_nearest_terms_box)
# combine the current hitset with resulting hitset using the current operator
if current_operator == '+':
result_hitset = result_hitset & current_hitset
elif current_operator == '-':
result_hitset = result_hitset - current_hitset
elif current_operator == '|':
result_hitset = result_hitset | current_hitset
else:
assert False, "Unknown operator in search_pattern_parenthesised()"
return result_hitset
# If searching with parenteses fails, perform search ignoring parentheses
except InvenioWebSearchQueryParserException:
print_warning(req, _("Nested or mismatched parentheses detected. Ignoring all parentheses in the query..."))
# remove the parentheses in the query. Current implementation removes all the parentheses,
# but it could be improved to romove only these that are not insede quotes
p = p.replace('(', ' ')
p = p.replace(')', ' ')
return search_pattern(req, p, f, m, ap, of, verbose, ln, display_nearest_terms_box=display_nearest_terms_box)
def search_unit(p, f=None, m=None):
"""Search for basic search unit defined by pattern 'p' and field
'f' and matching type 'm'. Return hitset of recIDs.
All the parameters are assumed to have been previously washed.
'p' is assumed to be already a ``basic search unit'' so that it
is searched as such and is not broken up in any way. Only
wildcard and span queries are being detected inside 'p'.
This function is suitable as a low-level API.
"""
## create empty output results set:
set = HitSet()
if not p: # sanity checking
return set
if m == 'a' or m == 'r':
# we are doing either phrase search or regexp search
index_id = get_index_id_from_field(f)
if index_id != 0:
set = search_unit_in_idxphrases(p, f, m)
else:
set = search_unit_in_bibxxx(p, f, m)
elif p.startswith("cited:"):
# we are doing search by the citation count
set = search_unit_by_times_cited(p[6:])
else:
# we are doing bibwords search by default
set = search_unit_in_bibwords(p, f)
return set
def search_unit_in_bibwords(word, f, decompress=zlib.decompress):
"""Searches for 'word' inside bibwordsX table for field 'f' and returns hitset of recIDs."""
set = HitSet() # will hold output result set
set_used = 0 # not-yet-used flag, to be able to circumvent set operations
# deduce into which bibwordsX table we will search:
stemming_language = get_index_stemming_language(get_index_id_from_field("anyfield"))
bibwordsX = "idxWORD%02dF" % get_index_id_from_field("anyfield")
if f:
index_id = get_index_id_from_field(f)
if index_id:
bibwordsX = "idxWORD%02dF" % index_id
stemming_language = get_index_stemming_language(index_id)
else:
return HitSet() # word index f does not exist
# wash 'word' argument and run query:
word = string.replace(word, '*', '%') # we now use '*' as the truncation character
words = string.split(word, "->", 1) # check for span query
if len(words) == 2:
word0 = re_word.sub('', words[0])
word1 = re_word.sub('', words[1])
if stemming_language:
word0 = lower_index_term(word0)
word1 = lower_index_term(word1)
word0 = stem(word0, stemming_language)
word1 = stem(word1, stemming_language)
res = run_sql("SELECT term,hitlist FROM %s WHERE term BETWEEN %%s AND %%s" % bibwordsX,
(wash_index_term(word0), wash_index_term(word1)))
else:
if f == 'journal':
pass # FIXME: quick hack for the journal index
else:
word = re_word.sub('', word)
if stemming_language:
word = lower_index_term(word)
word = stem(word, stemming_language)
if string.find(word, '%') >= 0: # do we have wildcard in the word?
if f == 'journal':
# FIXME: quick hack for the journal index
# FIXME: we can run a sanity check here for all indexes
res = ()
else:
res = run_sql("SELECT term,hitlist FROM %s WHERE term LIKE %%s" % bibwordsX,
(wash_index_term(word),))
else:
res = run_sql("SELECT term,hitlist FROM %s WHERE term=%%s" % bibwordsX,
(wash_index_term(word),))
# fill the result set:
for word, hitlist in res:
hitset_bibwrd = HitSet(hitlist)
# add the results:
if set_used:
set.union_update(hitset_bibwrd)
else:
set = hitset_bibwrd
set_used = 1
# okay, return result set:
return set
def search_unit_in_idxphrases(p, f, type):
"""Searches for phrase 'p' inside idxPHRASE*F table for field 'f' and returns hitset of recIDs found.
The search type is defined by 'type' (e.g. equals to 'r' for a regexp search)."""
set = HitSet() # will hold output result set
set_used = 0 # not-yet-used flag, to be able to circumvent set operations
# deduce in which idxPHRASE table we will search:
idxphraseX = "idxPHRASE%02dF" % get_index_id_from_field("anyfield")
if f:
index_id = get_index_id_from_field(f)
if index_id:
idxphraseX = "idxPHRASE%02dF" % index_id
else:
return HitSet() # phrase index f does not exist
# detect query type (exact phrase, partial phrase, regexp):
if type == 'r':
query_addons = "REGEXP %s"
query_params = (p,)
else:
p = string.replace(p, '*', '%') # we now use '*' as the truncation character
ps = string.split(p, "->", 1) # check for span query:
if len(ps) == 2:
query_addons = "BETWEEN %s AND %s"
query_params = (ps[0], ps[1])
else:
if string.find(p, '%') > -1:
query_addons = "LIKE %s"
query_params = (ps[0],)
else:
query_addons = "= %s"
query_params = (ps[0],)
# perform search:
res = run_sql("SELECT term,hitlist FROM %s WHERE term %s" % (idxphraseX, query_addons),
query_params)
# fill the result set:
for word, hitlist in res:
hitset_bibphrase = HitSet(hitlist)
# add the results:
if set_used:
set.union_update(hitset_bibphrase)
else:
set = hitset_bibphrase
set_used = 1
# okay, return result set:
return set
def search_unit_in_bibxxx(p, f, type):
"""Searches for pattern 'p' inside bibxxx tables for field 'f' and returns hitset of recIDs found.
The search type is defined by 'type' (e.g. equals to 'r' for a regexp search)."""
# FIXME: quick hack for the journal index
if f == 'journal':
return search_unit_in_bibwords(p, f)
p_orig = p # saving for eventual future 'no match' reporting
query_addons = "" # will hold additional SQL code for the query
query_params = () # will hold parameters for the query (their number may vary depending on TYPE argument)
# wash arguments:
f = string.replace(f, '*', '%') # replace truncation char '*' in field definition
if type == 'r':
query_addons = "REGEXP %s"
query_params = (p,)
else:
p = string.replace(p, '*', '%') # we now use '*' as the truncation character
ps = string.split(p, "->", 1) # check for span query:
if len(ps) == 2:
query_addons = "BETWEEN %s AND %s"
query_params = (ps[0], ps[1])
else:
if string.find(p, '%') > -1:
query_addons = "LIKE %s"
query_params = (ps[0],)
else:
query_addons = "= %s"
query_params = (ps[0],)
# construct 'tl' which defines the tag list (MARC tags) to search in:
tl = []
if str(f[0]).isdigit() and str(f[1]).isdigit():
tl.append(f) # 'f' seems to be okay as it starts by two digits
else:
# convert old ALEPH tag names, if appropriate: (TODO: get rid of this before entering this function)
if CFG_WEBSEARCH_FIELDS_CONVERT.has_key(string.lower(f)):
f = CFG_WEBSEARCH_FIELDS_CONVERT[string.lower(f)]
# deduce desired MARC tags on the basis of chosen 'f'
tl = get_field_tags(f)
if not tl:
# f index does not exist, nevermind
pass
# okay, start search:
l = [] # will hold list of recID that matched
for t in tl:
# deduce into which bibxxx table we will search:
digit1, digit2 = int(t[0]), int(t[1])
bx = "bib%d%dx" % (digit1, digit2)
bibx = "bibrec_bib%d%dx" % (digit1, digit2)
# construct and run query:
if t == "001":
res = run_sql("SELECT id FROM bibrec WHERE id %s" % query_addons,
query_params)
else:
query = "SELECT bibx.id_bibrec FROM %s AS bx LEFT JOIN %s AS bibx ON bx.id=bibx.id_bibxxx WHERE bx.value %s" % \
(bx, bibx, query_addons)
if len(t) != 6 or t[-1:]=='%':
# wildcard query, or only the beginning of field 't'
# is defined, so add wildcard character:
query += " AND bx.tag LIKE %s"
res = run_sql(query, query_params + (t + '%',))
else:
# exact query for 't':
query += " AND bx.tag=%s"
res = run_sql(query, query_params + (t,))
# fill the result set:
for id_bibrec in res:
if id_bibrec[0]:
l.append(id_bibrec[0])
# check no of hits found:
nb_hits = len(l)
# okay, return result set:
set = HitSet(l)
return set
def search_unit_in_bibrec(datetext1, datetext2, type='c'):
"""
Return hitset of recIDs found that were either created or modified
(according to 'type' arg being 'c' or 'm') from datetext1 until datetext2, inclusive.
Does not pay attention to pattern, collection, anything. Useful
to intersect later on with the 'real' query.
"""
set = HitSet()
if type.startswith("m"):
type = "modification_date"
else:
type = "creation_date" # by default we are searching for creation dates
res = run_sql("SELECT id FROM bibrec WHERE %s>=%%s AND %s<=%%s" % (type, type),
(datetext1, datetext2))
for row in res:
set += row[0]
return set
def search_unit_by_times_cited(p):
"""
Return histset of recIDs found that are cited P times.
Usually P looks like '10->23'.
"""
numstr = '"'+p+'"'
#this is sort of stupid but since we may need to
#get the records that do _not_ have cites, we have to
#know the ids of all records, too
#but this is needed only if bsu_p is 0 or 0 or 0->0
allrecs = []
if p == 0 or p == "0" or \
p.startswith("0->") or p.endswith("->0"):
allrecs = HitSet(run_sql_cached("SELECT id FROM bibrec", affected_tables=['bibrec']))
return get_records_with_num_cites(numstr, allrecs)
def intersect_results_with_collrecs(req, hitset_in_any_collection, colls, ap=0, of="hb", verbose=0, ln=CFG_SITE_LANG, display_nearest_terms_box=True):
"""Return dict of hitsets given by intersection of hitset with the collection universes."""
_ = gettext_set_language(ln)
# search stage 4: intersect with the collection universe:
if verbose and of.startswith("h"):
t1 = os.times()[4]
results = {}
results_nbhits = 0
for coll in colls:
results[coll] = hitset_in_any_collection & get_collection_reclist(coll)
results_nbhits += len(results[coll])
if results_nbhits == 0:
# no hits found, try to search in Home:
results_in_Home = hitset_in_any_collection & get_collection_reclist(CFG_SITE_NAME)
if len(results_in_Home) > 0:
# some hits found in Home, so propose this search:
if of.startswith("h") and display_nearest_terms_box:
url = websearch_templates.build_search_url(req.argd, cc=CFG_SITE_NAME, c=[])
print_warning(req, _("No match found in collection %(x_collection)s. Other public collections gave %(x_url_open)s%(x_nb_hits)d hits%(x_url_close)s.") %\
{'x_collection': '' + string.join([get_coll_i18nname(coll, ln, False) for coll in colls], ', ') + '',
'x_url_open': '' % (url),
'x_nb_hits': len(results_in_Home),
'x_url_close': ''})
results = {}
else:
# no hits found in Home, recommend different search terms:
if of.startswith("h") and display_nearest_terms_box:
print_warning(req, _("No public collection matched your query. "
"If you were looking for a non-public document, please choose "
"the desired restricted collection first."))
results = {}
if verbose and of.startswith("h"):
t2 = os.times()[4]
print_warning(req, "Search stage 4: intersecting with collection universe gave %d hits." % results_nbhits)
print_warning(req, "Search stage 4: execution took %.2f seconds." % (t2 - t1))
return results
def intersect_results_with_hitset(req, results, hitset, ap=0, aptext="", of="hb"):
"""Return intersection of search 'results' (a dict of hitsets
with collection as key) with the 'hitset', i.e. apply
'hitset' intersection to each collection within search
'results'.
If the final 'results' set is to be empty, and 'ap'
(approximate pattern) is true, and then print the `warningtext'
and return the original 'results' set unchanged. If 'ap' is
false, then return empty results set.
"""
if ap:
results_ap = copy.deepcopy(results)
else:
results_ap = {} # will return empty dict in case of no hits found
nb_total = 0
for coll in results.keys():
results[coll].intersection_update(hitset)
nb_total += len(results[coll])
if nb_total == 0:
if of.startswith("h"):
print_warning(req, aptext)
results = results_ap
return results
def create_similarly_named_authors_link_box(author_name, ln=CFG_SITE_LANG):
"""Return a box similar to ``Not satisfied...'' one by proposing
author searches for similar names. Namely, take AUTHOR_NAME
and the first initial of the firstame (after comma) and look
into author index whether authors with e.g. middle names exist.
Useful mainly for CERN Library that sometimes contains name
forms like Ellis-N, Ellis-Nick, Ellis-Nicolas all denoting the
same person. The box isn't proposed if no similarly named
authors are found to exist.
"""
# return nothing if not configured:
if CFG_WEBSEARCH_CREATE_SIMILARLY_NAMED_AUTHORS_LINK_BOX == 0:
return ""
# return empty box if there is no initial:
if re.match(r'[^ ,]+, [^ ]', author_name) is None:
return ""
# firstly find name comma initial:
author_name_to_search = re.sub(r'^([^ ,]+, +[^ ,]).*$', '\\1', author_name)
# secondly search for similar name forms:
similar_author_names = {}
for name in author_name_to_search, strip_accents(author_name_to_search):
for tag in get_field_tags("author"):
# deduce into which bibxxx table we will search:
digit1, digit2 = int(tag[0]), int(tag[1])
bx = "bib%d%dx" % (digit1, digit2)
bibx = "bibrec_bib%d%dx" % (digit1, digit2)
if len(tag) != 6 or tag[-1:]=='%':
# only the beginning of field 't' is defined, so add wildcard character:
res = run_sql("""SELECT bx.value FROM %s AS bx
WHERE bx.value LIKE %%s AND bx.tag LIKE %%s""" % bx,
(name + "%", tag + "%"))
else:
res = run_sql("""SELECT bx.value FROM %s AS bx
WHERE bx.value LIKE %%s AND bx.tag=%%s""" % bx,
(name + "%", tag))
for row in res:
similar_author_names[row[0]] = 1
# remove the original name and sort the list:
try:
del similar_author_names[author_name]
except KeyError:
pass
# thirdly print the box:
out = ""
if similar_author_names:
out_authors = similar_author_names.keys()
out_authors.sort()
tmp_authors = []
for out_author in out_authors:
nbhits = get_nbhits_in_bibxxx(out_author, "author")
if nbhits:
tmp_authors.append((out_author, nbhits))
out += websearch_templates.tmpl_similar_author_names(
authors=tmp_authors, ln=ln)
return out
def create_nearest_terms_box(urlargd, p, f, t='w', n=5, ln=CFG_SITE_LANG, intro_text_p=True):
"""Return text box containing list of 'n' nearest terms above/below 'p'
for the field 'f' for matching type 't' (words/phrases) in
language 'ln'.
Propose new searches according to `urlargs' with the new words.
If `intro_text_p' is true, then display the introductory message,
otherwise print only the nearest terms in the box content.
"""
# load the right message language
_ = gettext_set_language(ln)
out = ""
nearest_terms = []
if not p: # sanity check
p = "."
index_id = get_index_id_from_field(f)
# look for nearest terms:
if t == 'w':
nearest_terms = get_nearest_terms_in_bibwords(p, f, n, n)
if not nearest_terms:
return _("No word index is available for %s.") % \
('' + cgi.escape(get_field_i18nname(get_field_name(f) or f, ln, False)) + '')
else:
nearest_terms = []
if index_id:
nearest_terms = get_nearest_terms_in_idxphrase(p, index_id, n, n)
if not nearest_terms:
nearest_terms = get_nearest_terms_in_bibxxx(p, f, n, n)
if not nearest_terms:
return _("No phrase index is available for %s.") % \
('' + cgi.escape(get_field_i18nname(get_field_name(f) or f, ln, False)) + '')
terminfo = []
for term in nearest_terms:
if t == 'w':
hits = get_nbhits_in_bibwords(term, f)
else:
if index_id:
hits = get_nbhits_in_idxphrases(term, f)
else:
hits = get_nbhits_in_bibxxx(term, f)
argd = {}
argd.update(urlargd)
# check which fields contained the requested parameter, and replace it.
for (px, fx) in ('p', 'f'), ('p1', 'f1'), ('p2', 'f2'), ('p3', 'f3'):
if px in argd:
argd_px = argd[px]
if t == 'w':
# p was stripped of accents, to do the same:
argd_px = strip_accents(argd_px)
if f == argd[fx] or f == "anyfield" or f == "":
if string.find(argd_px, p) > -1:
argd[px] = string.replace(argd_px, p, term)
break
else:
if string.find(argd_px, f+':'+p) > -1:
argd[px] = string.replace(argd_px, f+':'+p, f+':'+term)
break
elif string.find(argd_px, f+':"'+p+'"') > -1:
argd[px] = string.replace(argd_px, f+':"'+p+'"', f+':"'+term+'"')
break
terminfo.append((term, hits, argd))
intro = ""
if intro_text_p: # add full leading introductory text
if f:
intro = _("Search term %(x_term)s inside index %(x_index)s did not match any record. Nearest terms in any collection are:") % \
{'x_term': "" + cgi.escape(p.startswith("%") and p.endswith("%") and p[1:-1] or p) + "",
'x_index': "" + cgi.escape(get_field_i18nname(get_field_name(f) or f, ln, False)) + ""}
else:
intro = _("Search term %s did not match any record. Nearest terms in any collection are:") % \
("" + cgi.escape(p.startswith("%") and p.endswith("%") and p[1:-1] or p) + "")
return websearch_templates.tmpl_nearest_term_box(p=p, ln=ln, f=f, terminfo=terminfo,
intro=intro)
def get_nearest_terms_in_bibwords(p, f, n_below, n_above):
"""Return list of +n -n nearest terms to word `p' in index for field `f'."""
nearest_words = [] # will hold the (sorted) list of nearest words to return
# deduce into which bibwordsX table we will search:
bibwordsX = "idxWORD%02dF" % get_index_id_from_field("anyfield")
if f:
index_id = get_index_id_from_field(f)
if index_id:
bibwordsX = "idxWORD%02dF" % index_id
else:
return nearest_words
# firstly try to get `n' closest words above `p':
res = run_sql("SELECT term FROM %s WHERE term<%%s ORDER BY term DESC LIMIT %%s" % bibwordsX,
(p, n_above))
for row in res:
nearest_words.append(row[0])
nearest_words.reverse()
# secondly insert given word `p':
nearest_words.append(p)
# finally try to get `n' closest words below `p':
res = run_sql("SELECT term FROM %s WHERE term>%%s ORDER BY term ASC LIMIT %%s" % bibwordsX,
(p, n_below))
for row in res:
nearest_words.append(row[0])
return nearest_words
def get_nearest_terms_in_idxphrase(p, index_id, n_below, n_above):
"""Browse (-n_above, +n_below) closest bibliographic phrases
for the given pattern p in the given field idxPHRASE table,
regardless of collection.
Return list of [phrase1, phrase2, ... , phrase_n]."""
idxphraseX = "idxPHRASE%02dF" % index_id
res_above = run_sql("SELECT term FROM %s WHERE term<%%s ORDER BY term DESC LIMIT %%s" % idxphraseX, (p, n_above))
res_above = map(lambda x: x[0], res_above)
res_above.reverse()
res_below = run_sql("SELECT term FROM %s WHERE term>=%%s ORDER BY term ASC LIMIT %%s" % idxphraseX, (p, n_below))
res_below = map(lambda x: x[0], res_below)
return res_above + res_below
def get_nearest_terms_in_idxphrase_with_collection(p, index_id, n_below, n_above, collection):
"""Browse (-n_above, +n_below) closest bibliographic phrases
for the given pattern p in the given field idxPHRASE table,
considering the collection (HitSet).
Return list of [(phrase1, hitset), (phrase2, hitset), ... , (phrase_n, hitset)]."""
idxphraseX = "idxPHRASE%02dF" % index_id
res_above = run_sql("SELECT term,hitlist FROM %s WHERE term<%%s ORDER BY term DESC LIMIT %%s" % idxphraseX, (p, n_above * 3))
res_above = [(term, HitSet(hitlist) & collection) for term, hitlist in res_above]
res_above = [(term, len(hitlist)) for term, hitlist in res_above if hitlist]
res_below = run_sql("SELECT term,hitlist FROM %s WHERE term>=%%s ORDER BY term ASC LIMIT %%s" % idxphraseX, (p, n_below * 3))
res_below = [(term, HitSet(hitlist) & collection) for term, hitlist in res_below]
res_below = [(term, len(hitlist)) for term, hitlist in res_below if hitlist]
res_above.reverse()
return res_above[-n_above:] + res_below[:n_below]
def get_nearest_terms_in_bibxxx(p, f, n_below, n_above):
"""Browse (-n_above, +n_below) closest bibliographic phrases
for the given pattern p in the given field f, regardless
of collection.
Return list of [phrase1, phrase2, ... , phrase_n]."""
## determine browse field:
if not f and string.find(p, ":") > 0: # does 'p' contain ':'?
f, p = string.split(p, ":", 1)
# FIXME: quick hack for the journal index
if f == 'journal':
return get_nearest_terms_in_bibwords(p, f, n_below, n_above)
## We are going to take max(n_below, n_above) as the number of
## values to ferch from bibXXx. This is needed to work around
## MySQL UTF-8 sorting troubles in 4.0.x. Proper solution is to
## use MySQL 4.1.x or our own idxPHRASE in the future.
index_id = get_index_id_from_field(f)
if index_id:
return get_nearest_terms_in_idxphrase(p, index_id, n_below, n_above)
n_fetch = 2*max(n_below, n_above)
## construct 'tl' which defines the tag list (MARC tags) to search in:
tl = []
if str(f[0]).isdigit() and str(f[1]).isdigit():
tl.append(f) # 'f' seems to be okay as it starts by two digits
else:
# deduce desired MARC tags on the basis of chosen 'f'
tl = get_field_tags(f)
## start browsing to fetch list of hits:
browsed_phrases = {} # will hold {phrase1: 1, phrase2: 1, ..., phraseN: 1} dict of browsed phrases (to make them unique)
# always add self to the results set:
browsed_phrases[p.startswith("%") and p.endswith("%") and p[1:-1] or p] = 1
for t in tl:
# deduce into which bibxxx table we will search:
digit1, digit2 = int(t[0]), int(t[1])
bx = "bib%d%dx" % (digit1, digit2)
bibx = "bibrec_bib%d%dx" % (digit1, digit2)
# firstly try to get `n' closest phrases above `p':
if len(t) != 6 or t[-1:]=='%': # only the beginning of field 't' is defined, so add wildcard character:
res = run_sql("""SELECT bx.value FROM %s AS bx
WHERE bx.value<%%s AND bx.tag LIKE %%s
ORDER BY bx.value DESC LIMIT %%s""" % bx,
(p, t + "%", n_fetch))
else:
res = run_sql("""SELECT bx.value FROM %s AS bx
WHERE bx.value<%%s AND bx.tag=%%s
ORDER BY bx.value DESC LIMIT %%s""" % bx,
(p, t, n_fetch))
for row in res:
browsed_phrases[row[0]] = 1
# secondly try to get `n' closest phrases equal to or below `p':
if len(t) != 6 or t[-1:]=='%': # only the beginning of field 't' is defined, so add wildcard character:
res = run_sql("""SELECT bx.value FROM %s AS bx
WHERE bx.value>=%%s AND bx.tag LIKE %%s
ORDER BY bx.value ASC LIMIT %%s""" % bx,
(p, t + "%", n_fetch))
else:
res = run_sql("""SELECT bx.value FROM %s AS bx
WHERE bx.value>=%%s AND bx.tag=%%s
ORDER BY bx.value ASC LIMIT %%s""" % bx,
(p, t, n_fetch))
for row in res:
browsed_phrases[row[0]] = 1
# select first n words only: (this is needed as we were searching
# in many different tables and so aren't sure we have more than n
# words right; this of course won't be needed when we shall have
# one ACC table only for given field):
phrases_out = browsed_phrases.keys()
phrases_out.sort(lambda x, y: cmp(string.lower(strip_accents(x)),
string.lower(strip_accents(y))))
# find position of self:
try:
idx_p = phrases_out.index(p)
except:
idx_p = len(phrases_out)/2
# return n_above and n_below:
return phrases_out[max(0, idx_p-n_above):idx_p+n_below]
def get_nbhits_in_bibwords(word, f):
"""Return number of hits for word 'word' inside words index for field 'f'."""
out = 0
# deduce into which bibwordsX table we will search:
bibwordsX = "idxWORD%02dF" % get_index_id_from_field("anyfield")
if f:
index_id = get_index_id_from_field(f)
if index_id:
bibwordsX = "idxWORD%02dF" % index_id
else:
return 0
if word:
res = run_sql("SELECT hitlist FROM %s WHERE term=%%s" % bibwordsX,
(word,))
for hitlist in res:
out += len(HitSet(hitlist[0]))
return out
def get_nbhits_in_idxphrases(word, f):
"""Return number of hits for word 'word' inside phrase index for field 'f'."""
out = 0
# deduce into which bibwordsX table we will search:
idxphraseX = "idxPHRASE%02dF" % get_index_id_from_field("anyfield")
if f:
index_id = get_index_id_from_field(f)
if index_id:
idxphraseX = "idxPHRASE%02dF" % index_id
else:
return 0
if word:
res = run_sql("SELECT hitlist FROM %s WHERE term=%%s" % idxphraseX,
(word,))
for hitlist in res:
out += len(HitSet(hitlist[0]))
return out
def get_nbhits_in_bibxxx(p, f):
"""Return number of hits for word 'word' inside words index for field 'f'."""
## determine browse field:
if not f and string.find(p, ":") > 0: # does 'p' contain ':'?
f, p = string.split(p, ":", 1)
# FIXME: quick hack for the journal index
if f == 'journal':
return get_nbhits_in_bibwords(p, f)
## construct 'tl' which defines the tag list (MARC tags) to search in:
tl = []
if str(f[0]).isdigit() and str(f[1]).isdigit():
tl.append(f) # 'f' seems to be okay as it starts by two digits
else:
# deduce desired MARC tags on the basis of chosen 'f'
tl = get_field_tags(f)
# start searching:
recIDs = {} # will hold dict of {recID1: 1, recID2: 1, ..., } (unique recIDs, therefore)
for t in tl:
# deduce into which bibxxx table we will search:
digit1, digit2 = int(t[0]), int(t[1])
bx = "bib%d%dx" % (digit1, digit2)
bibx = "bibrec_bib%d%dx" % (digit1, digit2)
if len(t) != 6 or t[-1:]=='%': # only the beginning of field 't' is defined, so add wildcard character:
res = run_sql("""SELECT bibx.id_bibrec FROM %s AS bibx, %s AS bx
WHERE bx.value=%%s AND bx.tag LIKE %%s
AND bibx.id_bibxxx=bx.id""" % (bibx, bx),
(p, t + "%"))
else:
res = run_sql("""SELECT bibx.id_bibrec FROM %s AS bibx, %s AS bx
WHERE bx.value=%%s AND bx.tag=%%s
AND bibx.id_bibxxx=bx.id""" % (bibx, bx),
(p, t))
for row in res:
recIDs[row[0]] = 1
return len(recIDs)
def get_mysql_recid_from_aleph_sysno(sysno):
"""Returns DB's recID for ALEPH sysno passed in the argument (e.g. "002379334CER").
Returns None in case of failure."""
out = None
res = run_sql("""SELECT bb.id_bibrec FROM bibrec_bib97x AS bb, bib97x AS b
WHERE b.value=%s AND b.tag='970__a' AND bb.id_bibxxx=b.id""",
(sysno,))
if res:
out = res[0][0]
return out
def guess_primary_collection_of_a_record(recID):
"""Return primary collection name a record recid belongs to, by
testing 980 identifier.
May lead to bad guesses when a collection is defined dynamically
via dbquery.
