Diffusion invenio-infoscience (maint-2.0) Browse

search_engine.py
c59bfe6f74cemaint-2.0

search_engine.py
View Options

			# -- 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 copy
			import string
			import os
			import re
			import time
			import urllib
			import urlparse
			import zlib

			## 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

			try:
			from mod_python import apache
			except ImportError, e:
			pass # ignore user personalisation, needed e.g. for command-line

			try:
			import invenio.template
			websearch_templates = invenio.template.load('websearch')
			except:
			pass

			## 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 <em>%s</em> 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:
			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("""<?xml version="1.0" encoding="UTF-8"?>\n""")
			else:
			# we are doing XML output:
			req.content_type = "text/xml"
			req.send_http_header()
			req.write("""<?xml version="1.0" encoding="UTF-8"?>\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 = """
			<script type='text/javascript'>
			jsMath = {
			Controls: {cookie: {printwarn: 0}}
			};
			</script>
			<script src='/jsMath/easy/invenio-jsmath.js' type='text/javascript'></script>
			"""
			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 += ' <a class="navtrail" href="%s/record/%s">%s</a>' % \
			(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 "\n"

			def create_page_title_search_pattern_info(p, p1, p2, p3):
			"""Create the search pattern bit for the page <title> 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 += """<select name="%sd">""" % name
			box += """<option value="">%s""" % _("any day")
			for day in range(1, 32):
			box += """<option value="%02d"%s>%02d""" % (day, is_selected(day, selected_day), day)
			box += """</select>"""
			# month
			box += """<select name="%sm">""" % name
			box += """<option value="">%s""" % _("any month")
			for mm, month in [(1, _("January")), (2, _("February")), (3, _("March")), (4, _("April")), \
			(5, _("May")), (6, _("June")), (7, _("July")), (8, _("August")), \
			(9, _("September")), (10, _("October")), (11, _("November")), (12, _("December"))]:
			box += """<option value="%02d"%s>%s""" % (mm, is_selected(mm, selected_month), month)
			box += """</select>"""
			# year
			box += """<select name="%sy">""" % name
			box += """<option value="">%s""" % _("any year")
			this_year = int(time.strftime("%Y", time.localtime()))
			for year in range(this_year-20, this_year+1):
			box += """<option value="%d"%s>%d""" % (year, is_selected(year, selected_year), year)
			box += """</select>"""
			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")
			})
			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
			if len(collection_reclist_cache.cache.keys()) == 1 or \
			aas == -1:
			show_colls = 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,
			)

			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 = """
			<select name="%s">
			<option value="a"%s>%s
			<option value="o"%s>%s
			<option value="n"%s>%s
			</select>
			""" % (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 = """
			<select name="%s">
			<option value="a"%s>%s
			<option value="o"%s>%s
			<option value="e"%s>%s
			<option value="p"%s>%s
			<option value="r"%s>%s
			</select>
			""" % (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):
			"""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 = []

			# 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]

			# 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

			# check if some real collections remain:
			if len(colls)==0:
			colls = [cc]

			# 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,))
			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'
			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)

			# second, let us decide on collection splitting:
			if split_colls == 0:
			# type A - no sons are wanted
			colls_out = colls_out_for_display
			# elif split_colls == 1:
			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:
			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)

			return (cc, colls_out_for_display, colls_out)

			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 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):
			"""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 ORDER BY cc.score DESC""",
			(coll,))
			for name, dbquery in res:
			if dbquery: # this is 'real' collection, so return it:
			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, """<p>No match close to <em>%s</em> found in given collections.
			#Please try different term.<p>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):
			"""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)<len(basic_search_units) and basic_search_units[idx_unit+1][0]=="\|"):
			# stage 2-1: this basic search unit is retained, since
			# either the hitset is non-empty, or the approximate
			# pattern treatment is switched off, or the search unit
			# was joined by an OR operator to preceding/following
			# units so we do not require that it exists
			basic_search_units_hitsets.append(basic_search_unit_hitset)
			else:
			# stage 2-2: no hits found for this search unit, try to replace non-alphanumeric chars inside pattern:
			if re.search(r'[^a-zA-Z0-9\s\:]', bsu_p):
			if bsu_p.startswith('"') and bsu_p.endswith('"'): # is it ACC query?
			bsu_pn = re.sub(r'[^a-zA-Z0-9\s\:]+', "*", bsu_p)
			else: # it is WRD query
			bsu_pn = re.sub(r'[^a-zA-Z0-9\s\:]+', " ", bsu_p)
			if verbose and of.startswith('h') and req:
			print_warning(req, "Trying (%s,%s,%s)" % (cgi.escape(bsu_pn), cgi.escape(bsu_f), cgi.escape(bsu_m)))
			basic_search_unit_hitset = search_pattern(req=None, p=bsu_pn, f=bsu_f, m=bsu_m, of="id", ln=ln)
			if len(basic_search_unit_hitset) > 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': "<em>" + cgi.escape(bsu_p) + "</em>",
			'x_query2': "<em>" + cgi.escape(bsu_pn) + "</em>"})
			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'):
			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'):
			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'):
			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):
			"""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)

