def convert_marc_xml(self, db_update_obj):
# check if it is a marc file
if self.source.endswith('.mrc') or self.source.endswith('.marc'):
# for each .mrc file create a sub-folder based on timestamp to store converted MARC/XML files
subfolder_name = '%s_%s' %(self.source.replace("Webapp/source/MARC/", "").replace("Webapp/source/BIBFRAME/", "").split('.')[0], datetime.fromtimestamp(time.time()).strftime('%Y-%m-%d %H:%M:%S'))
db_update_obj.master_file = subfolder_name
db_update_obj.save()
sufolder = os.path.join(self.folder, subfolder_name)
if not os.path.exists(sufolder):
os.makedirs(sufolder)
#create the BIBFRAME converter object
BIBFRAME = XML_BIBFRAME(subfolder_name)
output = os.path.join(sufolder, '')
with open(self.source, "rb") as marc_file:
reader = MARCReader(marc_file, to_unicode=True, force_utf8=False, utf8_handling='ignore')
for i, record in enumerate(reader):
#print ("converting record number %s to XML" %(str(i)))
if record.title():
ti = record.title()
ti = ti.replace("/", "")
ti = ti.replace(" ", "_")
ti = ti[0:50]
writer = XMLWriter(open(output + ti + '.xml','wb'))
writer.write(record)
writer.close()
else:
writer = XMLWriter(open(output + 'unknownTitle' + str(i) + '.xml','wb'))
marc_file.close()
#convert MARC/XML to BIBFRAME
db_update_obj.stage = "MARC-XML_to_BIBFRAME"
db_update_obj.save()
BIBFRAME.convert_to_BIBFRAME(i, db_update_obj)
# merge the BIBFRAME files into one (per the master MARC file) for ease of processing
merged = BIBFRAME.merger()
return merged
def writeRecordToFile(record, filename):
try:
writer = XMLWriter(open(filename, 'wb'))
writer.write(record)
except:
print "Error: Cannot write metadata to "+filename
writer.close() # Important!
def main():
files = parser.parse_args().files
for filename in files:
print filename
with open(filename) as fp:
reader = MARCReader(fp)
writer = XMLWriter(open(filename.split('.')[0]+'.xml', 'wb'))
for record in reader:
writer.write(record)
writer.close() # Important!
def upload_success(filename):
if request.method == 'POST':
file = "uploads/" + filename
if not os.path.exists("processed"):
os.makedirs("processed")
with open(file, "rb") as infile:
ts = datetime.fromtimestamp(time.time()).strftime('%H:%M:%S')
reader = MARCReader(infile,
to_unicode=True,
force_utf8=False,
utf8_handling='ignore')
files = []
for i, record in enumerate(reader):
print("converting record number " + str(i) + " to XML")
if record.title():
ti = record.title()
ti = ti.replace("/", "")
ti = ti.replace(" ", "_")
ti = ti[0:50]
writer = XMLWriter(open('processed/' + ti + '.xml', 'wb'))
writer.write(record)
writer.close()
files.append(ti)
else:
writer = XMLWriter(
open('processed/' + 'unknownTitle' + str(i) + '.xml',
'wb'))
files.append('unknownTitle' + str(i))
infile.close()
#print (file + " was successfully converted to XML")
xslt = ET.parse("marc2bibframe2-master/xsl/marc2bibframe2.xsl")
tf = datetime.fromtimestamp(time.time()).strftime('%H:%M:%S')
#print("walltime:", datetime.strptime(tf, '%H:%M:%S') - datetime.strptime(ts, '%H:%M:%S'))
for i, file in enumerate(files):
#print ("start processing record number " + str(i))
#print ("starting BIBFRAME transformation")
f = "processed/" + file + ".xml"
