bibupload.py
No OneTemporary
Actions

Subscribers

None

File Metadata

Created: Thu, May 30, 19:13

bibupload.py
View Options

	# -- coding: utf-8 --
	##
	## This file is part of Invenio.
	## Copyright (C) 2006, 2007, 2008, 2009, 2010, 2011, 2012 CERN.
	##
	## 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.
	##
	## 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 Invenio; if not, write to the Free Software Foundation, Inc.,
	## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.

	"""
	BibUpload: Receive MARC XML file and update the appropriate database
	tables according to options.
	"""

	__revision__ = "$Id$"

	import os
	import re
	import sys
	import time
	from datetime import datetime
	from zlib import compress
	import socket
	import marshal
	import copy
	import tempfile
	import urlparse
	import urllib2

	from invenio.config import CFG_OAI_ID_FIELD, \
	CFG_BIBUPLOAD_REFERENCE_TAG, \
	CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG, \
	CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG, \
	CFG_BIBUPLOAD_EXTERNAL_OAIID_PROVENANCE_TAG, \
	CFG_BIBUPLOAD_STRONG_TAGS, \
	CFG_BIBUPLOAD_CONTROLLED_PROVENANCE_TAGS, \
	CFG_BIBUPLOAD_SERIALIZE_RECORD_STRUCTURE, \
	CFG_BIBUPLOAD_DELETE_FORMATS, \
	CFG_SITE_URL, CFG_SITE_RECORD, \
	CFG_OAI_PROVENANCE_ALTERED_SUBFIELD

	from invenio.jsonutils import json, CFG_JSON_AVAILABLE
	from invenio.bibupload_config import CFG_BIBUPLOAD_CONTROLFIELD_TAGS, \
	CFG_BIBUPLOAD_SPECIAL_TAGS
	from invenio.dbquery import run_sql, \
	Error
	from invenio.bibrecord import create_records, \
	record_add_field, \
	record_delete_field, \
	record_xml_output, \
	record_get_field_instances, \
	record_get_field_value, \
	record_get_field_values, \
	field_get_subfield_values, \
	field_get_subfield_instances, \
	record_modify_subfield, \
	record_delete_subfield_from, \
	record_delete_fields, \
	record_add_subfield_into, \
	record_find_field, \
	record_extract_oai_id
	from invenio.search_engine import get_record
	from invenio.dateutils import convert_datestruct_to_datetext
	from invenio.errorlib import register_exception
	from invenio.intbitset import intbitset
	from invenio.config import CFG_WEBSUBMIT_FILEDIR
	from invenio.bibtask import task_init, write_message, \
	task_set_option, task_get_option, task_get_task_param, task_update_status, \
	task_update_progress, task_sleep_now_if_required, fix_argv_paths
	from invenio.bibdocfile import BibRecDocs, file_strip_ext, normalize_format, \
	get_docname_from_url, check_valid_url, download_url, \
	KEEP_OLD_VALUE, decompose_bibdocfile_url, InvenioWebSubmitFileError, \
	bibdocfile_url_p, CFG_BIBDOCFILE_AVAILABLE_FLAGS, guess_format_from_url

	from invenio.search_engine import search_pattern

	#Statistic variables
	stat = {}
	stat['nb_records_to_upload'] = 0
	stat['nb_records_updated'] = 0
	stat['nb_records_inserted'] = 0
	stat['nb_errors'] = 0
	stat['nb_holdingpen'] = 0
	stat['exectime'] = time.localtime()

	_WRITING_RIGHTS = None

	## Let's set a reasonable timeout for URL request (e.g. FFT)
	socket.setdefaulttimeout(40)

	_re_find_001 = re.compile('<controlfield\\s+tag=("001"\|\'001\')\\s>\\s(\\d)\\s</controlfield>', re.S)
	def bibupload_pending_recids():
	"""This function embed a bit of A.I. and is more a hack than an elegant
	algorithm. It should be updated in case bibupload/bibsched are modified
	in incompatible ways.
	This function return the intbitset of all the records that are being
	(or are scheduled to be) touched by other bibuploads.
	"""
	options = run_sql("""SELECT arguments FROM schTASK WHERE status<>'DONE' AND
	proc='bibupload' AND (status='RUNNING' OR status='CONTINUING' OR
	status='WAITING' OR status='SCHEDULED' OR status='ABOUT TO STOP' OR
	status='ABOUT TO SLEEP')""")
	ret = intbitset()
	xmls = []
	if options:
	for arguments in options:
	arguments = marshal.loads(arguments[0])
	for argument in arguments[1:]:
	if argument.startswith('/'):
	# XMLs files are recognizable because they're absolute
	# files...
	xmls.append(argument)
	for xmlfile in xmls:
	# Let's grep for the 001
	try:
	xml = open(xmlfile).read()
	ret += [int(group[1]) for group in _re_find_001.findall(xml)]
	except:
	continue
	return ret

	### bibupload engine functions:
	def bibupload(record, opt_tag=None, opt_mode=None,
	opt_stage_to_start_from=1, opt_notimechange=0, oai_rec_id = "", pretend=False):
	"""Main function: process a record and fit it in the tables
	bibfmt, bibrec, bibrec_bibxxx, bibxxx with proper record
	metadata.

	Return (error_code, recID) of the processed record.
	"""
	assert(opt_mode in ('insert', 'replace', 'replace_or_insert', 'reference',
	'correct', 'append', 'format', 'holdingpen', 'delete'))

	error = None
	now = datetime.now() # will hold record creation/modification date
	# If there are special tags to proceed check if it exists in the record
	if opt_tag is not None and not(record.has_key(opt_tag)):
	msg = " Failed: Tag not found, enter a valid tag to update."
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, -1, msg)

	# Extraction of the Record Id from 001, SYSNO or OAIID tags:
	rec_id = retrieve_rec_id(record, opt_mode, pretend=pretend)
	if rec_id == -1:
	msg = " Failed: either the record already exists and insert was " \
	"requested or the record does not exists and " \
	"replace/correct/append has been used"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, -1, msg)
	elif rec_id > 0:
	write_message(" -Retrieve record ID (found %s): DONE." % rec_id, verbose=2)
	if not record.has_key('001'):
	# Found record ID by means of SYSNO or OAIID, and the
	# input MARCXML buffer does not have this 001 tag, so we
	# should add it now:
	error = record_add_field(record, '001', controlfield_value=rec_id)
	if error is None:
	msg = " Failed: Error during adding the 001 controlfield " \
	"to the record"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	else:
	error = None
	write_message(" -Added tag 001: DONE.", verbose=2)

	write_message(" -Check if the xml marc file is already in the database: DONE" , verbose=2)

	# Reference mode check if there are reference tag
	if opt_mode == 'reference':
	error = extract_tag_from_record(record, CFG_BIBUPLOAD_REFERENCE_TAG)
	if error is None:
	msg = " Failed: No reference tags has been found..."
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, -1, msg)
	else:
	error = None
	write_message(" -Check if reference tags exist: DONE", verbose=2)

	record_deleted_p = False
	if opt_mode == 'insert' or \
	(opt_mode == 'replace_or_insert') and rec_id is None:
	insert_mode_p = True
	# Insert the record into the bibrec databases to have a recordId
	rec_id = create_new_record(pretend=pretend)
	write_message(" -Creation of a new record id (%d): DONE" % rec_id, verbose=2)

	# we add the record Id control field to the record
	error = record_add_field(record, '001', controlfield_value=rec_id)
	if error is None:
	msg = " Failed: Error during adding the 001 controlfield " \
	"to the record"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	else:
	error = None

	error = record_add_field(record, '005', controlfield_value=now.strftime("%Y%m%d%H%M%S.0"))
	if error is None:
	write_message(" Failed: Error during adding to 005 controlfield to record",verbose=1,stream=sys.stderr)
	return (1, int(rec_id))
	else:
	error=None

	elif opt_mode != 'insert' and opt_mode != 'format' and \
	opt_stage_to_start_from != 5:
	insert_mode_p = False
	# Update Mode
	# Retrieve the old record to update
	rec_old = get_record(rec_id)
	record_had_altered_bit = record_get_field_values(rec_old, CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[:3], CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[3], CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[4], CFG_OAI_PROVENANCE_ALTERED_SUBFIELD)
	# Also save a copy to restore previous situation in case of errors
	original_record = get_record(rec_id)

	if original_record.has_key('005'):
	record_delete_field(original_record,'005')

	if rec_old is None:
	msg = " Failed during the creation of the old record!"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	else:
	write_message(" -Retrieve the old record to update: DONE", verbose=2)

	if rec_old.has_key('005'):
	record_delete_field(rec_old,'005')

	# In Replace mode, take over old strong tags if applicable:
	if opt_mode == 'replace' or \
	opt_mode == 'replace_or_insert':
	copy_strong_tags_from_old_record(record, rec_old)

	# Delete tags to correct in the record
	if opt_mode == 'correct' or opt_mode == 'reference':
	delete_tags_to_correct(record, rec_old, opt_tag)
	write_message(" -Delete the old tags to correct in the old record: DONE",
	verbose=2)

	# Delete tags specified if in delete mode
	if opt_mode == 'delete':
	record = delete_tags(record, rec_old)
	write_message(" -Delete specified tags in the old record: DONE", verbose=2)

	# Append new tag to the old record and update the new record with the old_record modified
	if opt_mode == 'append' or opt_mode == 'correct' or \
	opt_mode == 'reference':
	record = append_new_tag_to_old_record(record, rec_old,
	opt_tag, opt_mode)
	write_message(" -Append new tags to the old record: DONE", verbose=2)

	# 005 tag should be added everytime the record is modified
	# If an exiting record is modified, its 005 tag should be overwritten with a new revision value
	if record.has_key('005'):
	record_delete_field(record, '005')
	write_message(" Deleted the existing 005 tag.", verbose=2)
	error = record_add_field(record, '005', controlfield_value=now.strftime("%Y%m%d%H%M%S.0"))
	if error is None:
	write_message(" Failed: Error during adding to 005 controlfield to record",verbose=1,stream=sys.stderr)
	return (1, int(rec_id))
	else:
	error=None
	write_message(" -Added tag 005: DONE. "+ str(record_get_field_value(record,'005','','')), verbose=2)

	# if record_had_altered_bit, this must be set to true, since the
	# record has been altered.
	if record_had_altered_bit:
	oai_provenance_fields = record_get_field_instances(record, CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[:3], CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[3], CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[4])
	for oai_provenance_field in oai_provenance_fields:
	for i, (code, dummy_value) in enumerate(oai_provenance_field[0]):
	if code == CFG_OAI_PROVENANCE_ALTERED_SUBFIELD:
	oai_provenance_field[0][i] = (code, 'true')

	# now we clear all the rows from bibrec_bibxxx from the old
	# record (they will be populated later (if needed) during
	# stage 4 below):
	delete_bibrec_bibxxx(rec_old, rec_id, pretend=pretend)

	record_deleted_p = True
	write_message(" -Clean bibrec_bibxxx: DONE", verbose=2)
	write_message(" -Stage COMPLETED", verbose=2)


	try:
	if not record_is_valid(record):
	msg = "ERROR: record is not valid"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, -1, msg)