			# 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 patter
			current_hitset = search_pattern(req, current_pattern, f, m, ap, of, verbose, ln)

			# 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)

			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):
			"""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"):
			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': '<em>' + string.join([get_coll_i18nname(coll, ln, False) for coll in colls], ', ') + '</em>',
			'x_url_open': '<a class="nearestterms" href="%s">' % (url),
			'x_nb_hits': len(results_in_Home),
			'x_url_close': '</a>'})
			results = {}
			else:
			# no hits found in Home, recommend different search terms:
			if of.startswith("h"):
			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.") % \
			('<em>' + cgi.escape(get_field_i18nname(get_field_name(f) or f, ln, False)) + '</em>')
			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.") % \
			('<em>' + cgi.escape(get_field_i18nname(get_field_name(f) or f, ln, False)) + '</em>')

			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:
			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': "<em>" + cgi.escape(p.startswith("%") and p.endswith("%") and p[1:-1] or p) + "</em>",
			'x_index': "<em>" + cgi.escape(get_field_i18nname(get_field_name(f) or f, ln, False)) + "</em>"}
			else:
			intro = _("Search term %s did not match any record. Nearest terms in any collection are:") % \
			("<em>" + cgi.escape(p.startswith("%") and p.endswith("%") and p[1:-1] or p) + "</em>")

			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:
			# 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='<br />', epilogue='<br />'):
			"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_results_overview(req, colls, results_final_nb_total, results_final_nb, cpu_time, ln=CFG_SITE_LANG, ec=[]):
			"""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,
			)

			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)
			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 == 'holdings':
			req.write(webstyle_templates.detailed_record_container_top(recIDs[irec],
			tabs,
			ln))
			req.write(format_record(recIDs[irec], 'HDHOLD', ln=ln, user_info=user_info, verbose=verbose))
			req.write(webstyle_templates.detailed_record_container_bottom(recIDs[irec],
			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 += " <record>\n"
			out += " <header>\n"
			for oai_id in get_fieldvalues(recID, CFG_OAI_ID_FIELD):
			out += " <identifier>%s</identifier>\n" % oai_id
			out += " <datestamp>%s</datestamp>\n" % get_modification_date(recID)
			out += " </header>\n"
			out += " <metadata>\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 += """ <record xmlns="http://www.loc.gov/MARC21/slim">\n"""
			out += " <controlfield tag=\"001\">%d</controlfield>\n" % int(recID)
			elif format.startswith("xm"):
			out += """ <record>\n"""
			out += " <controlfield tag=\"001\">%d</controlfield>\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 += "<datafield tag=\"%s\" ind1=\"%s\" ind2=\"%s\"><subfield code=\"%s\">%s</subfield></datafield>\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 += "<datafield tag=\"980\" ind1=\"\" ind2=\"\"><subfield code=\"c\">DELETED</subfield></datafield>\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 += """ <controlfield tag="%s" >%s</controlfield>\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 += """ </datafield>\n"""
			out += """ <datafield tag="%s" ind1="%s" ind2="%s">\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 += """ <subfield code="%s">%s</subfield>\n""" % \
			(encode_for_xml(field[-1:]), value)

			# all fields/subfields printed in this run, so close the tag:
			if field_number_old != -999:
			out += """ </datafield>\n"""
			i = 0 # Next loop should start looking at bib%0 and bibrec_bib00x
			# we are at the end of printing the record:
			out += " </record>\n"

			elif format == "xd" or format == "oai_dc":
			# XML Dublin Core format, possibly OAI -- select only some bibXXx fields:
			out += """ <dc xmlns="http://purl.org/dc/elements/1.1/"
			xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
			xsi:schemaLocation="http://purl.org/dc/elements/1.1/
			http://www.openarchives.org/OAI/1.1/dc.xsd">\n"""
			if record_exist_p == -1:
			out += ""
			else:
			for f in get_fieldvalues(recID, "041__a"):
			out += " <language>%s</language>\n" % f