dom = ET.parse(f)
transform = ET.XSLT(xslt)
newdom = transform(dom)
if not os.path.exists("BIB"):
os.makedirs("BIB")
with open("BIB/" + file + ".xml", "w+") as oo:
oo.write(str(newdom).replace('<?xml version="1.0"?>', ''))
#print ("starting enrichment process")
main("BIB/" + file + ".xml")
tf = datetime.fromtimestamp(time.time()).strftime('%H:%M:%S')
#print("walltime:", datetime.strptime(tf, '%H:%M:%S') - datetime.strptime(ts, '%H:%M:%S'))
return redirect(url_for('process_success', files=file))
return '''
def upload_success(filename):
if request.method == 'POST':
file = "uploads/" + filename
if not os.path.exists("processed"):
os.makedirs("processed")
with open(file, "rb") as infile:
ts = datetime.fromtimestamp(time.time()).strftime('%H:%M:%S')
reader = MARCReader(infile, to_unicode=True, force_utf8=False, utf8_handling='ignore')
files = []
for i, record in enumerate(reader):
print ("converting record number " + str(i) + " to XML")
if record.title():
ti = record.title()
ti = ti.replace("/", "")
ti = ti.replace(" ", "_")
ti = ti[0:50]
writer = XMLWriter(open('processed/' + ti + '.xml','wb'))
writer.write(record)
writer.close()
files.append(ti)
else:
writer = XMLWriter(open('processed/' + 'unknownTitle' + str(i) + '.xml','wb'))
files.append('unknownTitle' + str(i))
infile.close()
#print (file + " was successfully converted to XML")
xslt = ET.parse("marc2bibframe2-master/xsl/marc2bibframe2.xsl")
tf = datetime.fromtimestamp(time.time()).strftime('%H:%M:%S')
#print("walltime:", datetime.strptime(tf, '%H:%M:%S') - datetime.strptime(ts, '%H:%M:%S'))
for i, file in enumerate(files):
#print ("start processing record number " + str(i))
#print ("starting BIBFRAME transformation")
f = "processed/" + file + ".xml"
dom = ET.parse(f)
transform = ET.XSLT(xslt)
newdom = transform(dom)
if not os.path.exists("BIB"):
os.makedirs("BIB")
with open ("BIB/" + file + ".xml", "w+") as oo:
oo.write(str(newdom).replace('<?xml version="1.0"?>', ''))
#print ("starting enrichment process")
main("BIB/" + file + ".xml")
tf = datetime.fromtimestamp(time.time()).strftime('%H:%M:%S')
#print("walltime:", datetime.strptime(tf, '%H:%M:%S') - datetime.strptime(ts, '%H:%M:%S'))
return redirect(url_for('process_success',
files=file))
return'''
def convert_marc_xml(self, db_update_obj):
# check if it is a marc file
if self.source.endswith('.mrc') or self.source.endswith('.marc'):
# for each .mrc file create a sub-folder based on timestamp to store converted MARC/XML files
subfolder_name = '%s_%s' % (self.source.replace(
"Webapp/source/MARC/", "").replace(
"Webapp/source/BIBFRAME/",
"").split('.')[0], datetime.fromtimestamp(
time.time()).strftime('%Y-%m-%d %H:%M:%S'))
db_update_obj.master_file = subfolder_name
db_update_obj.save()
sufolder = os.path.join(self.folder, subfolder_name)
if not os.path.exists(sufolder):
os.makedirs(sufolder)
#create the BIBFRAME converter object
BIBFRAME = XML_BIBFRAME(subfolder_name)
output = os.path.join(sufolder, '')
with open(self.source, "rb") as marc_file:
reader = MARCReader(marc_file,
to_unicode=True,
force_utf8=False,
utf8_handling='ignore')
for i, record in enumerate(reader):