	# Have a look if we have FMT tags
	we_have_fmt_tags_p = extract_tag_from_record(record, 'FMT') is not None
	write_message("Stage 1: Start (Insert of FMT tags if exist).", verbose=2)
	if opt_stage_to_start_from <= 1 and we_have_fmt_tags_p:
	record = insert_fmt_tags(record, rec_id, opt_mode, pretend=pretend)
	if record is None:
	msg = " Stage 1 failed: Error while inserting FMT tags"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	elif record == 0:
	# Mode format finished
	stat['nb_records_updated'] += 1
	return (0, int(rec_id), "")
	write_message(" -Stage COMPLETED", verbose=2)
	else:
	write_message(" -Stage NOT NEEDED", verbose=2)

	# Have a look if we have FFT tags
	write_message("Stage 2: Start (Process FFT tags if exist).", verbose=2)
	record_had_FFT = False
	if opt_stage_to_start_from <= 2 and \
	extract_tag_from_record(record, 'FFT') is not None:
	record_had_FFT = True
	if not writing_rights_p():
	write_message(" Stage 2 failed: Error no rights to write fulltext files",
	verbose=1, stream=sys.stderr)
	task_update_status("ERROR")
	sys.exit(1)
	try:
	record = elaborate_fft_tags(record, rec_id, opt_mode, pretend=pretend)
	except Exception, e:
	register_exception()
	msg = " Stage 2 failed: Error while elaborating FFT tags: %s" % e
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	if record is None:
	msg = " Stage 2 failed: Error while elaborating FFT tags"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	write_message(" -Stage COMPLETED", verbose=2)
	else:
	write_message(" -Stage NOT NEEDED", verbose=2)

	# Have a look if we have FFT tags
	write_message("Stage 2B: Start (Synchronize 8564 tags).", verbose=2)
	has_bibdocs = run_sql("SELECT count(id_bibdoc) FROM bibrec_bibdoc JOIN bibdoc ON id_bibdoc=id WHERE id_bibrec=%s AND status<>'DELETED'", (rec_id, ))[0][0] > 0
	if opt_stage_to_start_from <= 2 and (has_bibdocs or record_had_FFT or extract_tag_from_record(record, '856') is not None):
	try:
	record = synchronize_8564(rec_id, record, record_had_FFT, pretend=pretend)
	except Exception, e:
	register_exception(alert_admin=True)
	msg = " Stage 2B failed: Error while synchronizing 8564 tags: %s" % e
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	if record is None:
	msg = " Stage 2B failed: Error while synchronizing 8564 tags"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	write_message(" -Stage COMPLETED", verbose=2)
	else:
	write_message(" -Stage NOT NEEDED", verbose=2)
	# Update of the BibFmt
	write_message("Stage 3: Start (Update bibfmt).", verbose=2)
	if opt_stage_to_start_from <= 3:
	# format the single record as xml
	rec_xml_new = record_xml_output(record)
	# Update bibfmt with the format xm of this record
	if opt_mode != 'format':
	modification_date = time.strftime('%Y-%m-%d %H:%M:%S', time.strptime(record_get_field_value(record,'005'),'%Y%m%d%H%M%S.0'))
	error = update_bibfmt_format(rec_id, rec_xml_new, 'xm', modification_date, pretend=pretend)
	if error == 1:
	msg = " Failed: error during update_bibfmt_format 'xm'"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	if CFG_BIBUPLOAD_SERIALIZE_RECORD_STRUCTURE:
	error = update_bibfmt_format(rec_id, marshal.dumps(record), 'recstruct', modification_date, pretend=pretend)
	if error == 1:
	msg = " Failed: error during update_bibfmt_format 'recstruct'"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	if not we_have_fmt_tags_p:
	# delete some formats like HB upon record change:
	for format_to_delete in CFG_BIBUPLOAD_DELETE_FORMATS:
	try:
	delete_bibfmt_format(rec_id, format_to_delete, pretend=pretend)
	except:
	# OK, some formats like HB could not have been deleted, no big deal
	pass
	# archive MARCXML format of this record for version history purposes:
	error = archive_marcxml_for_history(rec_id, pretend=pretend)
	if error == 1:
	msg = " Failed to archive MARCXML for history"
	write_message(msg, verbose=1, stream=sys.stderr)
	return (1, int(rec_id), msg)
	else:
	write_message(" -Archived MARCXML for history : DONE", verbose=2)
	write_message(" -Stage COMPLETED", verbose=2)

	# Update the database MetaData
	write_message("Stage 4: Start (Update the database with the metadata).",
	verbose=2)
	if opt_stage_to_start_from <= 4:
	if opt_mode in ('insert', 'replace', 'replace_or_insert',
	'append', 'correct', 'reference', 'delete'):
	update_database_with_metadata(record, rec_id, oai_rec_id, pretend=pretend)
	record_deleted_p = False
	else:
	write_message(" -Stage NOT NEEDED in mode %s" % opt_mode,
	verbose=2)
	write_message(" -Stage COMPLETED", verbose=2)
	else:
	write_message(" -Stage NOT NEEDED", verbose=2)

	# Finally we update the bibrec table with the current date
	write_message("Stage 5: Start (Update bibrec table with current date).",
	verbose=2)
	if opt_stage_to_start_from <= 5 and \
	opt_notimechange == 0 and \
	not insert_mode_p:
	write_message(" -Retrieved current localtime: DONE", verbose=2)
	update_bibrec_modif_date(now.strftime("%Y-%m-%d %H:%M:%S"), rec_id, pretend=pretend)
	write_message(" -Stage COMPLETED", verbose=2)
	else:
	write_message(" -Stage NOT NEEDED", verbose=2)

	# Increase statistics
	if insert_mode_p:
	stat['nb_records_inserted'] += 1
	else:
	stat['nb_records_updated'] += 1

	# Upload of this record finish
	write_message("Record "+str(rec_id)+" DONE", verbose=1)
	return (0, int(rec_id), "")
	finally:
	if record_deleted_p:
	## BibUpload has failed living the record deleted. We should
	## back the original record then.
	update_database_with_metadata(original_record, rec_id, oai_rec_id, pretend=pretend)
	write_message(" Restored original record", verbose=1, stream=sys.stderr)

	def record_is_valid(record):
	"""
	Check if the record is valid. Currently this simply checks if the record
	has exactly one rec_id.

	@param record: the record
	@type record: recstruct
	@return: True if the record is valid
	@rtype: bool
	"""
	rec_ids = record_get_field_values(record, tag="001")
	if len(rec_ids) != 1:
	write_message(" The record is not valid: it has not a single rec_id: %s" % (rec_ids), stream=sys.stderr)
	return False
	return True

	def find_record_ids_by_oai_id(oaiId):
	"""
	A method finding the records identifier provided the oai identifier
	returns a list of identifiers matching a given oai identifier
	"""
	# Is this record already in invenio (matching by oaiid)
	if oaiId:
	recids = search_pattern(p=oaiId, f=CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG, m='e')

	# Is this record already in invenio (matching by reportnumber i.e.
	# particularly 037. Idea: to avoid doubbles insertions)
	repnumber = oaiId.split(":")[-1]
	if repnumber:
	recids \|= search_pattern(p = repnumber,
	f = "reportnumber",
	m = 'e' )

	# Is this record already in invenio (matching by reportnumber i.e.
	# particularly 037. Idea: to avoid double insertions)
	repnumber = "arXiv:" + oaiId.split(":")[-1]
	recids \|= search_pattern(p = repnumber,
	f = "reportnumber",
	m = 'e' )

	return recids
	else:
	return intbitset()

	def insert_record_into_holding_pen(record, oai_id, pretend=False):
	query = "INSERT INTO bibHOLDINGPEN (oai_id, changeset_date, changeset_xml, id_bibrec) VALUES (%s, NOW(), %s, %s)"
	xml_record = record_xml_output(record)
	bibrec_ids = find_record_ids_by_oai_id(oai_id) # here determining the identifier of the record
	if len(bibrec_ids) > 0:
	bibrec_id = bibrec_ids.pop()
	else:
	# id not found by using the oai_id, let's use a wider search based
	# on any information we might have.
	bibrec_id = retrieve_rec_id(record, 'holdingpen', pretend=pretend)
	if bibrec_id is None:
	bibrec_id = 0

	if not pretend:
	run_sql(query, (oai_id, xml_record, bibrec_id))

	# record_id is logged as 0! ( We are not inserting into the main database)
	log_record_uploading(oai_id, task_get_task_param('task_id', 0), 0, 'H', pretend=pretend)
	stat['nb_holdingpen'] += 1

	def print_out_bibupload_statistics():
	"""Print the statistics of the process"""
	out = "Task stats: %(nb_input)d input records, %(nb_updated)d updated, " \
	"%(nb_inserted)d inserted, %(nb_errors)d errors, %(nb_holdingpen)d inserted to holding pen. " \
	"Time %(nb_sec).2f sec." % { \
	'nb_input': stat['nb_records_to_upload'],
	'nb_updated': stat['nb_records_updated'],
	'nb_inserted': stat['nb_records_inserted'],
	'nb_errors': stat['nb_errors'],
	'nb_holdingpen': stat['nb_holdingpen'],
	'nb_sec': time.time() - time.mktime(stat['exectime']) }
	write_message(out)

	def open_marc_file(path):
	"""Open a file and return the data"""
	try:
	# open the file containing the marc document
	marc_file = open(path,'r')
	marc = marc_file.read()
	marc_file.close()
	except IOError, erro:
	write_message("Error: %s" % erro, verbose=1, stream=sys.stderr)
	write_message("Exiting.", sys.stderr)
	if erro.errno == 2:
	# No such file or directory
	# Not scary
	task_update_status("CERROR")
	else:
	task_update_status("ERROR")
	sys.exit(1)
	return marc

	def xml_marc_to_records(xml_marc):
	"""create the records"""
	# Creation of the records from the xml Marc in argument
	recs = create_records(xml_marc, 1, 1)
	if recs == []:
	write_message("Error: Cannot parse MARCXML file.", verbose=1, stream=sys.stderr)
	write_message("Exiting.", sys.stderr)
	task_update_status("ERROR")
	sys.exit(1)
	elif recs[0][0] is None:
	write_message("Error: MARCXML file has wrong format: %s" % recs,
	verbose=1, stream=sys.stderr)
	write_message("Exiting.", sys.stderr)
	task_update_status("CERROR")
	sys.exit(1)
	else:
	recs = map((lambda x:x[0]), recs)
	return recs

	def find_record_format(rec_id, format):
	"""Look whether record REC_ID is formatted in FORMAT,
	i.e. whether FORMAT exists in the bibfmt table for this record.