In that case, return 'CFG_SITE_NAME'."""
out = CFG_SITE_NAME
dbcollids = get_fieldvalues(recID, "980__a")
if dbcollids:
dbquery = "collection:" + dbcollids[0]
res = run_sql("SELECT name FROM collection WHERE dbquery=%s", (dbquery,))
if res:
out = res[0][0]
return out
_re_collection_url = re.compile('/collection/(.+)')
def guess_collection_of_a_record(recID, referer=None):
"""Return collection name a record recid belongs to, by first testing
the referer URL if provided and otherwise returning the
primary collection."""
if referer:
dummy, hostname, path, dummy, query, dummy = urlparse.urlparse(referer)
g = _re_collection_url.match(path)
if g:
name = urllib.unquote_plus(g.group(1))
if recID in get_collection_reclist(name):
return name
elif path.startswith('/search'):
query = cgi.parse_qs(query)
for name in query.get('cc', []) + query.get('c', []):
if recID in get_collection_reclist(name):
return name
return guess_primary_collection_of_a_record(recID)
def get_all_collections_of_a_record(recID):
"""Return all the collection names a record belongs to.
Note this function is O(n_collections)."""
ret = []
for name in collection_reclist_cache.cache.keys():
if recID in get_collection_reclist(name):
ret.append(name)
return ret
def get_tag_name(tag_value, prolog="", epilog=""):
"""Return tag name from the known tag value, by looking up the 'tag' table.
Return empty string in case of failure.
Example: input='100__%', output=first author'."""
out = ""
res = run_sql_cached("SELECT name FROM tag WHERE value=%s", (tag_value,),
affected_tables=['tag',])
if res:
out = prolog + res[0][0] + epilog
return out
def get_fieldcodes():
"""Returns a list of field codes that may have been passed as 'search options' in URL.
Example: output=['subject','division']."""
out = []
res = run_sql_cached("SELECT DISTINCT(code) FROM field",
affected_tables=['field',])
for row in res:
out.append(row[0])
return out
def get_field_name(code):
"""Return the corresponding field_name given the field code.
e.g. reportnumber -> report number."""
res = run_sql_cached("SELECT name FROM field WHERE code=%s", (code, ),
affected_tables=['field',])
if res:
return res[0][0]
else:
return ""
def get_field_tags(field):
"""Returns a list of MARC tags for the field code 'field'.
Returns empty list in case of error.
Example: field='author', output=['100__%','700__%']."""
out = []
query = """SELECT t.value FROM tag AS t, field_tag AS ft, field AS f
WHERE f.code=%s AND ft.id_field=f.id AND t.id=ft.id_tag
ORDER BY ft.score DESC"""
res = run_sql(query, (field, ))
for val in res:
out.append(val[0])
return out
def get_fieldvalues(recIDs, tag, repetitive_values=True):
"""
Return list of field values for field TAG for the given record ID
or list of record IDs. (RECIDS can be both an integer or a list
of integers.)
If REPETITIVE_VALUES is set to True, then return all values even
if they are doubled. If set to False, then return unique values
only.
"""
out = []
if isinstance(recIDs, (int, long)):
recIDs =[recIDs,]
if not isinstance(recIDs, (list, tuple)):
return []
if len(recIDs) == 0:
return []
if tag == "001___":
# we have asked for tag 001 (=recID) that is not stored in bibXXx tables
out = [str(recID) for recID in recIDs]
else:
# we are going to look inside bibXXx tables
digits = tag[0:2]
try:
intdigits = int(digits)
if intdigits < 0 or intdigits > 99:
raise ValueError
except ValueError:
# invalid tag value asked for
return []
bx = "bib%sx" % digits
bibx = "bibrec_bib%sx" % digits
queryparam = []
for recID in recIDs:
queryparam.append(recID)
if not repetitive_values:
queryselect = "DISTINCT(bx.value)"
else:
queryselect = "bx.value"
query = "SELECT %s FROM %s AS bx, %s AS bibx WHERE bibx.id_bibrec IN (%s) " \
" AND bx.id=bibx.id_bibxxx AND bx.tag LIKE %%s " \
" ORDER BY bibx.field_number, bx.tag ASC" % \
(queryselect, bx, bibx, ("%s,"*len(queryparam))[:-1])
res = run_sql(query, tuple(queryparam) + (tag,))
for row in res:
out.append(row[0])
return out
def get_fieldvalues_alephseq_like(recID, tags_in):
"""Return buffer of ALEPH sequential-like textual format with fields found in the list TAGS_IN for record RECID."""
out = ""
if type(tags_in) is not list:
tags_in = [tags_in,]
if len(tags_in) == 1 and len(tags_in[0]) == 6:
## case A: one concrete subfield asked, so print its value if found
## (use with care: can false you if field has multiple occurrences)
out += string.join(get_fieldvalues(recID, tags_in[0]),"\n")
else:
## case B: print our "text MARC" format; works safely all the time
# find out which tags to output:
dict_of_tags_out = {}
if not tags_in:
for i in range(0, 10):
for j in range(0, 10):
dict_of_tags_out["%d%d%%" % (i, j)] = 1
else:
for tag in tags_in:
if len(tag) == 0:
for i in range(0, 10):
for j in range(0, 10):
dict_of_tags_out["%d%d%%" % (i, j)] = 1
elif len(tag) == 1:
for j in range(0, 10):
dict_of_tags_out["%s%d%%" % (tag, j)] = 1
elif len(tag) < 5:
dict_of_tags_out["%s%%" % tag] = 1
elif tag >= 6:
dict_of_tags_out[tag[0:5]] = 1
tags_out = dict_of_tags_out.keys()
tags_out.sort()
# search all bibXXx tables as needed:
for tag in tags_out:
digits = tag[0:2]
try:
intdigits = int(digits)
if intdigits < 0 or intdigits > 99:
raise ValueError
except ValueError:
# invalid tag value asked for
continue
if tag.startswith("001") or tag.startswith("00%"):
if out:
out += "\n"
out += "%09d %s %d" % (recID, "001__", recID)
bx = "bib%sx" % digits
bibx = "bibrec_bib%sx" % digits
query = "SELECT b.tag,b.value,bb.field_number FROM %s AS b, %s AS bb "\
"WHERE bb.id_bibrec=%%s AND b.id=bb.id_bibxxx AND b.tag LIKE %%s"\
"ORDER BY bb.field_number, b.tag ASC" % (bx, bibx)
res = run_sql(query, (recID, str(tag)+'%'))
# go through fields:
field_number_old = -999
field_old = ""
for row in res:
field, value, field_number = row[0], row[1], row[2]
ind1, ind2 = field[3], field[4]
if ind1 == "_":
ind1 = ""
if ind2 == "_":
ind2 = ""
# print field tag
if field_number != field_number_old or field[:-1] != field_old[:-1]:
if out:
out += "\n"
out += "%09d %s " % (recID, field[:5])
field_number_old = field_number
field_old = field
# print subfield value
if field[0:2] == "00" and field[-1:] == "_":
out += value
else:
out += "$$%s%s" % (field[-1:], value)
return out
def record_exists(recID):
"""Return 1 if record RECID exists.
Return 0 if it doesn't exist.
Return -1 if it exists but is marked as deleted."""
out = 0
res = run_sql("SELECT id FROM bibrec WHERE id=%s", (recID,), 1)
if res:
recID = int(recID)
# record exists; now check whether it isn't marked as deleted:
dbcollids = get_fieldvalues(recID, "980__%")
if ("DELETED" in dbcollids) or (CFG_CERN_SITE and "DUMMY" in dbcollids):
out = -1 # exists, but marked as deleted
else:
out = 1 # exists fine
return out
def record_public_p(recID):
"""Return 1 if the record is public, i.e. if it can be found in the Home collection.
Return 0 otherwise.
"""
return recID in get_collection_reclist(CFG_SITE_NAME)
def get_creation_date(recID, fmt="%Y-%m-%d"):
"Returns the creation date of the record 'recID'."
out = ""
res = run_sql("SELECT DATE_FORMAT(creation_date,%s) FROM bibrec WHERE id=%s", (fmt, recID), 1)
if res:
out = res[0][0]
return out
def get_modification_date(recID, fmt="%Y-%m-%d"):
"Returns the date of last modification for the record 'recID'."
out = ""
res = run_sql("SELECT DATE_FORMAT(modification_date,%s) FROM bibrec WHERE id=%s", (fmt, recID), 1)
if res:
out = res[0][0]
return out
def print_warning(req, msg, type='', prologue=' ', epilogue=' '):
"Prints warning message and flushes output."
if req and msg:
req.write(websearch_templates.tmpl_print_warning(
msg = msg,
type = type,
prologue = prologue,
epilogue = epilogue,
))
return
def print_search_info(p, f, sf, so, sp, rm, of, ot, collection=CFG_SITE_NAME, nb_found=-1, jrec=1, rg=10,
aas=0, ln=CFG_SITE_LANG, p1="", p2="", p3="", f1="", f2="", f3="", m1="", m2="", m3="", op1="", op2="",
sc=1, pl_in_url="",
d1y=0, d1m=0, d1d=0, d2y=0, d2m=0, d2d=0, dt="",
cpu_time=-1, middle_only=0):
"""Prints stripe with the information on 'collection' and 'nb_found' results and CPU time.
Also, prints navigation links (beg/next/prev/end) inside the results set.
If middle_only is set to 1, it will only print the middle box information (beg/netx/prev/end/etc) links.
This is suitable for displaying navigation links at the bottom of the search results page."""
out = ""
# sanity check:
if jrec < 1:
jrec = 1
if jrec > nb_found:
jrec = max(nb_found-rg+1, 1)
return websearch_templates.tmpl_print_search_info(
ln = ln,
collection = collection,
aas = aas,
collection_name = get_coll_i18nname(collection, ln, False),
collection_id = get_colID(collection),
middle_only = middle_only,
rg = rg,
nb_found = nb_found,
sf = sf,
so = so,
rm = rm,
of = of,
ot = ot,
p = p,
f = f,
p1 = p1,
p2 = p2,
p3 = p3,
f1 = f1,
f2 = f2,
f3 = f3,
m1 = m1,
m2 = m2,
m3 = m3,
op1 = op1,
op2 = op2,
pl_in_url = pl_in_url,
d1y = d1y,
d1m = d1m,
d1d = d1d,
d2y = d2y,
d2m = d2m,
d2d = d2d,
dt = dt,
jrec = jrec,
sc = sc,
sp = sp,
all_fieldcodes = get_fieldcodes(),
cpu_time = cpu_time,
)
def print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, collection=CFG_SITE_NAME, nb_found=-1, jrec=1, rg=10,
aas=0, ln=CFG_SITE_LANG, p1="", p2="", p3="", f1="", f2="", f3="", m1="", m2="", m3="", op1="", op2="",
sc=1, pl_in_url="",
d1y=0, d1m=0, d1d=0, d2y=0, d2m=0, d2d=0, dt="",
cpu_time=-1, middle_only=0):
"""Prints stripe with the information on 'collection' and 'nb_found' results and CPU time.
Also, prints navigation links (beg/next/prev/end) inside the results set.
If middle_only is set to 1, it will only print the middle box information (beg/netx/prev/end/etc) links.
This is suitable for displaying navigation links at the bottom of the search results page."""
out = ""
# sanity check:
if jrec < 1:
jrec = 1
if jrec > nb_found:
jrec = max(nb_found-rg+1, 1)
return websearch_templates.tmpl_print_hosted_search_info(
ln = ln,
collection = collection,
aas = aas,
collection_name = get_coll_i18nname(collection, ln, False),
collection_id = get_colID(collection),
middle_only = middle_only,
rg = rg,
nb_found = nb_found,
sf = sf,
so = so,
rm = rm,
of = of,
ot = ot,
p = p,
f = f,
p1 = p1,
p2 = p2,
p3 = p3,
f1 = f1,
f2 = f2,
f3 = f3,
m1 = m1,
m2 = m2,
m3 = m3,
op1 = op1,
op2 = op2,
pl_in_url = pl_in_url,
d1y = d1y,
d1m = d1m,
d1d = d1d,
d2y = d2y,
d2m = d2m,
d2d = d2d,
dt = dt,
jrec = jrec,
sc = sc,
sp = sp,
all_fieldcodes = get_fieldcodes(),
cpu_time = cpu_time,
)
def print_results_overview(req, colls, results_final_nb_total, results_final_nb, cpu_time, ln=CFG_SITE_LANG, ec=[], hosted_colls_potential_results_p=False):
"""Prints results overview box with links to particular collections below."""
out = ""
new_colls = []
for coll in colls:
new_colls.append({
'id': get_colID(coll),
'code': coll,
'name': get_coll_i18nname(coll, ln, False),
})
return websearch_templates.tmpl_print_results_overview(
ln = ln,
results_final_nb_total = results_final_nb_total,
results_final_nb = results_final_nb,
cpu_time = cpu_time,
colls = new_colls,
ec = ec,
hosted_colls_potential_results_p = hosted_colls_potential_results_p,
)
-def print_hosted_results(infos, of=None, ln=CFG_SITE_LANG, req=None, no_records_found=False, timeout=False):
+def print_hosted_results(url_and_engine, ln=CFG_SITE_LANG, of=None, req=None, no_records_found=False, search_timed_out=False, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Prints the full results of a hosted collection"""
if of.startswith("h"):
if no_records_found:
- return " No results found.."
- if timeout:
- return " The search engine did not respond in time"
+ return " No results found."
+ if search_timed_out:
+ return " The search engine did not respond in time."
return websearch_templates.tmpl_print_hosted_results(
+ url_and_engine=url_and_engine,
ln=ln,
- infos=infos,
of=of,
- req=req
+ req=req,
+ limit=limit
)
def sort_records(req, recIDs, sort_field='', sort_order='d', sort_pattern='', verbose=0, of='hb', ln=CFG_SITE_LANG):
"""Sort records in 'recIDs' list according sort field 'sort_field' in order 'sort_order'.
If more than one instance of 'sort_field' is found for a given record, try to choose that that is given by
'sort pattern', for example "sort by report number that starts by CERN-PS".
Note that 'sort_field' can be field code like 'author' or MARC tag like '100__a' directly."""
_ = gettext_set_language(ln)
## check arguments:
if not sort_field:
return recIDs
if len(recIDs) > CFG_WEBSEARCH_NB_RECORDS_TO_SORT:
if of.startswith('h'):
print_warning(req, _("Sorry, sorting is allowed on sets of up to %d records only. Using default sort order.") % CFG_WEBSEARCH_NB_RECORDS_TO_SORT, "Warning")
return recIDs
sort_fields = string.split(sort_field, ",")
recIDs_dict = {}
recIDs_out = []
## first deduce sorting MARC tag out of the 'sort_field' argument:
tags = []
for sort_field in sort_fields:
if sort_field and str(sort_field[0:2]).isdigit():
# sort_field starts by two digits, so this is probably a MARC tag already
tags.append(sort_field)
else:
# let us check the 'field' table
query = """SELECT DISTINCT(t.value) FROM tag AS t, field_tag AS ft, field AS f
WHERE f.code=%s AND ft.id_field=f.id AND t.id=ft.id_tag
ORDER BY ft.score DESC"""
res = run_sql(query, (sort_field, ))
if res:
for row in res:
tags.append(row[0])
else:
if of.startswith('h'):
print_warning(req, _("Sorry, %s does not seem to be a valid sort option. Choosing title sort instead.") % cgi.escape(sort_field), "Error")
tags.append("245__a")
if verbose >= 3:
print_warning(req, "Sorting by tags %s." % cgi.escape(repr(tags)))
if sort_pattern:
print_warning(req, "Sorting preferentially by %s." % cgi.escape(sort_pattern))
## check if we have sorting tag defined:
if tags:
# fetch the necessary field values:
for recID in recIDs:
val = "" # will hold value for recID according to which sort
vals = [] # will hold all values found in sorting tag for recID
for tag in tags:
vals.extend(get_fieldvalues(recID, tag))
if sort_pattern:
# try to pick that tag value that corresponds to sort pattern
bingo = 0
for v in vals:
if v.lower().startswith(sort_pattern.lower()): # bingo!
bingo = 1
val = v
break
if not bingo: # sort_pattern not present, so add other vals after spaces
val = sort_pattern + " " + string.join(vals)
else:
# no sort pattern defined, so join them all together
val = string.join(vals)
val = strip_accents(val.lower()) # sort values regardless of accents and case
if recIDs_dict.has_key(val):
recIDs_dict[val].append(recID)
else:
recIDs_dict[val] = [recID]
# sort them:
recIDs_dict_keys = recIDs_dict.keys()
recIDs_dict_keys.sort()
# now that keys are sorted, create output array:
for k in recIDs_dict_keys:
for s in recIDs_dict[k]:
recIDs_out.append(s)
# ascending or descending?
if sort_order == 'a':
recIDs_out.reverse()
# okay, we are done
return recIDs_out
else:
# good, no sort needed
return recIDs
def print_records(req, recIDs, jrec=1, rg=10, format='hb', ot='', ln=CFG_SITE_LANG, relevances=[], relevances_prologue="(", relevances_epilogue="%%)", decompress=zlib.decompress, search_pattern='', print_records_prologue_p=True, print_records_epilogue_p=True, verbose=0, tab=''):
"""
Prints list of records 'recIDs' formatted according to 'format' in
groups of 'rg' starting from 'jrec'.
Assumes that the input list 'recIDs' is sorted in reverse order,
so it counts records from tail to head.
A value of 'rg=-9999' means to print all records: to be used with care.
Print also list of RELEVANCES for each record (if defined), in
between RELEVANCE_PROLOGUE and RELEVANCE_EPILOGUE.
Print prologue and/or epilogue specific to 'format' if
'print_records_prologue_p' and/or print_records_epilogue_p' are
True.
"""
# load the right message language
_ = gettext_set_language(ln)
# sanity checking:
if req is None:
return
# get user_info (for formatting based on user)
if isinstance(req, cStringIO.OutputType):
user_info = {}
else:
user_info = collect_user_info(req)
if len(recIDs):
nb_found = len(recIDs)
if rg == -9999: # print all records
rg = nb_found
else:
rg = abs(rg)
if jrec < 1: # sanity checks
jrec = 1
if jrec > nb_found:
jrec = max(nb_found-rg+1, 1)
# will print records from irec_max to irec_min excluded:
irec_max = nb_found - jrec
irec_min = nb_found - jrec - rg
if irec_min < 0:
irec_min = -1
if irec_max >= nb_found:
irec_max = nb_found - 1
#req.write("%s:%d-%d" % (recIDs, irec_min, irec_max))
if format.startswith('x'):
# print header if needed
if print_records_prologue_p:
print_records_prologue(req, format)
# print records
recIDs_to_print = [recIDs[x] for x in range(irec_max, irec_min, -1)]
format_records(recIDs_to_print,
format,
ln=ln,
search_pattern=search_pattern,
record_separator="\n",
user_info=user_info,
req=req)
# print footer if needed
if print_records_epilogue_p:
print_records_epilogue(req, format)
elif format.startswith('t') or str(format[0:3]).isdigit():
# we are doing plain text output:
for irec in range(irec_max, irec_min, -1):
x = print_record(recIDs[irec], format, ot, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
req.write(x)
if x:
req.write('\n')
elif format == 'excel':
recIDs_to_print = [recIDs[x] for x in range(irec_max, irec_min, -1)]
create_excel(recIDs=recIDs_to_print, req=req, ln=ln, ot=ot)
else:
# we are doing HTML output:
if format == 'hp' or format.startswith("hb_") or format.startswith("hd_"):
# portfolio and on-the-fly formats:
for irec in range(irec_max, irec_min, -1):
req.write(print_record(recIDs[irec], format, ot, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose))
elif format.startswith("hb"):
# HTML brief format:
req.write(websearch_templates.tmpl_record_format_htmlbrief_header(
ln = ln))
for irec in range(irec_max, irec_min, -1):
row_number = jrec+irec_max-irec
recid = recIDs[irec]
if relevances and relevances[irec]:
relevance = relevances[irec]
else:
relevance = ''
record = print_record(recIDs[irec], format, ot, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
req.write(websearch_templates.tmpl_record_format_htmlbrief_body(
ln = ln,
recid = recid,
row_number = row_number,
relevance = relevance,
record = record,
relevances_prologue = relevances_prologue,
relevances_epilogue = relevances_epilogue,
))
req.write(websearch_templates.tmpl_record_format_htmlbrief_footer(
ln = ln))
elif format.startswith("hd"):
# HTML detailed format:
for irec in range(irec_max, irec_min, -1):
unordered_tabs = get_detailed_page_tabs(get_colID(guess_primary_collection_of_a_record(recIDs[irec])),
recIDs[irec], ln=ln)
ordered_tabs_id = [(tab_id, values['order']) for (tab_id, values) in unordered_tabs.iteritems()]
ordered_tabs_id.sort(lambda x,y: cmp(x[1],y[1]))
link_ln = ''
if ln != CFG_SITE_LANG:
link_ln = '?ln=%s' % ln
if CFG_WEBSEARCH_USE_ALEPH_SYSNOS:
recid_to_display = get_fieldvalues(recIDs[irec], CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG)[0]
else:
recid_to_display = recIDs[irec]
tabs = [(unordered_tabs[tab_id]['label'], \
'%s/record/%s/%s%s' % (CFG_SITE_URL, recid_to_display, tab_id, link_ln), \
tab_id == tab,
unordered_tabs[tab_id]['enabled']) \
for (tab_id, order) in ordered_tabs_id
if unordered_tabs[tab_id]['visible'] == True]
content = ''
# load content
if tab == 'usage':
req.write(webstyle_templates.detailed_record_container_top(recIDs[irec],
tabs,
ln))
r = calculate_reading_similarity_list(recIDs[irec], "downloads")
downloadsimilarity = None
downloadhistory = None
#if r:
# downloadsimilarity = r
if CFG_BIBRANK_SHOW_DOWNLOAD_GRAPHS:
downloadhistory = create_download_history_graph_and_box(recIDs[irec], ln)
r = calculate_reading_similarity_list(recIDs[irec], "pageviews")
viewsimilarity = None
if r: viewsimilarity = r
content = websearch_templates.tmpl_detailed_record_statistics(recIDs[irec],
ln,
downloadsimilarity=downloadsimilarity,
downloadhistory=downloadhistory,
viewsimilarity=viewsimilarity)
req.write(content)
req.write(webstyle_templates.detailed_record_container_bottom(recIDs[irec],
tabs,
ln))
elif tab == 'citations':
recid = recIDs[irec]
req.write(webstyle_templates.detailed_record_container_top(recid,
tabs,
ln))
req.write(websearch_templates.tmpl_detailed_record_citations_prologue(recid, ln))
# Citing
citinglist = []
r = calculate_cited_by_list(recid)
if r:
citinglist = r
req.write(websearch_templates.tmpl_detailed_record_citations_citing_list(recid,
ln,
citinglist=citinglist))
# Self-cited
selfcited = get_self_cited_by(recid)
req.write(websearch_templates.tmpl_detailed_record_citations_self_cited(recid,
ln, selfcited=selfcited, citinglist=citinglist))
# Co-cited
s = calculate_co_cited_with_list(recid)
cociting = None
if s:
cociting = s
req.write(websearch_templates.tmpl_detailed_record_citations_co_citing(recid,
ln,
cociting=cociting))
# Citation history
citationhistory = None
if r:
citationhistory = create_citation_history_graph_and_box(recid, ln)
#debug
if verbose > 3:
print_warning(req, "Citation graph debug: "+str(len(citationhistory)))
req.write(websearch_templates.tmpl_detailed_record_citations_citation_history(recid, ln, citationhistory))
req.write(websearch_templates.tmpl_detailed_record_citations_epilogue(recid, ln))
req.write(webstyle_templates.detailed_record_container_bottom(recid,
tabs,
ln))
elif tab == 'references':
req.write(webstyle_templates.detailed_record_container_top(recIDs[irec],
tabs,
ln))
req.write(format_record(recIDs[irec], 'HDREF', ln=ln, user_info=user_info, verbose=verbose))
req.write(webstyle_templates.detailed_record_container_bottom(recIDs[irec],
tabs,
ln))
elif tab == 'keywords':
from bibclassify_webinterface import \
record_get_keywords, get_sorting_options, \
generate_keywords, get_keywords_body
from invenio.webinterface_handler import wash_urlargd
form = req.form
argd = wash_urlargd(form, {
'generate': (str, 'no'),
'sort': (str, 'occurrences'),
'type': (str, 'tagcloud'),
'numbering': (str, 'off'),
})
recid = recIDs[irec]
req.write(webstyle_templates.detailed_record_container_top(recid,
tabs, ln))
if argd['generate'] == 'yes':
# The user asked to generate the keywords.
keywords = generate_keywords(req, recid)
else:
# Get the keywords contained in the MARC.
keywords = record_get_keywords(recid, argd)
if keywords:
req.write(get_sorting_options(argd, keywords))
elif argd['sort'] == 'related' and not keywords:
req.write('You may want to run BibIndex.')
# Output the keywords or the generate button.
get_keywords_body(keywords, req, recid, argd)
req.write(webstyle_templates.detailed_record_container_bottom(recid,
tabs, ln))
else:
# Metadata tab
req.write(webstyle_templates.detailed_record_container_top(recIDs[irec],
tabs,
ln,
show_short_rec_p=False))
creationdate = None
modificationdate = None
if record_exists(recIDs[irec]) == 1:
creationdate = get_creation_date(recIDs[irec])
modificationdate = get_modification_date(recIDs[irec])
content = print_record(recIDs[irec], format, ot, ln,
search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
content = websearch_templates.tmpl_detailed_record_metadata(
recID = recIDs[irec],
ln = ln,
format = format,
creationdate = creationdate,
modificationdate = modificationdate,
content = content)
req.write(content)
req.write(webstyle_templates.detailed_record_container_bottom(recIDs[irec],
tabs,
ln,
creationdate=creationdate,
modificationdate=modificationdate,
show_short_rec_p=False))
if len(tabs) > 0:
# Add the mini box at bottom of the page
if CFG_WEBCOMMENT_ALLOW_REVIEWS:
from invenio.webcomment import get_mini_reviews
reviews = get_mini_reviews(recid = recIDs[irec], ln=ln)
else:
reviews = ''
actions = format_record(recIDs[irec], 'HDACT', ln=ln, user_info=user_info, verbose=verbose)
files = format_record(recIDs[irec], 'HDFILE', ln=ln, user_info=user_info, verbose=verbose)
req.write(webstyle_templates.detailed_record_mini_panel(recIDs[irec],
ln,
format,
files=files,
reviews=reviews,
actions=actions))
else:
# Other formats
for irec in range(irec_max, irec_min, -1):
req.write(print_record(recIDs[irec], format, ot, ln,
search_pattern=search_pattern,
user_info=user_info, verbose=verbose))
else:
print_warning(req, _("Use different search terms."))
def print_records_prologue(req, format):
"""
Print the appropriate prologue for list of records in the given
format.
"""
prologue = "" # no prologue needed for HTML or Text formats
if format.startswith('xm'):
prologue = websearch_templates.tmpl_xml_marc_prologue()
elif format.startswith('xn'):
prologue = websearch_templates.tmpl_xml_nlm_prologue()
elif format.startswith('xw'):
prologue = websearch_templates.tmpl_xml_refworks_prologue()
elif format.startswith('xr'):
prologue = websearch_templates.tmpl_xml_rss_prologue()
elif format.startswith('xe'):
prologue = websearch_templates.tmpl_xml_endnote_prologue()
elif format.startswith('xo'):
prologue = websearch_templates.tmpl_xml_mods_prologue()
elif format.startswith('x'):
prologue = websearch_templates.tmpl_xml_default_prologue()
req.write(prologue)
def print_records_epilogue(req, format):
"""
Print the appropriate epilogue for list of records in the given
format.