#print ("converting record number %s to XML" %(str(i)))
if record.title():
ti = record.title()
ti = ti.replace("/", "")
ti = ti.replace(" ", "_")
ti = ti[0:50]
writer = XMLWriter(open(output + ti + '.xml', 'wb'))
writer.write(record)
writer.close()
else:
writer = XMLWriter(
open(output + 'unknownTitle' + str(i) + '.xml',
'wb'))
marc_file.close()
#convert MARC/XML to BIBFRAME
db_update_obj.stage = "MARC-XML_to_BIBFRAME"
db_update_obj.save()
BIBFRAME.convert_to_BIBFRAME(i, db_update_obj)
# merge the BIBFRAME files into one (per the master MARC file) for ease of processing
merged = BIBFRAME.merger()
return merged
def convert_marc_xml(self):
# for each .marc file in the "marc" folder, convert marc to MARC/XML and then to BIBFRAME
for index, files in enumerate(os.listdir(self.source)):
# check if it is a marc file
if files.endswith('.mrc') or files.endswith('.marc'):
# for each .mrc file create a sub-folder based on timestamp to store converted MARC/XML files
subfolder_name = '%s_%s' % (
files.split('.')[0], datetime.fromtimestamp(
time.time()).strftime('%Y-%m-%d %H:%M:%S'))
sufolder = os.path.join(self.folder, subfolder_name)
if not os.path.exists(sufolder):
os.makedirs(sufolder)
#create the BIBFRAME converter object
BIBFRAME = XML_BIBFRAME(subfolder_name)
file = os.path.join(self.source, files)
output = os.path.join(sufolder, '')
with open(file, "rb") as marc_file:
reader = MARCReader(marc_file,
to_unicode=True,
force_utf8=False,
utf8_handling='ignore')
for i, record in enumerate(reader):
print(
"converting record number %s of file number %s to XML"
% (str(i), str(index + 1)))
if record.title():
ti = record.title()
ti = ti.replace("/", "")
ti = ti.replace(" ", "_")
ti = ti[0:50]
writer = XMLWriter(open(output + ti + '.xml',
'wb'))
writer.write(record)
writer.close()
else:
writer = XMLWriter(
open(output + 'unknownTitle' + str(i) + '.xml',
'wb'))
marc_file.close()
#convert MARC/XML to BIBFRAME
BIBFRAME.convert_to_BIBFRAME(i)
# merge the BIBFRAME files into one (per the master MARC file) for ease of processing
BIBFRAME.merger()
def convert(self, filename):
"""Convert file into zero or more MARC records."""
g = Graph()
with open(filename, 'r') as fh:
g.parse(file=fh, format='n3')
# Get JSON-LD object in PyLD form
jld = pyld_jsonld_from_rdflib_graph(g)
# print(json.dumps(jld, indent=2, sort_keys=True))
# Manipulate the JSON in some way (say switch @context
# to a reference) and output
# -- no easy way to do this is rdflib
with open('cache/biblioteko_context.json', 'r') as cfh:
context = json.load(cfh)
with open('cache/biblioteko_frame.json', 'r') as ffh:
frame = json.load(ffh)
comp = jsonld.compact(jld, ctx=context)
comp = jsonld.compact(jsonld.frame(comp, frame),
ctx=context,
options={'graph': True})
if (self.dump_json):
sys.stderr.write("Framed and compacted JSON-LD:\n" +
json.dumps(comp, indent=2, sort_keys=True) + '\n')
self.jsonld = comp
memory = BytesIO()
writer = XMLWriter(memory)
for obj in self.jsonld['@graph']:
if (obj.get('type') == 'bf:Instance'):
writer.write(self.make_marc(obj))
else:
logging.info(
"Ignoring object %s type %s" %
(obj.get('id', 'NO-URI'), obj.get('type', 'NO-TYPE')))
writer.close(close_fh=False) # Important!