	Return the number of times it is formatted: 0 if not, 1 if yes,
	2 if found more than once (should never occur).
	"""
	out = 0
	query = """SELECT COUNT(id) FROM bibfmt WHERE id_bibrec=%s AND format=%s"""
	params = (rec_id, format)
	res = []
	try:
	res = run_sql(query, params)
	out = res[0][0]
	except Error, error:
	write_message(" Error during find_record_format() : %s " % error, verbose=1, stream=sys.stderr)
	return out

	def find_record_from_recid(rec_id):
	"""
	Try to find record in the database from the REC_ID number.
	Return record ID if found, None otherwise.
	"""
	try:
	res = run_sql("SELECT id FROM bibrec WHERE id=%s",
	(rec_id,))
	except Error, error:
	write_message(" Error during find_record_bibrec() : %s "
	% error, verbose=1, stream=sys.stderr)
	if res:
	return res[0][0]
	else:
	return None

	def find_record_from_sysno(sysno):
	"""
	Try to find record in the database from the external SYSNO number.
	Return record ID if found, None otherwise.
	"""
	bibxxx = 'bib'+CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[0:2]+'x'
	bibrec_bibxxx = 'bibrec_' + bibxxx
	try:
	res = run_sql("""SELECT bb.id_bibrec FROM %(bibrec_bibxxx)s AS bb,
	%(bibxxx)s AS b WHERE b.tag=%%s AND b.value=%%s
	AND bb.id_bibxxx=b.id""" % \
	{'bibxxx': bibxxx,
	'bibrec_bibxxx': bibrec_bibxxx},
	(CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG, sysno,))
	except Error, error:
	write_message(" Error during find_record_from_sysno(): %s " % error,
	verbose=1, stream=sys.stderr)
	if res:
	return res[0][0]
	else:
	return None

	def find_records_from_extoaiid(extoaiid, extoaisrc=None):
	"""
	Try to find records in the database from the external EXTOAIID number.
	Return list of record ID if found, None otherwise.
	"""
	assert(CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[:5] == CFG_BIBUPLOAD_EXTERNAL_OAIID_PROVENANCE_TAG[:5])
	bibxxx = 'bib'+CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[0:2]+'x'
	bibrec_bibxxx = 'bibrec_' + bibxxx
	try:
	write_message(' Looking for extoaiid="%s" with extoaisrc="%s"' % (extoaiid, extoaisrc), verbose=9)
	id_bibrecs = intbitset(run_sql("""SELECT bb.id_bibrec FROM %(bibrec_bibxxx)s AS bb,
	%(bibxxx)s AS b WHERE b.tag=%%s AND b.value=%%s
	AND bb.id_bibxxx=b.id""" % \
	{'bibxxx': bibxxx,
	'bibrec_bibxxx': bibrec_bibxxx},
	(CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG, extoaiid,)))
	write_message(' Partially found %s for extoaiid="%s"' % (id_bibrecs, extoaiid), verbose=9)
	ret = intbitset()
	for id_bibrec in id_bibrecs:
	record = get_record(id_bibrec)
	instances = record_get_field_instances(record, CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[0:3], CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[3], CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[4])
	write_message(' recid %s -> instances "%s"' % (id_bibrec, instances), verbose=9)
	for instance in instances:
	this_extoaisrc = field_get_subfield_values(instance, CFG_BIBUPLOAD_EXTERNAL_OAIID_PROVENANCE_TAG[5])
	this_extoaisrc = this_extoaisrc and this_extoaisrc[0] or None
	this_extoaiid = field_get_subfield_values(instance, CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[5])
	this_extoaiid = this_extoaiid and this_extoaiid[0] or None
	write_message(" this_extoaisrc -> %s, this_extoaiid -> %s" % (this_extoaisrc, this_extoaiid), verbose=9)
	if this_extoaiid == extoaiid:
	write_message(' recid %s -> provenance "%s"' % (id_bibrec, this_extoaisrc), verbose=9)
	if this_extoaisrc == extoaisrc:
	write_message('Found recid %s for extoaiid="%s" with provenance="%s"' % (id_bibrec, extoaiid, extoaisrc), verbose=9)
	ret.add(id_bibrec)
	break
	if this_extoaisrc is None:
	write_message('WARNING: Found recid %s for extoaiid="%s" that doesn\'t specify any provenance, while input record does.' % (id_bibrec, extoaiid), stream=sys.stderr)
	if extoaisrc is None:
	write_message('WARNING: Found recid %s for extoaiid="%s" that specify a provenance (%s), while input record does not have a provenance.' % (id_bibrec, extoaiid, this_extoaisrc), stream=sys.stderr)

	return ret
	except Error, error:
	write_message(" Error during find_records_from_extoaiid(): %s "
	% error, verbose=1, stream=sys.stderr)
	raise

	def find_record_from_oaiid(oaiid):
	"""
	Try to find record in the database from the OAI ID number and OAI SRC.
	Return record ID if found, None otherwise.
	"""
	bibxxx = 'bib'+CFG_OAI_ID_FIELD[0:2]+'x'
	bibrec_bibxxx = 'bibrec_' + bibxxx
	try:
	res = run_sql("""SELECT bb.id_bibrec FROM %(bibrec_bibxxx)s AS bb,
	%(bibxxx)s AS b WHERE b.tag=%%s AND b.value=%%s
	AND bb.id_bibxxx=b.id""" % \
	{'bibxxx': bibxxx,
	'bibrec_bibxxx': bibrec_bibxxx},
	(CFG_OAI_ID_FIELD, oaiid,))
	except Error, error:
	write_message(" Error during find_record_from_oaiid(): %s " % error,
	verbose=1, stream=sys.stderr)
	if res:
	return res[0][0]
	else:
	return None

	def extract_tag_from_record(record, tag_number):
	""" Extract the tag_number for record."""
	# first step verify if the record is not already in the database
	if record:
	return record.get(tag_number, None)
	return None

	def retrieve_rec_id(record, opt_mode, pretend=False):
	"""Retrieve the record Id from a record by using tag 001 or SYSNO or OAI ID
	tag. opt_mod is the desired mode."""

	rec_id = None

	# 1st step: we look for the tag 001
	tag_001 = extract_tag_from_record(record, '001')
	if tag_001 is not None:
	# We extract the record ID from the tag
	rec_id = tag_001[0][3]
	# if we are in insert mode => error
	if opt_mode == 'insert':
	write_message(" Failed: tag 001 found in the xml" \
	" submitted, you should use the option replace," \
	" correct or append to replace an existing" \
	" record. (-h for help)",
	verbose=1, stream=sys.stderr)
	return -1
	else:
	# we found the rec id and we are not in insert mode => continue
	# we try to match rec_id against the database:
	if find_record_from_recid(rec_id) is not None:
	# okay, 001 corresponds to some known record
	return int(rec_id)
	elif opt_mode in ('replace', 'replace_or_insert'):
	if task_get_option('force'):
	# we found the rec_id but it's not in the system and we are
	# requested to replace records. Therefore we create on the fly
	# a empty record allocating the recid.
	write_message(" Warning: tag 001 found in the xml with"
	" value %(rec_id)s, but rec_id %(rec_id)s does"
	" not exist. Since the mode replace was"
	" requested the rec_id %(rec_id)s is allocated"
	" on-the-fly." % {"rec_id" : rec_id},
	stream=sys.stderr)
	return create_new_record(rec_id=rec_id, pretend=pretend)
	else:
	# Since --force was not used we are going to raise an error
	write_message(" Failed: tag 001 found in the xml"
	" submitted with value %(rec_id)s. The"
	" corresponding record however does not"
	" exists. If you want to really create"
	" such record, please use the --force"
	" parameter when calling bibupload." % {
	"rec_id": rec_id}, stream=sys.stderr)
	return -1
	else:
	# The record doesn't exist yet. We shall have try to check
	# the SYSNO or OAI id later.
	write_message(" -Tag 001 value not found in database.",
	verbose=9)
	rec_id = None
	else:
	write_message(" -Tag 001 not found in the xml marc file.", verbose=9)

	if rec_id is None:
	# 2nd step we look for the SYSNO
	sysnos = record_get_field_values(record,
	CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[0:3],
	CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[3:4] != "_" and \
	CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[3:4] or "",
	CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[4:5] != "_" and \
	CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[4:5] or "",
	CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[5:6])
	if sysnos:
	sysno = sysnos[0] # there should be only one external SYSNO
	write_message(" -Checking if SYSNO " + sysno + \
	" exists in the database", verbose=9)
	# try to find the corresponding rec id from the database
	rec_id = find_record_from_sysno(sysno)
	if rec_id is not None:
	# rec_id found
	pass
	else:
	# The record doesn't exist yet. We will try to check
	# external and internal OAI ids later.
	write_message(" -Tag SYSNO value not found in database.",
	verbose=9)
	rec_id = None
	else:
	write_message(" -Tag SYSNO not found in the xml marc file.",
	verbose=9)

	if rec_id is None:
	# 2nd step we look for the external OAIID
	extoai_fields = record_get_field_instances(record,
	CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[0:3],
	CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[3:4] != "_" and \
	CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[3:4] or "",
	CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[4:5] != "_" and \
	CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[4:5] or "")
	if extoai_fields:
	for field in extoai_fields:
	extoaiid = field_get_subfield_values(field, CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[5:6])
	extoaisrc = field_get_subfield_values(field, CFG_BIBUPLOAD_EXTERNAL_OAIID_PROVENANCE_TAG[5:6])
	if extoaiid:
	extoaiid = extoaiid[0]
	if extoaisrc:
	extoaisrc = extoaisrc[0]
	else:
	extoaisrc = None
	write_message(" -Checking if EXTOAIID %s (%s) exists in the database" % (extoaiid, extoaisrc), verbose=9)
	# try to find the corresponding rec id from the database
	try:
	rec_ids = find_records_from_extoaiid(extoaiid, extoaisrc)
	except Error, e:
	write_message(e, verbose=1, stream=sys.stderr)
	return -1
	if rec_ids:
	# rec_id found
	rec_id = rec_ids.pop()
	break
	else:
	# The record doesn't exist yet. We will try to check
	# OAI id later.
	write_message(" -Tag EXTOAIID value not found in database.",
	verbose=9)
	rec_id = None
	else:
	write_message(" -Tag EXTOAIID not found in the xml marc file.", verbose=9)

	if rec_id is None:
	# 4th step we look for the OAI ID
	oaiidvalues = record_get_field_values(record,
	CFG_OAI_ID_FIELD[0:3],
	CFG_OAI_ID_FIELD[3:4] != "_" and \
	CFG_OAI_ID_FIELD[3:4] or "",
	CFG_OAI_ID_FIELD[4:5] != "_" and \
	CFG_OAI_ID_FIELD[4:5] or "",
	CFG_OAI_ID_FIELD[5:6])
	if oaiidvalues:
	oaiid = oaiidvalues[0] # there should be only one OAI ID
	write_message(" -Check if local OAI ID " + oaiid + \
	" exist in the database", verbose=9)

	# try to find the corresponding rec id from the database
	rec_id = find_record_from_oaiid(oaiid)
	if rec_id is not None:
	# rec_id found
	pass
	else:
	write_message(" -Tag OAI ID value not found in database.",
	verbose=9)
	rec_id = None
	else:
	write_message(" -Tag SYSNO not found in the xml marc file.",
	verbose=9)

	# Now we should have detected rec_id from SYSNO or OAIID
	# tags. (None otherwise.)
	if rec_id:
	if opt_mode == 'insert':
	write_message(" Failed : Record found in the database," \
	" you should use the option replace," \
	" correct or append to replace an existing" \
	" record. (-h for help)",
	verbose=1, stream=sys.stderr)
	return -1
	else:
	if opt_mode != 'insert' and \
	opt_mode != 'replace_or_insert':
	write_message(" Failed : Record not found in the database."\
	" Please insert the file before updating it."\
	" (-h for help)", verbose=1, stream=sys.stderr)
	return -1

	return rec_id and int(rec_id) or None

	### Insert functions

	def create_new_record(rec_id=None, pretend=False):
	"""
	Create new record in the database