"""
epilogue = "" # no epilogue needed for HTML or Text formats
if format.startswith('xm'):
epilogue = websearch_templates.tmpl_xml_marc_epilogue()
elif format.startswith('xn'):
epilogue = websearch_templates.tmpl_xml_nlm_epilogue()
elif format.startswith('xw'):
epilogue = websearch_templates.tmpl_xml_refworks_epilogue()
elif format.startswith('xr'):
epilogue = websearch_templates.tmpl_xml_rss_epilogue()
elif format.startswith('xe'):
epilogue = websearch_templates.tmpl_xml_endnote_epilogue()
elif format.startswith('xo'):
epilogue = websearch_templates.tmpl_xml_mods_epilogue()
elif format.startswith('x'):
epilogue = websearch_templates.tmpl_xml_default_epilogue()
req.write(epilogue)
def get_record(recid):
"""Directly the record object corresponding to the recid."""
from marshal import loads, dumps
from zlib import compress, decompress
if CFG_BIBUPLOAD_SERIALIZE_RECORD_STRUCTURE:
value = run_sql('SELECT value FROM bibfmt WHERE id_bibrec=%s AND FORMAT=\'recstruct\'', (recid, ))
if value:
try:
return loads(decompress(value[0][0]))
except:
### In case of corruption, let's rebuild it!
pass
return create_record(print_record(recid, 'xm'))[0]
def print_record(recID, format='hb', ot='', ln=CFG_SITE_LANG, decompress=zlib.decompress,
search_pattern=None, user_info=None, verbose=0):
"""Prints record 'recID' formatted accoding to 'format'."""
if format == 'recstruct':
return get_record(recID)
_ = gettext_set_language(ln)
out = ""
# sanity check:
record_exist_p = record_exists(recID)
if record_exist_p == 0: # doesn't exist
return out
# New Python BibFormat procedure for formatting
# Old procedure follows further below
# We must still check some special formats, but these
# should disappear when BibFormat improves.
if not (CFG_BIBFORMAT_USE_OLD_BIBFORMAT \
or format.lower().startswith('t') \
or format.lower().startswith('hm') \
or str(format[0:3]).isdigit() \
or ot):
# Unspecified format is hd
if format == '':
format = 'hd'
if record_exist_p == -1 and get_output_format_content_type(format) == 'text/html':
# HTML output displays a default value for deleted records.
# Other format have to deal with it.
out += _("The record has been deleted.")
else:
out += call_bibformat(recID, format, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
# at the end of HTML brief mode, print the "Detailed record" functionality:
if format.lower().startswith('hb') and \
format.lower() != 'hb_p':
out += websearch_templates.tmpl_print_record_brief_links(
ln = ln,
recID = recID,
)
return out
# Old PHP BibFormat procedure for formatting
# print record opening tags, if needed:
if format == "marcxml" or format == "oai_dc":
out += " \n"
out += " \n"
for oai_id in get_fieldvalues(recID, CFG_OAI_ID_FIELD):
out += " %s\n" % oai_id
out += " %s\n" % get_modification_date(recID)
out += " \n"
out += " \n"
if format.startswith("xm") or format == "marcxml":
# look for detailed format existence:
query = "SELECT value FROM bibfmt WHERE id_bibrec=%s AND format=%s"
res = run_sql(query, (recID, format), 1)
if res and record_exist_p == 1:
# record 'recID' is formatted in 'format', so print it
out += "%s" % decompress(res[0][0])
else:
# record 'recID' is not formatted in 'format' -- they are not in "bibfmt" table; so fetch all the data from "bibXXx" tables:
if format == "marcxml":
out += """ \n"""
out += " %d\n" % int(recID)
elif format.startswith("xm"):
out += """ \n"""
out += " %d\n" % int(recID)
if record_exist_p == -1:
# deleted record, so display only OAI ID and 980:
oai_ids = get_fieldvalues(recID, CFG_OAI_ID_FIELD)
if oai_ids:
out += "%s\n" % \
(CFG_OAI_ID_FIELD[0:3], CFG_OAI_ID_FIELD[3:4], CFG_OAI_ID_FIELD[4:5], CFG_OAI_ID_FIELD[5:6], oai_ids[0])
out += "DELETED\n"
else:
# controlfields
query = "SELECT b.tag,b.value,bb.field_number FROM bib00x AS b, bibrec_bib00x AS bb "\
"WHERE bb.id_bibrec=%s AND b.id=bb.id_bibxxx AND b.tag LIKE '00%%' "\
"ORDER BY bb.field_number, b.tag ASC"
res = run_sql(query, (recID, ))
for row in res:
field, value = row[0], row[1]
value = encode_for_xml(value)
out += """ %s\n""" % \
(encode_for_xml(field[0:3]), value)
# datafields
i = 1 # Do not process bib00x and bibrec_bib00x, as
# they are controlfields. So start at bib01x and
# bibrec_bib00x (and set i = 0 at the end of
# first loop)
for digit1 in range(0, 10):
for digit2 in range(i, 10):
bx = "bib%d%dx" % (digit1, digit2)
bibx = "bibrec_bib%d%dx" % (digit1, digit2)
query = "SELECT b.tag,b.value,bb.field_number FROM %s AS b, %s AS bb "\
"WHERE bb.id_bibrec=%%s AND b.id=bb.id_bibxxx AND b.tag LIKE %%s"\
"ORDER BY bb.field_number, b.tag ASC" % (bx, bibx)
res = run_sql(query, (recID, str(digit1)+str(digit2)+'%'))
field_number_old = -999
field_old = ""
for row in res:
field, value, field_number = row[0], row[1], row[2]
ind1, ind2 = field[3], field[4]
if ind1 == "_" or ind1 == "":
ind1 = " "
if ind2 == "_" or ind2 == "":
ind2 = " "
# print field tag
if field_number != field_number_old or field[:-1] != field_old[:-1]:
if field_number_old != -999:
out += """ \n"""
out += """ \n""" % \
(encode_for_xml(field[0:3]), encode_for_xml(ind1), encode_for_xml(ind2))
field_number_old = field_number
field_old = field
# print subfield value
value = encode_for_xml(value)
out += """ %s\n""" % \
(encode_for_xml(field[-1:]), value)
# all fields/subfields printed in this run, so close the tag:
if field_number_old != -999:
out += """ \n"""
i = 0 # Next loop should start looking at bib%0 and bibrec_bib00x
# we are at the end of printing the record:
out += " \n"
elif format == "xd" or format == "oai_dc":
# XML Dublin Core format, possibly OAI -- select only some bibXXx fields:
out += """ \n"""
if record_exist_p == -1:
out += ""
else:
for f in get_fieldvalues(recID, "041__a"):
out += " %s\n" % f
for f in get_fieldvalues(recID, "100__a"):
out += " %s\n" % encode_for_xml(f)
for f in get_fieldvalues(recID, "700__a"):
out += " %s\n" % encode_for_xml(f)
for f in get_fieldvalues(recID, "245__a"):
out += " %s\n" % encode_for_xml(f)
for f in get_fieldvalues(recID, "65017a"):
out += " %s\n" % encode_for_xml(f)
for f in get_fieldvalues(recID, "8564_u"):
out += " %s\n" % encode_for_xml(f)
for f in get_fieldvalues(recID, "520__a"):
out += " %s\n" % encode_for_xml(f)
out += " %s\n" % get_creation_date(recID)
out += " \n"
elif len(format) == 6 and str(format[0:3]).isdigit():
# user has asked to print some fields only
if format == "001":
out += "%s\n" % (format, recID, format)
else:
vals = get_fieldvalues(recID, format)
for val in vals:
out += "%s\n" % (format, val, format)
elif format.startswith('t'):
## user directly asked for some tags to be displayed only
if record_exist_p == -1:
out += get_fieldvalues_alephseq_like(recID, ["001", CFG_OAI_ID_FIELD, "980"])
else:
out += get_fieldvalues_alephseq_like(recID, ot)
elif format == "hm":
if record_exist_p == -1:
out += "\n
"
elif format == "hd":
# HTML detailed format
if record_exist_p == -1:
out += _("The record has been deleted.")
else:
# look for detailed format existence:
query = "SELECT value FROM bibfmt WHERE id_bibrec=%s AND format=%s"
res = run_sql(query, (recID, format), 1)
if res:
# record 'recID' is formatted in 'format', so print it
out += "%s" % decompress(res[0][0])
else:
# record 'recID' is not formatted in 'format', so try to call BibFormat on the fly or use default format:
out_record_in_format = call_bibformat(recID, format, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
if out_record_in_format:
out += out_record_in_format
else:
out += websearch_templates.tmpl_print_record_detailed(
ln = ln,
recID = recID,
)
elif format.startswith("hb_") or format.startswith("hd_"):
# underscore means that HTML brief/detailed formats should be called on-the-fly; suitable for testing formats
if record_exist_p == -1:
out += _("The record has been deleted.")
else:
out += call_bibformat(recID, format, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
elif format.startswith("hx"):
# BibTeX format, called on the fly:
if record_exist_p == -1:
out += _("The record has been deleted.")
else:
out += call_bibformat(recID, format, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
elif format.startswith("hs"):
# for citation/download similarity navigation links:
if record_exist_p == -1:
out += _("The record has been deleted.")
else:
out += '' % websearch_templates.build_search_url(recid=recID, ln=ln)
# firstly, title:
titles = get_fieldvalues(recID, "245__a")
if titles:
for title in titles:
out += "%s" % title
else:
# usual title not found, try conference title:
titles = get_fieldvalues(recID, "111__a")
if titles:
for title in titles:
out += "%s" % title
else:
# just print record ID:
out += "%s %d" % (get_field_i18nname("record ID", ln, False), recID)
out += ""
# secondly, authors:
authors = get_fieldvalues(recID, "100__a") + get_fieldvalues(recID, "700__a")
if authors:
out += " - %s" % authors[0]
if len(authors) > 1:
out += " et al"
# thirdly publication info:
publinfos = get_fieldvalues(recID, "773__s")
if not publinfos:
publinfos = get_fieldvalues(recID, "909C4s")
if not publinfos:
publinfos = get_fieldvalues(recID, "037__a")
if not publinfos:
publinfos = get_fieldvalues(recID, "088__a")
if publinfos:
out += " - %s" % publinfos[0]
else:
# fourthly publication year (if not publication info):
years = get_fieldvalues(recID, "773__y")
if not years:
years = get_fieldvalues(recID, "909C4y")
if not years:
years = get_fieldvalues(recID, "260__c")
if years:
out += " (%s)" % years[0]
else:
# HTML brief format by default
if record_exist_p == -1:
out += _("The record has been deleted.")
else:
query = "SELECT value FROM bibfmt WHERE id_bibrec=%s AND format=%s"
res = run_sql(query, (recID, format))
if res:
# record 'recID' is formatted in 'format', so print it
out += "%s" % decompress(res[0][0])
else:
# record 'recID' is not formatted in 'format', so try to call BibFormat on the fly: or use default format:
if CFG_WEBSEARCH_CALL_BIBFORMAT:
out_record_in_format = call_bibformat(recID, format, ln, search_pattern=search_pattern,
user_info=user_info, verbose=verbose)
if out_record_in_format:
out += out_record_in_format
else:
out += websearch_templates.tmpl_print_record_brief(
ln = ln,
recID = recID,
)
else:
out += websearch_templates.tmpl_print_record_brief(
ln = ln,
recID = recID,
)
# at the end of HTML brief mode, print the "Detailed record" functionality:
if format == 'hp' or format.startswith("hb_") or format.startswith("hd_"):
pass # do nothing for portfolio and on-the-fly formats
else:
out += websearch_templates.tmpl_print_record_brief_links(
ln = ln,
recID = recID,
)
# print record closing tags, if needed:
if format == "marcxml" or format == "oai_dc":
out += " \n"
out += " \n"
return out
def call_bibformat(recID, format="HD", ln=CFG_SITE_LANG, search_pattern=None, user_info=None, verbose=0):
"""
Calls BibFormat and returns formatted record.
BibFormat will decide by itself if old or new BibFormat must be used.
"""
keywords = []
if search_pattern is not None:
units = create_basic_search_units(None, str(search_pattern), None)
keywords = [unit[1] for unit in units if unit[0] != '-']
return format_record(recID,
of=format,
ln=ln,
search_pattern=keywords,
user_info=user_info,
verbose=verbose)
def log_query(hostname, query_args, uid=-1):
"""
Log query into the query and user_query tables.
Return id_query or None in case of problems.
"""
id_query = None
if uid >= 0:
# log the query only if uid is reasonable
res = run_sql("SELECT id FROM query WHERE urlargs=%s", (query_args,), 1)
try:
id_query = res[0][0]
except:
id_query = run_sql("INSERT INTO query (type, urlargs) VALUES ('r', %s)", (query_args,))
if id_query:
run_sql("INSERT INTO user_query (id_user, id_query, hostname, date) VALUES (%s, %s, %s, %s)",
(uid, id_query, hostname,
time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())))
return id_query
def log_query_info(action, p, f, colls, nb_records_found_total=-1):
"""Write some info to the log file for later analysis."""
try:
log = open(CFG_LOGDIR + "/search.log", "a")
log.write(time.strftime("%Y%m%d%H%M%S#", time.localtime()))
log.write(action+"#")
log.write(p+"#")
log.write(f+"#")
for coll in colls[:-1]:
log.write("%s," % coll)
log.write("%s#" % colls[-1])
log.write("%d" % nb_records_found_total)
log.write("\n")
log.close()
except:
pass
return
### CALLABLES
def perform_request_search(req=None, cc=CFG_SITE_NAME, c=None, p="", f="", rg=10, sf="", so="d", sp="", rm="", of="id", ot="", aas=0,
p1="", f1="", m1="", op1="", p2="", f2="", m2="", op2="", p3="", f3="", m3="", sc=0, jrec=0,
recid=-1, recidb=-1, sysno="", id=-1, idb=-1, sysnb="", action="", d1="",
d1y=0, d1m=0, d1d=0, d2="", d2y=0, d2m=0, d2d=0, dt="", verbose=0, ap=0, ln=CFG_SITE_LANG, ec=None, tab=""):
"""Perform search or browse request, without checking for
authentication. Return list of recIDs found, if of=id.
Otherwise create web page.
The arguments are as follows:
req - mod_python Request class instance.
cc - current collection (e.g. "ATLAS"). The collection the
user started to search/browse from.
c - collection list (e.g. ["Theses", "Books"]). The
collections user may have selected/deselected when
starting to search from 'cc'.
p - pattern to search for (e.g. "ellis and muon or kaon").
f - field to search within (e.g. "author").
rg - records in groups of (e.g. "10"). Defines how many hits
per collection in the search results page are
displayed.
sf - sort field (e.g. "title").
so - sort order ("a"=ascending, "d"=descending).
sp - sort pattern (e.g. "CERN-") -- in case there are more
values in a sort field, this argument tells which one
to prefer
rm - ranking method (e.g. "jif"). Defines whether results
should be ranked by some known ranking method.
of - output format (e.g. "hb"). Usually starting "h" means
HTML output (and "hb" for HTML brief, "hd" for HTML
detailed), "x" means XML output, "t" means plain text
output, "id" means no output at all but to return list
of recIDs found. (Suitable for high-level API.)
ot - output only these MARC tags (e.g. "100,700,909C0b").
Useful if only some fields are to be shown in the
output, e.g. for library to control some fields.
aas - advanced search ("0" means no, "1" means yes). Whether
search was called from within the advanced search
interface.
p1 - first pattern to search for in the advanced search
interface. Much like 'p'.
f1 - first field to search within in the advanced search
interface. Much like 'f'.
m1 - first matching type in the advanced search interface.
("a" all of the words, "o" any of the words, "e" exact
phrase, "p" partial phrase, "r" regular expression).
op1 - first operator, to join the first and the second unit
in the advanced search interface. ("a" add, "o" or,
"n" not).
p2 - second pattern to search for in the advanced search
interface. Much like 'p'.
f2 - second field to search within in the advanced search
interface. Much like 'f'.
m2 - second matching type in the advanced search interface.
("a" all of the words, "o" any of the words, "e" exact
phrase, "p" partial phrase, "r" regular expression).
op2 - second operator, to join the second and the third unit
in the advanced search interface. ("a" add, "o" or,
"n" not).
p3 - third pattern to search for in the advanced search
interface. Much like 'p'.
f3 - third field to search within in the advanced search
interface. Much like 'f'.
m3 - third matching type in the advanced search interface.
("a" all of the words, "o" any of the words, "e" exact
phrase, "p" partial phrase, "r" regular expression).
sc - split by collection ("0" no, "1" yes). Governs whether
we want to present the results in a single huge list,
or splitted by collection.
jrec - jump to record (e.g. "234"). Used for navigation
inside the search results.
recid - display record ID (e.g. "20000"). Do not
search/browse but go straight away to the Detailed
record page for the given recID.
recidb - display record ID bis (e.g. "20010"). If greater than
'recid', then display records from recid to recidb.
Useful for example for dumping records from the
database for reformatting.
sysno - display old system SYS number (e.g. ""). If you
migrate to CDS Invenio from another system, and store your
old SYS call numbers, you can use them instead of recid
if you wish so.
id - the same as recid, in case recid is not set. For
backwards compatibility.
idb - the same as recid, in case recidb is not set. For
backwards compatibility.
sysnb - the same as sysno, in case sysno is not set. For
backwards compatibility.
action - action to do. "SEARCH" for searching, "Browse" for
browsing. Default is to search.
d1 - first datetime in full YYYY-mm-dd HH:MM:DD format
(e.g. "1998-08-23 12:34:56"). Useful for search limits
on creation/modification date (see 'dt' argument
below). Note that 'd1' takes precedence over d1y, d1m,
d1d if these are defined.
d1y - first date's year (e.g. "1998"). Useful for search
limits on creation/modification date.
d1m - first date's month (e.g. "08"). Useful for search
limits on creation/modification date.
d1d - first date's day (e.g. "23"). Useful for search
limits on creation/modification date.
d2 - second datetime in full YYYY-mm-dd HH:MM:DD format
(e.g. "1998-09-02 12:34:56"). Useful for search limits
on creation/modification date (see 'dt' argument
below). Note that 'd2' takes precedence over d2y, d2m,
d2d if these are defined.
d2y - second date's year (e.g. "1998"). Useful for search
limits on creation/modification date.
d2m - second date's month (e.g. "09"). Useful for search
limits on creation/modification date.
d2d - second date's day (e.g. "02"). Useful for search
limits on creation/modification date.
dt - first and second date's type (e.g. "c"). Specifies
whether to search in creation dates ("c") or in
modification dates ("m"). When dt is not set and d1*
and d2* are set, the default is "c".
verbose - verbose level (0=min, 9=max). Useful to print some
internal information on the searching process in case
something goes wrong.
ap - alternative patterns (0=no, 1=yes). In case no exact
match is found, the search engine can try alternative
patterns e.g. to replace non-alphanumeric characters by
a boolean query. ap defines if this is wanted.
ln - language of the search interface (e.g. "en"). Useful
for internationalization.
ec - list of external search engines to search as well
(e.g. "SPIRES HEP").
"""
selected_external_collections_infos = None
# wash output format:
of = wash_output_format(of)
+ # raise an exception when trying to print out html or xml from the cli
+ if of.startswith("h") or of.startswith("x"):
+ assert req
+
# for every search engine request asking for an HTML output, we
# first regenerate cache of collection and field I18N names if
# needed; so that later we won't bother checking timestamps for
# I18N names at all:
if of.startswith("h"):
collection_i18nname_cache.recreate_cache_if_needed()
field_i18nname_cache.recreate_cache_if_needed()
# wash all arguments requiring special care
try:
(cc, colls_to_display, colls_to_search, hosted_colls, wash_colls_debug) = wash_colls(cc, c, sc, verbose) # which colls to search and to display?
except InvenioWebSearchUnknownCollectionError, exc:
colname = exc.colname
if of.startswith("h"):
page_start(req, of, cc, aas, ln, getUid(req),
websearch_templates.tmpl_collection_not_found_page_title(colname, ln))
req.write(websearch_templates.tmpl_collection_not_found_page_body(colname, ln))
return page_end(req, of, ln)
elif of == "id":
return []
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
else:
return page_end(req, of, ln)
p = wash_pattern(p)
f = wash_field(f)
p1 = wash_pattern(p1)
f1 = wash_field(f1)
p2 = wash_pattern(p2)
f2 = wash_field(f2)
p3 = wash_pattern(p3)
f3 = wash_field(f3)
datetext1, datetext2 = wash_dates(d1, d1y, d1m, d1d, d2, d2y, d2m, d2d)
# wash ranking method:
if not is_method_valid(None, rm):
rm = ""
_ = gettext_set_language(ln)
# backwards compatibility: id, idb, sysnb -> recid, recidb, sysno (if applicable)
if sysnb != "" and sysno == "":
sysno = sysnb
if id > 0 and recid == -1:
recid = id
if idb > 0 and recidb == -1:
recidb = idb
# TODO deduce passed search limiting criterias (if applicable)
pl, pl_in_url = "", "" # no limits by default
if action != "browse" and req and not isinstance(req, cStringIO.OutputType) \
and req.args: # we do not want to add options while browsing or while calling via command-line
fieldargs = cgi.parse_qs(req.args)
for fieldcode in get_fieldcodes():
if fieldargs.has_key(fieldcode):
for val in fieldargs[fieldcode]:
pl += "+%s:\"%s\" " % (fieldcode, val)
pl_in_url += "&%s=%s" % (urllib.quote(fieldcode), urllib.quote(val))
# deduce recid from sysno argument (if applicable):
if sysno: # ALEPH SYS number was passed, so deduce DB recID for the record:
recid = get_mysql_recid_from_aleph_sysno(sysno)
if recid is None:
recid = 0 # use recid 0 to indicate that this sysno does not exist
# deduce collection we are in (if applicable):
if recid > 0:
referer = None
if req:
referer = req.headers_in.get('Referer')
cc = guess_collection_of_a_record(recid, referer)
# deduce user id (if applicable):
try:
uid = getUid(req)
except:
uid = 0
## 0 - start output
if recid >= 0: # recid can be 0 if deduced from sysno and if such sysno does not exist
## 1 - detailed record display
title, description, keywords = \
websearch_templates.tmpl_record_page_header_content(req, recid, ln)
if req is not None and not req.header_only:
page_start(req, of, cc, aas, ln, uid, title, description, keywords, recid, tab)
# Default format is hb but we are in detailed -> change 'of'
if of == "hb":
of = "hd"
if record_exists(recid):
if recidb <= recid: # sanity check
recidb = recid + 1
if of == "id":
return [recidx for recidx in range(recid, recidb) if record_exists(recidx)]
else:
print_records(req, range(recid, recidb), -1, -9999, of, ot, ln, search_pattern=p, verbose=verbose, tab=tab)
if req and of.startswith("h"): # register detailed record page view event
client_ip_address = str(req.remote_ip)
register_page_view_event(recid, uid, client_ip_address)
else: # record does not exist
if of == "id":
return []
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
elif of.startswith("h"):
if req.header_only:
raise apache.SERVER_RETURN, apache.HTTP_NOT_FOUND
else:
print_warning(req, _("Requested record does not seem to exist."))
elif action == "browse":
## 2 - browse needed
of = 'hb'
page_start(req, of, cc, aas, ln, uid, _("Browse"), p=create_page_title_search_pattern_info(p, p1, p2, p3))
req.write(create_search_box(cc, colls_to_display, p, f, rg, sf, so, sp, rm, of, ot, aas, ln, p1, f1, m1, op1,
p2, f2, m2, op2, p3, f3, m3, sc, pl, d1y, d1m, d1d, d2y, d2m, d2d, dt, jrec, ec, action))
try:
if aas == 1 or (p1 or p2 or p3):
browse_pattern(req, colls_to_search, p1, f1, rg, ln)
browse_pattern(req, colls_to_search, p2, f2, rg, ln)
browse_pattern(req, colls_to_search, p3, f3, rg, ln)
else:
browse_pattern(req, colls_to_search, p, f, rg, ln)
except:
register_exception(req=req, alert_admin=True)
if of.startswith("h"):
req.write(create_error_box(req, verbose=verbose, ln=ln))
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
elif rm and p.startswith("recid:"):
## 3-ter - similarity search or citation search needed
if req and not req.header_only:
page_start(req, of, cc, aas, ln, uid, _("Search Results"), p=create_page_title_search_pattern_info(p, p1, p2, p3))
if of.startswith("h"):
req.write(create_search_box(cc, colls_to_display, p, f, rg, sf, so, sp, rm, of, ot, aas, ln, p1, f1, m1, op1,
p2, f2, m2, op2, p3, f3, m3, sc, pl, d1y, d1m, d1d, d2y, d2m, d2d, dt, jrec, ec, action))
if record_exists(p[6:]) != 1:
# record does not exist
if of.startswith("h"):
if req.header_only:
raise apache.SERVER_RETURN, apache.HTTP_NOT_FOUND
else:
print_warning(req, "Requested record does not seem to exist.")
if of == "id":
return []
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
else:
# record well exists, so find similar ones to it
t1 = os.times()[4]
results_similar_recIDs, results_similar_relevances, results_similar_relevances_prologue, results_similar_relevances_epilogue, results_similar_comments = \
rank_records(rm, 0, get_collection_reclist(cc), string.split(p), verbose)
if results_similar_recIDs:
t2 = os.times()[4]
cpu_time = t2 - t1
if of.startswith("h"):
req.write(print_search_info(p, f, sf, so, sp, rm, of, ot, cc, len(results_similar_recIDs),
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time))
print_warning(req, results_similar_comments)
print_records(req, results_similar_recIDs, jrec, rg, of, ot, ln,
results_similar_relevances, results_similar_relevances_prologue, results_similar_relevances_epilogue, search_pattern=p, verbose=verbose)
elif of=="id":
return results_similar_recIDs
elif of.startswith("x"):
print_records(req, results_similar_recIDs, jrec, rg, of, ot, ln,
results_similar_relevances, results_similar_relevances_prologue, results_similar_relevances_epilogue, search_pattern=p, verbose=verbose)
else:
# rank_records failed and returned some error message to display:
if of.startswith("h"):
print_warning(req, results_similar_relevances_prologue)
print_warning(req, results_similar_relevances_epilogue)
print_warning(req, results_similar_comments)
if of == "id":
return []
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
elif p.startswith("cocitedwith:"): #WAS EXPERIMENTAL
## 3-terter - cited by search needed
page_start(req, of, cc, aas, ln, uid, _("Search Results"), p=create_page_title_search_pattern_info(p, p1, p2, p3))
if of.startswith("h"):
req.write(create_search_box(cc, colls_to_display, p, f, rg, sf, so, sp, rm, of, ot, aas, ln, p1, f1, m1, op1,
p2, f2, m2, op2, p3, f3, m3, sc, pl, d1y, d1m, d1d, d2y, d2m, d2d, dt, jrec, ec, action))
recID = p[12:]
if record_exists(recID) != 1:
# record does not exist
if of.startswith("h"):
print_warning(req, "Requested record does not seem to exist.")