xml = memory.getvalue().decode(encoding='UTF-8')
# Dumb semi pretty print
xml = re.sub(r'(</\w+>)', r'\1\n', xml)
xml = re.sub(r'(\?>)', r'\1\n', xml)
xml = re.sub(r'(<record>)', r'\n\1\n', xml)
return (xml)
def marc_record_to_xml_string(record):
b = BytesIO()
writer = XMLWriter(b)
writer.write(record)
writer.close(close_fh=False)
# Transform record from bytes to string.
b.seek(0,0)
bytestr = b.read()
marc = bytestr.decode('UTF-8')
# Remove the XML declaration and collection stuff.
marc = re.sub(r'^.*<collection[^>]*>','',marc)
marc = re.sub(r'</collection>$','',marc)
# Remove tab characters.
marc = re.sub(r'\t','',marc)
# Verify cleaning worked:
if not marc.startswith('<record'):
print("Error: record failed to create clean XML.")
return False
else:
return marc
def create_MARC_XML(self):
"""
Creates a MARC XML format file from the given dataframe
:return:
"""
df = self.marc_data
# MARCXML file
output_file = self.data_path_processed / (
self.collection_id + "_final_" + self.dt_now + ".xml")
writer = XMLWriter(open(output_file, "wb"))
# MarcEdit MRK file
output_file_mrk = self.data_path_processed / (
self.collection_id + "_finalMRK_" + self.dt_now + ".txt")
mrk_file = open(output_file_mrk, "w", encoding="utf8")
start_time = time.time()
counter = 1
for index, row in df.iterrows():
record = Record()
# add control field
record.add_field(Field(tag="001", data=str(index)))
for col in df:
# if field is empty, skip
if str(row[col]) == "":
continue
# leader
elif col == "LDR":
l = record.leader
l.record_status = "c" # c - Corrected or revised
l.type_of_record = "p" # p - Mixed materials
# Bibliographic level
if row["351"] == "File Record" or "Item Record":
l.bibliographic_level = "c"
else:
l.bibliographic_level = "d"
l.coding_scheme = "a" # flag saying this record is utf8
l.cataloging_form = "a"
continue
# 008
elif col == "008":
field = Field(tag="008", data=row[col])
record.add_field(field)
continue
# extract field name
field = col[:3]
# extract indicators
if col.find("_") == -1:
col_name = "{:<5}".format(col)
ind = [col_name[3], col_name[4]]
else:
col_name = "{:<5}".format(col[:col.find("_")])
ind = [col_name[3], col_name[4]]
# extract sub-fields
subfields_data = list()
subfields_prep = list(filter(None, str(row[col]).split("$$")))
for subfield in subfields_prep:
if subfield == "":
continue
subfields_data.append(subfield[0])
subfields_data.append(subfield[1:])
# print('field:', field)
# if not ind:
# print('no indicators')
# else:
# print('indicators:', ind)
# # print('subfields:', subfields_data)
record.add_field(
Field(tag=field, indicators=ind, subfields=subfields_data))
counter += 1
# mrk_file.write(record.as_marc())
writer.write(record)
writer.close()
mrk_file.close()
run_time = time.time() - start_time
return counter, run_time
subfields = [ 'w', wval,
'a', unicode(label) ]
else:
subfields = [ 'a', unicode(label) ]
rec.add_field(
Field(
tag='550',
indicators = [' ', ' '],
subfields = subfields
)
)
# skos:note -> 680
for note in sorted(g.objects(conc, SKOS.note), key=lambda v:v.language):
rec.add_field(
Field(
tag='680',
indicators = [' ', ' '],
subfields=[ 'i', unicode(note),
'9', format_language(note.language) ]
)
)
# TODO 750 links to ysa/allars/lcsh - need to look up labels for the URIs
writer.write(rec)
writer.close()
def convert(cs, language, g):
vocId = cs.get("vocabulary_code")