	@param rec_id: if specified the new record will have this rec_id.
	@type rec_id: int
	@return: the allocated rec_id
	@rtype: int

	@note: in case of errors will be returned None
	"""
	if rec_id is not None:
	try:
	rec_id = int(rec_id)
	except (ValueError, TypeError), error:
	write_message(" Error during the creation_new_record function : %s "
	% error, verbose=1, stream=sys.stderr)
	return None
	if run_sql("SELECT id FROM bibrec WHERE id=%s", (rec_id, )):
	write_message(" Error during the creation_new_record function : the requested rec_id %s already exists." % rec_id)
	return None
	if pretend:
	if rec_id:
	return rec_id
	else:
	return run_sql("SELECT max(id)+1 FROM bibrec")[0][0]
	try:
	if rec_id is not None:
	return run_sql("INSERT INTO bibrec (id, creation_date, modification_date) VALUES (%s, NOW(), NOW())", (rec_id, ))
	else:
	return run_sql("INSERT INTO bibrec (creation_date, modification_date) VALUES (NOW(), NOW())")
	except Error, error:
	write_message(" Error during the creation_new_record function : %s " % error, verbose=1, stream=sys.stderr)
	return None

	def insert_bibfmt(id_bibrec, marc, format, modification_date='1970-01-01 00:00:00', pretend=False):
	"""Insert the format in the table bibfmt"""
	# compress the marc value
	pickled_marc = compress(marc)
	try:
	time.strptime(modification_date, "%Y-%m-%d %H:%M:%S")
	except ValueError:
	modification_date = '1970-01-01 00:00:00'

	query = """INSERT INTO bibfmt (id_bibrec, format, last_updated, value)
	VALUES (%s, %s, %s, %s)"""
	try:
	if not pretend:
	row_id = run_sql(query, (id_bibrec, format, modification_date, pickled_marc))
	return row_id
	else:
	return 1
	except Error, error:
	write_message(" Error during the insert_bibfmt function : %s "
	% error, verbose=1, stream=sys.stderr)
	return None

	def insert_record_bibxxx(tag, value, pretend=False):
	"""Insert the record into bibxxx"""
	# determine into which table one should insert the record
	table_name = 'bib'+tag[0:2]+'x'

	# check if the tag, value combination exists in the table
	query = """SELECT id,value FROM %s """ % table_name
	query += """ WHERE tag=%s AND value=%s"""
	params = (tag, value)
	try:
	res = run_sql(query, params)
	except Error, error:
	write_message(" Error during the insert_record_bibxxx function : %s "
	% error, verbose=1, stream=sys.stderr)

	# Note: compare now the found values one by one and look for
	# string binary equality (e.g. to respect lowercase/uppercase
	# match), regardless of the charset etc settings. Ideally we
	# could use a BINARY operator in the above SELECT statement, but
	# we would have to check compatibility on various MySQLdb versions
	# etc; this approach checks all matched values in Python, not in
	# MySQL, which is less cool, but more conservative, so it should
	# work better on most setups.
	for row in res:
	row_id = row[0]
	row_value = row[1]
	if row_value == value:
	return (table_name, row_id)

	# We got here only when the tag,value combination was not found,
	# so it is now necessary to insert the tag,value combination into
	# bibxxx table as new.
	query = """INSERT INTO %s """ % table_name
	query += """ (tag, value) values (%s , %s)"""
	params = (tag, value)
	try:
	if not pretend:
	row_id = run_sql(query, params)
	else:
	return (table_name, 1)
	except Error, error:
	write_message(" Error during the insert_record_bibxxx function : %s "
	% error, verbose=1, stream=sys.stderr)
	return (table_name, row_id)

	def insert_record_bibrec_bibxxx(table_name, id_bibxxx,
	field_number, id_bibrec, pretend=False):
	"""Insert the record into bibrec_bibxxx"""
	# determine into which table one should insert the record
	full_table_name = 'bibrec_'+ table_name

	# insert the proper row into the table
	query = """INSERT INTO %s """ % full_table_name
	query += """(id_bibrec,id_bibxxx, field_number) values (%s , %s, %s)"""
	params = (id_bibrec, id_bibxxx, field_number)
	try:
	if not pretend:
	res = run_sql(query, params)
	else:
	return 1
	except Error, error:
	write_message(" Error during the insert_record_bibrec_bibxxx"
	" function 2nd query : %s " % error, verbose=1, stream=sys.stderr)
	return res

	def synchronize_8564(rec_id, record, record_had_FFT, pretend=False):
	"""
	Synchronize 8564_ tags and BibDocFile tables.

	This function directly manipulate the record parameter.

	@type rec_id: positive integer
	@param rec_id: the record identifier.
	@param record: the record structure as created by bibrecord.create_record
	@type record_had_FFT: boolean
	@param record_had_FFT: True if the incoming bibuploaded-record used FFT
	@return: the manipulated record (which is also modified as a side effect)
	"""
	def merge_marc_into_bibdocfile(field, pretend=False):
	"""
	Internal function that reads a single field and store its content
	in BibDocFile tables.
	@param field: the 8564_ field containing a BibDocFile URL.
	"""
	write_message('Merging field: %s' % (field, ), verbose=9)
	url = field_get_subfield_values(field, 'u')[:1] or field_get_subfield_values(field, 'q')[:1]
	description = field_get_subfield_values(field, 'y')[:1]
	comment = field_get_subfield_values(field, 'z')[:1]
	if url:
	recid, docname, format = decompose_bibdocfile_url(url[0])
	if recid != rec_id:
	write_message("INFO: URL %s is not pointing to a fulltext owned by this record (%s)" % (url, recid), stream=sys.stderr)
	else:
	try:
	bibdoc = BibRecDocs(recid).get_bibdoc(docname)
	if description and not pretend:
	bibdoc.set_description(description[0], format)
	if comment and not pretend:
	bibdoc.set_comment(comment[0], format)
	except InvenioWebSubmitFileError:
	## Apparently the referenced docname doesn't exist anymore.
	## Too bad. Let's skip it.
	write_message("WARNING: docname %s does not seem to exist for record %s. Has it been renamed outside FFT?" % (docname, recid), stream=sys.stderr)

	def merge_bibdocfile_into_marc(field, subfields):
	"""
	Internal function that reads BibDocFile table entries referenced by
	the URL in the given 8564_ field and integrate the given information
	directly with the provided subfields.

	@param field: the 8564_ field containing a BibDocFile URL.
	@param subfields: the subfields corresponding to the BibDocFile URL
	generated after BibDocFile tables.
	"""
	write_message('Merging subfields %s into field %s' % (subfields, field), verbose=9)
	subfields = dict(subfields) ## We make a copy not to have side-effects
	subfield_to_delete = []
	for subfield_position, (code, value) in enumerate(field_get_subfield_instances(field)):
	## For each subfield instance already existing...
	if code in subfields:
	## ...We substitute it with what is in BibDocFile tables
	record_modify_subfield(record, '856', code, subfields[code],
	subfield_position, field_position_global=field[4])
	del subfields[code]
	else:
	## ...We delete it otherwise
	subfield_to_delete.append(subfield_position)

	subfield_to_delete.sort()

	for counter, position in enumerate(subfield_to_delete):
	## FIXME: Very hackish algorithm. Since deleting a subfield
	## will alterate the position of following subfields, we
	## are taking note of this and adjusting further position
	## by using a counter.
	record_delete_subfield_from(record, '856', position - counter,
	field_position_global=field[4])

	subfields = subfields.items()
	subfields.sort()
	for code, value in subfields:
	## Let's add non-previously existing subfields
	record_add_subfield_into(record, '856', code, value,
	field_position_global=field[4])

	def get_bibdocfile_managed_info():
	"""
	Internal function to eturns a dictionary of
	BibDocFile URL -> wanna-be subfields.

	@rtype: mapping
	@return: BibDocFile URL -> wanna-be subfields dictionary
	"""
	ret = {}
	bibrecdocs = BibRecDocs(rec_id)
	latest_files = bibrecdocs.list_latest_files(list_hidden=False)
	for afile in latest_files:
	url = afile.get_url()
	ret[url] = {'u' : url}
	description = afile.get_description()
	comment = afile.get_comment()
	subformat = afile.get_subformat()
	if description:
	ret[url]['y'] = description
	if comment:
	ret[url]['z'] = comment
	if subformat:
	ret[url]['x'] = subformat

	return ret

	write_message("Synchronizing MARC of recid '%s' with:\n%s" % (rec_id, record), verbose=9)
	tags856s = record_get_field_instances(record, '856', '%', '%')
	write_message("Original 856%% instances: %s" % tags856s, verbose=9)
	tags8564s_to_add = get_bibdocfile_managed_info()
	write_message("BibDocFile instances: %s" % tags8564s_to_add, verbose=9)
	positions_tags8564s_to_remove = []

	for local_position, field in enumerate(tags856s):
	if field[1] == '4' and field[2] == ' ':
	write_message('Analysing %s' % (field, ), verbose=9)
	for url in field_get_subfield_values(field, 'u') + field_get_subfield_values(field, 'q'):
	if url in tags8564s_to_add:
	if record_had_FFT:
	merge_bibdocfile_into_marc(field, tags8564s_to_add[url])
	else:
	merge_marc_into_bibdocfile(field, pretend=pretend)
	del tags8564s_to_add[url]
	break
	elif bibdocfile_url_p(url) and decompose_bibdocfile_url(url)[0] == rec_id:
	positions_tags8564s_to_remove.append(local_position)
	write_message("%s to be deleted and re-synchronized" % (field, ), verbose=9)
	break

	record_delete_fields(record, '856', positions_tags8564s_to_remove)

	tags8564s_to_add = tags8564s_to_add.values()
	tags8564s_to_add.sort()
	for subfields in tags8564s_to_add:
	subfields = subfields.items()
	subfields.sort()
	record_add_field(record, '856', '4', ' ', subfields=subfields)

	write_message('Final record: %s' % record, verbose=9)
	return record

	def elaborate_fft_tags(record, rec_id, mode, pretend=False):
	"""
	Process FFT tags that should contain $a with file pathes or URLs
	to get the fulltext from. This function enriches record with
	proper 8564 URL tags, downloads fulltext files and stores them
	into var/data structure where appropriate.

	CFG_BIBUPLOAD_WGET_SLEEP_TIME defines time to sleep in seconds in
	between URL downloads.