if of == "id":
return []
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
else:
# record well exists, so find co-cited ones:
t1 = os.times()[4]
results_cocited_recIDs = map(lambda x: x[0], calculate_co_cited_with_list(int(recID)))
if results_cocited_recIDs:
t2 = os.times()[4]
cpu_time = t2 - t1
if of.startswith("h"):
req.write(print_search_info(p, f, sf, so, sp, rm, of, ot, CFG_SITE_NAME, len(results_cocited_recIDs),
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time))
print_records(req, results_cocited_recIDs, jrec, rg, of, ot, ln, search_pattern=p, verbose=verbose)
elif of=="id":
return results_cocited_recIDs
elif of.startswith("x"):
print_records(req, results_cocited_recIDs, jrec, rg, of, ot, ln, search_pattern=p, verbose=verbose)
else:
# cited rank_records failed and returned some error message to display:
if of.startswith("h"):
print_warning(req, "nothing found")
if of == "id":
return []
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
else:
## 3 - common search needed
query_in_cache = False
query_representation_in_cache = repr((p,f,colls_to_search))
page_start(req, of, cc, aas, ln, uid, p=create_page_title_search_pattern_info(p, p1, p2, p3))
if of.startswith("h") and verbose and wash_colls_debug:
print_warning(req, "wash_colls debugging info : %s" % wash_colls_debug)
# search into the hosted collections only if the output format is html or xml
if hosted_colls and (of.startswith("h") or of.startswith("x")) and not p.startswith("recid:"):
# hosted_colls_results : the hosted collections' searches that did not timeout
# hosted_colls_timeouts : the hosted collections' searches that timed out and will be searched later on again
(hosted_colls_results, hosted_colls_timeouts) = calculate_hosted_collections_results(req, [p, p1, p2, p3], f, hosted_colls, verbose, ln, CFG_HOSTED_COLLECTION_TIMEOUT_ANTE_SEARCH)
# successful searches
if hosted_colls_results:
hosted_colls_true_results = []
for result in hosted_colls_results:
# if the number of results is None or 0 (or False) then just do nothing
if result[1] == None or result[1] == False:
# these are the searches the returned no or zero results
if verbose:
print_warning(req, "Hosted collections (perform_search_request): %s returned no results" % result[0][1].name)
else:
# these are the searches that actually returned results on time
hosted_colls_true_results.append(result)
if verbose:
print_warning(req, "Hosted collections (perform_search_request): %s returned %s results in %s seconds" % (result[0][1].name, result[1], result[2]))
else:
if verbose:
print_warning(req, "Hosted collections (perform_search_request): there were no hosted collections results to be printed at this time")
if hosted_colls_timeouts:
if verbose:
for timeout in hosted_colls_timeouts:
print_warning(req, "Hosted collections (perform_search_request): %s timed out and will be searched again later" % timeout[0][1].name)
# we need to know for later use if there were any hosted collections to be searched even if they weren't in the end
elif hosted_colls and ((not (of.startswith("h") or of.startswith("x"))) or p.startswith("recid:")):
(hosted_colls_results, hosted_colls_timeouts) = (None, None)
else:
if verbose:
print_warning(req, "Hosted collections (perform_search_request): there were no hosted collections to be searched")
## let's define some useful boolean variables:
# True means there are actual or potential hosted collections results to be printed
hosted_colls_actual_or_potential_results_p = not (not hosted_colls or not ((hosted_colls_results and hosted_colls_true_results) or hosted_colls_timeouts))
# True means there are hosted collections timeouts to take care of later
# (useful for more accurate printing of results later)
hosted_colls_potential_results_p = not (not hosted_colls or not hosted_colls_timeouts)
# True means we only have hosted collections to deal with
only_hosted_colls_actual_or_potential_results_p = not colls_to_search and hosted_colls_actual_or_potential_results_p
if of.startswith("h"):
req.write(create_search_box(cc, colls_to_display, p, f, rg, sf, so, sp, rm, of, ot, aas, ln, p1, f1, m1, op1,
p2, f2, m2, op2, p3, f3, m3, sc, pl, d1y, d1m, d1d, d2y, d2m, d2d, dt, jrec, ec, action))
t1 = os.times()[4]
results_in_any_collection = HitSet()
if aas == 1 or (p1 or p2 or p3):
## 3A - advanced search
try:
results_in_any_collection = search_pattern_parenthesised(req, p1, f1, m1, ap=ap, of=of, verbose=verbose, ln=ln)
if len(results_in_any_collection) == 0:
if of.startswith("h"):
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
if p2:
results_tmp = search_pattern_parenthesised(req, p2, f2, m2, ap=ap, of=of, verbose=verbose, ln=ln)
if op1 == "a": # add
results_in_any_collection.intersection_update(results_tmp)
elif op1 == "o": # or
results_in_any_collection.union_update(results_tmp)
elif op1 == "n": # not
results_in_any_collection.difference_update(results_tmp)
else:
if of.startswith("h"):
print_warning(req, "Invalid set operation %s." % cgi.escape(op1), "Error")
if len(results_in_any_collection) == 0:
if of.startswith("h"):
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
if p3:
results_tmp = search_pattern_parenthesised(req, p3, f3, m3, ap=ap, of=of, verbose=verbose, ln=ln)
if op2 == "a": # add
results_in_any_collection.intersection_update(results_tmp)
elif op2 == "o": # or
results_in_any_collection.union_update(results_tmp)
elif op2 == "n": # not
results_in_any_collection.difference_update(results_tmp)
else:
if of.startswith("h"):
print_warning(req, "Invalid set operation %s." % cgi.escape(op2), "Error")
except:
register_exception(req=req, alert_admin=True)
if of.startswith("h"):
req.write(create_error_box(req, verbose=verbose, ln=ln))
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
else:
## 3B - simple search
if search_results_cache.cache.has_key(query_representation_in_cache):
# query is not in the cache already, so reuse it:
query_in_cache = True
results_in_any_collection = search_results_cache.cache[query_representation_in_cache]
if verbose and of.startswith("h"):
print_warning(req, "Search stage 0: query found in cache, reusing cached results.")
else:
try:
# added the display_nearest_terms_box parameter to avoid printing out the "Nearest terms in any collection"
# recommendations when there are results only in the hosted collections. Also added the if clause to avoid
# searching in case we know we only have actual or potential hosted collections results
if not only_hosted_colls_actual_or_potential_results_p:
results_in_any_collection = search_pattern_parenthesised(req, p, f, ap=ap, of=of, verbose=verbose, ln=ln, display_nearest_terms_box=not hosted_colls_actual_or_potential_results_p)
except:
register_exception(req=req, alert_admin=True)
if of.startswith("h"):
req.write(create_error_box(req, verbose=verbose, ln=ln))
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
return page_end(req, of, ln)
if len(results_in_any_collection) == 0 and not hosted_colls_actual_or_potential_results_p:
if of.startswith("h"):
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
# store this search query results into search results cache if needed:
if CFG_WEBSEARCH_SEARCH_CACHE_SIZE and not query_in_cache:
if len(search_results_cache.cache) > CFG_WEBSEARCH_SEARCH_CACHE_SIZE:
search_results_cache.clear()
search_results_cache.cache[query_representation_in_cache] = results_in_any_collection
if verbose and of.startswith("h"):
print_warning(req, "Search stage 3: storing query results in cache.")
# search stage 4: intersection with collection universe:
try:
# added the display_nearest_terms_box parameter to avoid printing out the "Nearest terms in any collection"
# recommendations when there results only in the hosted collections. Also added the if clause to avoid
# searching in case we know since the last stage that we have no results in any collection
if len(results_in_any_collection) != 0:
results_final = intersect_results_with_collrecs(req, results_in_any_collection, colls_to_search, ap, of, verbose, ln, display_nearest_terms_box=not hosted_colls_actual_or_potential_results_p)
else:
results_final = {}
except:
register_exception(req=req, alert_admin=True)
if of.startswith("h"):
req.write(create_error_box(req, verbose=verbose, ln=ln))
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
return page_end(req, of, ln)
if results_final == {} and not hosted_colls_actual_or_potential_results_p:
if of.startswith("h"):
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
if of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
# search stage 5: apply search option limits and restrictions:
if datetext1 != "" and results_final != {}:
if verbose and of.startswith("h"):
print_warning(req, "Search stage 5: applying time etc limits, from %s until %s..." % (datetext1, datetext2))
try:
results_final = intersect_results_with_hitset(req,
results_final,
search_unit_in_bibrec(datetext1, datetext2, dt),
ap,
aptext= _("No match within your time limits, "
"discarding this condition..."),
of=of)
except:
register_exception(req=req, alert_admin=True)
if of.startswith("h"):
req.write(create_error_box(req, verbose=verbose, ln=ln))
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
return page_end(req, of, ln)
if results_final == {} and not hosted_colls_actual_or_potential_results_p:
if of.startswith("h"):
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
#if of.startswith("x"):
# # Print empty, but valid XML
# print_records_prologue(req, of)
# print_records_epilogue(req, of)
return page_end(req, of, ln)
if pl and results_final != {}:
pl = wash_pattern(pl)
if verbose and of.startswith("h"):
print_warning(req, "Search stage 5: applying search pattern limit %s..." % cgi.escape(pl))
try:
results_final = intersect_results_with_hitset(req,
results_final,
search_pattern_parenthesised(req, pl, ap=0, ln=ln),
ap,
aptext=_("No match within your search limits, "
"discarding this condition..."),
of=of)
except:
register_exception(req=req, alert_admin=True)
if of.startswith("h"):
req.write(create_error_box(req, verbose=verbose, ln=ln))
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
return page_end(req, of, ln)
if results_final == {} and not hosted_colls_actual_or_potential_results_p:
if of.startswith("h"):
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
if of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
return page_end(req, of, ln)
t2 = os.times()[4]
cpu_time = t2 - t1
## search stage 6: display results:
results_final_nb_total = 0
results_final_nb = {} # will hold number of records found in each collection
# (in simple dict to display overview more easily)
for coll in results_final.keys():
results_final_nb[coll] = len(results_final[coll])
#results_final_nb_total += results_final_nb[coll]
# Now let us calculate results_final_nb_total more precisely,
# in order to get the total number of "distinct" hits across
# searched collections; this is useful because a record might
# have been attributed to more than one primary collection; so
# we have to avoid counting it multiple times. The price to
# pay for this accuracy of results_final_nb_total is somewhat
# increased CPU time.
if results_final.keys() == 1:
# only one collection; no need to union them
results_final_for_all_selected_colls = results_final.values()[0]
results_final_nb_total = results_final_nb.values()[0]
else:
# okay, some work ahead to union hits across collections:
results_final_for_all_selected_colls = HitSet()
for coll in results_final.keys():
results_final_for_all_selected_colls.union_update(results_final[coll])
results_final_nb_total = len(results_final_for_all_selected_colls)
#if hosted_colls and (of.startswith("h") or of.startswith("x")):
if hosted_colls_actual_or_potential_results_p:
if hosted_colls_results:
for result in hosted_colls_true_results:
colls_to_search.append(result[0][1].name)
results_final_nb[result[0][1].name] = result[1]
results_final_nb_total += result[1]
cpu_time += result[2]
if hosted_colls_timeouts:
for timeout in hosted_colls_timeouts:
colls_to_search.append(timeout[1].name)
# use -963 as a special number to identify the collections that timed out
results_final_nb[timeout[1].name] = -963
# we continue past this point only if there is a hosted collection that has timed out and might offer potential results
if results_final_nb_total ==0 and not hosted_colls_potential_results_p:
if of.startswith("h"):
print_warning(req, "No match found, please enter different search terms.")
elif of.startswith("x"):
# Print empty, but valid XML
print_records_prologue(req, of)
print_records_epilogue(req, of)
else:
# yes, some hits found: good!
# collection list may have changed due to not-exact-match-found policy so check it out:
for coll in results_final.keys():
if coll not in colls_to_search:
colls_to_search.append(coll)
# print results overview:
if of == "id":
# we have been asked to return list of recIDs
recIDs = list(results_final_for_all_selected_colls)
if sf: # do we have to sort?
recIDs = sort_records(req, recIDs, sf, so, sp, verbose, of)
elif rm: # do we have to rank?
results_final_for_all_colls_rank_records_output = rank_records(rm, 0, results_final_for_all_selected_colls,
string.split(p) + string.split(p1) +
string.split(p2) + string.split(p3), verbose)
if results_final_for_all_colls_rank_records_output[0]:
recIDs = results_final_for_all_colls_rank_records_output[0]
return recIDs
elif of.startswith("h"):
if of not in ['hcs']:
# added the hosted_colls_potential_results_p parameter to help print out the overview more accurately
req.write(print_results_overview(req, colls_to_search, results_final_nb_total, results_final_nb, cpu_time, ln, ec, hosted_colls_potential_results_p=hosted_colls_potential_results_p))
selected_external_collections_infos = print_external_results_overview(req, cc, [p, p1, p2, p3], f, ec, verbose, ln)
# print number of hits found for XML outputs:
if of.startswith("x"):
req.write("\n" % results_final_nb_total)
# print records:
if of in ['hcs']:
# feed the current search to be summarized:
from invenio.search_engine_summarizer import summarize_records
summarize_records(results_final_for_all_selected_colls, 'hcs', ln, p, f, req)
else:
if len(colls_to_search)>1:
cpu_time = -1 # we do not want to have search time printed on each collection
print_records_prologue(req, of)
for coll in colls_to_search:
if results_final.has_key(coll) and len(results_final[coll]):
if of.startswith("h"):
req.write(print_search_info(p, f, sf, so, sp, rm, of, ot, coll, results_final_nb[coll],
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time))
results_final_recIDs = list(results_final[coll])
results_final_relevances = []
results_final_relevances_prologue = ""
results_final_relevances_epilogue = ""
if sf: # do we have to sort?
results_final_recIDs = sort_records(req, results_final_recIDs, sf, so, sp, verbose, of)
elif rm: # do we have to rank?
results_final_recIDs_ranked, results_final_relevances, results_final_relevances_prologue, results_final_relevances_epilogue, results_final_comments = \
rank_records(rm, 0, results_final[coll],
string.split(p) + string.split(p1) +
string.split(p2) + string.split(p3), verbose)
if of.startswith("h"):
print_warning(req, results_final_comments)
if results_final_recIDs_ranked:
results_final_recIDs = results_final_recIDs_ranked
else:
# rank_records failed and returned some error message to display:
print_warning(req, results_final_relevances_prologue)
print_warning(req, results_final_relevances_epilogue)
print_records(req, results_final_recIDs, jrec, rg, of, ot, ln,
results_final_relevances,
results_final_relevances_prologue,
results_final_relevances_epilogue,
search_pattern=p,
print_records_prologue_p=False,
print_records_epilogue_p=False,
verbose=verbose)
if of.startswith("h"):
req.write(print_search_info(p, f, sf, so, sp, rm, of, ot, coll, results_final_nb[coll],
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time, 1))
#if hosted_colls and (of.startswith("h") or of.startswith("x")):
if hosted_colls_actual_or_potential_results_p:
if hosted_colls_results:
# TODO: add a verbose message here
for result in hosted_colls_true_results:
if of.startswith("h"):
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, result[0][1].name, results_final_nb[result[0][1].name],
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time))
- req.write(print_hosted_results(result[0], of, ln, req))
+ req.write(print_hosted_results(url_and_engine=result[0], ln=ln, of=of, req=req, limit=rg))
if of.startswith("h"):
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, result[0][1].name, results_final_nb[result[0][1].name],
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time, 1))
if hosted_colls_timeouts:
# TODO: add a verbose message here
# TODO: check if verbose messages still work when dealing with (re)calculations of timeouts
(hosted_colls_timeouts_results, hosted_colls_timeouts_timeouts) = do_calculate_hosted_collections_results(req, ln, None, verbose, None, hosted_colls_timeouts, CFG_HOSTED_COLLECTION_TIMEOUT_POST_SEARCH)
if hosted_colls_timeouts_results:
hosted_colls_timeouts_true_results = []
for result in hosted_colls_timeouts_results:
if result[1] == None or result[1] == False:
## these are the searches the returned no or zero results
## also print a nearest terms box, in case this is the only
## collection being searched and it returns no results?
if of.startswith("h"):
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, result[0][1].name, -963,
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time))
- req.write(print_hosted_results(result[0], of, ln, req, no_records_found=True))
+ req.write(print_hosted_results(url_and_engine=result[0], ln=ln, of=of, req=req, no_records_found=True, limit=rg))
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, result[0][1].name, -963,
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time, 1))
else:
# these are the searches that actually returned results on time
if of.startswith("h"):
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, result[0][1].name, result[1],
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time))
- req.write(print_hosted_results(result[0], of, ln, req))
+ req.write(print_hosted_results(url_and_engine=result[0], ln=ln, of=of, req=req, limit=rg))
if of.startswith("h"):
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, result[0][1].name, result[1],
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time, 1))
if hosted_colls_timeouts_timeouts:
for timeout in hosted_colls_timeouts_timeouts:
if of.startswith("h"):
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, timeout[0][1].name, -963,
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time))
- req.write(print_hosted_results(timeout[0], of, ln, req, timeout=True))
+ req.write(print_hosted_results(url_and_engine=timeout[0], ln=ln, of=of, req=req, search_timed_out=True, limit=rg))
req.write(print_hosted_search_info(p, f, sf, so, sp, rm, of, ot, timeout[0][1].name, -963,
jrec, rg, aas, ln, p1, p2, p3, f1, f2, f3, m1, m2, m3, op1, op2,
sc, pl_in_url,
d1y, d1m, d1d, d2y, d2m, d2d, dt, cpu_time, 1))
print_records_epilogue(req, of)
if f == "author" and of.startswith("h"):
req.write(create_similarly_named_authors_link_box(p, ln))
# log query:
try:
id_query = log_query(req.remote_host, req.args, uid)
if of.startswith("h") and id_query:
if not of in ['hcs']:
# display alert/RSS teaser for non-summary formats:
req.write(websearch_templates.tmpl_alert_rss_teaser_box_for_query(id_query, ln=ln))
except:
# do not log query if req is None (used by CLI interface)
pass
log_query_info("ss", p, f, colls_to_search, results_final_nb_total)
# External searches
if of.startswith("h"):
if not of in ['hcs']:
perform_external_collection_search(req, cc, [p, p1, p2, p3], f, ec, verbose, ln, selected_external_collections_infos)
return page_end(req, of, ln)
def perform_request_cache(req, action="show"):
"""Manipulates the search engine cache."""
req.content_type = "text/html"
req.send_http_header()
req.write("")
out = ""
out += "
Search Cache
"
# clear cache if requested:
if action == "clear":
search_results_cache.clear()
req.write(out)
# show collection reclist cache:
out = "
Collection reclist cache
"
out += "- collection table last updated: %s" % get_table_update_time('collection')
out += " - reclist cache timestamp: %s" % collection_reclist_cache.timestamp
out += " - reclist cache contents:"
out += "
"
for coll in collection_reclist_cache.cache.keys():
if collection_reclist_cache.cache[coll]:
out += "%s (%d) " % (coll, len(collection_reclist_cache.cache[coll]))
out += "
"
req.write(out)
# show search results cache:
out = "
Search Cache
"
out += "- search cache usage: %d queries cached (max. ~%d)" % \
(len(search_results_cache.cache), CFG_WEBSEARCH_SEARCH_CACHE_SIZE)
if len(search_results_cache.cache):
out += " - search cache contents:"
out += "
"
for query, hitset in search_results_cache.cache.items():
out += " %s ... %s" % (query, hitset)
out += """
"
req.write(out)
# show field i18nname cache:
out = "
Field I18N names cache
"
out += "- fieldname table last updated: %s" % get_table_update_time('fieldname')
out += " - i18nname cache timestamp: %s" % field_i18nname_cache.timestamp
out += " - i18nname cache contents:"
out += "
"
for field in field_i18nname_cache.cache.keys():
for ln in field_i18nname_cache.cache[field].keys():
out += "%s, %s = %s " % (field, ln, field_i18nname_cache.cache[field][ln])
out += "
"
req.write(out)
# show collection i18nname cache:
out = "
Collection I18N names cache
"
out += "- collectionname table last updated: %s" % get_table_update_time('collectionname')
out += " - i18nname cache timestamp: %s" % collection_i18nname_cache.timestamp
out += " - i18nname cache contents:"
out += "
"
for coll in collection_i18nname_cache.cache.keys():
for ln in collection_i18nname_cache.cache[coll].keys():
out += "%s, %s = %s " % (coll, ln, collection_i18nname_cache.cache[coll][ln])
out += "
"
req.write(out)
req.write("")
return "\n"
def perform_request_log(req, date=""):
"""Display search log information for given date."""
req.content_type = "text/html"
req.send_http_header()
req.write("")
req.write("
Search Log
")
if date: # case A: display stats for a day
yyyymmdd = string.atoi(date)
req.write("
Date: %d
" % yyyymmdd)
req.write("""
""")
req.write("
%s
%s
%s
%s
%s
%s
" % ("No.", "Time", "Pattern", "Field", "Collection", "Number of Hits"))
# read file:
p = os.popen("grep ^%d %s/search.log" % (yyyymmdd, CFG_LOGDIR), 'r')
lines = p.readlines()
p.close()
# process lines:
i = 0
for line in lines:
try:
datetime, aas, p, f, c, nbhits = string.split(line,"#")
i += 1
req.write("
")
else: # case B: display summary stats per day
yyyymm01 = int(time.strftime("%Y%m01", time.localtime()))
yyyymmdd = int(time.strftime("%Y%m%d", time.localtime()))
req.write("""
""")
req.write("
%s
%s
" % ("Day", "Number of Queries"))
for day in range(yyyymm01, yyyymmdd + 1):
p = os.popen("grep -c ^%d %s/search.log" % (day, CFG_LOGDIR), 'r')
for line in p.readlines():
req.write("""
')
num_results_regex = re.compile(r' (?:Books|Journals) ([0-9]+?) ')
def __init__(self, host="www-lib.kek.jp", path="cgi-bin/"):
super(KISSBooksExternalCollectionResultsParser, self).__init__(host, path)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
self.buffer_decode_from('Shift_JIS')
self.buffer = self.buffer.replace('\n', ' ')
html = ""
results_to_parse = self.line.finditer(self.buffer)
for result in results_to_parse:
if len(self.results) >= CFG_EXTERNAL_COLLECTION_MAXRESULTS:
return
data = result.group()
title_match = self.title.match(data)
if title_match:
- self.add_html_result(html)
+ self.add_html_result(html, limit)
num = title_match.group(1)
url = title_match.group(2)
title = title_match.group(3)
html = num + ') " + title + " "
else:
info_line_match = self.info_line.match(data)
if info_line_match:
info = info_line_match.group(1)
html += info + ' '
- self.add_html_result(html)
+ self.add_html_result(html, limit)
class GoogleExternalCollectionResultsParser(ExternalCollectionResultsParser):
"""Parser for Google"""
num_results_regex = re.compile(r'of about ([0-9,]+?)')
def __init__(self, host = "www.google.com", path=""):
super(GoogleExternalCollectionResultsParser, self).__init__(host, path)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
elements = self.buffer.split("
")
if len(elements) <= 1:
return
for element in elements[1:]:
end_index = element.find('')
if end_index != -1:
element = element[:end_index + 8]
- self.add_html_result(element)
+ self.add_html_result(element, limit)
class GoogleScholarExternalCollectionResultsParser(GoogleExternalCollectionResultsParser):
"""Parser for Google Scholar."""
def __init__(self, host = "scholar.google.com", path=""):
super(GoogleScholarExternalCollectionResultsParser, self).__init__(host, path)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
elements = self.buffer.split("
")
if len(elements) <= 1:
return
for element in elements[1:-1]:
end_index = element.find('')
if end_index != -1:
element = element[:end_index + 8]
- self.add_html_result(element + ' ')
+ self.add_html_result(element + ' ', limit)
class GoogleBooksExternalCollectionResultsParser(GoogleExternalCollectionResultsParser):
"""Parser for Google Books."""
num_results_regex = re.compile(r' with ([0-9]+?) pages on ')
def __init__(self, host = "books.google.com", path=""):
super(GoogleBooksExternalCollectionResultsParser, self).__init__(host, path)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
elements = self.buffer.split('
')
if len(elements) <= 1:
return
for element in elements[1:-1]:
- self.add_html_result(element)
+ self.add_html_result(element, limit)
class SPIRESExternalCollectionResultsParser(ExternalCollectionResultsParser):
"""Parser for SPIRES."""
num_results_regex = re.compile(r'Paper [0-9]+ to [0-9]+ of ([0-9]+)')
def __init__(self, host="www.slac.stanford.edu", path="spires/find/hep/"):
super(SPIRESExternalCollectionResultsParser, self).__init__(host, path)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
elements = self.buffer.split('
')
if len(elements) <= 2:
return
for element in elements[1:-1]:
- self.add_html_result(element)
+ self.add_html_result(element, limit)
class SCIRUSExternalCollectionResultsParser(ExternalCollectionResultsParser):
"""Parser for SCIRUS."""
num_results_regex = re.compile(r'([0-9,]+) total ')
result_separator = re.compile(r'
)')
def __init__(self, host='www.scirus.com', path='srsapp/'):
super(SCIRUSExternalCollectionResultsParser, self).__init__(host, path)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
data = self.buffer.replace('\n', ' ')
for element in self.result_separator.finditer(data):
data = element.group(1)
parsed_line = self.result_decode.match(data)
if parsed_line is not None:
link = parsed_line.group(1)
date = parsed_line.group(2)
comments = parsed_line.group(3)
similar = parsed_line.group(4)
html = "%(link)s - %(date)s %(comments)s %(similar)s " % {'link' : link,
'date' : date, 'comments' : comments, 'similar' : similar}
else:
html = self.cleaning.sub("", data) + ' '
- self.add_html_result(html)
+ self.add_html_result(html, limit)
class CiteSeerExternalCollectionResultsParser(ExternalCollectionResultsParser):
"""Parser for CiteSeer."""
num_results_regex = re.compile(r' (?:More than |)([0-9]+)(?: documents found.| results)')
result_separator = re.compile(r'.*?', re.DOTALL)
def __init__(self, host='', path=''):
super(CiteSeerExternalCollectionResultsParser, self).__init__(host, path)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
for element in self.result_separator.finditer(self.buffer):
- self.add_html_result(element.group() + ' ')
+ self.add_html_result(element.group() + ' ', limit)
class CDSInvenioHTMLExternalCollectionResultsParser(ExternalCollectionResultsParser):
"""HTML brief (hb) Parser for Invenio"""
def __init__(self, params):
self.buffer = ""
self.results = []
self.clean()
for (name, value) in params.iteritems():
setattr(self, name, value)
self.num_results_regex = re.compile(self.num_results_regex_str)
self.nbrecs_regex = re.compile(self.nbrecs_regex_str, re.IGNORECASE)
- def parse(self, of=None, req=None):
+ def parse(self, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records."""
# the patterns :
# level_a : select only the results
level_a_pat = re.compile(r'', re.DOTALL + re.MULTILINE + re.IGNORECASE)
# level_b : purge html from the basket input fields
level_b_pat = re.compile(r']*?/>', re.DOTALL + re.MULTILINE + re.IGNORECASE)
# level_c : separate the results from one another
level_c_pat = re.compile(r'(
.*?
)', re.DOTALL + re.MULTILINE + re.IGNORECASE)
# the long way :
#level_a_res = level_a_pat.search(self.buffer)
#level_ab_res = level_a_res.group(1)
#level_b_res = level_b_pat.sub('', level_ab_res)
#level_c_res = level_c_pat.finditer(level_b_res)
# the short way :
try:
results = level_c_pat.finditer(level_b_pat.sub('', level_a_pat.search(self.buffer).group(1)))
for result in results:
# each result is placed in each own table since it already has its rows and cells defined
- self.add_html_result('
' + result.group(1) + '
')
+ self.add_html_result('
' + result.group(1) + '
', limit)
except AttributeError:
# in case there were no results found an Attribute error is raised
pass
class CDSInvenioXMLExternalCollectionResultsParser(ExternalCollectionResultsParser):
"""XML (xm) parser for Invenio"""
def __init__(self, params):
self.buffer = ""
self.results = []
self.clean()
for (name, value) in params.iteritems():
setattr(self, name, value)
self.num_results_regex = re.compile(self.num_results_regex_str)
self.nbrecs_regex = re.compile(self.nbrecs_regex_str, re.IGNORECASE)
- def parse(self, of='hb', req=None):
+ def parse(self, of='hb', req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Parse buffer to extract records. Format the records using the selected output format."""
(recids, records) = self.parse_and_extract_records(of, req)
if req and cgi.parse_qs(req.args).has_key('jrec'):
counter = int(cgi.parse_qs(req.args)['jrec'][0]) - 1
else:
counter = 0
for recid in recids:
counter += 1
if of == 'hb':
html = """
%(counter)s.
%(record)s
""" % {'recid': recid,
'counter': counter,
'record': records[recid]}
elif of == 'hd':
# HTML detailed (hd) is not supported yet
# TODO: either disable the hd output format or print it out correctly
html = """"""
elif of == 'xm':
html = records[recid]
- self.add_html_result(html)
+ else:
+ html = None
+ if html:
+ self.add_html_result(html, limit)
def parse_and_extract_records(self, of='hb', req=None):
"""Parse the buffer and return a list of the recids and a
dictionary with key:value pairs like the following
recid:formated record with the selected output format"""
# the patterns :
# separate the records from one another
record_pat = re.compile(r'(.*?)', re.DOTALL + re.MULTILINE + re.IGNORECASE)
# extract the recid
recid_pat = re.compile(r'([0-9]+?)', re.DOTALL + re.MULTILINE + re.IGNORECASE)
if not of:
of='hb'
try:
results = record_pat.finditer(self.buffer)
records = {}
recids = []
for result in results:
xml_record = result.group(1)
recid = recid_pat.search(xml_record).group(1)
recids.append(recid)
if of != 'xm':
records[recid] = format_record(None, of, xml_record=xml_record)
elif of == 'xm':
records[recid] = xml_record
return (recids,records)
except AttributeError:
# in case there were no results found an Attribute error is raised
return ([], {})
diff --git a/modules/websearch/lib/websearch_external_collections_searcher.py b/modules/websearch/lib/websearch_external_collections_searcher.py
index 30e6df527..87d93f07d 100644
--- a/modules/websearch/lib/websearch_external_collections_searcher.py
+++ b/modules/websearch/lib/websearch_external_collections_searcher.py
@@ -1,604 +1,620 @@
# -*- coding: utf-8 -*-
## This file is part of CDS Invenio.
## Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 CERN.