# variable for a bit complicated constants and casting/converting them to appropiate types
helper_variables = {
"vocCode" : (cs.get("vocabulary_code") + "/" + LANGUAGES[language] \
if cs.getboolean("multilanguage", fallback=False) \
else vocId),
"groupingClasses" : [URIRef(x) for x in cs.get("groupingClasses", fallback=",".join(GROUPINGCLASSES)).split(",")],
"groupingClassesDefault" : [URIRef(x) for x in cs.parser.get("DEFAULT", "groupingClasses", fallback=",".join(GROUPINGCLASSES)).split(",")],
'modificationDates': cs.get("modificationDates", fallback=None),
'keepModified' : cs.get("keepModifiedAfter", fallback=None),
'keepGroupingClasses' : cs.getboolean("keepGroupingClasses", fallback=False),
'defaultOutputFileName' : "yso2marc-" + cs.name.lower() + "-" + language + ".mrcx"
}
if helper_variables['keepModified']:
helper_variables['keepModifiedLimit'] = False \
if cs.get("keepModifiedAfter", fallback=KEEPMODIFIEDAFTER).lower() == "all" \
else datetime.date(datetime.strptime(cs.get("keepModifiedAfter"), "%Y-%m-%d"))
if cs.get("output", fallback=None):
parts = cs.get("languages").split(",")
if len(parts) > 1:
output = cs.get("output")
if len(output.split(".")) > 1:
helper_variables["outputFileName"] = ".".join(
output.split(".")
[:-1]) + "-" + language + "." + output.split(".")[-1]
else:
helper_variables["outputFileName"] = output + "-" + language
if not "outputFileName" in helper_variables:
helper_variables["outputFileName"] = cs.get(
"output", fallback=helper_variables["defaultOutputFileName"])
#modified_dates on dict-objekti, joka sisältää tietueen id:n avaimena ja
#arvona tuplen, jossa on tietueen viimeinen muokkauspäivämäärä ja tietueen sisältö MD5-tiivisteenä
if helper_variables['modificationDates']:
if os.path.isfile(helper_variables['modificationDates']):
with open(helper_variables['modificationDates'],
'rb') as pickle_file:
try:
modified_dates = pickle.load(pickle_file)
except EOFError:
logging.error(
"The file %s for modification dates is empty " %
helper_variables['modificationDates'])
sys.exit(2)
else:
modified_dates = {}
logging.info(
"Processing vocabulary with vocabulary code '%s' in language '%s'" %
(vocId, language))
incrementor = 0
writer_records_counter = 0
ET_namespaces = {"marcxml": "http://www.loc.gov/MARC21/slim"}
handle = open(
cs.get("output", fallback=helper_variables["defaultOutputFileName"]),
"wb")
writer = XMLWriter(handle)
# listataan preflabelit, jotta voidaan karsia alt_labelit, jotka toisessa käsitteessä pref_labelina
pref_labels = set()
for conc in g.subjects(RDF.type, SKOS.Concept):
pref_label = g.preferredLabel(conc, lang=language)
if pref_label:
pref_labels.add(str(pref_label[0][1]))
# vain nämä mts-käsiteryhmät otetaan mukaan, ryhmän nimestä ei tehdä MARC21-tietuetta
ids = {"occupations": ['m2332'], "titles": ['m121', 'm3764']}
uris = {}
for key in ids:
uris[key] = set()
for id in ids[key]:
uris[key].add(MTS + id)
for group in g.subjects(RDF.type, ISOTHES.ConceptGroup):
for key in uris:
if any(str(group).endswith(uri) for uri in uris[key]):
get_member_groups(g, group, uris[key])
concs = []
if helper_variables['keepModified']:
concs = []
for uri in modified_dates:
if modified_dates[uri][0] >= helper_variables['keepModifiedLimit']:
concs.append(URIRef(uri))
else:
for conc in g.subjects(RDF.type, SKOS.Concept):
concs.append(conc)