	Note: if an FFT tag contains multiple $a subfields, we upload them
	into different 856 URL tags in the metadata. See regression test
	case test_multiple_fft_insert_via_http().
	"""

	# Let's define some handy sub procedure.
	def _add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags, pretend=False):
	"""Adds a new format for a given bibdoc. Returns True when everything's fine."""
	write_message('Add new format to %s url: %s, format: %s, docname: %s, doctype: %s, newname: %s, description: %s, comment: %s, flags: %s' % (repr(bibdoc), url, format, docname, doctype, newname, description, comment, flags), verbose=9)
	try:
	if not url: # Not requesting a new url. Just updating comment & description
	return _update_description_and_comment(bibdoc, docname, format, description, comment, flags, pretend=pretend)
	try:
	if not pretend:
	bibdoc.add_file_new_format(url, description=description, comment=comment, flags=flags)
	except StandardError, e:
	write_message("('%s', '%s', '%s', '%s', '%s', '%s', '%s', '%s') not inserted because format already exists (%s)." % (url, format, docname, doctype, newname, description, comment, flags, e), stream=sys.stderr)
	raise
	except Exception, e:
	write_message("Error in adding '%s' as a new format because of: %s" % (url, e), stream=sys.stderr)
	raise
	return True

	def _add_new_version(bibdoc, url, format, docname, doctype, newname, description, comment, flags, pretend=False):
	"""Adds a new version for a given bibdoc. Returns True when everything's fine."""
	write_message('Add new version to %s url: %s, format: %s, docname: %s, doctype: %s, newname: %s, description: %s, comment: %s, flags: %s' % (repr(bibdoc), url, format, docname, doctype, newname, description, comment, flags))
	try:
	if not url:
	return _update_description_and_comment(bibdoc, docname, format, description, comment, flags, pretend=pretend)
	try:
	if not pretend:
	bibdoc.add_file_new_version(url, description=description, comment=comment, flags=flags)
	except StandardError, e:
	write_message("('%s', '%s', '%s', '%s', '%s', '%s', '%s', '%s') not inserted because '%s'." % (url, format, docname, doctype, newname, description, comment, flags, e), stream=sys.stderr)
	raise
	except Exception, e:
	write_message("Error in adding '%s' as a new version because of: %s" % (url, e), stream=sys.stderr)
	raise
	return True

	def _update_description_and_comment(bibdoc, docname, format, description, comment, flags, pretend=False):
	"""Directly update comments and descriptions."""
	write_message('Just updating description and comment for %s with format %s with description %s, comment %s and flags %s' % (docname, format, description, comment, flags), verbose=9)
	try:
	if not pretend:
	bibdoc.set_description(description, format)
	bibdoc.set_comment(comment, format)
	for flag in CFG_BIBDOCFILE_AVAILABLE_FLAGS:
	if flag in flags:
	bibdoc.set_flag(flag, format)
	else:
	bibdoc.unset_flag(flag, format)
	except StandardError, e:
	write_message("('%s', '%s', '%s', '%s', '%s') description and comment not updated because '%s'." % (docname, format, description, comment, flags, e))
	raise
	return True

	if mode == 'delete':
	raise StandardError('FFT tag specified but bibupload executed in --delete mode')

	tuple_list = extract_tag_from_record(record, 'FFT')
	if tuple_list: # FFT Tags analysis
	write_message("FFTs: "+str(tuple_list), verbose=9)
	docs = {} # docnames and their data

	for fft in record_get_field_instances(record, 'FFT', ' ', ' '):
	# Let's discover the type of the document
	# This is a legacy field and will not be enforced any particular
	# check on it.
	doctype = field_get_subfield_values(fft, 't')
	if doctype:
	doctype = doctype[0]
	else: # Default is Main
	doctype = 'Main'

	# Let's discover the url.
	url = field_get_subfield_values(fft, 'a')
	if url:
	url = url[0]
	try:
	check_valid_url(url)
	except StandardError, e:
	raise StandardError, "fft '%s' specifies in $a a location ('%s') with problems: %s" % (fft, url, e)
	else:
	url = ''

	# Let's discover the description
	description = field_get_subfield_values(fft, 'd')
	if description != []:
	description = description[0]
	else:
	if mode == 'correct' and doctype != 'FIX-MARC':
	## If the user require to correct, and do not specify
	## a description this means she really want to
	## modify the description.
	description = ''
	else:
	description = KEEP_OLD_VALUE

	# Let's discover the desired docname to be created/altered
	name = field_get_subfield_values(fft, 'n')
	if name:
	## Let's remove undesired extensions
	name = file_strip_ext(name[0] + '.pdf')
	else:
	if url:
	name = get_docname_from_url(url)
	elif mode != 'correct' and doctype != 'FIX-MARC':
	raise StandardError, "Warning: fft '%s' doesn't specifies either a location in $a or a docname in $n" % str(fft)
	else:
	continue

	# Let's discover the desired new docname in case we want to change it
	newname = field_get_subfield_values(fft, 'm')
	if newname:
	newname = file_strip_ext(newname[0] + '.pdf')
	else:
	newname = name

	# Let's discover the desired format
	format = field_get_subfield_values(fft, 'f')
	if format:
	format = normalize_format(format[0])
	else:
	if url:
	format = guess_format_from_url(url)
	else:
	format = ""

	# Let's discover the icon
	icon = field_get_subfield_values(fft, 'x')
	if icon != []:
	icon = icon[0]
	if icon != KEEP_OLD_VALUE:
	try:
	check_valid_url(icon)
	except StandardError, e:
	raise StandardError, "fft '%s' specifies in $x an icon ('%s') with problems: %s" % (fft, icon, e)
	else:
	icon = ''

	# Let's discover the comment
	comment = field_get_subfield_values(fft, 'z')
	if comment != []:
	comment = comment[0]
	else:
	if mode == 'correct' and doctype != 'FIX-MARC':
	## See comment on description
	comment = ''
	else:
	comment = KEEP_OLD_VALUE

	# Let's discover the restriction
	restriction = field_get_subfield_values(fft, 'r')
	if restriction != []:
	restriction = restriction[0]
	else:
	if mode == 'correct' and doctype != 'FIX-MARC':
	## See comment on description
	restriction = ''
	else:
	restriction = KEEP_OLD_VALUE

	version = field_get_subfield_values(fft, 'v')
	if version:
	version = version[0]
	else:
	version = ''

	flags = field_get_subfield_values(fft, 'o')
	for flag in flags:
	if flag not in CFG_BIBDOCFILE_AVAILABLE_FLAGS:
	raise StandardError, "fft '%s' specifies a non available flag: %s" % (fft, flag)

	if docs.has_key(name): # new format considered
	(doctype2, newname2, restriction2, version2, urls) = docs[name]
	if doctype2 != doctype:
	raise StandardError, "fft '%s' specifies a different doctype from previous fft with docname '%s'" % (str(fft), name)
	if newname2 != newname:
	raise StandardError, "fft '%s' specifies a different newname from previous fft with docname '%s'" % (str(fft), name)
	if restriction2 != restriction:
	raise StandardError, "fft '%s' specifies a different restriction from previous fft with docname '%s'" % (str(fft), name)
	if version2 != version:
	raise StandardError, "fft '%x' specifies a different version than the previous fft with docname '%s'" % (str(fft), name)
	for (url2, format2, description2, comment2, flags2) in urls:
	if format == format2:
	raise StandardError, "fft '%s' specifies a second file '%s' with the same format '%s' from previous fft with docname '%s'" % (str(fft), url, format, name)
	if url or format:
	urls.append((url, format, description, comment, flags))
	if icon:
	urls.append((icon, icon[len(file_strip_ext(icon)):] + ';icon', description, comment, flags))
	else:
	if url or format:
	docs[name] = (doctype, newname, restriction, version, [(url, format, description, comment, flags)])
	if icon:
	docs[name][4].append((icon, icon[len(file_strip_ext(icon)):] + ';icon', description, comment, flags))
	elif icon:
	docs[name] = (doctype, newname, restriction, version, [(icon, icon[len(file_strip_ext(icon)):] + ';icon', description, comment, flags)])
	else:
	docs[name] = (doctype, newname, restriction, version, [])

	write_message('Result of FFT analysis:\n\tDocs: %s' % (docs,), verbose=9)

	# Let's remove all FFT tags
	record_delete_field(record, 'FFT', ' ', ' ')

	# Preprocessed data elaboration
	bibrecdocs = BibRecDocs(rec_id)

	## Let's pre-download all the URLs to see if, in case of mode 'correct' or 'append'
	## we can avoid creating a new revision.
	for docname, (doctype, newname, restriction, version, urls) in docs.items():
	downloaded_urls = []
	try:
	bibdoc = bibrecdocs.get_bibdoc(docname)
	except InvenioWebSubmitFileError:
	## A bibdoc with the given docname does not exists.
	## So there is no chance we are going to revise an existing
	## format with an identical file :-)
	bibdoc = None

	new_revision_needed = False
	for url, format, description, comment, flags in urls:
	if url:
	try:
	downloaded_url = download_url(url, format)
	write_message("%s saved into %s" % (url, downloaded_url), verbose=9)
	except Exception, err:
	write_message("Error in downloading '%s' because of: %s" % (url, err), stream=sys.stderr)
	raise
	if mode == 'correct' and bibdoc is not None and not new_revision_needed:
	downloaded_urls.append((downloaded_url, format, description, comment, flags))
	if not bibdoc.check_file_exists(downloaded_url):
	new_revision_needed = True
	else:
	write_message("WARNING: %s is already attached to bibdoc %s for recid %s" % (url, docname, rec_id), stream=sys.stderr)
	elif mode == 'append' and bibdoc is not None:
	if not bibdoc.check_file_exists(downloaded_url):
	downloaded_urls.append((downloaded_url, format, description, comment, flags))
	else:
	write_message("WARNING: %s is already attached to bibdoc %s for recid %s" % (url, docname, rec_id), stream=sys.stderr)
	else:
	downloaded_urls.append((downloaded_url, format, description, comment, flags))
	else:
	downloaded_urls.append(('', format, description, comment, flags))
	if mode == 'correct' and bibdoc is not None and not new_revision_needed:
	## Since we don't need a new revision (because all the files
	## that are being uploaded are different)
	## we can simply remove the urls but keep the other information
	write_message("No need to add a new revision for docname %s for recid %s" % (docname, rec_id), verbose=2)
	docs[docname] = (doctype, newname, restriction, version, [('', format, description, comment, flags) for (dummy, format, description, comment, flags) in downloaded_urls])
	for downloaded_url, dummy, dummy, dummy, dummy in downloaded_urls:
	## Let's free up some space :-)
	if downloaded_url and os.path.exists(downloaded_url):
	os.remove(downloaded_url)
	else:
	if downloaded_urls or mode != 'append':
	docs[docname] = (doctype, newname, restriction, version, downloaded_urls)
	else:
	## In case we are in append mode and there are no urls to append
	## we discard the whole FFT
	del docs[docname]

	if mode == 'replace': # First we erase previous bibdocs
	if not pretend:
	for bibdoc in bibrecdocs.list_bibdocs():
	bibdoc.delete()
	bibrecdocs.build_bibdoc_list()

