def parse_xml_to_array(xml_file, strict=False, normalize_form=None):
"""
parse an xml file and return the records as an array. If you would
like the parser to explicitly check the namespaces for the MARCSlim
namespace use the strict=True option.
Valid values for normalize_form are 'NFC', 'NFKC', 'NFD', and 'NFKD'. See
unicodedata.normalize info.
"""
handler = XmlHandler(strict, normalize_form)
parse_xml(xml_file, handler)
return handler.records
def map_xml(function, *files):
"""
map a function onto the file, so that for each record that is
parsed the function will get called with the extracted record
def do_it(r):
print(r)
map_xml(do_it, 'marc.xml')
"""
handler = XmlHandler()
handler.process_record = function
for xml_file in files:
parse_xml(xml_file, handler)
def __call__(self, element):
print element[0][1].text
handler = marcxml.XmlHandler()
marcxml.parse_xml(StringIO(tostring(element[0])), handler)
return handler.records[0]
def __call__(self, element):
handler = marcxml.XmlHandler(strict=True, normalize_form='NFC')
marcxml.parse_xml(StringIO(tostring(element[0], encoding='UTF-8')), handler)
return handler.records[0]
def __call__(self, element):
handler = marcxml.XmlHandler(strict=True, normalize_form='NFC')
marcxml.parse_xml(StringIO(tostring(element[0], encoding='UTF-8')), handler)
return handler.records[0]
def __call__(self, element):
#print element[0][1].text
handler = marcxml.XmlHandler()
marcxml.parse_xml(StringIO(tostring(element[0], encoding='UTF-8')), handler)
return handler.records[0]
def MarcXML(xml):
handler = marcxml.XmlHandler()
data = tostring(xml, encoding='UTF-8')
marcxml.parse_xml(io.BytesIO(data), handler)
return handler.records[0]
def __call__(self, element):
# print element[0][1].text
handler = marcxml.XmlHandler(strict=True, normalize_form="NFC")
marcxml.parse_xml(StringIO(tostring(element[0], encoding="UTF-8")), handler)
return handler.records[0]
def __call__(self, element): handler = marcxml.XmlHandler() #marcxml.parse_xml( StringIO( """<?xml version="1.0" encoding="UTF-8"?>\n""" + tostring(element[0]) ), handler) marcxml.parse_xml( StringIO( tostring(element[0]) ), handler) return handler.records[0]
from OAIClient import *
from oaipmh import metadata
from oaipmh.datestamp import datestamp_to_datetime
from cStringIO import StringIO
from lxml.etree import tostring, XPathEvaluator
from pymarc import marcxml
from pymarc.record import *
import datetime
import sys
import properties
class MARCXMLReader(object):
"""Returns the PyMARC record from the OAI structure for MARC XML"""
def __call__(self, element):
handler = marcxml.XmlHandler()
marcxml.parse_xml( StringIO( tostring(element[0]) ), handler)
return handler.records[0]
if len(sys.argv) != 4:
print "Usage: deletebib.py <YYYYMMDD> <YYYYMMDD> <sigel1>[,sigel2,sigel3,...,sigeln]"
exit(1)
marcxml_reader = MARCXMLReader()
oaipmh_host = 'http://data.libris.kb.se'
from_time = sys.argv[1]
until = sys.argv[2]
sigels = sys.argv[3].split(',')
user = properties.user
password = properties.password
def __call__(self, element):
handler = marcxml.XmlHandler()
marcxml.parse_xml(StringIO(tostring(element[0])), handler)
return marcxml.record_to_xml(handler.records[0], namespace=True)