			for f in get_fieldvalues(recID, "100__a"):
			out += " <creator>%s</creator>\n" % encode_for_xml(f)

			for f in get_fieldvalues(recID, "700__a"):
			out += " <creator>%s</creator>\n" % encode_for_xml(f)

			for f in get_fieldvalues(recID, "245__a"):
			out += " <title>%s</title>\n" % encode_for_xml(f)

			for f in get_fieldvalues(recID, "65017a"):
			out += " <subject>%s</subject>\n" % encode_for_xml(f)

			for f in get_fieldvalues(recID, "8564_u"):
			out += " <identifier>%s</identifier>\n" % encode_for_xml(f)

			for f in get_fieldvalues(recID, "520__a"):
			out += " <description>%s</description>\n" % encode_for_xml(f)

			out += " <date>%s</date>\n" % get_creation_date(recID)
			out += " </dc>\n"

			elif len(format) == 6 and str(format[0:3]).isdigit():
			# user has asked to print some fields only
			if format == "001":
			out += "<!--%s-begin-->%s<!--%s-end-->\n" % (format, recID, format)
			else:
			vals = get_fieldvalues(recID, format)
			for val in vals:
			out += "<!--%s-begin-->%s<!--%s-end-->\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 += "<pre>" + cgi.escape(get_fieldvalues_alephseq_like(recID, ["001", CFG_OAI_ID_FIELD, "980"])) + "</pre>"
			else:
			out += "<pre>" + cgi.escape(get_fieldvalues_alephseq_like(recID, ot)) + "</pre>"

			elif format.startswith("h") and ot:
			## user directly asked for some tags to be displayed only
			if record_exist_p == -1:
			out += "<pre>" + get_fieldvalues_alephseq_like(recID, ["001", CFG_OAI_ID_FIELD, "980"]) + "</pre>"
			else:
			out += "<pre>" + get_fieldvalues_alephseq_like(recID, ot) + "</pre>"

			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 += '<a href="%s">' % websearch_templates.build_search_url(recid=recID, ln=ln)
			# firstly, title:
			titles = get_fieldvalues(recID, "245__a")
			if titles:
			for title in titles:
			out += "<strong>%s</strong>" % title
			else:
			# usual title not found, try conference title:
			titles = get_fieldvalues(recID, "111__a")
			if titles:
			for title in titles:
			out += "<strong>%s</strong>" % title
			else:
			# just print record ID:
			out += "<strong>%s %d</strong>" % (get_field_i18nname("record ID", ln, False), recID)
			out += "</a>"
			# secondly, authors:
			authors = get_fieldvalues(recID, "100__a") + get_fieldvalues(recID, "700__a")
			if authors:
			out += " - %s" % authors[0]
			if len(authors) > 1:
			out += " <em>et al</em>"
			# 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 += " </metadata>\n"
			out += " </record>\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

			def wash_url_argument(var, new_type):
			"""Wash list argument into 'new_type', that can be 'list',
			'str', or 'int'. Useful for washing mod_python passed
			arguments, that are all lists of strings (URL args may be
			multiple), but we sometimes want only to take the first value,
			and sometimes to represent it as string or numerical value."""
			out = []
			if new_type == 'list': # return lst
			if type(var) is list:
			out = var
			else:
			out = [var]
			elif new_type == 'str': # return str
			if type(var) is list:
			try:
			out = "%s" % var[0]
			except:
			out = ""
			elif type(var) is str:
			out = var
			else:
			out = "%s" % var
			elif new_type == 'int': # return int
			if type(var) is list:
			try:
			out = string.atoi(var[0])
			except:
			out = 0
			elif type(var) is int:
			out = var
			elif type(var) is str:
			try:
			out = string.atoi(var)
			except:
			out = 0
			else:
			out = 0
			return out