	for docname, (doctype, newname, restriction, version, urls) in docs.iteritems():
	write_message("Elaborating olddocname: '%s', newdocname: '%s', doctype: '%s', restriction: '%s', urls: '%s', mode: '%s'" % (docname, newname, doctype, restriction, urls, mode), verbose=9)
	if mode in ('insert', 'replace'): # new bibdocs, new docnames, new marc
	if newname in bibrecdocs.get_bibdoc_names():
	write_message("('%s', '%s') not inserted because docname already exists." % (newname, urls), stream=sys.stderr)
	raise StandardError
	try:
	if not pretend:
	bibdoc = bibrecdocs.add_bibdoc(doctype, newname)
	bibdoc.set_status(restriction)
	else:
	bibdoc = None
	except Exception, e:
	write_message("('%s', '%s', '%s') not inserted because: '%s'." % (doctype, newname, urls, e), stream=sys.stderr)
	raise StandardError
	for (url, format, description, comment, flags) in urls:
	assert(_add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags, pretend=pretend))
	elif mode == 'replace_or_insert': # to be thought as correct_or_insert
	for bibdoc in bibrecdocs.list_bibdocs():
	if bibdoc.get_docname() == docname:
	if doctype not in ('PURGE', 'DELETE', 'EXPUNGE', 'REVERT', 'FIX-ALL', 'FIX-MARC', 'DELETE-FILE'):
	if newname != docname:
	try:
	if not pretend:
	bibdoc.change_name(newname)
	## Let's refresh the list of bibdocs.
	bibrecdocs.build_bibdoc_list()
	except StandardError, e:
	write_message(e, stream=sys.stderr)
	raise
	found_bibdoc = False
	for bibdoc in bibrecdocs.list_bibdocs():
	if bibdoc.get_docname() == newname:
	found_bibdoc = True
	if doctype == 'PURGE':
	if not pretend:
	bibdoc.purge()
	elif doctype == 'DELETE':
	if not pretend:
	bibdoc.delete()
	elif doctype == 'EXPUNGE':
	if not pretend:
	bibdoc.expunge()
	elif doctype == 'FIX-ALL':
	if not pretend:
	bibrecdocs.fix(docname)
	elif doctype == 'FIX-MARC':
	pass
	elif doctype == 'DELETE-FILE':
	if urls:
	for (url, format, description, comment, flags) in urls:
	if not pretend:
	bibdoc.delete_file(format, version)
	elif doctype == 'REVERT':
	try:
	if not pretend:
	bibdoc.revert(version)
	except Exception, e:
	write_message('(%s, %s) not correctly reverted: %s' % (newname, version, e), stream=sys.stderr)
	raise
	else:
	if restriction != KEEP_OLD_VALUE:
	if not pretend:
	bibdoc.set_status(restriction)
	# Since the docname already existed we have to first
	# bump the version by pushing the first new file
	# then pushing the other files.
	if urls:
	(first_url, first_format, first_description, first_comment, first_flags) = urls[0]
	other_urls = urls[1:]
	assert(_add_new_version(bibdoc, first_url, first_format, docname, doctype, newname, first_description, first_comment, first_flags, pretend=pretend))
	for (url, format, description, comment, flags) in other_urls:
	assert(_add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags, pretend=pretend))
	## Let's refresh the list of bibdocs.
	bibrecdocs.build_bibdoc_list()
	if not found_bibdoc:
	if not pretend:
	bibdoc = bibrecdocs.add_bibdoc(doctype, newname)
	bibdoc.set_status(restriction)
	for (url, format, description, comment, flags) in urls:
	assert(_add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags))
	elif mode == 'correct':
	for bibdoc in bibrecdocs.list_bibdocs():
	if bibdoc.get_docname() == docname:
	if doctype not in ('PURGE', 'DELETE', 'EXPUNGE', 'REVERT', 'FIX-ALL', 'FIX-MARC', 'DELETE-FILE'):
	if newname != docname:
	try:
	if not pretend:
	bibdoc.change_name(newname)
	## Let's refresh the list of bibdocs.
	bibrecdocs.build_bibdoc_list()
	except StandardError, e:
	write_message('Error in renaming %s to %s: %s' % (docname, newname, e), stream=sys.stderr)
	raise
	found_bibdoc = False
	for bibdoc in bibrecdocs.list_bibdocs():
	if bibdoc.get_docname() == newname:
	found_bibdoc = True
	if doctype == 'PURGE':
	if not pretend:
	bibdoc.purge()
	elif doctype == 'DELETE':
	if not pretend:
	bibdoc.delete()
	elif doctype == 'EXPUNGE':
	if not pretend:
	bibdoc.expunge()
	elif doctype == 'FIX-ALL':
	if not pretend:
	bibrecdocs.fix(newname)
	elif doctype == 'FIX-MARC':
	pass
	elif doctype == 'DELETE-FILE':
	if urls:
	for (url, format, description, comment, flags) in urls:
	if not pretend:
	bibdoc.delete_file(format, version)
	elif doctype == 'REVERT':
	try:
	if not pretend:
	bibdoc.revert(version)
	except Exception, e:
	write_message('(%s, %s) not correctly reverted: %s' % (newname, version, e), stream=sys.stderr)
	raise
	else:
	if restriction != KEEP_OLD_VALUE:
	if not pretend:
	bibdoc.set_status(restriction)
	if urls:
	(first_url, first_format, first_description, first_comment, first_flags) = urls[0]
	other_urls = urls[1:]
	assert(_add_new_version(bibdoc, first_url, first_format, docname, doctype, newname, first_description, first_comment, first_flags, pretend=pretend))
	for (url, format, description, comment, flags) in other_urls:
	assert(_add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags, pretend=pretend))
	## Let's refresh the list of bibdocs.
	bibrecdocs.build_bibdoc_list()
	if not found_bibdoc:
	if doctype in ('PURGE', 'DELETE', 'EXPUNGE', 'FIX-ALL', 'FIX-MARC', 'DELETE-FILE', 'REVERT'):
	write_message("('%s', '%s', '%s') not performed because '%s' docname didn't existed." % (doctype, newname, urls, docname), stream=sys.stderr)
	raise StandardError
	else:
	if not pretend:
	bibdoc = bibrecdocs.add_bibdoc(doctype, newname)
	bibdoc.set_status(restriction)
	for (url, format, description, comment, flags) in urls:
	assert(_add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags))
	elif mode == 'append':
	try:
	found_bibdoc = False
	for bibdoc in bibrecdocs.list_bibdocs():
	if bibdoc.get_docname() == docname:
	found_bibdoc = True
	for (url, format, description, comment, flags) in urls:
	assert(_add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags, pretend=pretend))
	if not found_bibdoc:
	try:
	if not pretend:
	bibdoc = bibrecdocs.add_bibdoc(doctype, docname)
	bibdoc.set_status(restriction)
	for (url, format, description, comment, flags) in urls:
	assert(_add_new_format(bibdoc, url, format, docname, doctype, newname, description, comment, flags))
	except Exception, e:
	register_exception()
	write_message("('%s', '%s', '%s') not appended because: '%s'." % (doctype, newname, urls, e), stream=sys.stderr)
	raise
	except:
	register_exception()
	raise
	return record

	def insert_fmt_tags(record, rec_id, opt_mode, pretend=False):
	"""Process and insert FMT tags"""

	fmt_fields = record_get_field_instances(record, 'FMT')
	if fmt_fields:
	for fmt_field in fmt_fields:
	# Get the d, f, g subfields of the FMT tag
	try:
	d_value = field_get_subfield_values(fmt_field, "d")[0]
	except IndexError:
	d_value = ""
	try:
	f_value = field_get_subfield_values(fmt_field, "f")[0]
	except IndexError:
	f_value = ""
	try:
	g_value = field_get_subfield_values(fmt_field, "g")[0]
	except IndexError:
	g_value = ""
	# Update the format
	if not pretend:
	res = update_bibfmt_format(rec_id, g_value, f_value, d_value, pretend=pretend)
	if res == 1:
	write_message(" Failed: Error during update_bibfmt", verbose=1, stream=sys.stderr)

	# If we are in format mode, we only care about the FMT tag
	if opt_mode == 'format':
	return 0
	# We delete the FMT Tag of the record
	record_delete_field(record, 'FMT')
	write_message(" -Delete field FMT from record : DONE", verbose=2)
	return record

	elif opt_mode == 'format':
	write_message(" Failed: Format updated failed : No tag FMT found", verbose=1, stream=sys.stderr)
	return None
	else:
	return record


	### Update functions

	def update_bibrec_modif_date(now, bibrec_id, pretend=False):
	"""Update the date of the record in bibrec table """
	query = """UPDATE bibrec SET modification_date=%s WHERE id=%s"""
	params = (now, bibrec_id)
	try:
	if not pretend:
	run_sql(query, params)
	write_message(" -Update record modification date : DONE" , verbose=2)
	except Error, error:
	write_message(" Error during update_bibrec_modif_date function : %s" % error,
	verbose=1, stream=sys.stderr)

	def update_bibfmt_format(id_bibrec, format_value, format_name, modification_date=None, pretend=False):
	"""Update the format in the table bibfmt"""
	if modification_date is None:
	modification_date = time.strftime('%Y-%m-%d %H:%M:%S')
	else:
	try:
	time.strptime(modification_date, "%Y-%m-%d %H:%M:%S")
	except ValueError:
	modification_date = '1970-01-01 00:00:00'

	# We check if the format is already in bibFmt
	nb_found = find_record_format(id_bibrec, format_name)
	if nb_found == 1:
	# we are going to update the format
	# compress the format_value value
	pickled_format_value = compress(format_value)
	# update the format:
	query = """UPDATE bibfmt SET last_updated=%s, value=%s WHERE id_bibrec=%s AND format=%s"""
	params = (modification_date, pickled_format_value, id_bibrec, format_name)
	try:
	if not pretend:
	row_id = run_sql(query, params)
	if not pretend and row_id is None:
	write_message(" Failed: Error during update_bibfmt_format function", verbose=1, stream=sys.stderr)
	return 1
	else:
	write_message(" -Update the format %s in bibfmt : DONE" % format_name , verbose=2)
	return 0
	except Error, error:
	write_message(" Error during the update_bibfmt_format function : %s " % error, verbose=1, stream=sys.stderr)

	elif nb_found > 1:
	write_message(" Failed: Same format %s found several time in bibfmt for the same record." % format_name, verbose=1, stream=sys.stderr)
	return 1
	else:
	# Insert the format information in BibFMT
	res = insert_bibfmt(id_bibrec, format_value, format_name, modification_date, pretend=pretend)
	if res is None:
	write_message(" Failed: Error during insert_bibfmt", verbose=1, stream=sys.stderr)
	return 1
	else:
	write_message(" -Insert the format %s in bibfmt : DONE" % format_name , verbose=2)
	return 0

	def delete_bibfmt_format(id_bibrec, format_name, pretend=False):
	"""
	Delete format FORMAT_NAME from bibfmt table for record ID_BIBREC.
	"""
	if not pretend:
	run_sql("DELETE FROM bibfmt WHERE id_bibrec=%s and format=%s", (id_bibrec, format_name))
	return 0

	def archive_marcxml_for_history(recID, pretend=False):
	"""
	Archive current MARCXML format of record RECID from BIBFMT table
	into hstRECORD table. Useful to keep MARCXML history of records.