##
## CDS Invenio is free software; you can redistribute it and/or
## modify it under the terms of the GNU General Public License as
## published by the Free Software Foundation; either version 2 of the
## License, or (at your option) any later version.
##
## CDS Invenio is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with CDS Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
"""CDS Invenio external search Engines."""
__revision__ = "$Id$"
import sys
import urllib
import cgi
import types
from invenio.config import CFG_SITE_LANG
from invenio.websearch_external_collections_config import CFG_EXTERNAL_COLLECTIONS, CFG_EXTERNAL_COLLECTION_MAXRESULTS
from invenio.websearch_external_collections_parser import CDSIndicoCollectionResutsParser, \
GoogleExternalCollectionResultsParser, \
KISSExternalCollectionResultsParser, GoogleScholarExternalCollectionResultsParser, \
GoogleBooksExternalCollectionResultsParser, KISSBooksExternalCollectionResultsParser, \
SPIRESExternalCollectionResultsParser, SCIRUSExternalCollectionResultsParser, \
CiteSeerExternalCollectionResultsParser
def format_basic(basic):
"""Format a basic query"""
if basic[3] == "w":
return basic[1]
else:
return '"' + basic[1] + '"'
def only_field(basic_search_units, fieldname):
"""Check if in the basic search units, there is only on field representated."""
for search_unit in basic_search_units:
if search_unit[2] != fieldname:
return False
return True
class ExternalSearchEngine(object):
"""Global class for interfaces to external search engines."""
lang_translator = None
def __init__(self, configuration):
self.search_url = ""
self.combiner = " "
self.name = None
self.parser_params = None
self.parser = None
self.fetch_format = ""
self.selected_by_default = False
for (name, value) in configuration.iteritems():
setattr(self, name, value)
if self.parser_params:
setattr(self, 'parser', self.parser_params['parser'](self.parser_params))
if 'fetch_format' in self.parser_params.keys():
self.fetch_format = self.parser_params['fetch_format']
def build_units(self, basic_search_units):
""" Build the research units for basic_search_units provided"""
units = []
for search_unit in basic_search_units:
unit = self.build_search_unit_unit(search_unit)
if unit is not None:
units.append(unit)
return units
def build_search_unit_unit(self, basic):
"""Build a search string from a search unit. This is the base
version that just keep keywords with "+". """
if basic[0] == "+":
return basic[1]
return None
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build a URL for a specific set of search_units."""
units = self.build_units(basic_search_units)
if len(units) == 0:
return None
request = self.combine_units(units)
url_request = urllib.quote(request)
return self.search_url + url_request
def combine_units(self, units):
"""Combine the units to make a boolean AND query."""
return self.combiner.join(units)
def __repr__(self):
return 'ec:' + self.name
class SortedFieldsSearchEngine(ExternalSearchEngine):
"""Class for search engines that used separate query box for fields."""
def __init__(self, configuration):
self.fields = []
self.fields_content = {}
self.search_url = ""
self.converter = {}
super(SortedFieldsSearchEngine, self).__init__(configuration)
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build a search URL. Reuse the search pattern found in req only with Invenio-based search engines"""
self.clear_fields()
self.fill_fields(basic_search_units)
def clear_fields(self):
"""Clear fields to be able to build a new URL."""
self.fields_content = {}
for field_name in self.fields:
self.fields_content[field_name] = []
self.fields_content["default"] = []
def fill_fields(self, basic_search_units):
"""Fill fields with the apropriate research terms."""
for search_unit in basic_search_units:
self.add_search_unit(search_unit)
def add_search_unit(self, search_unit):
"""Add a search unit to fields to search."""
if search_unit[0] == "-":
return
search = format_basic(search_unit)
field_name = search_unit[2]
if field_name in self.fields:
self.fields_content[field_name].append(search)
else:
self.fields_content["default"].append(search)
# CERN
class CDSIndicoSearchEngine(ExternalSearchEngine):
"""Global class for CDS Search Engines."""
index_translator = {'title' : 'title', 'author': 'speaker', 'fulltext': 'fulltext'}
lang_translator = {
'ca': 'ca', 'cs': 'cs', 'de': 'de', 'el': 'el', 'en': 'en', 'es': 'es',
'fr': 'fr', 'it': 'it', 'ja': 'ja', 'no': 'no', 'pl': 'pl', 'pt': 'pt',
'ru': 'ru', 'sk': 'sk', 'sv': 'sv', 'uk': 'uk', 'default' : 'en'}
def __init__(self, configuration):
super(CDSIndicoSearchEngine, self).__init__(configuration)
self.base_url = 'http://indicosearch.cern.ch/'
self.search_url = 'http://indicosearchpublic.cern.ch/search?cc=INDICOPUBLIC&p='
self.parser = CDSIndicoCollectionResutsParser()
def build_search_unit_unit(self, basic):
"""Build a search string from a search unit.
This will also translate index name using the index_translator
dictionary."""
operator = basic[0]
pattern = basic[1]
index = basic[2]
search_type = basic[3]
if self.index_translator.has_key(index):
index = self.index_translator[index]
else:
index = None
if index:
return operator + index + ':"' + pattern + '"'
else:
if search_type == 'w':
return operator + pattern
else:
return operator + '"' + pattern + '"'
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build an URL for a specific set of search_units."""
url = super(CDSIndicoSearchEngine, self).build_search_url(basic_search_units, None, lang)
if not url:
return None
if self.lang_translator.has_key(lang):
dest_lang = self.lang_translator[lang]
else:
dest_lang = self.lang_translator['default']
return url + '&ln=' + dest_lang + '&rg=' + str(CFG_EXTERNAL_COLLECTION_MAXRESULTS)
class CERNEDMSSearchEngine(SortedFieldsSearchEngine):
"""CERN EDMS"""
def __init__(self, configuration):
super(CERNEDMSSearchEngine, self).__init__(configuration)
self.base_url = "http://edms.cern.ch/cedar/plsql/fullsearch.doc_search"
self.search_url = "http://edms.cern.ch/cedar/plsql/fullsearch.doc_search?p_search_type=ADVANCED&"
self.search_url_simple = "http://edms.cern.ch/cedar/plsql/fullsearch.doc_search?p_search_type=BASE&p_free_text="
self.fields = ["author", "keyword", "abstract", "title", "reportnumber"]
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build an URL for CERN EDMS."""
super(CERNEDMSSearchEngine, self).build_search_url(basic_search_units)
if len(self.fields_content["default"]) > 0:
free_text = self.bind_fields(["author", "keyword", "abstract", "title", "reportnumber", "default"])
return self.search_url_simple + free_text
else:
authors = self.bind_fields(["author"])
title = self.bind_fields(["title", "abstract", "keyword"])
reportnumber = self.bind_fields(["reportnumber"])
url_parts = []
if authors != '':
url_parts.append('p_author=' + authors)
if title != "":
url_parts.append('p_title=' + title)
if reportnumber != "":
url_parts.append('p_document_id=' + reportnumber)
if len(url_parts) == 0:
return None
return self.search_url + "&".join(url_parts)
def bind_fields(self, fieldname_list):
"""Combine some fields together."""
result = []
for key in fieldname_list:
content = self.fields_content[key]
if len(content) > 0:
result.append(" ".join(content))
return urllib.quote(" ".join(result))
class CERNAgendaSearchEngine(ExternalSearchEngine):
"""CERN Agenda"""
def __init__(self, configuration):
super(CERNAgendaSearchEngine, self).__init__(configuration)
self.base_url = "http://agenda.cern.ch"
self.search_url_author = "http://agenda.cern.ch/search.php?field=speaker&search=Search&keywords="
self.search_url_title = "http://agenda.cern.ch/search.php?field=title&search=Search&keywords="
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build an url for searching on CERN Agenda. This will only work if there is only author
or title tags."""
if only_field(basic_search_units, "author"):
self.search_url = self.search_url_author
elif only_field(basic_search_units, "title"):
self.search_url = self.search_url_title
else:
return None
return super(CERNAgendaSearchEngine, self).build_search_url(basic_search_units)
# Google
class GoogleSearchEngine(ExternalSearchEngine):
"""Search engine class for Google """
def __init__(self, configuration):
super(GoogleSearchEngine, self).__init__(configuration)
self.base_url = "http://www.google.com"
self.search_url = "http://www.google.com/search?q="
self.parser = GoogleExternalCollectionResultsParser()
def build_search_unit_unit(self, search_unit):
"""Build a part of the google query."""
return self.build_search_unit_unit_google(search_unit, "")
def build_search_unit_unit_google(self, search_unit, author_tag):
"""Parse a unit and return it in a google query form."""
sign = search_unit[0].replace("+", "")
if search_unit[2] == "author":
if search_unit[1].find(",") >= 0 and search_unit[3] != "p":
(lastname, firstname) = search_unit[1].split(",", 1)
return sign + author_tag + '"%s %s" OR ' % (lastname, firstname) + \
sign + author_tag + '"%s %s"' % (firstname, lastname)
else:
return sign + author_tag + search_unit[1]
if search_unit[3] == "w":
return sign + search_unit[1]
else:
return sign + '"' + search_unit[1] + '"'
class GoogleBooksSearchEngine(GoogleSearchEngine):
"""Interface for searching on Google Books."""
def __init__(self, configuration):
super(GoogleBooksSearchEngine, self).__init__(configuration)
self.base_url = "http://books.google.com"
self.search_url = "http://books.google.com/books?q="
self.parser = GoogleBooksExternalCollectionResultsParser()
class GoogleScholarSearchEngine(GoogleSearchEngine):
"""Interface for searching on Google Scholar."""
def __init__(self, configuration):
super(GoogleScholarSearchEngine, self).__init__(configuration)
self.base_url = 'http://scholar.google.com'
self.search_url = 'http://scholar.google.com/scholar?q='
self.parser = GoogleScholarExternalCollectionResultsParser()
def build_search_unit_unit(self, search_unit):
"""Build a unit for Google Scholar. Is different from Google one's
because there is an author tag for authors."""
return self.build_search_unit_unit_google(search_unit, "author:")
# Kiss
class KissSearchEngine(SortedFieldsSearchEngine):
"""Search interface for KEK Information Service System.
Not to be used directly but with Kiss*SearchEngine. """
def __init__(self, configuration):
super(KissSearchEngine, self).__init__(configuration)
self.converter = { "author": "AU=", "year": "YR=",
"title": "TI=", "reportnumber": "RP=" }
self.fields = self.converter.keys()
self.parser = KISSExternalCollectionResultsParser()
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build an URL for a search."""
super(KissSearchEngine, self).build_search_url(basic_search_units)
url_parts = []
for key in self.fields:
if len(self.fields_content[key]) > 0:
field_request = " and ".join(self.fields_content[key])
url_part = self.converter[key] + urllib.quote(field_request)
url_parts.append(url_part)
if len(url_parts) == 0:
return None
return self.search_url + "&".join(url_parts)
def add_search_unit(self, search_unit):
"""Add a search unit to fields to search."""
if search_unit[0] == "+":
search = search_unit[1]
field_name = search_unit[2]
if field_name == "author":
self.add_author(search, search_unit[3])
elif field_name == "year" or field_name == "reportnumber":
self.fields_content[field_name].append(search)
else:
self.fields_content["title"].append("'" + search + "'")
def add_author(self, author_name, unit_type):
"""Handle an author unit. """
if author_name.find(",") >= 0 and unit_type != "p":
(lastname, firstname) = author_name.split(",", 1)
if firstname:
self.fields_content["author"].append("'%s, %c'" % (lastname, firstname[0]))
else:
self.fields_content["author"].append("'%s'" % (lastname))
else:
self.fields_content["author"].append("'" + author_name + "'")
class KissForPreprintsSearchEngine(KissSearchEngine):
"""Interface for seaching on Kiss for Preprints"""
def __init__(self, configuration):
super(KissForPreprintsSearchEngine, self).__init__(configuration)
self.base_url = "http://www-lib.kek.jp/KISS/kiss_prepri.html"
self.search_url = "http://www-lib.kek.jp/cgi-bin/kiss_prepri.v8?"
class KissForBooksAndJournalsSearchEngine(KissSearchEngine):
"""Interface for seaching on Kiss for Books and Journals"""
def __init__(self, configuration):
super(KissForBooksAndJournalsSearchEngine, self).__init__(configuration)
self.base_url = "http://www-lib.kek.jp/KISS/kiss_book.html"
self.search_url = "http://www-lib.kek.jp/cgi-bin/kiss_book.v8?DSP=1&"
self.parser = KISSBooksExternalCollectionResultsParser()
# Scirus
class ScirusSearchEngine(ExternalSearchEngine):
"""Interface for the Scirus search engine."""
def __init__(self, configuration):
super(ScirusSearchEngine, self).__init__(configuration)
self.base_url = "http://www.scirus.com/srsapp/"
self.search_url = "http://www.scirus.com/srsapp/search?q="
self.parser = SCIRUSExternalCollectionResultsParser()
def build_search_unit_unit(self, search_unit):
"""Build a unit for a search unit"""
sign = search_unit[0].replace("+", "")
search = self.normalize_unit(search_unit)
if search_unit[2] == "author":
return sign + "au:" + search
if search_unit[2] == "title":
return sign + "ti:" + search
if search_unit[2] == "keyword":
return sign + "keyword:" + search
if search_unit[3] == "w":
return sign + search
def normalize_unit(self, search_unit):
""" Add double quote if needed. """
if search_unit[3] == "a":
return '"' + search_unit[1] + '"'
else:
return search_unit[1]
# Spires
class SPIRESSearchEngine(ExternalSearchEngine):
"""Interface for the Spires Search engine."""
def __init__(self, configuration):
super(SPIRESSearchEngine, self).__init__(configuration)
self.base_url = "http://www.slac.stanford.edu/spires/hep/"
self.search_url = "http://www.slac.stanford.edu/spires/find/hep/?rawcmd=find+"
self.combiner = " and "
self.parser = SPIRESExternalCollectionResultsParser()
def build_search_unit_unit(self, basic):
"""Build a search string from a search unit. This is the base
version that just keep keywords with "+". """
word = format_basic(basic)
if basic[0] == "-":
sign = "not "
else:
sign = ""
if basic[2] == "author":
return sign + "a " + word
if basic[2] == "title":
return sign + "t " + word
if basic[2] == "keyword":
return sign + "k " + word
if basic[2] == "reportnumber":
return sign + "r " + word
if basic[0] == "+":
return "a " + word + " or t " + word + " or k " + word
else:
return "not a " + word + " and not t " + word + " and not k " + word
class SPIRESBooksSearchEngine(SPIRESSearchEngine):
"""SPIRES Books"""
def __init__(self, configuration):
super(SPIRESBooksSearchEngine, self).__init__(configuration)
self.base_url = "http://www.slac.stanford.edu/library/catalog/"
self.search_url = "http://www.slac.stanford.edu/spires/find/books/?rawcmd=find+"
self.parser = None
class SPIRESJournalsSearchEngine(SPIRESSearchEngine):
"""SPIRES Journals"""
def __init__(self, configuration):
super(SPIRESJournalsSearchEngine, self).__init__(configuration)
self.base_url = "http://www.slac.stanford.edu/spires/find/tserials/"
self.search_url = "http://www.slac.stanford.edu/spires/find/tserials/?rawcmd=find+"
# Misc
class AmazonSearchEngine(ExternalSearchEngine):
"""Interface for searching books on Amazon."""
def __init__(self, configuration):
super(AmazonSearchEngine, self).__init__(configuration)
self.base_url = "http://www.amazon.com"
self.search_url_general = "http://www.amazon.com/exec/obidos/external-search/?tag=cern&keyword="
self.search_url_author = "http://www.amazon.com/exec/obidos/external-search/?tag=cern&field-author="
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build an URL for Amazon"""
if only_field(basic_search_units, "author"):
self.search_url = self.search_url_author
else:
self.search_url = self.search_url_general
return super(AmazonSearchEngine, self).build_search_url(basic_search_units)
class CiteseerSearchEngine(ExternalSearchEngine):
"""Interface for searching on CiteSeer."""
def __init__(self, configuration):
super(CiteseerSearchEngine, self).__init__(configuration)
self.base_url = "http://citeseer.ist.psu.edu"
self.search_url = "http://citeseer.ist.psu.edu/cs?q="
self.parser = CiteSeerExternalCollectionResultsParser()
class INSPECSearchEngine(ExternalSearchEngine):
"""INSPEC"""
def __init__(self, configuration):
super(INSPECSearchEngine, self).__init__(configuration)
self.base_url = "http://www.datastarweb.com/cern/"
self.search_url = "http://www.datastarweb.com/cern/?dblabel=inzz&query="
self.combiner = " AND "
def build_search_unit_unit(self, basic):
"""Build a search string from a search unit. This is the base
version that just keep keywords with "+". """
word = format_basic(basic)
if basic[0] == "-":
return None
if basic[2] == "author":
return word + ".au."
if basic[2] == "title":
return word + ".ti."
if basic[2] == "abstract":
return word + ".ab."
if basic[2] == "year":
return word + ".yr."
return word + ".ti. OR " + word + ".ab."
class NEBISSearchEngine(ExternalSearchEngine):
"""NEBIS"""
def __init__(self, configuration):
super(NEBISSearchEngine, self).__init__(configuration)
self.base_url = "http://opac.nebis.ch"
self.search_url_general = "http://opac.nebis.ch/F/?func=find-b&find_code=WRD&REQUEST="
self.search_url_author = "http://opac.nebis.ch/F/?func=find-b&find_code=WAU&REQUEST="
self.search_url_title = "http://opac.nebis.ch/F/?func=find-b&find_code=WTI&REQUEST="
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build an URL for NEBIS"""
if only_field(basic_search_units, "author"):
self.search_url = self.search_url_author
elif only_field(basic_search_units, "title"):
self.search_url = self.search_url_title
else:
self.search_url = self.search_url_general
return super(NEBISSearchEngine, self).build_search_url(basic_search_units)
# Invenio based
class CDSInvenioSearchEngine(ExternalSearchEngine):
"""Generic search engine class for Invenio based sites"""
def __init__(self, configuration):
super(CDSInvenioSearchEngine, self).__init__(configuration)
- def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG):
+ def build_search_url(self, basic_search_units, req_args=None, lang=CFG_SITE_LANG, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Build a URL for an Invenio based site"""
if req_args:
search_url_params = ""
if type(req_args) is list:
# req_args is by definition a string. It is a list
# only when we manually pass it as a list of recids.
conjunction = " or "
search_url_recids = conjunction.join(['recid:%s'] * len(req_args))
params = tuple(req_args)
search_url_recids %= params
req_args = "p=" + search_url_recids
req_args_dict = cgi.parse_qs(req_args)
if req_args_dict.has_key('p'):
search_url_params += urllib.quote(req_args_dict['p'][0])
if req_args_dict.has_key('jrec'):
search_url_params += '&jrec=' + req_args_dict['jrec'][0]
if req_args_dict.has_key('rg'):
search_url_params += '&rg=' + req_args_dict['rg'][0]
+ else:
+ search_url_params += '&rg=' + str(limit)
+ if req_args_dict.has_key('d1d'):
+ search_url_params += '&d1d=' + req_args_dict['d1d'][0]
+ if req_args_dict.has_key('d1m'):
+ search_url_params += '&d1m=' + req_args_dict['d1m'][0]
+ if req_args_dict.has_key('d1y'):
+ search_url_params += '&d1y=' + req_args_dict['d1y'][0]
+ if req_args_dict.has_key('d2d'):
+ search_url_params += '&d2d=' + req_args_dict['d2d'][0]
+ if req_args_dict.has_key('d2m'):
+ search_url_params += '&d2m=' + req_args_dict['d2m'][0]
+ if req_args_dict.has_key('d2y'):
+ search_url_params += '&d2y=' + req_args_dict['d2y'][0]
+ if req_args_dict.has_key('ap'):
+ search_url_params += '&ap=' + req_args_dict['ap'][0]
search_url_params += '&of=' + self.fetch_format
return self.search_url + search_url_params
else:
units = self.build_units(basic_search_units)
if len(units) == 0:
return None
request = self.combine_units(units)
url_request = urllib.quote(request)
- return self.search_url + url_request + '&of=' + self.fetch_format
+ return self.search_url + url_request + '&rg=' + str(limit) + '&of=' + self.fetch_format
def build_search_unit_unit(self, basic):
"""Build a search string from a search unit. Reconstructs original user query"""
# TO DO: correct & improve the print out
# adding the semicolon in case a specific field is chosen
if basic[2] != "": basic[2] = basic[2] + ":"
# adding the single quotes in case a multi word values is searched for
if basic[3] == "a": basic[1] = "'" + basic[1] + "'"
return basic[0] + " " + basic[2] + basic[1]
def build_record_urls(self, recids):
"""Given a list of records this function returns a dictionary with
recid:external_url key:value pairs"""
if type(recids) is not list:
recids = [recids]
recids_urls = []
for recid in recids:
recids_urls.append((recid, self.record_url + recid))
return tuple(recids_urls)
external_collections_dictionary = {}
def build_external_collections_dictionary():
"""Build the dictionary of the external collections."""
for (name, configuration) in CFG_EXTERNAL_COLLECTIONS.iteritems():
engine_name = configuration.pop('engine', 'External') + 'SearchEngine'
configuration['name'] = name
if globals().has_key(engine_name):
external_collections_dictionary[name] = globals()[engine_name](configuration)
else:
sys.stderr.write("Error : not found " + engine_name + "\n")
build_external_collections_dictionary()
diff --git a/modules/websearch/lib/websearch_templates.py b/modules/websearch/lib/websearch_templates.py
index aef98b310..2ece98057 100644
--- a/modules/websearch/lib/websearch_templates.py
+++ b/modules/websearch/lib/websearch_templates.py
@@ -1,4048 +1,4049 @@
# -*- coding: utf-8 -*-
## This file is part of CDS Invenio.
## Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 CERN.
##
## CDS Invenio is free software; you can redistribute it and/or
## modify it under the terms of the GNU General Public License as
## published by the Free Software Foundation; either version 2 of the
## License, or (at your option) any later version.
##
## CDS Invenio is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with CDS Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
# pylint: disable-msg=C0301
__revision__ = "$Id$"
import time
import cgi
import gettext
import string
import re
import locale
from urllib import quote, urlencode
from xml.sax.saxutils import escape as xml_escape
from invenio.config import \
CFG_WEBSEARCH_LIGHTSEARCH_PATTERN_BOX_WIDTH, \
CFG_WEBSEARCH_SIMPLESEARCH_PATTERN_BOX_WIDTH, \
CFG_WEBSEARCH_ADVANCEDSEARCH_PATTERN_BOX_WIDTH, \
CFG_WEBSEARCH_AUTHOR_ET_AL_THRESHOLD, \
CFG_WEBSEARCH_USE_ALEPH_SYSNOS, \
CFG_WEBSEARCH_SPLIT_BY_COLLECTION, \
CFG_BIBRANK_SHOW_READING_STATS, \
CFG_BIBRANK_SHOW_DOWNLOAD_STATS, \
CFG_BIBRANK_SHOW_DOWNLOAD_GRAPHS, \
CFG_BIBRANK_SHOW_CITATION_LINKS, \
CFG_BIBRANK_SHOW_CITATION_STATS, \
CFG_BIBRANK_SHOW_CITATION_GRAPHS, \
CFG_WEBSEARCH_INSTANT_BROWSE_RSS, \
CFG_WEBSEARCH_RSS_TTL, \
CFG_SITE_LANG, \
CFG_SITE_NAME, \
CFG_SITE_NAME_INTL, \
CFG_VERSION, \
CFG_SITE_URL, \
CFG_SITE_SUPPORT_EMAIL, \
CFG_SITE_ADMIN_EMAIL, \
CFG_INSPIRE_SITE, \
CFG_WEBSEARCH_DEFAULT_SEARCH_INTERFACE, \
CFG_WEBSEARCH_ENABLED_SEARCH_INTERFACES, \
CFG_WEBSEARCH_MAX_RECORDS_IN_GROUPS, \
CFG_BIBINDEX_CHARS_PUNCTUATION, \
CFG_WEBCOMMENT_ALLOW_COMMENTS, \
CFG_WEBCOMMENT_ALLOW_REVIEWS, \
CFG_WEBSEARCH_SHOW_COMMENT_COUNT, \
CFG_WEBSEARCH_SHOW_REVIEW_COUNT
from invenio.dbquery import run_sql
from invenio.messages import gettext_set_language
#from invenio.search_engine_config import CFG_EXPERIMENTAL_FEATURES
from invenio.urlutils import make_canonical_urlargd, drop_default_urlargd, create_html_link, create_url
from invenio.htmlutils import nmtoken_from_string
from invenio.webinterface_handler import wash_urlargd
from invenio.bibrank_citation_searcher import get_cited_by_count
from invenio.intbitset import intbitset
from invenio.websearch_external_collections import external_collection_get_state, get_external_collection_engine
from invenio.websearch_external_collections_utils import get_collection_id
+from invenio.websearch_external_collections_config import CFG_EXTERNAL_COLLECTION_MAXRESULTS
_RE_PUNCTUATION = re.compile(CFG_BIBINDEX_CHARS_PUNCTUATION)
_RE_SPACES = re.compile(r"\s+")
def get_fieldvalues(recID, tag):
"""Return list of field values for field TAG inside record RECID.