			### 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)

			# 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) = wash_colls(cc, c, sc) # 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 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.get_remote_host(apache.REMOTE_NOLOOKUP))
			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 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"):
			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:
			results_in_any_collection = search_pattern_parenthesised(req, p, f, ap=ap, of=of, verbose=verbose, ln=ln)
			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:
			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:
			results_final = intersect_results_with_collrecs(req, results_in_any_collection, colls_to_search, ap, of, verbose, ln)
			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 == {}:
			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 != "":
			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 == {}:
			if of.startswith("h"):
			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 pl:
			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 == {}:
			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 results_final_nb_total == 0:
			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']:
			req.write(print_results_overview(req, colls_to_search, results_final_nb_total, results_final_nb, cpu_time, ln, ec))
			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("<!-- Search-Engine-Total-Number-Of-Results: %s -->\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))
			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.get_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("<html>")
			out = ""
			out += "<h1>Search Cache</h1>"
			# clear cache if requested:
			if action == "clear":
			search_results_cache.clear()
			req.write(out)
			# show collection reclist cache:
			out = "<h3>Collection reclist cache</h3>"
			out += "- collection table last updated: %s" % get_table_update_time('collection')
			out += "<br />- reclist cache timestamp: %s" % collection_reclist_cache.timestamp
			out += "<br />- reclist cache contents:"
			out += "<blockquote>"
			for coll in collection_reclist_cache.cache.keys():
			if collection_reclist_cache.cache[coll]:
			out += "%s (%d)<br />" % (coll, len(collection_reclist_cache.cache[coll]))
			out += "</blockquote>"
			req.write(out)
			# show search results cache:
			out = "<h3>Search Cache</h3>"
			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 += "<br />- search cache contents:"
			out += "<blockquote>"
			for query, hitset in search_results_cache.cache.items():
			out += "<br />%s ... %s" % (query, hitset)
			out += """<p><a href="%s/search/cache?action=clear">clear search results cache</a>""" % CFG_SITE_URL
			out += "</blockquote>"
			req.write(out)
			# show field i18nname cache:
			out = "<h3>Field I18N names cache</h3>"
			out += "- fieldname table last updated: %s" % get_table_update_time('fieldname')
			out += "<br />- i18nname cache timestamp: %s" % field_i18nname_cache.timestamp
			out += "<br />- i18nname cache contents:"
			out += "<blockquote>"
			for field in field_i18nname_cache.cache.keys():
			for ln in field_i18nname_cache.cache[field].keys():
			out += "%s, %s = %s<br />" % (field, ln, field_i18nname_cache.cache[field][ln])
			out += "</blockquote>"
			req.write(out)
			# show collection i18nname cache:
			out = "<h3>Collection I18N names cache</h3>"
			out += "- collectionname table last updated: %s" % get_table_update_time('collectionname')
			out += "<br />- i18nname cache timestamp: %s" % collection_i18nname_cache.timestamp
			out += "<br />- i18nname cache contents:"
			out += "<blockquote>"
			for coll in collection_i18nname_cache.cache.keys():
			for ln in collection_i18nname_cache.cache[coll].keys():
			out += "%s, %s = %s<br />" % (coll, ln, collection_i18nname_cache.cache[coll][ln])
			out += "</blockquote>"
			req.write(out)
			req.write("</html>")
			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("<html>")
			req.write("<h1>Search Log</h1>")
			if date: # case A: display stats for a day
			yyyymmdd = string.atoi(date)
			req.write("<p><big><strong>Date: %d</strong></big><p>" % yyyymmdd)
			req.write("""<table border="1">""")
			req.write("<tr><td><strong>%s</strong></td><td><strong>%s</strong></td><td><strong>%s</strong></td><td><strong>%s</strong></td><td><strong>%s</strong></td><td><strong>%s</strong></td></tr>" % ("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("<tr><td align=\"right\">#%d</td><td>%s:%s:%s</td><td>%s</td><td>%s</td><td>%s</td><td>%s</td></tr>" \
			% (i, datetime[8:10], datetime[10:12], datetime[12:], p, f, c, nbhits))
			except:
			pass # ignore eventual wrong log lines
			req.write("</table>")
			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("""<table border="1">""")
			req.write("<tr><td><strong>%s</strong></td><td><strong>%s</strong></tr>" % ("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("""<tr><td>%s</td><td align="right"><a href="%s/search/log?date=%d">%s</a></td></tr>""" % \
			(day, CFG_SITE_URL, day, line))
			p.close()
			req.write("</table>")
			req.write("</html>")
			return "\n"


			def get_most_popular_field_values(recids, tags, exclude_values=None, count_repetitive_values=True):
			"""
			Analyze RECIDS and look for TAGS and return most popular values
			and the frequency with which they occur sorted according to
			descending frequency.

			If a value is found in EXCLUDE_VALUES, then do not count it.

			If COUNT_REPETITIVE_VALUES is True, then we count every occurrence
			of value in the tags. If False, then we count the value only once
			regardless of the number of times it may appear in a record.
			(But, if the same value occurs in another record, we count it, of
			course.)