	Return 0 if everything went fine. Return 1 otherwise.
	"""
	try:
	res = run_sql("SELECT id_bibrec, value, last_updated FROM bibfmt WHERE format='xm' AND id_bibrec=%s",
	(recID,))
	if res and not pretend:
	run_sql("""INSERT INTO hstRECORD (id_bibrec, marcxml, job_id, job_name, job_person, job_date, job_details)
	VALUES (%s,%s,%s,%s,%s,%s,%s)""",
	(res[0][0], res[0][1], task_get_task_param('task_id', 0), 'bibupload', task_get_task_param('user','UNKNOWN'), res[0][2],
	'mode: ' + task_get_option('mode','UNKNOWN') + '; file: ' + task_get_option('file_path','UNKNOWN') + '.'))
	except Error, error:
	write_message(" Error during archive_marcxml_for_history: %s " % error,
	verbose=1, stream=sys.stderr)
	return 1
	return 0

	def update_database_with_metadata(record, rec_id, oai_rec_id = "oai", pretend=False):
	"""Update the database tables with the record and the record id given in parameter"""
	for tag in record.keys():
	# check if tag is not a special one:
	if tag not in CFG_BIBUPLOAD_SPECIAL_TAGS:
	# for each tag there is a list of tuples representing datafields
	tuple_list = record[tag]
	# this list should contain the elements of a full tag [tag, ind1, ind2, subfield_code]
	tag_list = []
	tag_list.append(tag)
	for single_tuple in tuple_list:
	# these are the contents of a single tuple
	subfield_list = single_tuple[0]
	ind1 = single_tuple[1]
	ind2 = single_tuple[2]
	# append the ind's to the full tag
	if ind1 == '' or ind1 == ' ':
	tag_list.append('_')
	else:
	tag_list.append(ind1)
	if ind2 == '' or ind2 == ' ':
	tag_list.append('_')
	else:
	tag_list.append(ind2)
	datafield_number = single_tuple[4]

	if tag in CFG_BIBUPLOAD_SPECIAL_TAGS:
	# nothing to do for special tags (FFT, FMT)
	pass
	elif tag in CFG_BIBUPLOAD_CONTROLFIELD_TAGS and tag != "001":
	value = single_tuple[3]
	# get the full tag
	full_tag = ''.join(tag_list)

	# update the tables
	write_message(" insertion of the tag "+full_tag+" with the value "+value, verbose=9)
	# insert the tag and value into into bibxxx
	(table_name, bibxxx_row_id) = insert_record_bibxxx(full_tag, value, pretend=pretend)
	#print 'tname, bibrow', table_name, bibxxx_row_id;
	if table_name is None or bibxxx_row_id is None:
	write_message(" Failed : during insert_record_bibxxx", verbose=1, stream=sys.stderr)
	# connect bibxxx and bibrec with the table bibrec_bibxxx
	res = insert_record_bibrec_bibxxx(table_name, bibxxx_row_id, datafield_number, rec_id, pretend=pretend)
	if res is None:
	write_message(" Failed : during insert_record_bibrec_bibxxx", verbose=1, stream=sys.stderr)
	else:
	# get the tag and value from the content of each subfield
	for subfield in subfield_list:
	subtag = subfield[0]
	value = subfield[1]
	tag_list.append(subtag)
	# get the full tag
	full_tag = ''.join(tag_list)
	# update the tables
	write_message(" insertion of the tag "+full_tag+" with the value "+value, verbose=9)
	# insert the tag and value into into bibxxx
	(table_name, bibxxx_row_id) = insert_record_bibxxx(full_tag, value, pretend=pretend)
	if table_name is None or bibxxx_row_id is None:
	write_message(" Failed : during insert_record_bibxxx", verbose=1, stream=sys.stderr)
	# connect bibxxx and bibrec with the table bibrec_bibxxx
	res = insert_record_bibrec_bibxxx(table_name, bibxxx_row_id, datafield_number, rec_id, pretend=pretend)
	if res is None:
	write_message(" Failed : during insert_record_bibrec_bibxxx", verbose=1, stream=sys.stderr)
	# remove the subtag from the list
	tag_list.pop()
	tag_list.pop()
	tag_list.pop()
	tag_list.pop()
	write_message(" -Update the database with metadata : DONE", verbose=2)

	log_record_uploading(oai_rec_id, task_get_task_param('task_id', 0), rec_id, 'P', pretend=pretend)

	def append_new_tag_to_old_record(record, rec_old, opt_tag, opt_mode):
	"""Append new tags to a old record"""

	def _append_tag(tag):
	# Reference mode append only reference tag
	if opt_mode == 'reference':
	if tag == CFG_BIBUPLOAD_REFERENCE_TAG:
	for single_tuple in record[tag]:
	# We retrieve the information of the tag
	subfield_list = single_tuple[0]
	ind1 = single_tuple[1]
	ind2 = single_tuple[2]
	# We add the datafield to the old record
	write_message(" Adding tag: %s ind1=%s ind2=%s code=%s" % (tag, ind1, ind2, subfield_list), verbose=9)
	newfield_number = record_add_field(rec_old, tag, ind1,
	ind2, subfields=subfield_list)
	if newfield_number is None:
	write_message(" Error when adding the field"+tag, verbose=1, stream=sys.stderr)
	else:
	if tag in CFG_BIBUPLOAD_CONTROLFIELD_TAGS:
	if tag == '001':
	pass
	else:
	# if it is a controlfield,just access the value
	for single_tuple in record[tag]:
	controlfield_value = single_tuple[3]
	# add the field to the old record
	newfield_number = record_add_field(rec_old, tag,
	controlfield_value=controlfield_value)
	if newfield_number is None:
	write_message(" Error when adding the field"+tag, verbose=1, stream=sys.stderr)
	else:
	# For each tag there is a list of tuples representing datafields
	for single_tuple in record[tag]:
	# We retrieve the information of the tag
	subfield_list = single_tuple[0]
	ind1 = single_tuple[1]
	ind2 = single_tuple[2]
	if '%s%s%s' % (tag, ind1 == ' ' and '_' or ind1, ind2 == ' ' and '_' or ind2) in (CFG_BIBUPLOAD_EXTERNAL_OAIID_TAG[:5], CFG_BIBUPLOAD_EXTERNAL_SYSNO_TAG[:5]):
	## We don't want to append the external identifier
	## if it is already existing.
	if record_find_field(rec_old, tag, single_tuple)[0] is not None:
	write_message(" Not adding tag: %s ind1=%s ind2=%s subfields=%s: it's already there" % (tag, ind1, ind2, subfield_list), verbose=9)
	continue
	# We add the datafield to the old record
	write_message(" Adding tag: %s ind1=%s ind2=%s subfields=%s" % (tag, ind1, ind2, subfield_list), verbose=9)
	newfield_number = record_add_field(rec_old, tag, ind1,
	ind2, subfields=subfield_list)
	if newfield_number is None:
	write_message(" Error when adding the field"+tag, verbose=1, stream=sys.stderr)

	if opt_tag is not None:
	_append_tag(opt_tag)
	else:
	# Go through each tag in the appended record
	for tag in record:
	_append_tag(tag)
	return rec_old

	def copy_strong_tags_from_old_record(record, rec_old):
	"""
	Look for strong tags in RECORD and REC_OLD. If no strong tags are
	found in RECORD, then copy them over from REC_OLD. This function
	modifies RECORD structure on the spot.
	"""
	for strong_tag in CFG_BIBUPLOAD_STRONG_TAGS:
	if not record_get_field_instances(record, strong_tag):
	strong_tag_old_field_instances = record_get_field_instances(rec_old, strong_tag)
	if strong_tag_old_field_instances:
	for strong_tag_old_field_instance in strong_tag_old_field_instances:
	sf_vals, fi_ind1, fi_ind2, controlfield, dummy = strong_tag_old_field_instance
	record_add_field(record, strong_tag, fi_ind1, fi_ind2, controlfield, sf_vals)
	return

	### Delete functions
	def delete_tags(record, rec_old):
	"""
	Returns a record structure with all the fields in rec_old minus the
	fields in record.

	@param record: The record containing tags to delete.
	@type record: record structure

	@param rec_old: The original record.
	@type rec_old: record structure

	@return: The modified record.
	@rtype: record structure
	"""
	returned_record = copy.deepcopy(rec_old)
	for tag, fields in record.iteritems():
	if tag in ('001', ):
	continue
	for field in fields:
	local_position = record_find_field(returned_record, tag, field)[1]
	if local_position is not None:
	record_delete_field(returned_record, tag, field_position_local=local_position)
	return returned_record

	def delete_tags_to_correct(record, rec_old, opt_tag):
	"""
	Delete tags from REC_OLD which are also existing in RECORD. When
	deleting, pay attention not only to tags, but also to indicators,
	so that fields with the same tags but different indicators are not
	deleted.
	"""
	## Some fields are controlled via provenance information.
	## We should re-add saved fields at the end.
	fields_to_readd = {}
	for tag in CFG_BIBUPLOAD_CONTROLLED_PROVENANCE_TAGS:
	if tag[:3] in record:
	tmp_field_instances = record_get_field_instances(record, tag[:3], tag[3], tag[4]) ## Let's discover the provenance that will be updated
	provenances_to_update = []
	for instance in tmp_field_instances:
	for code, value in instance[0]:
	if code == tag[5]:
	if value not in provenances_to_update:
	provenances_to_update.append(value)
	break
	else:
	## The provenance is not specified.
	## let's add the special empty provenance.
	if '' not in provenances_to_update:
	provenances_to_update.append('')
	potential_fields_to_readd = record_get_field_instances(rec_old, tag[:3], tag[3], tag[4]) ## Let's take all the field corresponding to tag
	## Let's save apart all the fields that should be updated, but
	## since they have a different provenance not mentioned in record
	## they should be preserved.
	fields = []
	for sf_vals, ind1, ind2, dummy_cf, dummy_line in potential_fields_to_readd:
	for code, value in sf_vals:
	if code == tag[5]:
	if value not in provenances_to_update:
	fields.append(sf_vals)
	break
	else:
	if '' not in provenances_to_update:
	## Empty provenance, let's protect in any case
	fields.append(sf_vals)
	fields_to_readd[tag] = fields

	# browse through all the tags from the MARCXML file:
	for tag in record:
	# do we have to delete only a special tag or any tag?
	if opt_tag is None or opt_tag == tag:
	# check if the tag exists in the old record too:
	if tag in rec_old and tag != '001':
	# the tag does exist, so delete all record's tag+ind1+ind2 combinations from rec_old
	for dummy_sf_vals, ind1, ind2, dummy_cf, field_number in record[tag]:
	write_message(" Delete tag: " + tag + " ind1=" + ind1 + " ind2=" + ind2, verbose=9)
	record_delete_field(rec_old, tag, ind1, ind2)