FIXME: should be imported commonly for search_engine too."""
out = []
if tag == "001___":
# we have asked for recID that is not stored in bibXXx tables
out.append(str(recID))
else:
# we are going to look inside bibXXx tables
digit = tag[0:2]
bx = "bib%sx" % digit
bibx = "bibrec_bib%sx" % digit
query = "SELECT bx.value FROM %s AS bx, %s AS bibx WHERE bibx.id_bibrec='%s' AND bx.id=bibx.id_bibxxx AND bx.tag LIKE '%s'" \
"ORDER BY bibx.field_number, bx.tag ASC" % (bx, bibx, recID, tag)
res = run_sql(query)
for row in res:
out.append(row[0])
return out
class Template:
# This dictionary maps CDS Invenio language code to locale codes (ISO 639)
tmpl_localemap = {
'bg': 'bg_BG',
'ca': 'ca_ES',
'de': 'de_DE',
'el': 'el_GR',
'en': 'en_US',
'es': 'es_ES',
'pt': 'pt_BR',
'fr': 'fr_FR',
'it': 'it_IT',
'ro': 'ro_RO',
'ru': 'ru_RU',
'rw': 'rw_RW',
'sk': 'sk_SK',
'cs': 'cs_CZ',
'no': 'no_NO',
'sv': 'sv_SE',
'uk': 'uk_UA',
'ja': 'ja_JA',
'pl': 'pl_PL',
'hr': 'hr_HR',
'zh_CN': 'zh_CN',
'zh_TW': 'zh_TW',
'hu': 'hu_HU',
'af': 'af_ZA',
'gl': 'gl_ES'
}
tmpl_default_locale = "en_US" # which locale to use by default, useful in case of failure
# Type of the allowed parameters for the web interface for search results
search_results_default_urlargd = {
'cc': (str, CFG_SITE_NAME),
'c': (list, []),
'p': (str, ""), 'f': (str, ""),
'rg': (int, 10),
'sf': (str, ""),
'so': (str, "d"),
'sp': (str, ""),
'rm': (str, ""),
'of': (str, "hb"),
'ot': (list, []),
'aas': (int, CFG_WEBSEARCH_DEFAULT_SEARCH_INTERFACE),
'as': (int, CFG_WEBSEARCH_DEFAULT_SEARCH_INTERFACE),
'p1': (str, ""), 'f1': (str, ""), 'm1': (str, ""), 'op1':(str, ""),
'p2': (str, ""), 'f2': (str, ""), 'm2': (str, ""), 'op2':(str, ""),
'p3': (str, ""), 'f3': (str, ""), 'm3': (str, ""),
'sc': (int, 0),
'jrec': (int, 0),
'recid': (int, -1), 'recidb': (int, -1), 'sysno': (str, ""),
'id': (int, -1), 'idb': (int, -1), 'sysnb': (str, ""),
'action': (str, "search"),
'action_search': (str, ""),
'action_browse': (str, ""),
'd1': (str, ""),
'd1y': (int, 0), 'd1m': (int, 0), 'd1d': (int, 0),
'd2': (str, ""),
'd2y': (int, 0), 'd2m': (int, 0), 'd2d': (int, 0),
'dt': (str, ""),
'ap': (int, 1),
'verbose': (int, 0),
'ec': (list, []),
}
# ...and for search interfaces
search_interface_default_urlargd = {
'aas': (int, CFG_WEBSEARCH_DEFAULT_SEARCH_INTERFACE),
'as': (int, CFG_WEBSEARCH_DEFAULT_SEARCH_INTERFACE),
'verbose': (int, 0)}
# ...and for RSS feeds
rss_default_urlargd = {'c' : (list, []),
'cc' : (str, ""),
'p' : (str, ""),
'f' : (str, ""),
'p1' : (str, ""),
'f1' : (str, ""),
'm1' : (str, ""),
'op1': (str, ""),
'p2' : (str, ""),
'f2' : (str, ""),
'm2' : (str, ""),
'op2': (str, ""),
'p3' : (str, ""),
'f3' : (str, ""),
'm3' : (str, "")}
tmpl_openurl_accepted_args = {
'id' : (list, []),
'genre' : (str, ''),
'aulast' : (str, ''),
'aufirst' : (str, ''),
'auinit' : (str, ''),
'auinit1' : (str, ''),
'auinitm' : (str, ''),
'issn' : (str, ''),
'eissn' : (str, ''),
'coden' : (str, ''),
'isbn' : (str, ''),
'sici' : (str, ''),
'bici' : (str, ''),
'title' : (str, ''),
'stitle' : (str, ''),
'atitle' : (str, ''),
'volume' : (str, ''),
'part' : (str, ''),
'issue' : (str, ''),
'spage' : (str, ''),
'epage' : (str, ''),
'pages' : (str, ''),
'artnum' : (str, ''),
'date' : (str, ''),
'ssn' : (str, ''),
'quarter' : (str, ''),
'url_ver' : (str, ''),
'ctx_ver' : (str, ''),
'rft_val_fmt' : (str, ''),
'rft_id' : (list, []),
'rft.atitle' : (str, ''),
'rft.title' : (str, ''),
'rft.jtitle' : (str, ''),
'rft.stitle' : (str, ''),
'rft.date' : (str, ''),
'rft.volume' : (str, ''),
'rft.issue' : (str, ''),
'rft.spage' : (str, ''),
'rft.epage' : (str, ''),
'rft.pages' : (str, ''),
'rft.artnumber' : (str, ''),
'rft.issn' : (str, ''),
'rft.eissn' : (str, ''),
'rft.aulast' : (str, ''),
'rft.aufirst' : (str, ''),
'rft.auinit' : (str, ''),
'rft.auinit1' : (str, ''),
'rft.auinitm' : (str, ''),
'rft.ausuffix' : (str, ''),
'rft.au' : (list, []),
'rft.aucorp' : (str, ''),
'rft.isbn' : (str, ''),
'rft.coden' : (str, ''),
'rft.sici' : (str, ''),
'rft.genre' : (str, 'unknown'),
'rft.chron' : (str, ''),
'rft.ssn' : (str, ''),
'rft.quarter' : (int, ''),
'rft.part' : (str, ''),
'rft.btitle' : (str, ''),
'rft.isbn' : (str, ''),
'rft.atitle' : (str, ''),
'rft.place' : (str, ''),
'rft.pub' : (str, ''),
'rft.edition' : (str, ''),
'rft.tpages' : (str, ''),
'rft.series' : (str, ''),
}
tmpl_opensearch_rss_url_syntax = "%(CFG_SITE_URL)s/rss?p={searchTerms}&jrec={startIndex}&rg={count}&ln={language}&startIndex" % {'CFG_SITE_URL': CFG_SITE_URL}
tmpl_opensearch_html_url_syntax = "%(CFG_SITE_URL)s/search?p={searchTerms}&jrec={startIndex}&rg={count}&ln={language}&startIndex" % {'CFG_SITE_URL': CFG_SITE_URL}
def tmpl_openurl2invenio(self, openurl_data):
""" Return an Invenio url corresponding to a search with the data
included in the openurl form map.
"""
def isbn_to_isbn13_isbn10(isbn):
isbn = isbn.replace(' ', '').replace('-', '')
if len(isbn) == 10 and isbn.isdigit():
## We already have isbn10
return ('', isbn)
if len(isbn) != 13 and isbn.isdigit():
return ('', '')
isbn13, isbn10 = isbn, isbn[3:-1]
checksum = 0
weight = 10
for char in isbn10:
checksum += int(char) * weight
weight -= 1
checksum = 11 - (checksum % 11)
if checksum == 10:
isbn10 += 'X'
if checksum == 11:
isbn10 += '0'
else:
isbn10 += str(checksum)
return (isbn13, isbn10)
from invenio.search_engine import perform_request_search
doi = ''
pmid = ''
bibcode = ''
oai = ''
issn = ''
isbn = ''
for elem in openurl_data['id']:
if elem.startswith('doi:'):
doi = elem[len('doi:'):]
elif elem.startswith('pmid:'):
pmid = elem[len('pmid:'):]
elif elem.startswith('bibcode:'):
bibcode = elem[len('bibcode:'):]
elif elem.startswith('oai:'):
oai = elem[len('oai:'):]
for elem in openurl_data['rft_id']:
if elem.startswith('info:doi/'):
doi = elem[len('info:doi/'):]
elif elem.startswith('info:pmid/'):
pmid = elem[len('info:pmid/'):]
elif elem.startswith('info:bibcode/'):
bibcode = elem[len('info:bibcode/'):]
elif elem.startswith('info:oai/'):
oai = elem[len('info:oai/')]
elif elem.startswith('urn:ISBN:'):
isbn = elem[len('urn:ISBN:'):]
elif elem.startswith('urn:ISSN:'):
issn = elem[len('urn:ISSN:'):]
## Building author query
aulast = openurl_data['rft.aulast'] or openurl_data['aulast']
aufirst = openurl_data['rft.aufirst'] or openurl_data['aufirst']
auinit = openurl_data['rft.auinit'] or \
openurl_data['auinit'] or \
openurl_data['rft.auinit1'] + ' ' + openurl_data['rft.auinitm'] or \
openurl_data['auinit1'] + ' ' + openurl_data['auinitm'] or aufirst[:1]
auinit = auinit.upper()
if aulast and aufirst:
author_query = 'author:"%s, %s" or author:"%s, %s"' % (aulast, aufirst, aulast, auinit)
elif aulast and auinit:
author_query = 'author:"%s, %s"' % (aulast, auinit)
else:
author_query = ''
## Building title query
title = openurl_data['rft.atitle'] or \
openurl_data['atitle'] or \
openurl_data['rft.btitle'] or \
openurl_data['rft.title'] or \
openurl_data['title']
if title:
title_query = 'title:"%s"' % title
title_query_cleaned = 'title:"%s"' % _RE_SPACES.sub(' ', _RE_PUNCTUATION.sub(' ', title))
else:
title_query = ''
## Building journal query
jtitle = openurl_data['rft.stitle'] or \
openurl_data['stitle'] or \
openurl_data['rft.jtitle'] or \
openurl_data['title']
if jtitle:
journal_query = 'journal:"%s"' % jtitle
else:
journal_query = ''
## Building isbn query
isbn = isbn or openurl_data['rft.isbn'] or \
openurl_data['isbn']
isbn13, isbn10 = isbn_to_isbn13_isbn10(isbn)
if isbn13:
isbn_query = 'isbn:"%s" or isbn:"%s"' % (isbn13, isbn10)
elif isbn10:
isbn_query = 'isbn:"%s"' % isbn10
else:
isbn_query = ''
## Building issn query
issn = issn or openurl_data['rft.eissn'] or \
openurl_data['eissn'] or \
openurl_data['rft.issn'] or \
openurl_data['issn']
if issn:
issn_query = 'issn:"%s"' % issn
else:
issn_query = ''
## Building coden query
coden = openurl_data['rft.coden'] or openurl_data['coden']
if coden:
coden_query = 'coden:"%s"' % coden
else:
coden_query = ''
## Building doi query
if False: #doi: #FIXME Temporaly disabled until doi field is properly setup
doi_query = 'doi:"%s"' % doi
else:
doi_query = ''
## Trying possible searches
if doi_query:
if perform_request_search(p=doi_query):
return '%s/search?%s' % (CFG_SITE_URL, urlencode({
'p' : doi_query,
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hd'}))
if isbn_query:
if perform_request_search(p=isbn_query):
return '%s/search?%s' % (CFG_SITE_URL, urlencode({
'p' : isbn_query,
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hd'}))
if coden_query:
if perform_request_search(p=coden_query):
return '%s/search?%s' % (CFG_SITE_URL, urlencode({
'p' : coden_query,
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hd'}))
if author_query and title_query:
if perform_request_search(p='%s and %s' % (title_query, author_query)):
return '%s/search?%s' % (CFG_SITE_URL, urlencode({
'p' : '%s and %s' % (title_query, author_query),
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hd'}))
if title_query:
result = len(perform_request_search(p=title_query))
if result == 1:
return '%s/search?%s' % (CFG_SITE_URL, urlencode({
'p' : title_query,
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hd'}))
elif result > 1:
return '%s/search?%s' % (CFG_SITE_URL, urlencode({
'p' : title_query,
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hb'}))
## Nothing worked, let's return a search that the user can improve
if author_query and title_query:
return '%s/search%s' % (CFG_SITE_URL, make_canonical_urlargd({
'p' : '%s and %s' % (title_query_cleaned, author_query),
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hb'}, {}))
elif title_query:
return '%s/search%s' % (CFG_SITE_URL, make_canonical_urlargd({
'p' : title_query_cleaned,
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hb'}, {}))
else:
## Mmh. Too few information provided.
return '%s/search%s' % (CFG_SITE_URL, make_canonical_urlargd({
'p' : 'recid:-1',
'sc' : CFG_WEBSEARCH_SPLIT_BY_COLLECTION,
'of' : 'hb'}, {}))
def tmpl_opensearch_description(self, ln):
""" Returns the OpenSearch description file of this site.
"""
_ = gettext_set_language(ln)
return """%(short_name)s%(long_name)s%(description)sUTF-8UTF-8*%(CFG_SITE_ADMIN_EMAIL)sPowered by CDS Invenio%(CFG_SITE_URL)s""" % \
{'CFG_SITE_URL': CFG_SITE_URL,
'short_name': CFG_SITE_NAME_INTL.get(ln, CFG_SITE_NAME)[:16],
'long_name': CFG_SITE_NAME_INTL.get(ln, CFG_SITE_NAME),
'description': (_("Search on %(x_CFG_SITE_NAME_INTL)s") % \
{'x_CFG_SITE_NAME_INTL': CFG_SITE_NAME_INTL.get(ln, CFG_SITE_NAME)})[:1024],
'CFG_SITE_ADMIN_EMAIL': CFG_SITE_ADMIN_EMAIL,
'rss_search_syntax': self.tmpl_opensearch_rss_url_syntax,
'html_search_syntax': self.tmpl_opensearch_html_url_syntax
}
def build_search_url(self, known_parameters={}, **kargs):
""" Helper for generating a canonical search
url. 'known_parameters' is the list of query parameters you
inherit from your current query. You can then pass keyword
arguments to modify this query.
build_search_url(known_parameters, of="xm")
The generated URL is absolute.
"""
parameters = {}
parameters.update(known_parameters)
parameters.update(kargs)
# Now, we only have the arguments which have _not_ their default value
parameters = drop_default_urlargd(parameters, self.search_results_default_urlargd)
# Treat `as' argument specially:
if parameters.has_key('aas'):
parameters['as'] = parameters['aas']
del parameters['aas']
# Asking for a recid? Return a /record/ URL
if 'recid' in parameters:
target = "%s/record/%s" % (CFG_SITE_URL, parameters['recid'])
del parameters['recid']
target += make_canonical_urlargd(parameters, self.search_results_default_urlargd)
return target
return "%s/search%s" % (CFG_SITE_URL, make_canonical_urlargd(parameters, self.search_results_default_urlargd))
def build_search_interface_url(self, known_parameters={}, **kargs):
""" Helper for generating a canonical search interface URL."""
parameters = {}
parameters.update(known_parameters)
parameters.update(kargs)
c = parameters['c']
del parameters['c']
# Now, we only have the arguments which have _not_ their default value
parameters = drop_default_urlargd(parameters, self.search_results_default_urlargd)
# Treat `as' argument specially:
if parameters.has_key('aas'):
parameters['as'] = parameters['aas']
del parameters['aas']
if c and c != CFG_SITE_NAME:
base = CFG_SITE_URL + '/collection/' + quote(c)
else:
base = CFG_SITE_URL
return create_url(base, parameters)
def build_rss_url(self, known_parameters, **kargs):
"""Helper for generating a canonical RSS URL"""
parameters = {}
parameters.update(known_parameters)
parameters.update(kargs)
# Keep only interesting parameters
argd = wash_urlargd(parameters, self.rss_default_urlargd)
if argd:
# Handle 'c' differently since it is a list
c = argd.get('c', [])
del argd['c']
# Create query, and drop empty params
args = make_canonical_urlargd(argd, self.rss_default_urlargd)
if c != []:
# Add collections
c = [quote(coll) for coll in c]
if args == '':
args += '?'
else:
args += '&'
args += 'c=' + '&c='.join(c)
return CFG_SITE_URL + '/rss' + args
def tmpl_record_page_header_content(self, req, recid, ln):
""" Provide extra information in the header of /record pages """
_ = gettext_set_language(ln)
title = get_fieldvalues(recid, "245__a")
if title:
title = cgi.escape(title[0])
else:
title = _("Record") + ' #%d' % recid
keywords = ', '.join(get_fieldvalues(recid, "6531_a"))
description = ' '.join(get_fieldvalues(recid, "520__a"))
description += "\n"
description += '; '.join(get_fieldvalues(recid, "100__a") + get_fieldvalues(recid, "700__a"))
return [cgi.escape(x, True) for x in (title, description, keywords)]
def tmpl_navtrail_links(self, aas, ln, dads):
"""
Creates the navigation bar at top of each search page (*Home > Root collection > subcollection > ...*)
Parameters:
- 'aas' *int* - Should we display an advanced search box?
- 'ln' *string* - The language to display
- 'separator' *string* - The separator between two consecutive collections
- 'dads' *list* - A list of parent links, eachone being a dictionary of ('name', 'longname')
"""
out = []
for url, name in dads:
args = {'c': url, 'as': aas, 'ln': ln}
out.append(create_html_link(self.build_search_interface_url(**args), {}, cgi.escape(name), {'class': 'navtrail'}))
return ' > '.join(out)
def tmpl_webcoll_body(self, ln, collection, te_portalbox,
searchfor, np_portalbox, narrowsearch,
focuson, instantbrowse, ne_portalbox):
""" Creates the body of the main search page.
Parameters:
- 'ln' *string* - language of the page being generated
- 'collection' - collection id of the page being generated
- 'te_portalbox' *string* - The HTML code for the portalbox on top of search
- 'searchfor' *string* - The HTML code for the search for box
- 'np_portalbox' *string* - The HTML code for the portalbox on bottom of search
- 'narrowsearch' *string* - The HTML code for the search categories (left bottom of page)
- 'focuson' *string* - The HTML code for the "focuson" categories (right bottom of page)
- 'ne_portalbox' *string* - The HTML code for the bottom of the page
"""
if not narrowsearch:
narrowsearch = instantbrowse
body = '''
""" % {'ne_portalbox' : ne_portalbox}
return body
def tmpl_portalbox(self, title, body):
"""Creates portalboxes based on the parameters
Parameters:
- 'title' *string* - The title of the box
- 'body' *string* - The HTML code for the body of the box
"""
out = """
%(title)s
%(body)s
""" % {'title' : cgi.escape(title), 'body' : body}
return out
def tmpl_searchfor_light(self, ln, collection_id, collection_name, record_count,
example_search_queries): # EXPERIMENTAL
"""Produces light *Search for* box for the current collection.
Parameters:
- 'ln' *string* - *str* The language to display
- 'collection_id' - *str* The collection id
- 'collection_name' - *str* The collection name in current language
- 'example_search_queries' - *list* List of search queries given as example for this collection
"""
# load the right message language
_ = gettext_set_language(ln)
out = '''
'''
argd = drop_default_urlargd({'ln': ln, 'sc': CFG_WEBSEARCH_SPLIT_BY_COLLECTION},
self.search_results_default_urlargd)
# Only add non-default hidden values
for field, value in argd.items():
out += self.tmpl_input_hidden(field, value)
header = _("Search %s records for:") % \
self.tmpl_nbrecs_info(record_count, "", "")
asearchurl = self.build_search_interface_url(c=collection_id,
aas=max(CFG_WEBSEARCH_ENABLED_SEARCH_INTERFACES),
ln=ln)
# Build example of queries for this collection
example_search_queries_links = [create_html_link(self.build_search_url(p=example_query,
ln=ln,
aas=-1,
c=collection_id),
{},
cgi.escape(example_query),
{'class': 'examplequery'}) \
for example_query in example_search_queries]
example_query_html = ''
if len(example_search_queries) > 0:
example_query_link = example_search_queries_links[0]
# offers more examples if possible
more = ''
if len(example_search_queries_links) > 1:
more = '''
%(more_example_queries)s
''' % {'more_example_queries': ' '.join(example_search_queries_links[1:]),
'show_less':_("less"),
'show_more':_("more")}
example_query_html += '''
%(example)s :: %(more)s
''' % {'example': _("Example: %(x_sample_search_query)s") % \
{'x_sample_search_query': example_query_link},
'more': more}
# display options to search in current collection or everywhere
search_in = ''
if collection_name != CFG_SITE_NAME_INTL.get(ln, CFG_SITE_NAME):
search_in += '''
''' % {'search_in_collection_name': _("Search in %(x_collection_name)s") % \
{'x_collection_name': collection_name},
'collection_id': collection_id,
'root_collection_name': CFG_SITE_NAME,
'search_everywhere': _("Search everywhere")}
# print commentary start:
out += '''
%(search_in)s
''' % {'ln' : ln,
'sizepattern' : CFG_WEBSEARCH_LIGHTSEARCH_PATTERN_BOX_WIDTH,
'langlink': ln != CFG_SITE_LANG and '?ln=' + ln or '',
'siteurl' : CFG_SITE_URL,
'asearch' : create_html_link(asearchurl, {}, _('Advanced Search')),
'header' : header,
'msg_search' : _('Search'),
'msg_browse' : _('Browse'),
'msg_search_tips' : _('Search Tips'),
'search_in': search_in,
'example_query_html': example_query_html}
return out
def tmpl_searchfor_simple(self, ln, collection_id, collection_name, record_count, middle_option):
"""Produces simple *Search for* box for the current collection.
Parameters:
- 'ln' *string* - *str* The language to display
- 'collection_id' - *str* The collection id
- 'collection_name' - *str* The collection name in current language
- 'record_count' - *str* Number of records in this collection
- 'middle_option' *string* - HTML code for the options (any field, specific fields ...)
"""
# load the right message language
_ = gettext_set_language(ln)
out = '''
'''
argd = drop_default_urlargd({'ln': ln, 'cc': collection_id, 'sc': CFG_WEBSEARCH_SPLIT_BY_COLLECTION},
self.search_results_default_urlargd)
# Only add non-default hidden values
for field, value in argd.items():
out += self.tmpl_input_hidden(field, value)
header = _("Search %s records for:") % \
self.tmpl_nbrecs_info(record_count, "", "")
asearchurl = self.build_search_interface_url(c=collection_id,
aas=max(CFG_WEBSEARCH_ENABLED_SEARCH_INTERFACES),
ln=ln)
# print commentary start:
out += '''
''' % {'ln' : ln,
'sizepattern' : CFG_WEBSEARCH_SIMPLESEARCH_PATTERN_BOX_WIDTH,
'langlink': ln != CFG_SITE_LANG and '?ln=' + ln or '',
'siteurl' : CFG_SITE_URL,
'asearch' : create_html_link(asearchurl, {}, _('Advanced Search')),
'header' : header,
'middle_option' : middle_option,
'msg_search' : _('Search'),
'msg_browse' : _('Browse'),
'msg_search_tips' : _('Search Tips')}
return out
def tmpl_searchfor_advanced(self,
ln, # current language
collection_id,
collection_name,
record_count,
middle_option_1, middle_option_2, middle_option_3,
searchoptions,
sortoptions,
rankoptions,
displayoptions,
formatoptions
):
"""
Produces advanced *Search for* box for the current collection.
Parameters:
- 'ln' *string* - The language to display
- 'middle_option_1' *string* - HTML code for the first row of options (any field, specific fields ...)
- 'middle_option_2' *string* - HTML code for the second row of options (any field, specific fields ...)
- 'middle_option_3' *string* - HTML code for the third row of options (any field, specific fields ...)
- 'searchoptions' *string* - HTML code for the search options
- 'sortoptions' *string* - HTML code for the sort options
- 'rankoptions' *string* - HTML code for the rank options
- 'displayoptions' *string* - HTML code for the display options
- 'formatoptions' *string* - HTML code for the format options
"""
# load the right message language
_ = gettext_set_language(ln)
out = '''
'''
argd = drop_default_urlargd({'ln': ln, 'aas': 1, 'cc': collection_id, 'sc': CFG_WEBSEARCH_SPLIT_BY_COLLECTION},
self.search_results_default_urlargd)
# Only add non-default hidden values
for field, value in argd.items():
out += self.tmpl_input_hidden(field, value)
header = _("Search %s records for") % \
self.tmpl_nbrecs_info(record_count, "", "")
header += ':'
ssearchurl = self.build_search_interface_url(c=collection_id, aas=min(CFG_WEBSEARCH_ENABLED_SEARCH_INTERFACES), ln=ln)
out += '''
""" % {
'added' : _("Added/modified since:"),
'until' : _("until:"),
'added_or_modified': self.tmpl_inputdatetype(ln=ln),
'date_added' : self.tmpl_inputdate("d1", ln=ln),
'date_until' : self.tmpl_inputdate("d2", ln=ln),
'msg_sort' : _("Sort by:"),
'msg_display' : _("Display results:"),
'msg_format' : _("Output format:"),
'sortoptions' : sortoptions,
'rankoptions' : rankoptions,
'displayoptions' : displayoptions,
'formatoptions' : formatoptions
}
return out
def tmpl_matchtype_box(self, name='m', value='', ln='en'):
"""Returns HTML code for the 'match type' selection box.
Parameters:
- 'name' *string* - The name of the produced select
- 'value' *string* - The selected value (if any value is already selected)
- 'ln' *string* - the language to display
"""
# load the right message language
_ = gettext_set_language(ln)
out = """
""" % {'name' : name,
'sela' : self.tmpl_is_selected('a', value),
'opta' : _("All of the words:"),
'selo' : self.tmpl_is_selected('o', value),
'opto' : _("Any of the words:"),
'sele' : self.tmpl_is_selected('e', value),
'opte' : _("Exact phrase:"),
'selp' : self.tmpl_is_selected('p', value),
'optp' : _("Partial phrase:"),
'selr' : self.tmpl_is_selected('r', value),
'optr' : _("Regular expression:")
}
return out
def tmpl_is_selected(self, var, fld):
"""
Checks if *var* and *fld* are equal, and if yes, returns ' selected="selected"'. Useful for select boxes.
Parameters:
- 'var' *string* - First value to compare
- 'fld' *string* - Second value to compare
"""
if var == fld:
return ' selected="selected"'
else:
return ""
def tmpl_andornot_box(self, name='op', value='', ln='en'):
"""
Returns HTML code for the AND/OR/NOT selection box.
Parameters:
- 'name' *string* - The name of the produced select
- 'value' *string* - The selected value (if any value is already selected)
- 'ln' *string* - the language to display
"""
# load the right message language
_ = gettext_set_language(ln)
out = """
""" % {'name' : name,
'sela' : self.tmpl_is_selected('a', value), 'opta' : _("AND"),
'selo' : self.tmpl_is_selected('o', value), 'opto' : _("OR"),
'seln' : self.tmpl_is_selected('n', value), 'optn' : _("AND NOT")
}
return out
def tmpl_inputdate(self, name, ln, sy = 0, sm = 0, sd = 0):
"""
Produces *From Date*, *Until Date* kind of selection box. Suitable for search options.
Parameters:
- 'name' *string* - The base name of the produced selects
- 'ln' *string* - the language to display
"""
# load the right message language
_ = gettext_set_language(ln)
box = """
"""
# month
box += """
"""
# year
box += """
"""
return box
def tmpl_inputdatetype(self, dt='', ln=CFG_SITE_LANG):
"""
Produces input date type selection box to choose
added-or-modified date search option.
Parameters:
- 'dt' *string - date type (c=created, m=modified)
- 'ln' *string* - the language to display
"""
# load the right message language
_ = gettext_set_language(ln)
box = """
""" % { 'added': _("Added since:"),
'modified': _("Modified since:"),
'sel': self.tmpl_is_selected(dt, 'm'),
}
return box
def tmpl_narrowsearch(self, aas, ln, type, father,
has_grandchildren, sons, display_grandsons,
grandsons):
"""
Creates list of collection descendants of type *type* under title *title*.
If aas==1, then links to Advanced Search interfaces; otherwise Simple Search.
Suitable for 'Narrow search' and 'Focus on' boxes.
Parameters:
- 'aas' *bool* - Should we display an advanced search box?
- 'ln' *string* - The language to display
- 'type' *string* - The type of the produced box (virtual collections or normal collections)
- 'father' *collection* - The current collection
- 'has_grandchildren' *bool* - If the current collection has grand children
- 'sons' *list* - The list of the sub-collections (first level)
- 'display_grandsons' *bool* - If the grand children collections should be displayed (2 level deep display)
- 'grandsons' *list* - The list of sub-collections (second level)
"""
# load the right message language
_ = gettext_set_language(ln)
title = {'r': _("Narrow by collection:"),
'v': _("Focus on:")}[type]
if has_grandchildren:
style_prolog = ""
style_epilog = ""
else:
style_prolog = ""
style_epilog = ""
out = """
%(title)s
""" % {'title' : title,
'narrowsearchbox': {'r': 'narrowsearchbox',
'v': 'focusonsearchbox'}[type]}
# iterate through sons:
i = 0
for son in sons:
out += """
""" % \
{ 'narrowsearchbox': {'r': 'narrowsearchbox',
'v': 'focusonsearchbox'}[type]}
if type == 'r':
if son.restricted_p() and son.restricted_p() != father.restricted_p():
out += """
""" % {'name' : cgi.escape(son.name) }
# hosted collections are checked by default only when configured so
elif str(son.dbquery).startswith("hostedcollection:"):
external_collection_engine = get_external_collection_engine(str(son.name))
if external_collection_engine and external_collection_engine.selected_by_default:
out += """""" % {'name' : cgi.escape(son.name) }
elif external_collection_engine and not external_collection_engine.selected_by_default:
out += """""" % {'name' : cgi.escape(son.name) }
else:
# strangely, the external collection engine was never found. In that case,
# why was the hosted collection here in the first place?
out += """""" % {'name' : cgi.escape(son.name) }
else:
out += """""" % {'name' : cgi.escape(son.name) }
else:
out += ''
out += """
%(link)s%(recs)s """ % {
'link': create_html_link(self.build_search_interface_url(c=son.name, ln=ln, aas=aas),
{}, style_prolog + cgi.escape(son.get_name(ln)) + style_epilog),
'recs' : self.tmpl_nbrecs_info(son.nbrecs, ln=ln)}
# the following prints the "external collection" arrow just after the name and
# number of records of the hosted collection
# 1) we might want to make the arrow work as an anchor to the hosted collection as well.
# That would probably require a new separate function under invenio.urlutils
# 2) we might want to place the arrow between the name and the number of records of the hosted collection
# That would require to edit/separate the above out += ...
if type == 'r':
if str(son.dbquery).startswith("hostedcollection:"):
out += """""" % \
{ 'siteurl' : CFG_SITE_URL, 'name' : cgi.escape(son.name), }
if son.restricted_p():
out += """ [%(msg)s] """ % { 'msg' : _("restricted") }
if display_grandsons and len(grandsons[i]):
# iterate trough grandsons:
out += """ """
for grandson in grandsons[i]:
out += """ %(link)s%(nbrec)s """ % {
'link': create_html_link(self.build_search_interface_url(c=grandson.name, ln=ln, aas=aas),
{},
cgi.escape(grandson.get_name(ln))),
'nbrec' : self.tmpl_nbrecs_info(grandson.nbrecs, ln=ln)}
# the following prints the "external collection" arrow just after the name and
# number of records of the hosted collection
# Some relatives comments have been made just above
if type == 'r':
if str(grandson.dbquery).startswith("hostedcollection:"):
out += """""" % \
{ 'siteurl' : CFG_SITE_URL, 'name' : cgi.escape(grandson.name), }
out += """
"""
i += 1
out += "
"
return out
def tmpl_searchalso(self, ln, engines_list, collection_id):
_ = gettext_set_language(ln)
box_name = _("Search also:")
html = """
%(box_name)s
""" % locals()
for engine in engines_list:
internal_name = engine.name
name = _(internal_name)
base_url = engine.base_url
if external_collection_get_state(engine, collection_id) == 3:
checked = ' checked="checked"'
else:
checked = ''
html += """
"""
return html
def tmpl_nbrecs_info(self, number, prolog=None, epilog=None, ln=CFG_SITE_LANG):
"""
Return information on the number of records.