			Example:
			>>> get_most_popular_field_values(range(11,20), '980__a')
			(('PREPRINT', 10), ('THESIS', 7), ...)
			>>> get_most_popular_field_values(range(11,20), ('100__a', '700__a'))
			(('Ellis, J', 10), ('Ellis, N', 7), ...)
			>>> get_most_popular_field_values(range(11,20), ('100__a', '700__a'), ('Ellis, J'))
			(('Ellis, N', 7), ...)
			"""

			def _get_most_popular_field_values_helper_sorter(val1, val2):
			"Compare VAL1 and VAL2 according to, firstly, frequency, then secondly, alphabetically."
			compared_via_frequencies = cmp(valuefreqdict[val2], valuefreqdict[val1])
			if compared_via_frequencies == 0:
			return cmp(val1.lower(), val2.lower())
			else:
			return compared_via_frequencies

			valuefreqdict = {}
			## sanity check:
			if not exclude_values:
			exclude_values = []
			if isinstance(tags, str):
			tags = (tags,)
			## find values to count:
			vals_to_count = []
			if count_repetitive_values:
			# counting technique A: can look up many records at once: (very fast)
			for tag in tags:
			vals_to_count.extend(get_fieldvalues(recids, tag))
			else:
			# counting technique B: must count record-by-record: (slow)
			for recid in recids:
			vals_in_rec = []
			for tag in tags:
			for val in get_fieldvalues(recid, tag, False):
			vals_in_rec.append(val)
			# do not count repetitive values within this record
			# (even across various tags, so need to unify again):
			dtmp = {}
			for val in vals_in_rec:
			dtmp[val] = 1
			vals_in_rec = dtmp.keys()
			vals_to_count.extend(vals_in_rec)
			## are we to exclude some of found values?
			for val in vals_to_count:
			if val not in exclude_values:
			if valuefreqdict.has_key(val):
			valuefreqdict[val] += 1
			else:
			valuefreqdict[val] = 1
			## sort by descending frequency of values:
			out = ()
			vals = valuefreqdict.keys()
			vals.sort(_get_most_popular_field_values_helper_sorter)
			for val in vals:
			out += (val, valuefreqdict[val]),
			return out

			def profile(p="", f="", c=CFG_SITE_NAME):
			"""Profile search time."""
			import profile
			import pstats
			profile.run("perform_request_search(p='%s',f='%s', c='%s')" % (p, f, c), "perform_request_search_profile")
			p = pstats.Stats("perform_request_search_profile")
			p.strip_dirs().sort_stats("cumulative").print_stats()
			return 0

			## test cases:
			#print wash_colls(CFG_SITE_NAME,"Library Catalogue", 0)
			#print wash_colls("Periodicals & Progress Reports",["Periodicals","Progress Reports"], 0)
			#print wash_field("wau")
			#print print_record(20,"tm","001,245")
			#print create_opft_search_units(None, "PHE-87-13","reportnumber")
			#print ":"+wash_pattern("* and % doo * %")+":\n"
			#print ":"+wash_pattern("*")+":\n"
			#print ":"+wash_pattern("ellis* ell* e*%")+":\n"
			#print run_sql("SELECT name,dbquery from collection")
			#print get_index_id("author")
			#print get_coll_ancestors("Theses")
			#print get_coll_sons("Articles & Preprints")
			#print get_coll_real_descendants("Articles & Preprints")
			#print get_collection_reclist("Theses")
			#print log(sys.stdin)
			#print search_unit_in_bibrec('2002-12-01','2002-12-12')
			#print type(wash_url_argument("-1",'int'))
			#print get_nearest_terms_in_bibxxx("ellis", "author", 5, 5)
			#print call_bibformat(68, "HB_FLY")
			#print get_fieldvalues(10, "980__a")
			#print get_fieldvalues_alephseq_like(10,"001___")
			#print get_fieldvalues_alephseq_like(10,"980__a")
			#print get_fieldvalues_alephseq_like(10,"foo")
			#print get_fieldvalues_alephseq_like(10,"-1")
			#print get_fieldvalues_alephseq_like(10,"99")
			#print get_fieldvalues_alephseq_like(10,["001", "980"])

			## profiling:
			#profile("of the this")
			#print perform_request_search(p="ellis")

R3600/modules/websearch/lib/search_engine.pyc59bfe6f74cemaint-2.0

search_engine.pyView Options

R3600/modules/websearch/lib/search_engine.py
c59bfe6f74cemaint-2.0

search_engine.py
View Options