	## Ok, we readd necessary fields!
	for tag, fields in fields_to_readd.iteritems():
	for sf_vals in fields:
	write_message(" Adding tag: " + tag[:3] + " ind1=" + tag[3] + " ind2=" + tag[4] + " code=" + str(sf_vals), verbose=9)
	record_add_field(rec_old, tag[:3], tag[3], tag[4], subfields=sf_vals)

	def delete_bibrec_bibxxx(record, id_bibrec, pretend=False):
	"""Delete the database record from the table bibxxx given in parameters"""
	# we clear all the rows from bibrec_bibxxx from the old record
	for tag in record.keys():
	if tag not in CFG_BIBUPLOAD_SPECIAL_TAGS:
	# for each name construct the bibrec_bibxxx table name
	table_name = 'bibrec_bib'+tag[0:2]+'x'
	# delete all the records with proper id_bibrec
	query = """DELETE FROM `%s` where id_bibrec = %s"""
	params = (table_name, id_bibrec)
	if not pretend:
	try:
	run_sql(query % params)
	except Error, error:
	write_message(" Error during the delete_bibrec_bibxxx function : %s " % error, verbose=1, stream=sys.stderr)

	def main():
	"""Main that construct all the bibtask."""
	task_init(authorization_action='runbibupload',
	authorization_msg="BibUpload Task Submission",
	description="""Receive MARC XML file and update appropriate database
	tables according to options.
	Examples:
	$ bibupload -i input.xml
	""",
	help_specific_usage=""" -a, --append\t\tnew fields are appended to the existing record
	-c, --correct\t\tfields are replaced by the new ones in the existing record, except
	\t\t\twhen overridden by CFG_BIBUPLOAD_CONTROLLED_PROVENANCE_TAGS
	-f, --format\t\ttakes only the FMT fields into account. Does not update
	-i, --insert\t\tinsert the new record in the database
	-r, --replace\t\tthe existing record is entirely replaced by the new one,
	\t\t\texcept for fields in CFG_BIBUPLOAD_STRONG_TAGS
	-z, --reference\tupdate references (update only 999 fields)
	-d, --delete\t\tspecified fields are deleted in existing record
	-S, --stage=STAGE\tstage to start from in the algorithm (0: always done; 1: FMT tags;
	\t\t\t2: FFT tags; 3: BibFmt; 4: Metadata update; 5: time update)
	-n, --notimechange\tdo not change record last modification date when updating
	-o, --holdingpen\tInsert record into holding pen instead of the normal database
	--pretend\t\tdo not really insert/append/correct/replace the input file
	--force\t\twhen --replace, use provided 001 tag values, even if the matching
	\t\t\trecord does not exist (thus allocating it on-the-fly)
	--callback-url\tSend via a POST request a JSON-serialized answer (see admin guide), in
	\t\t\torder to provide a feedback to an external service about the outcome of the operation.
	--nonce\t\twhen used together with --callback add the nonce value in the JSON message.
	""",
	version=__revision__,
	specific_params=("ircazdS:fno",
	[
	"insert",
	"replace",
	"correct",
	"append",
	"reference",
	"delete",
	"stage=",
	"format",
	"notimechange",
	"holdingpen",
	"pretend",
	"force",
	"callback-url=",
	"nonce="
	]),
	task_submit_elaborate_specific_parameter_fnc=task_submit_elaborate_specific_parameter,
	task_run_fnc=task_run_core)

	def task_submit_elaborate_specific_parameter(key, value, opts, args):
	""" Given the string key it checks it's meaning, eventually using the
	value. Usually it fills some key in the options dict.
	It must return True if it has elaborated the key, False, if it doesn't
	know that key.
	eg:
	if key in ['-n', '--number']:
	task_get_option(\1) = value
	return True
	return False
	"""

	# No time change option
	if key in ("-n", "--notimechange"):
	task_set_option('notimechange', 1)

	# Insert mode option
	elif key in ("-i", "--insert"):
	if task_get_option('mode') == 'replace':
	# if also replace found, then set to replace_or_insert
	task_set_option('mode', 'replace_or_insert')
	else:
	task_set_option('mode', 'insert')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	# Replace mode option
	elif key in ("-r", "--replace"):
	if task_get_option('mode') == 'insert':
	# if also insert found, then set to replace_or_insert
	task_set_option('mode', 'replace_or_insert')
	else:
	task_set_option('mode', 'replace')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))
	# Holding pen mode option
	elif key in ("-o", "--holdingpen"):
	write_message("Holding pen mode", verbose=3)
	task_set_option('mode', 'holdingpen')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))
	# Correct mode option
	elif key in ("-c", "--correct"):
	task_set_option('mode', 'correct')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	# Append mode option
	elif key in ("-a", "--append"):
	task_set_option('mode', 'append')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	# Reference mode option
	elif key in ("-z", "--reference"):
	task_set_option('mode', 'reference')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	elif key in ("-d", "--delete"):
	task_set_option('mode', 'delete')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	# Format mode option
	elif key in ("-f", "--format"):
	task_set_option('mode', 'format')
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	elif key in ("--pretend",):
	task_set_option('pretend', True)
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	elif key in ("--force",):
	task_set_option('force', True)
	fix_argv_paths([args[0]])
	task_set_option('file_path', os.path.abspath(args[0]))

	# Stage
	elif key in ("-S", "--stage"):
	try:
	value = int(value)
	except ValueError:
	print >> sys.stderr, """The value specified for --stage must be a valid integer, not %s""" % value
	return False
	if not (0 <= value <= 5):
	print >> sys.stderr, """The value specified for --stage must be comprised between 0 and 5"""
	return False
	task_set_option('stage_to_start_from', value)

	elif key in ("--callback-url", ):
	task_set_option('callback_url', value)
	elif key in ("--nonce", ):
	task_set_option('nonce', value)
	else:
	return False
	return True


	def task_submit_check_options():
	""" Reimplement this method for having the possibility to check options
	before submitting the task, in order for example to provide default
	values. It must return False if there are errors in the options.
	"""
	if task_get_option('mode') is None:
	write_message("Please specify at least one update/insert mode!")
	return False

	if task_get_option('file_path') is None:
	write_message("Missing filename! -h for help.")
	return False
	return True

	def writing_rights_p():
	"""Return True in case bibupload has the proper rights to write in the
	fulltext file folder."""
	global _WRITING_RIGHTS
	if _WRITING_RIGHTS is not None:
	return _WRITING_RIGHTS
	try:
	if not os.path.exists(CFG_WEBSUBMIT_FILEDIR):
	os.makedirs(CFG_WEBSUBMIT_FILEDIR)
	fd, filename = tempfile.mkstemp(suffix='.txt', prefix='test', dir=CFG_WEBSUBMIT_FILEDIR)
	test = os.fdopen(fd, 'w')
	test.write('TEST')
	test.close()
	if open(filename).read() != 'TEST':
	raise IOError("Can not successfully write and readback %s" % filename)
	os.remove(filename)
	except:
	register_exception(alert_admin=True)
	return False
	return True

	def post_results_to_callback_url(results, callback_url):
	if not CFG_JSON_AVAILABLE:
	from warnings import warn
	warn("--callback-url used but simplejson/json not available")
	return
	json_results = json.dumps(results)
	## <scheme>://<netloc>/<path>?<query>#<fragment>
	scheme, netloc, path, query, fragment = urlparse.urlsplit(callback_url)
	## See: http://stackoverflow.com/questions/111945/is-there-any-way-to-do-http-put-in-python
	if scheme == 'http':
	opener = urllib2.build_opener(urllib2.HTTPHandler)
	elif scheme == 'https':
	opener = urllib2.build_opener(urllib2.HTTPSHandler)
	else:
	raise ValueError("Scheme not handled %s for callback_url %s" % (scheme, callback_url))
	request = urllib2.Request(callback_url, data=json_results)
	request.add_header('Content-Type', 'application/json')
	request.add_header('User-Agent', '')
	request.get_method = lambda: 'POST'
	return opener.open(request)

	def task_run_core():
	""" Reimplement to add the body of the task."""
	error = 0
	write_message("Input file '%s', input mode '%s'." %
	(task_get_option('file_path'), task_get_option('mode')))
	write_message("STAGE 0:", verbose=2)

	if task_get_option('file_path') is not None:
	write_message("start preocessing", verbose=3)
	task_update_progress("Reading XML input")
	recs = xml_marc_to_records(open_marc_file(task_get_option('file_path')))
	stat['nb_records_to_upload'] = len(recs)
	write_message(" -Open XML marc: DONE", verbose=2)
	task_sleep_now_if_required(can_stop_too=True)
	write_message("Entering records loop", verbose=3)
	callback_url = task_get_option('callback_url')
	results_for_callback = {'results': []}
	if recs is not None:
	# We proceed each record by record
	for record in recs:
	record_id = record_extract_oai_id(record)
	task_sleep_now_if_required(can_stop_too=True)
	if task_get_option("mode") == "holdingpen":
	#inserting into the holding pen
	write_message("Inserting into holding pen", verbose=3)
	insert_record_into_holding_pen(record, record_id)
	else:
	write_message("Inserting into main database", verbose=3)
	error = bibupload(
	record,
	opt_tag=task_get_option('tag'),
	opt_mode=task_get_option('mode'),
	opt_stage_to_start_from=task_get_option('stage_to_start_from'),
	opt_notimechange=task_get_option('notimechange'),
	oai_rec_id=record_id,
	pretend=task_get_option('pretend'))
	if error[0] == 1:
	if record:
	write_message(record_xml_output(record),
	stream=sys.stderr)
	else:
	write_message("Record could not have been parsed",
	stream=sys.stderr)
	stat['nb_errors'] += 1
	if callback_url:
	results_for_callback['results'].append({'recid': error[1], 'success': False, 'error_message': error[2]})
	elif error[0] == 2:
	if record:
	write_message(record_xml_output(record),
	stream=sys.stderr)
	else:
	write_message("Record could not have been parsed",
	stream=sys.stderr)
	if callback_url:
	results_for_callback['results'].append({'recid': error[1], 'success': False, 'error_message': error[2]})
	elif error[0] == 0:
	if callback_url:
	from invenio.search_engine import print_record
	results_for_callback['results'].append({'recid': error[1], 'success': True, "marcxml": print_record(error[1], 'xm'), 'url': "%s/%s/%s" % (CFG_SITE_URL, CFG_SITE_RECORD, error[1])})
	else:
	if callback_url:
	results_for_callback['results'].append({'recid': error[1], 'success': False, 'error_message': error[2]})

	task_update_progress("Done %d out of %d." % \
	(stat['nb_records_inserted'] + \
	stat['nb_records_updated'],
	stat['nb_records_to_upload']))
	else:
	write_message(" Error bibupload failed: No record found",
	verbose=1, stream=sys.stderr)
	callback_url = task_get_option("callback_url")
	if callback_url:
	nonce = task_get_option("nonce")
	if nonce:
	results_for_callback["nonce"] = nonce
	post_results_to_callback_url(results_for_callback, callback_url)

	if task_get_task_param('verbose') >= 1:
	# Print out the statistics
	print_out_bibupload_statistics()

	# Check if they were errors
	return not stat['nb_errors'] >= 1

	def log_record_uploading(oai_rec_id, task_id, bibrec_id, insertion_db, pretend=False):
	if oai_rec_id != "" and oai_rec_id != None:
	query = """UPDATE oaiHARVESTLOG SET date_inserted=NOW(), inserted_to_db=%s, id_bibrec=%s WHERE oai_id = %s AND bibupload_task_id = %s ORDER BY date_harvested LIMIT 1"""
	try:
	if not pretend:
	run_sql(query, (str(insertion_db), str(bibrec_id), str(oai_rec_id), str(task_id), ))
	except Error, error:
	write_message(" Error during the log_record_uploading function : %s "
	% error, verbose=1, stream=sys.stderr)
	if __name__ == "__main__":
	main()

bibupload.pyNo OneTemporaryActions

File Metadata

bibupload.pyView Options

Event Timeline

bibupload.py
No OneTemporary
Actions

bibupload.py
View Options