Parameters:
- 'number' *string* - The number of records
- 'prolog' *string* (optional) - An HTML code to prefix the number (if **None**, will be
'(')
- 'epilog' *string* (optional) - An HTML code to append to the number (if **None**, will be
')')
"""
if number is None:
number = 0
if prolog is None:
prolog = ''' ('''
if epilog is None:
epilog = ''')'''
return prolog + self.tmpl_nice_number(number, ln) + epilog
def tmpl_box_restricted_content(self, ln):
"""
Displays a box containing a *restricted content* message
Parameters:
- 'ln' *string* - The language to display
"""
# load the right message language
_ = gettext_set_language(ln)
return _("This collection is restricted. If you are authorized to access it, please click on the Search button.")
def tmpl_box_hosted_collection(self, ln):
"""
Displays a box containing a *hosted collection* message
Parameters:
- 'ln' *string* - The language to display
"""
# load the right message language
_ = gettext_set_language(ln)
return _("This is a hosted external collection. Please click on the Search button to see its latest additions.")
def tmpl_box_no_records(self, ln):
"""
Displays a box containing a *no content* message
Parameters:
- 'ln' *string* - The language to display
"""
# load the right message language
_ = gettext_set_language(ln)
return _("This collection does not contain any document yet.")
def tmpl_instant_browse(self, aas, ln, recids, more_link = None):
"""
Formats a list of records (given in the recids list) from the database.
Parameters:
- 'aas' *int* - Advanced Search interface or not (0 or 1)
- 'ln' *string* - The language to display
- 'recids' *list* - the list of records from the database
- 'more_link' *string* - the "More..." link for the record. If not given, will not be displayed
"""
# load the right message language
_ = gettext_set_language(ln)
body = '''
''' % {'header' : _("Latest additions:"),
'body' : body,
}
def tmpl_searchwithin_select(self, ln, fieldname, selected, values):
"""
Produces 'search within' selection box for the current collection.
Parameters:
- 'ln' *string* - The language to display
- 'fieldname' *string* - the name of the select box produced
- 'selected' *string* - which of the values is selected
- 'values' *list* - the list of values in the select
"""
out = '"""
return out
def tmpl_select(self, fieldname, values, selected=None, css_class=''):
"""
Produces a generic select box
Parameters:
- 'css_class' *string* - optional, a css class to display this select with
- 'fieldname' *list* - the name of the select box produced
- 'selected' *string* - which of the values is selected
- 'values' *list* - the list of values in the select
"""
if css_class != '':
class_field = ' class="%s"' % css_class
else:
class_field = ''
out = '"""
return out
def tmpl_record_links(self, recid, ln):
"""
Displays the *More info* and *Find similar* links for a record
Parameters:
- 'ln' *string* - The language to display
- 'recid' *string* - the id of the displayed record
"""
# load the right message language
_ = gettext_set_language(ln)
out = ''' %(detailed)s - %(similar)s''' % {
'detailed': create_html_link(self.build_search_url(recid=recid, ln=ln),
{},
_("Detailed record"), {'class': "moreinfo"}),
'similar': create_html_link(self.build_search_url(p="recid:%d" % recid, rm='wrd', ln=ln),
{},
_("Similar records"),
{'class': "moreinfo"})}
if CFG_BIBRANK_SHOW_CITATION_LINKS:
num_timescited = get_cited_by_count(recid)
if num_timescited:
out += ''' - %s ''' % \
create_html_link(self.build_search_url(p='recid:%d' % recid, rm='citation', ln=ln),
{}, _("Cited by %i records") % num_timescited, {'class': "moreinfo"})
return out
def tmpl_record_body(self, titles, authors, dates, rns, abstracts, urls_u, urls_z, ln):
"""
Displays the "HTML basic" format of a record
Parameters:
- 'authors' *list* - the authors (as strings)
- 'dates' *list* - the dates of publication
- 'rns' *list* - the quicknotes for the record
- 'abstracts' *list* - the abstracts for the record
- 'urls_u' *list* - URLs to the original versions of the record
- 'urls_z' *list* - Not used
"""
out = ""
for title in titles:
out += "%(title)s " % {
'title' : cgi.escape(title)
}
if authors:
out += " / "
for author in authors[:CFG_WEBSEARCH_AUTHOR_ET_AL_THRESHOLD]:
out += '%s; ' % \
create_html_link(self.build_search_url(p=author, f='author', ln=ln),
{}, cgi.escape(author))
if len(authors) > CFG_WEBSEARCH_AUTHOR_ET_AL_THRESHOLD:
out += "et al"
for date in dates:
out += " %s." % cgi.escape(date)
for rn in rns:
out += """ [%(rn)s]""" % {'rn' : cgi.escape(rn)}
for abstract in abstracts:
out += " %(abstract)s [...]" % {'abstract' : cgi.escape(abstract[:1+string.find(abstract, '.')]) }
for idx in range(0, len(urls_u)):
out += """ %(name)s""" % {
'url' : urls_u[idx],
'name' : urls_u[idx]
}
return out
def tmpl_search_in_bibwords(self, p, f, ln, nearest_box):
"""
Displays the *Words like current ones* links for a search
Parameters:
- 'p' *string* - Current search words
- 'f' *string* - the fields in which the search was done
- 'nearest_box' *string* - the HTML code for the "nearest_terms" box - most probably from a create_nearest_terms_box call
"""
# load the right message language
_ = gettext_set_language(ln)
out = '
'
if f:
out += _("Words nearest to %(x_word)s inside %(x_field)s in any collection are:") % {'x_word': '' + cgi.escape(p) + '',
'x_field': '' + cgi.escape(f) + ''}
else:
out += _("Words nearest to %(x_word)s in any collection are:") % {'x_word': '' + cgi.escape(p) + ''}
out += ' ' + nearest_box + '
'
return out
def tmpl_nearest_term_box(self, p, ln, f, terminfo, intro):
"""
Displays the *Nearest search terms* box
Parameters:
- 'p' *string* - Current search words
- 'f' *string* - a collection description (if the search has been completed in a collection)
- 'ln' *string* - The language to display
- 'terminfo': tuple (term, hits, argd) for each near term
- 'intro' *string* - the intro HTML to prefix the box with
"""
out = '''
'''
for term, hits, argd in terminfo:
if hits:
hitsinfo = str(hits)
else:
hitsinfo = '-'
term = cgi.escape(term)
if term == p: # print search word for orientation:
nearesttermsboxbody_class = "nearesttermsboxbodyselected"
if hits > 0:
term = create_html_link(self.build_search_url(argd), {},
term, {'class': "nearesttermsselected"})
else:
nearesttermsboxbody_class = "nearesttermsboxbody"
term = create_html_link(self.build_search_url(argd), {},
term, {'class': "nearestterms"})
out += '''\
"
def tmpl_browse_pattern(self, f, fn, ln, browsed_phrases_in_colls, colls, rg):
"""
Displays the *Nearest search terms* box
Parameters:
- 'f' *string* - field (*not* i18nized)
- 'fn' *string* - field name (i18nized)
- 'ln' *string* - The language to display
- 'browsed_phrases_in_colls' *array* - the phrases to display
- 'colls' *array* - the list of collection parameters of the search (c's)
- 'rg' *int* - the number of records
"""
# load the right message language
_ = gettext_set_language(ln)
out = """
%(hits)s
%(fn)s
""" % {
'hits' : _("Hits"),
'fn' : cgi.escape(fn)
}
if len(browsed_phrases_in_colls) == 1:
# one hit only found:
phrase, nbhits = browsed_phrases_in_colls[0][0], browsed_phrases_in_colls[0][1]
query = {'c': colls,
'ln': ln,
'p': '"%s"' % phrase.replace('"', '\\"'),
'f': f,
'rg' : rg}
out += """
%(nbhits)s
%(link)s
""" % {'nbhits': nbhits,
'link': create_html_link(self.build_search_url(query),
{}, cgi.escape(phrase))}
elif len(browsed_phrases_in_colls) > 1:
# first display what was found but the last one:
for phrase, nbhits in browsed_phrases_in_colls[:-1]:
query = {'c': colls,
'ln': ln,
'p': '"%s"' % phrase.replace('"', '\\"'),
'f': f,
'rg' : rg}
out += """
%(nbhits)s
%(link)s
""" % {'nbhits' : nbhits,
'link': create_html_link(self.build_search_url(query),
{},
cgi.escape(phrase))}
# now display last hit as "previous term":
phrase, nbhits = browsed_phrases_in_colls[0]
query_previous = {'c': colls,
'ln': ln,
'p': '"%s"' % phrase.replace('"', '\\"'),
'f': f,
'rg' : rg}
# now display last hit as "next term":
phrase, nbhits = browsed_phrases_in_colls[-1]
query_next = {'c': colls,
'ln': ln,
'p': '"%s"' % phrase.replace('"', '\\"'),
'f': f,
'rg' : rg}
out += """
"""
return out
def tmpl_search_box(self, ln, aas, cc, cc_intl, ot, sp,
action, fieldslist, f1, f2, f3, m1, m2, m3,
p1, p2, p3, op1, op2, rm, p, f, coll_selects,
d1y, d2y, d1m, d2m, d1d, d2d, dt, sort_fields,
sf, so, ranks, sc, rg, formats, of, pl, jrec, ec,
show_colls=True, show_title=True):
"""
Displays the *Nearest search terms* box
Parameters:
- 'ln' *string* - The language to display
- 'aas' *bool* - Should we display an advanced search box? -1 -> 1, from simpler to more advanced
- 'cc_intl' *string* - the i18nized current collection name, used for display
- 'cc' *string* - the internal current collection name
- 'ot', 'sp' *string* - hidden values
- 'action' *string* - the action demanded by the user
- 'fieldslist' *list* - the list of all fields available, for use in select within boxes in advanced search
- 'p, f, f1, f2, f3, m1, m2, m3, p1, p2, p3, op1, op2, op3, rm' *strings* - the search parameters
- 'coll_selects' *array* - a list of lists, each containing the collections selects to display
- 'd1y, d2y, d1m, d2m, d1d, d2d' *int* - the search between dates
- 'dt' *string* - the dates' types (creation dates, modification dates)
- 'sort_fields' *array* - the select information for the sort fields
- 'sf' *string* - the currently selected sort field
- 'so' *string* - the currently selected sort order ("a" or "d")
- 'ranks' *array* - ranking methods
- 'rm' *string* - selected ranking method
- 'sc' *string* - split by collection or not
- 'rg' *string* - selected results/page
- 'formats' *array* - available output formats
- 'of' *string* - the selected output format
- 'pl' *string* - `limit to' search pattern
- show_colls *bool* - propose coll selection box?
- show_title *bool* show cc_intl in page title?
"""
# load the right message language
_ = gettext_set_language(ln)
# These are hidden fields the user does not manipulate
# directly
if aas == -1:
argd = drop_default_urlargd({
'ln': ln, 'aas': aas,
'ot': ot, 'sp': sp, 'ec': ec,
}, self.search_results_default_urlargd)
else:
argd = drop_default_urlargd({
'cc': cc, 'ln': ln, 'aas': aas,
'ot': ot, 'sp': sp, 'ec': ec,
}, self.search_results_default_urlargd)
out = ""
if show_title:
# display cc name if asked for
out += '''
%(ccname)s
''' % {'ccname' : cgi.escape(cc_intl),}
out += '''
"""
return out
def tmpl_input_hidden(self, name, value):
"Produces the HTML code for a hidden field "
if isinstance(value, list):
list_input = [self.tmpl_input_hidden(name, val) for val in value]
return "\n".join(list_input)
# # Treat `as', `aas' arguments specially:
if name == 'aas':
name = 'as'
return """""" % {
'name' : cgi.escape(str(name), 1),
'value' : cgi.escape(str(value), 1),
}
def _add_mark_to_field(self, value, fields, ln, chars=1):
"""Adds the current value as a MARC tag in the fields array
Useful for advanced search"""
# load the right message language
_ = gettext_set_language(ln)
out = fields
if value and str(value[0:chars]).isdigit():
out.append({'value' : value,
'text' : str(value) + " " + _("MARC tag")
})
return out
def tmpl_search_pagestart(self, ln) :
"page start for search page. Will display after the page header"
return """
"""
def tmpl_search_pageend(self, ln) :
"page end for search page. Will display just before the page footer"
return """
"""
def tmpl_print_warning(self, msg, type, prologue, epilogue):
"""Prints warning message and flushes output.
Parameters:
- 'msg' *string* - The message string
- 'type' *string* - the warning type
- 'prologue' *string* - HTML code to display before the warning
- 'epilogue' *string* - HTML code to display after the warning
"""
out = '\n%s' % (prologue)
if type:
out += '%s: ' % type
out += '%s%s' % (msg, epilogue)
return out
def tmpl_print_search_info(self, ln, middle_only,
collection, collection_name, collection_id,
aas, sf, so, rm, rg, nb_found, of, ot, p, f, f1,
f2, f3, m1, m2, m3, op1, op2, p1, p2,
p3, d1y, d1m, d1d, d2y, d2m, d2d, dt,
all_fieldcodes, cpu_time, pl_in_url,
jrec, sc, sp):
"""Prints stripe with the information on 'collection' and 'nb_found' results and CPU time.
Also, prints navigation links (beg/next/prev/end) inside the results set.
If middle_only is set to 1, it will only print the middle box information (beg/netx/prev/end/etc) links.
This is suitable for displaying navigation links at the bottom of the search results page.
Parameters:
- 'ln' *string* - The language to display
- 'middle_only' *bool* - Only display parts of the interface
- 'collection' *string* - the collection name
- 'collection_name' *string* - the i18nized current collection name
- 'aas' *bool* - if we display the advanced search interface
- 'sf' *string* - the currently selected sort format
- 'so' *string* - the currently selected sort order ("a" or "d")
- 'rm' *string* - selected ranking method
- 'rg' *int* - selected results/page
- 'nb_found' *int* - number of results found
- 'of' *string* - the selected output format
- 'ot' *string* - hidden values
- 'p' *string* - Current search words
- 'f' *string* - the fields in which the search was done
- 'f1, f2, f3, m1, m2, m3, p1, p2, p3, op1, op2' *strings* - the search parameters
- 'jrec' *int* - number of first record on this page
- 'd1y, d2y, d1m, d2m, d1d, d2d' *int* - the search between dates
- 'dt' *string* the dates' type (creation date, modification date)
- 'all_fieldcodes' *array* - all the available fields
- 'cpu_time' *float* - the time of the query in seconds
"""
# load the right message language
_ = gettext_set_language(ln)
out = ""
# left table cells: print collection name
if not middle_only:
out += '''
"
out += ""
return out
def tmpl_print_hosted_search_info(self, ln, middle_only,
collection, collection_name, collection_id,
aas, sf, so, rm, rg, nb_found, of, ot, p, f, f1,
f2, f3, m1, m2, m3, op1, op2, p1, p2,
p3, d1y, d1m, d1d, d2y, d2m, d2d, dt,
all_fieldcodes, cpu_time, pl_in_url,
jrec, sc, sp):
"""Prints stripe with the information on 'collection' and 'nb_found' results and CPU time.
Also, prints navigation links (beg/next/prev/end) inside the results set.
If middle_only is set to 1, it will only print the middle box information (beg/netx/prev/end/etc) links.
This is suitable for displaying navigation links at the bottom of the search results page.
Parameters:
- 'ln' *string* - The language to display
- 'middle_only' *bool* - Only display parts of the interface
- 'collection' *string* - the collection name
- 'collection_name' *string* - the i18nized current collection name
- 'aas' *bool* - if we display the advanced search interface
- 'sf' *string* - the currently selected sort format
- 'so' *string* - the currently selected sort order ("a" or "d")
- 'rm' *string* - selected ranking method
- 'rg' *int* - selected results/page
- 'nb_found' *int* - number of results found
- 'of' *string* - the selected output format
- 'ot' *string* - hidden values
- 'p' *string* - Current search words
- 'f' *string* - the fields in which the search was done
- 'f1, f2, f3, m1, m2, m3, p1, p2, p3, op1, op2' *strings* - the search parameters
- 'jrec' *int* - number of first record on this page
- 'd1y, d2y, d1m, d2m, d1d, d2d' *int* - the search between dates
- 'dt' *string* the dates' type (creation date, modification date)
- 'all_fieldcodes' *array* - all the available fields
- 'cpu_time' *float* - the time of the query in seconds
"""
# load the right message language
_ = gettext_set_language(ln)
out = ""
# left table cells: print collection name
if not middle_only:
out += '''
""" % { 'siteurl' : CFG_SITE_URL }
# middle table cell: print beg/next/prev/end arrows:
if not middle_only:
# in case we have a hosted collection that timed out do not print its number of records, as it is yet unknown
if nb_found != -963:
out += """
# %(recs_found)s """ % {
# 'recs_found' : _("%s records found") % ('' + self.tmpl_nice_number(nb_found, ln) + '')
# }
else:
out += ""
# we do not care about timed out hosted collections here, because the bumber of records found will never be bigger
# than rg anyway, since it's negative
if nb_found > rg:
out += "" + cgi.escape(collection_name) + " : " + _("%s records found") % ('' + self.tmpl_nice_number(nb_found, ln) + '') + " "
if nb_found > rg: # navig.arrows are needed, since we have many hits
query = {'p': p, 'f': f,
'cc': collection,
'sf': sf, 'so': so,
'sp': sp, 'rm': rm,
'of': of, 'ot': ot,
'aas': aas, 'ln': ln,
'p1': p1, 'p2': p2, 'p3': p3,
'f1': f1, 'f2': f2, 'f3': f3,
'm1': m1, 'm2': m2, 'm3': m3,
'op1': op1, 'op2': op2,
'sc': 0,
'd1y': d1y, 'd1m': d1m, 'd1d': d1d,
'd2y': d2y, 'd2m': d2m, 'd2d': d2d,
'dt': dt,
}
# @todo here
def img(gif, txt):
return '' % {
'txt': txt, 'gif': gif, 'siteurl': CFG_SITE_URL}
if jrec-rg > 1:
out += create_html_link(self.build_search_url(query, jrec=1, rg=rg),
{}, img('sb', _("begin")),
{'class': 'img'})
if jrec > 1:
out += create_html_link(self.build_search_url(query, jrec=max(jrec-rg, 1), rg=rg),
{}, img('sp', _("previous")),
{'class': 'img'})
if jrec+rg-1 < nb_found:
out += "%d - %d" % (jrec, jrec+rg-1)
else:
out += "%d - %d" % (jrec, nb_found)
if nb_found >= jrec+rg:
out += create_html_link(self.build_search_url(query,
jrec=jrec+rg,
rg=rg),
{}, img('sn', _("next")),
{'class':'img'})
if nb_found >= jrec+rg+rg:
out += create_html_link(self.build_search_url(query,
jrec=nb_found-rg+1,
rg=rg),
{}, img('se', _("end")),
{'class': 'img'})
# still in the navigation part
cc = collection
sc = 0
for var in ['p', 'cc', 'f', 'sf', 'so', 'of', 'rg', 'aas', 'ln', 'p1', 'p2', 'p3', 'f1', 'f2', 'f3', 'm1', 'm2', 'm3', 'op1', 'op2', 'sc', 'd1y', 'd1m', 'd1d', 'd2y', 'd2m', 'd2d', 'dt']:
out += self.tmpl_input_hidden(name = var, value = vars()[var])
for var in ['ot', 'sp', 'rm']:
if vars()[var]:
out += self.tmpl_input_hidden(name = var, value = vars()[var])
if pl_in_url:
fieldargs = cgi.parse_qs(pl_in_url)
for fieldcode in all_fieldcodes:
# get_fieldcodes():
if fieldargs.has_key(fieldcode):
for val in fieldargs[fieldcode]:
out += self.tmpl_input_hidden(name = fieldcode, value = val)
out += """ %(jump)s """ % {
'jump' : _("jump to record:"),
'jrec' : jrec,
}
if not middle_only:
out += "
"
else:
out += ""
# right table cell: cpu time info
if not middle_only:
if cpu_time > -1:
out += """
%(time)s
""" % {
'time' : _("Search took %s seconds.") % ('%.2f' % cpu_time),
}
out += "
"
else:
out += ""
out += "
"
return out
def tmpl_nice_number(self, number, ln=CFG_SITE_LANG, thousands_separator=',', max_ndigits_after_dot=None):
"""
Return nicely printed number NUMBER in language LN using
given THOUSANDS_SEPARATOR character.
If max_ndigits_after_dot is specified and the number is float, the
number is rounded by taking in consideration up to max_ndigits_after_dot
digit after the dot.
This version does not pay attention to locale. See
tmpl_nice_number_via_locale().
"""
if type(number) is float:
if max_ndigits_after_dot is not None:
number = round(number, max_ndigits_after_dot)
int_part, frac_part = str(number).split('.')
return '%s.%s' % (self.tmpl_nice_number(int(int_part), ln, thousands_separator), frac_part)
else:
chars_in = list(str(number))
number = len(chars_in)
chars_out = []
for i in range(0, number):
if i % 3 == 0 and i != 0:
chars_out.append(thousands_separator)
chars_out.append(chars_in[number-i-1])
chars_out.reverse()
return ''.join(chars_out)
def tmpl_nice_number_via_locale(self, number, ln=CFG_SITE_LANG):
"""
Return nicely printed number NUM in language LN using the locale.
See also version tmpl_nice_number().
"""
if number is None:
return None
# Temporarily switch the numeric locale to the requested one, and format the number
# In case the system has no locale definition, use the vanilla form
ol = locale.getlocale(locale.LC_NUMERIC)
try:
locale.setlocale(locale.LC_NUMERIC, self.tmpl_localemap.get(ln, self.tmpl_default_locale))
except locale.Error:
return str(number)
try:
number = locale.format('%d', number, True)
except TypeError:
return str(number)
locale.setlocale(locale.LC_NUMERIC, ol)
return number
def tmpl_record_format_htmlbrief_header(self, ln):
"""Returns the header of the search results list when output
is html brief. Note that this function is called for each collection
results when 'split by collection' is enabled.
See also: tmpl_record_format_htmlbrief_footer(..),
tmpl_record_format_htmlbrief_body(..)
Parameters:
- 'ln' *string* - The language to display
"""
# load the right message language
_ = gettext_set_language(ln)
out = """
""" % {
'siteurl' : CFG_SITE_URL,
}
return out
def tmpl_record_format_htmlbrief_footer(self, ln):
"""Returns the footer of the search results list when output
is html brief. Note that this function is called for each collection
results when 'split by collection' is enabled.
See also: tmpl_record_format_htmlbrief_header(..),
tmpl_record_format_htmlbrief_body(..)
Parameters:
- 'ln' *string* - The language to display
"""
# load the right message language
_ = gettext_set_language(ln)
out = """
""" % {
'basket' : _("ADD TO BASKET")
}
return out
def tmpl_record_format_htmlbrief_body(self, ln, recid,
row_number, relevance,
record, relevances_prologue,
relevances_epilogue):
"""Returns the html brief format of one record. Used in the
search results list for each record.
See also: tmpl_record_format_htmlbrief_header(..),
tmpl_record_format_htmlbrief_footer(..)
Parameters:
- 'ln' *string* - The language to display
- 'row_number' *int* - The position of this record in the list
- 'recid' *int* - The recID
- 'relevance' *string* - The relevance of the record
- 'record' *string* - The formatted record
- 'relevances_prologue' *string* - HTML code to prepend the relevance indicator
- 'relevances_epilogue' *string* - HTML code to append to the relevance indicator (used mostly for formatting)
"""
# load the right message language
_ = gettext_set_language(ln)
out = """
%(number)s.
""" % {'recid': recid,
'number': row_number}
if relevance:
out += """
""" % record
return out
def tmpl_print_results_overview(self, ln, results_final_nb_total, cpu_time, results_final_nb, colls, ec, hosted_colls_potential_results_p=False):
"""Prints results overview box with links to particular collections below.
Parameters:
- 'ln' *string* - The language to display
- 'results_final_nb_total' *int* - The total number of hits for the query
- 'colls' *array* - The collections with hits, in the format:
- 'coll[code]' *string* - The code of the collection (canonical name)
- 'coll[name]' *string* - The display name of the collection
- 'results_final_nb' *array* - The number of hits, indexed by the collection codes:
- 'cpu_time' *string* - The time the query took
- 'url_args' *string* - The rest of the search query
- 'ec' *array* - selected external collections
- 'hosted_colls_potential_results_p' *boolean* - check if there are any hosted collections searches
that timed out during the pre-search
"""
if len(colls) == 1 and not ec:
# if one collection only and no external collections, print nothing:
return ""
# load the right message language
_ = gettext_set_language(ln)
# first find total number of hits:
# if there were no hosted collections that timed out during the pre-search print out the exact number of records found
if not hosted_colls_potential_results_p:
out = """
%(founds)s
""" % {
'founds' : _("%(x_fmt_open)sResults overview:%(x_fmt_close)s Found %(x_nb_records)s records in %(x_nb_seconds)s seconds.") %\
{'x_fmt_open': '',
'x_fmt_close': '',
'x_nb_records': '' + self.tmpl_nice_number(results_final_nb_total, ln) + '',
'x_nb_seconds': '%.2f' % cpu_time}
}
# if there were (only) hosted_collections that timed out during the pre-search print out a fuzzier message
else:
if results_final_nb_total == 0:
out = """
""" % {
'founds' : _("%(x_fmt_open)sResults overview:%(x_fmt_close)s Found at least %(x_nb_records)s records in %(x_nb_seconds)s seconds.") %\
{'x_fmt_open': '',
'x_fmt_close': '',
'x_nb_records': '' + self.tmpl_nice_number(results_final_nb_total, ln) + '',
'x_nb_seconds': '%.2f' % cpu_time}
}
# then print hits per collection:
for coll in colls:
if results_final_nb.has_key(coll['code']) and results_final_nb[coll['code']] > 0:
out += '''%(coll_name)s,
%(number)s ''' % {
'coll' : coll['id'],
'coll_name' : cgi.escape(coll['name']),
'number' : _("%s records found") % ('' + self.tmpl_nice_number(results_final_nb[coll['code']], ln) + '')
}
# the following is used for hosted collections that have timed out,
# i.e. for which we don't know the exact number of results yet.
elif results_final_nb.has_key(coll['code']) and results_final_nb[coll['code']] == -963:
out += '''%(coll_name)s''' % {
'coll' : coll['id'],
'coll_name' : cgi.escape(coll['name']),
'number' : _("%s records found") % ('' + self.tmpl_nice_number(results_final_nb[coll['code']], ln) + '')
}
out += "
"
return out
- def tmpl_print_hosted_results(self, ln, infos, of=None, req=None):
+ def tmpl_print_hosted_results(self, url_and_engine, ln, of=None, req=None, limit=CFG_EXTERNAL_COLLECTION_MAXRESULTS):
"""Print results of a given search engine.