#luotujen käsitteiden tunnukset, joilla voidaan selvittää modification_dates-listan avulla poistetut käsitteet
created_concepts = set()
for concept in concs:
#vain ammateista ja arvonimistä luodaan MARC21-tietueet
if not (concept in uris['occupations'] or concept in uris['titles']):
continue
created_concepts.add(str(concept))
incrementor += 1
if incrementor % 1000 == 0:
logging.info("Processing %sth concept" % (incrementor))
#skipataan ryhmittelevät käsitteet
if not helper_variables['keepGroupingClasses']:
if any(conceptType in helper_variables["groupingClasses"]
for conceptType in g.objects(concept, RDF.type)):
continue
rec = Record()
rec.leader = cs.get("leaderNew", fallback=LEADERNEW)
# 024 muut standarditunnukset - käsitteen URI tallennetaan tähän
rec.add_field(
Field(tag='024',
indicators=['7', ' '],
subfields=['a', concept, '2', "uri"]))
# 040 luetteloiva organisaatio
rec.add_field(
Field(tag='040',
indicators=[' ', ' '],
subfields=[
'a',
cs.get("creatorAgency", fallback=CREATOR_AGENCY), 'b',
LANGUAGES[language], 'f', helper_variables["vocCode"]
]))
valueProps = sorted(getValues(g,
concept,
SKOS.prefLabel,
language=language),
key=lambda o: o.value)
if len(valueProps) == 0:
logging.warning(
"Could not find preflabel for concept %s in language %s. Skipping the whole concept."
% (concept, language))
continue
elif len(valueProps) != 1:
logging.warning(
"Multiple prefLabels detected for concept %s in language %s. Choosing the first."
% (concept, language))
if concept in uris['occupations']:
tag = "174"
subfield_code = "a"
elif concept in uris['titles']:
tag = "168"
subfield_code = "d"
rec.add_field(
Field(tag=tag,
indicators=[' ', ' '],
subfields=[
subfield_code,
decomposedÅÄÖtoUnicodeCharacters(
unicodedata.normalize(NORMALIZATION_FORM,
str(valueProps[0].value)))
]))
# skos:altLabel -> 467, 474
# 450 katso-viittaus
# jätetään tuottamatta 45X-kentät, jotka ovat toisessa käsitteessä 15X-kenttinä, paitsi altLabelein kohdalla
seen_values = set()
for valueProp in sorted(getValues(g,
concept, [SKOS.altLabel],
language=language),
key=lambda o: str(o.value)):
if valueProp.prop != SKOS.altLabel and str(
valueProp.value) in pref_labels:
continue
if valueProp.prop == SKOS.hiddenLabel:
if str(valueProp.value) in seen_values:
continue
seen_values.add(str(valueProp.value))
if concept in uris['occupations']:
tag = "474"
subfield_code = "a"
elif concept in uris['titles']:
tag = "468"
subfield_code = "d"
rec.add_field(
Field(tag=tag,
indicators=[' ', ' '],
subfields=[
subfield_code,
decomposedÅÄÖtoUnicodeCharacters(
unicodedata.normalize(NORMALIZATION_FORM,
str(valueProp.value)))
]))
valueProps = getValues(g, concept, [
SKOS.prefLabel, SKOS.exactMatch, SKOS.closeMatch, SKOS.broadMatch,
SKOS.narrowMatch, SKOS.relatedMatch
])
fields = list(
) # kerätään kentät tähän muuttujaan, joka sitten lopuksi järjestetään
for valueProp in valueProps:
if valueProp.prop == SKOS.prefLabel:
# suodatetaan samankieliset, jotka menivät jo 1xx-kenttiin
# valueProp.value sisältää tässä poikkeuksellisesti jo halutun literaalin
# (vrt. kun muissa on solmu)
if valueProp.value.language == language:
continue
else:
# otetaan vain viittaukset samaan sanastoon
continue
if concept in uris['occupations']:
tag = "774"
subfield_code = "a"
elif concept in uris['titles']:
tag = "768"
subfield_code = "d"
sub2 = "mts" + "/" + LANGUAGES[valueProp.value.language]
fields.append(
Field(tag=tag,