"""
_ = gettext_set_language(ln)
- url = infos[0]
- engine = infos[1]
- name = _(engine.name)
+ #url = url_and_engine[0]
+ engine = url_and_engine[1]
+ #name = _(engine.name)
db_id = get_collection_id(engine.name)
- base_url = engine.base_url
+ #base_url = engine.base_url
out = ""
- results = engine.parser.parse_and_get_results(None, of=of, req=req, parseonly=True)
+ results = engine.parser.parse_and_get_results(None, of=of, req=req, limit=limit, parseonly=True)
if len(results) != 0:
if of == 'hb':
out += """
""" % {
'siteurl' : CFG_SITE_URL,
'col_db_id' : db_id,
}
else:
if of == 'hb':
out += """
"""
for result in results:
out += result.html
if len(results) != 0:
if of == 'hb':
out += """
""" % {
'basket' : _("ADD TO BASKET")
}
else:
if of == 'hb':
out += """
"""
# we have already checked if there are results or no, maybe the following if should be removed?
if not results:
if of.startswith("h"):
out = _('No results found...') + ' '
return out
def tmpl_print_searchresultbox(self, header, body):
"""print a nicely formatted box for search results """
#_ = gettext_set_language(ln)
# first find total number of hits:
out = '
'+header+'
'+body+'
'
return out
def tmpl_search_no_boolean_hits(self, ln, nearestterms):
"""No hits found, proposes alternative boolean queries
Parameters:
- 'ln' *string* - The language to display
- 'nearestterms' *array* - Parts of the interface to display, in the format:
- 'nearestterms[nbhits]' *int* - The resulting number of hits
- 'nearestterms[url_args]' *string* - The search parameters
- 'nearestterms[p]' *string* - The search terms
"""
# load the right message language
_ = gettext_set_language(ln)
out = _("Boolean query returned no hits. Please combine your search terms differently.")
out += '''
'''
for term, hits, argd in nearestterms:
out += '''\
"""
return out
def tmpl_similar_author_names(self, authors, ln):
"""No hits found, proposes alternative boolean queries
Parameters:
- 'authors': a list of (name, hits) tuples
- 'ln' *string* - The language to display
"""
# load the right message language
_ = gettext_set_language(ln)
out = '''
%(similar)s
''' % {
'similar' : _("See also: similar author names")
}
for author, hits in authors:
out += '''\
"""
return out
def tmpl_print_record_detailed(self, recID, ln):
"""Displays a detailed on-the-fly record
Parameters:
- 'ln' *string* - The language to display
- 'recID' *int* - The record id
"""
# okay, need to construct a simple "Detailed record" format of our own:
out = "
"
# secondly, title:
titles = get_fieldvalues(recID, "245__a")
for title in titles:
out += "
%s
" % cgi.escape(title)
# thirdly, authors:
authors = get_fieldvalues(recID, "100__a") + get_fieldvalues(recID, "700__a")
if authors:
out += "
"
for author in authors:
out += '%s; ' % create_html_link(self.build_search_url(
ln=ln,
p=author,
f='author'),
{}, cgi.escape(author))
out += "
"
# fourthly, date of creation:
dates = get_fieldvalues(recID, "260__c")
for date in dates:
out += "
%s
" % date
# fifthly, abstract:
abstracts = get_fieldvalues(recID, "520__a")
for abstract in abstracts:
out += """
Abstract: %s
""" % abstract
# fifthly bis, keywords:
keywords = get_fieldvalues(recID, "6531_a")
if len(keywords):
out += """
Keyword(s):"""
for keyword in keywords:
out += '%s; ' % create_html_link(
self.build_search_url(ln=ln,
p=keyword,
f='keyword'),
{}, cgi.escape(keyword))
out += '
'
# fifthly bis bis, published in:
prs_p = get_fieldvalues(recID, "909C4p")
prs_v = get_fieldvalues(recID, "909C4v")
prs_y = get_fieldvalues(recID, "909C4y")
prs_n = get_fieldvalues(recID, "909C4n")
prs_c = get_fieldvalues(recID, "909C4c")
for idx in range(0, len(prs_p)):
out += """
Publ. in: %s""" % prs_p[idx]
if prs_v and prs_v[idx]:
out += """%s""" % prs_v[idx]
if prs_y and prs_y[idx]:
out += """(%s)""" % prs_y[idx]
if prs_n and prs_n[idx]:
out += """, no.%s""" % prs_n[idx]
if prs_c and prs_c[idx]:
out += """, p.%s""" % prs_c[idx]
out += """.
"""
# sixthly, fulltext link:
urls_z = get_fieldvalues(recID, "8564_z")
urls_u = get_fieldvalues(recID, "8564_u")
for idx in range(0, len(urls_u)):
link_text = "URL"
try:
if urls_z[idx]:
link_text = urls_z[idx]
except IndexError:
pass
out += """
""" % (link_text, urls_u[idx], urls_u[idx])
# print some white space at the end:
out += "
"
return out
def tmpl_print_record_list_for_similarity_boxen(self, title, recID_score_list, ln=CFG_SITE_LANG):
"""Print list of records in the "hs" (HTML Similarity) format for similarity boxes.
RECID_SCORE_LIST is a list of (recID1, score1), (recID2, score2), etc.
"""
from invenio.search_engine import print_record, record_public_p
recID_score_list_to_be_printed = []
# firstly find 5 first public records to print:
nb_records_to_be_printed = 0
nb_records_seen = 0
while nb_records_to_be_printed < 5 and nb_records_seen < len(recID_score_list) and nb_records_seen < 50:
# looking through first 50 records only, picking first 5 public ones
(recID, score) = recID_score_list[nb_records_seen]
nb_records_seen += 1
if record_public_p(recID):
nb_records_to_be_printed += 1
recID_score_list_to_be_printed.append([recID, score])
# secondly print them:
out = '''
%(title)s
''' % { 'title': cgi.escape(title) }
for recid, score in recID_score_list_to_be_printed:
out += '''
"""
return out
def tmpl_print_record_brief(self, ln, recID):
"""Displays a brief record on-the-fly
Parameters:
- 'ln' *string* - The language to display
- 'recID' *int* - The record id
"""
out = ""
# record 'recID' does not exist in format 'format', so print some default format:
# firstly, title:
titles = get_fieldvalues(recID, "245__a")
# secondly, authors:
authors = get_fieldvalues(recID, "100__a") + get_fieldvalues(recID, "700__a")
# thirdly, date of creation:
dates = get_fieldvalues(recID, "260__c")
# thirdly bis, report numbers:
rns = get_fieldvalues(recID, "037__a")
rns = get_fieldvalues(recID, "088__a")
# fourthly, beginning of abstract:
abstracts = get_fieldvalues(recID, "520__a")
# fifthly, fulltext link:
urls_z = get_fieldvalues(recID, "8564_z")
urls_u = get_fieldvalues(recID, "8564_u")
## unAPI identifier
out = '\n' % recID
out += self.tmpl_record_body(
titles = titles,
authors = authors,
dates = dates,
rns = rns,
abstracts = abstracts,
urls_u = urls_u,
urls_z = urls_z,
ln=ln)
return out
def tmpl_print_record_brief_links(self, ln, recID):
"""Displays links for brief record on-the-fly
Parameters:
- 'ln' *string* - The language to display
- 'recID' *int* - The record id
"""
from invenio.webcommentadminlib import get_nb_reviews, get_nb_comments
# load the right message language
_ = gettext_set_language(ln)
out = '
'
if CFG_WEBSEARCH_USE_ALEPH_SYSNOS:
alephsysnos = get_fieldvalues(recID, "970__a")
if len(alephsysnos)>0:
alephsysno = alephsysnos[0]
out += '%s' % \
create_html_link(self.build_search_url(recid=alephsysno,
ln=ln),
{}, _("Detailed record"),
{'class': "moreinfo"})
else:
out += '%s' % \
create_html_link(self.build_search_url(recid=recID, ln=ln),
{},
_("Detailed record"),
{'class': "moreinfo"})
else:
out += '%s' % \
create_html_link(self.build_search_url(recid=recID, ln=ln),
{}, _("Detailed record"),
{'class': "moreinfo"})
out += ' - %s' % \
create_html_link(self.build_search_url(p="recid:%d" % recID,
rm="wrd",
ln=ln),
{}, _("Similar records"),
{'class': "moreinfo"})
if CFG_BIBRANK_SHOW_CITATION_LINKS:
num_timescited = get_cited_by_count(recID)
if num_timescited:
out += ' - %s' % \
create_html_link(self.build_search_url(p="recid:%d" % recID,
rm="citation",
ln=ln),
{}, num_timescited > 1 and _("Cited by %i records") % num_timescited
or _("Cited by 1 record"),
{'class': "moreinfo"})
else:
out+=""
if CFG_WEBCOMMENT_ALLOW_COMMENTS and CFG_WEBSEARCH_SHOW_COMMENT_COUNT:
num_comments = get_nb_comments(recID)
if num_comments:
out += ' - %s' % \
create_html_link(CFG_SITE_URL + '/record/' + str(recID)
+ '/comments?ln=%s' % ln, {}, num_comments > 1 and _("%i comments")
% (num_comments) or _("1 comment"),
{'class': "moreinfo"})
else:
out+=""
if CFG_WEBCOMMENT_ALLOW_REVIEWS and CFG_WEBSEARCH_SHOW_REVIEW_COUNT:
num_reviews = get_nb_reviews(recID)
if num_reviews:
out += ' - %s' % \
create_html_link(CFG_SITE_URL + '/record/' + str(recID)
+ '/reviews?ln=%s' % ln, {}, num_reviews > 1 and _("%i reviews")
% (num_reviews) or _("1 review"), {'class': "moreinfo"})
else:
out+=""
out+='
'
return out
def tmpl_xml_rss_prologue(self, current_url=None,
previous_url=None, next_url=None,
first_url=None, last_url=None,
nb_found=None, jrec=None, rg=None):
"""Creates XML RSS 2.0 prologue."""
out = """%(sitename)s
%(siteurl)s
%(sitename)s latest documents%(sitelang)s%(timestamp)sCDS Invenio %(version)s%(sitesupportemail)s%(timetolive)s%(previous_link)s%(next_link)s%(current_link)s%(total_results)s%(start_index)s%(total_results)s
%(siteurl)s/img/cds.png%(sitename)s
%(siteurl)s
Search Search this site:p
%(siteurl)s/search
""" % {'sitename': CFG_SITE_NAME,
'siteurl': CFG_SITE_URL,
'sitelang': CFG_SITE_LANG,
'search_syntax': self.tmpl_opensearch_rss_url_syntax,
'timestamp': time.strftime("%a, %d %b %Y %H:%M:%S GMT", time.gmtime()),
'version': CFG_VERSION,
'sitesupportemail': CFG_SITE_SUPPORT_EMAIL,
'timetolive': CFG_WEBSEARCH_RSS_TTL,
'current_link': (current_url and \
'\n\n' % current_url) or '',
'previous_link': (previous_url and \
'\n' % previous_url) or '',
'next_link': (next_url and \
'\n' % next_url) or '',
'first_link': (first_url and \
'\n' % first_url) or '',
'last_link': (last_url and \
'\n' % last_url) or '',
'total_results': (nb_found and \
'\n%i' % nb_found) or '',
'start_index': (jrec and \
'\n%i' % jrec) or '',
'total_results': (rg and \
'\n%i' % rg) or '',
}
return out
def tmpl_xml_rss_epilogue(self):
"""Creates XML RSS 2.0 epilogue."""
out = """\
\n"""
return out
def tmpl_xml_nlm_prologue(self):
"""Creates XML NLM prologue."""
out = """\n"""
return out
def tmpl_xml_nlm_epilogue(self):
"""Creates XML NLM epilogue."""
out = """\n"""
return out
def tmpl_xml_refworks_prologue(self):
"""Creates XML RefWorks prologue."""
out = """\n"""
return out
def tmpl_xml_refworks_epilogue(self):
"""Creates XML RefWorks epilogue."""
out = """\n"""
return out
def tmpl_xml_endnote_prologue(self):
"""Creates XML EndNote prologue."""
out = """\n"""
return out
def tmpl_xml_endnote_epilogue(self):
"""Creates XML EndNote epilogue."""
out = """\n"""
return out
def tmpl_xml_marc_prologue(self):
"""Creates XML MARC prologue."""
out = """\n"""
return out
def tmpl_xml_marc_epilogue(self):
"""Creates XML MARC epilogue."""
out = """\n"""
return out
def tmpl_xml_mods_prologue(self):
"""Creates XML MODS prologue."""
out = """\n"""
return out
def tmpl_xml_mods_epilogue(self):
"""Creates XML MODS epilogue."""
out = """\n"""
return out
def tmpl_xml_default_prologue(self):
"""Creates XML default format prologue. (Sanity calls only.)"""
out = """\n"""
return out
def tmpl_xml_default_epilogue(self):
"""Creates XML default format epilogue. (Sanity calls only.)"""
out = """\n"""
return out
def tmpl_collection_not_found_page_title(self, colname, ln=CFG_SITE_LANG):
"""
Create page title for cases when unexisting collection was asked for.
"""
_ = gettext_set_language(ln)
out = _("Collection %s Not Found") % cgi.escape(colname)
return out
def tmpl_collection_not_found_page_body(self, colname, ln=CFG_SITE_LANG):
"""
Create page body for cases when unexisting collection was asked for.
"""
_ = gettext_set_language(ln)
out = """
%(title)s
%(sorry)s
%(you_may_want)s
""" % { 'title': self.tmpl_collection_not_found_page_title(colname, ln),
'sorry': _("Sorry, collection %s does not seem to exist.") % \
('' + cgi.escape(colname) + ''),
'you_may_want': _("You may want to start browsing from %s.") % \
('' + \
cgi.escape(CFG_SITE_NAME_INTL.get(ln, CFG_SITE_NAME)) + '')}
return out
def tmpl_alert_rss_teaser_box_for_query(self, id_query, ln):
"""Propose teaser for setting up this query as alert or RSS feed.
Parameters:
- 'id_query' *int* - ID of the query we make teaser for
- 'ln' *string* - The language to display
"""
# load the right message language
_ = gettext_set_language(ln)
# get query arguments:
res = run_sql("SELECT urlargs FROM query WHERE id=%s", (id_query,))
argd = {}
if res:
argd = cgi.parse_qs(res[0][0])
rssurl = self.build_rss_url(argd)
alerturl = CFG_SITE_URL + '/youralerts/input?ln=%s&idq=%s' % (ln, id_query)
out = '''
%(similar)s
%(msg_alert)s
''' % {
'similar' : _("Interested in being notified about new results for this query?"),
'msg_alert': _("""Set up a personal %(x_url1_open)semail alert%(x_url1_close)s
or subscribe to the %(x_url2_open)sRSS feed%(x_url2_close)s.""") % \
{'x_url1_open': ' ' % (alerturl, CFG_SITE_URL) + ' ' % (alerturl),
'x_url1_close': '',
'x_url2_open': ' ' % (rssurl, CFG_SITE_URL) + ' ' % rssurl,
'x_url2_close': '',
}}
return out
def tmpl_detailed_record_metadata(self, recID, ln, format,
content,
creationdate=None,
modificationdate=None):
"""Returns the main detailed page of a record
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
- 'format' *string* - The format in used to print the record
- 'content' *string* - The main content of the page
- 'creationdate' *string* - The creation date of the printed record
- 'modificationdate' *string* - The last modification date of the printed record
"""
_ = gettext_set_language(ln)
## unAPI identifier
out = '\n' % recID
out += content
return out
def tmpl_detailed_record_statistics(self, recID, ln,
downloadsimilarity,
downloadhistory, viewsimilarity):
"""Returns the statistics page of a record
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
- downloadsimilarity *string* - downloadsimilarity box
- downloadhistory *string* - downloadhistory box
- viewsimilarity *string* - viewsimilarity box
"""
# load the right message language
_ = gettext_set_language(ln)
out = ''
if CFG_BIBRANK_SHOW_DOWNLOAD_STATS and downloadsimilarity is not None:
similar = self.tmpl_print_record_list_for_similarity_boxen (
_("People who downloaded this document also downloaded:"), downloadsimilarity, ln)
out = '
'
out += ' '
if CFG_BIBRANK_SHOW_READING_STATS and viewsimilarity is not None:
out += self.tmpl_print_record_list_for_similarity_boxen (
_("People who viewed this page also viewed:"), viewsimilarity, ln)
if CFG_BIBRANK_SHOW_DOWNLOAD_GRAPHS and downloadhistory is not None:
out += downloadhistory + ' '
return out
def tmpl_detailed_record_citations_prologue(self, recID, ln):
"""Returns the prologue of the citations page of a record
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
"""
return '
'
def tmpl_detailed_record_citations_epilogue(self, recID, ln):
"""Returns the epilogue of the citations page of a record
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
"""
return '
'
def tmpl_detailed_record_citations_citing_list(self, recID, ln,
citinglist):
"""Returns the list of record citing this one
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
- citinglist *list* - a list of tuples [(x1,y1),(x2,y2),..] where x is doc id and y is number of citations
"""
# load the right message language
_ = gettext_set_language(ln)
out = ''
if CFG_BIBRANK_SHOW_CITATION_STATS and citinglist is not None:
similar = self.tmpl_print_record_list_for_similarity_boxen(
_("Cited by: %s records") % len (citinglist), citinglist, ln)
out += '''
%(similar)s %(more)s
''' % {
'more': create_html_link(
self.build_search_url(p='recid:%d' % \
recID, #XXXX
rm='citation', ln=ln),
{}, _("more")),
'similar': similar}
return out
def tmpl_detailed_record_citations_citation_history(self, recID, ln,
citationhistory):
"""Returns the citations history graph of this record
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
- citationhistory *string* - citationhistory box
"""
# load the right message language
_ = gettext_set_language(ln)
out = ''
if CFG_BIBRANK_SHOW_CITATION_GRAPHS and citationhistory is not None:
out = '
%s
' % citationhistory
else:
out = ""
else:
out+= "no citationhistory -->"
return out
def tmpl_detailed_record_citations_co_citing(self, recID, ln,
cociting):
"""Returns the list of cocited records
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
- cociting *string* - cociting box
"""
# load the right message language
_ = gettext_set_language(ln)
out = ''
if CFG_BIBRANK_SHOW_CITATION_STATS and cociting is not None:
similar = self.tmpl_print_record_list_for_similarity_boxen (
_("Co-cited with: %s records") % len (cociting), cociting, ln)
out = '''
%(similar)s %(more)s
''' % { 'more': create_html_link(self.build_search_url(p='cocitedwith:%d' % recID, ln=ln),
{}, _("more")),
'similar': similar }
return out
def tmpl_detailed_record_citations_self_cited(self, recID, ln,
selfcited, citinglist):
"""Returns the list of self-citations for this record
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
- selfcited list - a list of self-citations for recID
"""
# load the right message language
_ = gettext_set_language(ln)
out = ''
if CFG_BIBRANK_SHOW_CITATION_GRAPHS and selfcited is not None:
sc_scorelist = [] #a score list for print..
for s in selfcited:
#copy weight from citations
weight = 0
for c in citinglist:
(crec, score) = c
if crec == s:
weight = score
tmp = [s, weight]
sc_scorelist.append(tmp)
scite = self.tmpl_print_record_list_for_similarity_boxen (
_(".. of which self-citations: %s records") % len (selfcited), sc_scorelist, ln)
out = '
'+scite+'
'
return out
def tmpl_author_information(self, req, pubs, authorname, num_downloads, aff_pubdict,
citedbylist, kwtuples, authors, vtuples, ln):
"""Prints stuff about the author given as authorname.
1. Author name + his/her institutes. Each institute I has a link
to papers where the auhtor has I as institute.
2. Publications, number: link to search by author.
3. Keywords
4. Author collabs
5. Publication venues like journals
The parameters are data structures needed to produce 1-6, as follows:
req - request
pubs - list of recids, probably the records that have the author as an author
authorname - evident
num_downloads - evident
aff_pubdict - a dictionary where keys are inst names and values lists of recordids
citedbylist - list of recs that cite pubs
kwtuples - keyword tuples like ('HIGGS BOSON',[3,4]) where 3 and 4 are recids
authors - a list of authors that have collaborated with authorname
"""
from invenio.search_engine import perform_request_search
_ = gettext_set_language(ln)
#make a authoraff string that looks like CERN (1), Caltech (2) etc
authoraff = ""
for a in aff_pubdict.keys():
recids = "+or+".join(map(str, aff_pubdict[a]))
print_a = a
if (print_a == ' '):
print_a = _("unknown")
if authoraff:
authoraff += ' '
authoraff += ""+print_a+' ('+str(len(aff_pubdict[a]))+")"
#print a "general" banner about the author
req.write("
" + authorname + "
")
#print affiliations
line1 = "" + _("Affiliations:") + ""
line2 = authoraff
req.write(self.tmpl_print_searchresultbox(line1, line2))
# print frequent keywords:
keywstr = ""
if (kwtuples):
for (kw, freq) in kwtuples:
if keywstr:
keywstr += ' '
#create a link in author=x, keyword=y
searchstr = create_html_link(self.build_search_url(
p='author:"' + authorname + '" ' +
'keyword:"' + kw + '"'),
{}, kw+" ("+str(freq)+")",)
keywstr = keywstr+" "+searchstr
banner = self.tmpl_print_searchresultbox("" + _("Frequent keywords:") + "", keywstr)
req.write(banner)
# print frequent co-authors:
collabstr = ""
if (authors):
for c in authors:
c = c.strip()
if collabstr:
collabstr += ' '
#do not add this person him/herself in the list
if not c == authorname:
commpubs = intbitset(pubs) & intbitset(perform_request_search(p="author:\"%s\" author:\"%s\"" % (authorname, c)))
collabstr = collabstr + " "+c+" ("+str(len(commpubs))+")"
banner = self.tmpl_print_searchresultbox("" + _("Frequent co-authors:") + "", collabstr)
req.write(banner)
# print frequently publishes in journals:
if (vtuples):
pubinfo = ""
for t in vtuples:
(journal, num) = t
pubinfo += create_html_link(self.build_search_url(p='author:"' + authorname + '" ' + \
'journal:"' + journal + '"'),
{}, journal + " ("+str(num)+") ")
banner = self.tmpl_print_searchresultbox("" + _("Frequently publishes in:") + "", pubinfo)
req.write(banner)
# print papers:
searchstr = create_html_link(self.build_search_url(p=authorname,
f='author'),
{}, "All papers ("+str(len(pubs))+")",)
line1 = "" + _("Papers:") + ""
line2 = searchstr
if num_downloads:
line2 + " ("+_("downloaded")+" "
line2 += str(num_downloads)+" "+_("times")+")"
from invenio.search_engine import perform_request_search
if CFG_INSPIRE_SITE:
CFG_COLLS = ['Book',
'Conference',
'Introductory',
'Lectures',
'Preprint',
'Published',
'Report',
'Review',
'Thesis']
else:
CFG_COLLS = ['Article',
'Book',
'Preprint',]
for coll in CFG_COLLS:
collsearch = intbitset(pubs) & intbitset(perform_request_search(p="collection:"+coll))
if len(collsearch) > 0:
num = len(collsearch)
line2 += " " + create_html_link(self.build_search_url(p='author:"' + authorname + '" ' + \
'collection:' + coll),
{}, coll + " ("+str(num)+")",)
banner = self.tmpl_print_searchresultbox(line1, line2)
req.write(banner)
# print citations:
if len(citedbylist):
line1 = "" + _("Citations:") + ""
line2 = ""
req.write(self.tmpl_print_searchresultbox(line1, line2))
# they will be printed after that
def tmpl_detailed_record_references(self, recID, ln, content):
"""Returns the discussion page of a record
Parameters:
- 'recID' *int* - The ID of the printed record
- 'ln' *string* - The language to display
- 'content' *string* - The main content of the page
"""
# load the right message language
_ = gettext_set_language(ln)
out = ''
if content is not None:
out += content
return out
def tmpl_citesummary_prologue(self, d_total_recs, l_colls, searchpattern, searchfield, ln=CFG_SITE_LANG):
"""HTML citesummary format, prologue. A part of HCS format suite."""
_ = gettext_set_language(ln)
out = """
%(msg_title)s
""" % \
{'msg_title': _("Citation summary results"),}
for coll, colldef in l_colls:
out += '
%s
' % coll
out += '
'
out += """
%(msg_recs)s
""" % \
{'msg_recs': _("Total number of citable papers analyzed:"),}
for coll, colldef in l_colls:
link_url = CFG_SITE_URL + '/search?p='
if searchpattern:
p = searchpattern
if searchfield:
if " " in searchpattern:
p = searchfield + ':"' + searchpattern + '"'
else:
p = searchfield + ':' + searchpattern
link_url += quote(p)
if colldef:
link_url += ' ' + quote(colldef)
link_url += '&rm=citation';
link_text = self.tmpl_nice_number(d_total_recs[coll], ln)
out += '
'
return out
def tmpl_citesummary_overview(self, d_total_cites, d_avg_cites, l_colls, ln=CFG_SITE_LANG):
"""HTML citesummary format, overview. A part of HCS format suite."""
_ = gettext_set_language(ln)
out = """
%(msg_cites)s
""" % \
{'msg_cites': _("Total number of citations:"),}
for coll, colldef in l_colls:
out += '
%s
' % self.tmpl_nice_number(d_total_cites[coll], ln)
out += '
'
out += """
%(msg_avgcit)s
""" % \
{'msg_avgcit': _("Average citations per paper:"),}
for coll, colldef in l_colls:
out += '
%.1f
' % d_avg_cites[coll]
out += '
'
out += """
%(msg_breakdown)s
""" % \
{'msg_breakdown': _("Breakdown of papers by citations:"),}
return out
def tmpl_citesummary_breakdown_by_fame(self, d_cites, low, high, fame, l_colls, searchpattern, searchfield, ln=CFG_SITE_LANG):
"""HTML citesummary format, breakdown by fame. A part of HCS format suite."""
_ = gettext_set_language(ln)
out = """
%(fame)s
""" % \
{'fame': fame,}
for coll, colldef in l_colls:
link_url = CFG_SITE_URL + '/search?p='
if searchpattern:
p = searchpattern
if searchfield:
if " " in searchpattern:
p = searchfield + ':"' + searchpattern + '"'
else:
p = searchfield + ':' + searchpattern
link_url += quote(p) + ' '
if colldef:
link_url += quote(colldef) + ' '
if low == 0 and high == 0:
link_url += quote('cited:0')
else:
link_url += quote('cited:%i->%i' % (low, high))
link_url += '&rm=citation';
link_text = self.tmpl_nice_number(d_cites[coll], ln)
out += '
'
return out
def tmpl_citesummary_epilogue(self, ln=CFG_SITE_LANG):
"""HTML citesummary format, epilogue. A part of HCS format suite."""
_ = gettext_set_language(ln)
out = """
"""
return out
def tmpl_unapi(self, formats, identifier=None):
"""
Provide a list of object format available from the unAPI service
for the object identified by IDENTIFIER
"""
out = '\n'
if identifier:
out += '\n' % (identifier)
else:
out += "\n"
for format_name, format_type in formats.iteritems():
docs = ''
if format_name == 'xn':
docs = 'http://www.nlm.nih.gov/databases/dtd/'
format_type = 'application/xml'
format_name = 'nlm'
elif format_name == 'xm':
docs = 'http://www.loc.gov/standards/marcxml/schema/MARC21slim.xsd'
format_type = 'application/xml'
format_name = 'marcxml'
elif format_name == 'xr':
format_type = 'application/rss+xml'
docs = 'http://www.rssboard.org/rss-2-0/'
elif format_name == 'xw':
format_type = 'application/xml'
docs = 'http://www.refworks.com/RefWorks/help/RefWorks_Tagged_Format.htm'
elif format_name == 'xoaidc':
format_type = 'application/xml'
docs = 'http://www.openarchives.org/OAI/2.0/oai_dc.xsd'
elif format_name == 'xe':
format_type = 'application/xml'
docs= 'http://www.endnote.com/support/'
format_name = 'endnote'
elif format_name == 'xd':
format_type = 'application/xml'
docs = 'http://dublincore.org/schemas/'
format_name = 'dc'
elif format_name == 'xo':
format_type = 'application/xml'
docs = 'http://www.loc.gov/standards/mods/v3/mods-3-3.xsd'
format_name = 'mods'
if docs:
out += '\n' % (xml_escape(format_name), xml_escape(format_type), xml_escape(docs))
else:
out += '\n' % (xml_escape(format_name), xml_escape(format_type))
out += ""
return out