indicators=[' ', ' '],
subfields=[
subfield_code,
decomposedÅÄÖtoUnicodeCharacters(
unicodedata.normalize(NORMALIZATION_FORM,
str(valueProp.value))),
'4', 'EQ', '2', sub2, '0', concept
]))
# sort fields and add them
for sorted_field in sorted(fields,
key=lambda o: (o.tag, o.value().lower())):
rec.add_field(sorted_field)
writer_records_counter += 1
writer.write(rec)
if helper_variables['modificationDates']:
md5 = hashlib.md5()
md5.update(str.encode(str(rec)))
hash = md5.hexdigest()
if str(concept) in modified_dates:
if not hash == modified_dates[str(concept)][1]:
modified_dates[str(concept)] = (date.today(), hash)
else:
modified_dates[str(concept)] = (date.today(), hash)
#tuotetaan poistetut käsitteet, kun haetaan muuttuneet käsitteet
#jos tietue on modified_dates-parametrillä määritettyssä tiedostossa, mutta ei graafissa, tulkitana poistetuksi tietueeksi
#mts:ssä ei ole deprekointipäiviä
#
if helper_variables['keepModified']:
concs = []
for conc in g.subjects(RDF.type, SKOS.Concept):
if conc in uris['occupations'] or conc in uris['titles']:
concs.append(str(conc))
for conc in modified_dates:
if conc not in concs:
#jos ei ole hajautussummaa (tuplen 2. arvo), luodaan deprekoitu käsite
if modified_dates[conc][1]:
rec = Record()
rec.leader = cs.get("leaderDeleted0",
fallback=LEADERDELETED0)
rec.add_field(
Field(tag='024',
indicators=['7', ' '],
subfields=['a', conc, '2', "uri"]))
modified_dates[conc] = (date.today(), "")
writer_records_counter += 1
writer.write(rec)
if handle is not sys.stdout:
writer.close()
if helper_variables['modificationDates']:
with open(helper_variables['modificationDates'], 'wb') as output:
pickle.dump(modified_dates, output, pickle.HIGHEST_PROTOCOL)
#jos luodaan kaikki käsitteet, tuotetaan tuotetaan lopuksi käsitteet laveassa XML-muodossa
#if not helper_variables['keepModified']:
parser = ET.XMLParser(remove_blank_text=True, strip_cdata=False)
file_path = helper_variables["outputFileName"]
tree = ET.parse(file_path, parser)
e = tree.getroot()
handle = open(
cs.get("output", fallback=helper_variables["defaultOutputFileName"]),
"wb")
handle.write(
ET.tostring(e,
encoding='UTF-8',
pretty_print=True,
xml_declaration=True))
if handle is not sys.stdout:
handle.close()
# lokitetaan vähän tietoa konversiosta
logging.info("Processed %s concepts. Wrote %s MARCXML records." %
(incrementor, writer_records_counter))
if cs.get("outputSpecified", fallback=None) == None:
outputChannel = sys.stdout.buffer
with open(
cs.get("output",
fallback=helper_variables['defaultOutputFileName']),
"rb") as f:
shutil.copyfileobj(f, outputChannel)
if cs.get("outputSpecified", fallback=None) == None:
os.remove(
cs.get("output",
fallback=helper_variables['defaultOutputFileName']))
logging.info("Conversion completed: %s" %
datetime.now().replace(microsecond=0).isoformat())
def compare_records(args):
"""
input_file_1: Vertailtavien MARC-tietueiden tiedostonimi
input_file_2: Tiedostonimi MARC-tietueille, joista tallennetaan muokatut ja uudet
output_file: muokatuista ja uusista tietueista muodostetun MARCXML-tiedoston nimi
pickle_file: tallettaa muutospäivämäärät pickle-tiedostoon date_1 ja date_2 parametrien mukaan
date_1: alkuperäinen päivämäärä
date_2: muutospäivämäärä
"""
#git rev-list -1 --before="2019-08-23 23:59" master
#git log
input_file_1 = args.first_input_file
input_file_2 = args.second_input_file
mrcx_file = args.output_mrcx
pickle_file = args.output_pkl
date_1 = args.original_date
date_2 = args.modified_date
modified_records = 0
new_records = 0
all_records = {}
loglevel = logging.INFO
logger = logging.getLogger()
logger.setLevel(loglevel)
if date_1:
old_date = datetime.date(datetime.strptime(date_1, "%Y-%m-%d"))
else:
old_date = date.fromtimestamp(os.path.getmtime(input_file_1))
if date_2:
new_date = datetime.date(datetime.strptime(date_2, "%Y-%m-%d"))
else:
new_date = date.fromtimestamp(os.path.getmtime(input_file_2))
writer = XMLWriter(open(mrcx_file, "wb"))
records = parse_xml_to_array(input_file_1)
old_records_dict = {}
for record in records:
md5 = hashlib.md5()
md5.update(str.encode(str(record)))
for field in record.get_fields('024'):
old_records_dict.update({field['a']: md5.hexdigest()})
records = parse_xml_to_array(input_file_2)
for record in records:
record_id = None
modified = False
modified_date = old_date
for field in record.get_fields('024'):
record_id = field['a']
if record_id:
if record_id in old_records_dict:
md5 = hashlib.md5()
md5.update(str.encode(str(record)))
hash = md5.hexdigest()
old_hash = old_records_dict[record_id]
if not old_hash == hash:
modified = True
modified_records += 1
else:
modified = True
new_records += 1
else:
logging.warning("Record id missing")
if modified:
writer.write(record)
modified_date = new_date
all_records[record_id] = (modified_date, hash)
logging.info("Number of modified records: %s"%modified_records)
logging.info("Number of new records: %s"%new_records)
if pickle_file:
with open(pickle_file, 'wb') as output:
pickle.dump(all_records, output, pickle.HIGHEST_PROTOCOL)
output.close()
writer.close()
parser = ET.XMLParser(remove_blank_text=True,strip_cdata=False)
tree = ET.parse(mrcx_file, parser)
e = tree.getroot()
handle = open(mrcx_file, "wb")
handle.write(ET.tostring(e, encoding='UTF-8', pretty_print=True, xml_declaration=True))
def writeMARCXML(record, filename):
# Write out record in MARCXML format
writer = XMLWriter(open(filename,'wb'))
writer.write(record)
writer.close() # Important!
def main():
global inputfile, target, mapping
parser = argparse.ArgumentParser(
description='Process and map classification authority records for BCUR.'
)
parser.add_argument('-i',
'--inputfiles',
type=str,
nargs='+',
help='one or more file(s) to be processed',
required=True)
parser.add_argument('-o',
'--outputfile',
type=str,
nargs=1,
help='name of the output file',
required=True)
parser.add_argument('-m',
'--map',
type=str,
nargs=1,
help='map target code',
required=True,
choices=valid_targets)
args = parser.parse_args()
targetcode = args.map[0]
# For musi and musg records, found records are copied as-is, no field mapping.
if targetcode in ('musi', 'musg'):
mapping = False
outputfile = args.outputfile[0]
# Open a new XML document in which target records will be stored
global writer
writer = XMLWriter(open(outputfile, 'wb'))
# Record start processing time
tstart = datetime.datetime.now()
# Print header row in case log is opened as CSV
print("Notice,Champ,Contenu du champ,Message")
# Loop through the list of input files and call the mapping function
for infile in args.inputfiles:
inputfile = infile
target = targetcode
if mapping:
print(
f"----- Traitement du fichier {inputfile} avec mapping {target} -----"
)
else:
print(
f"----- Traitement du fichier {inputfile} sans mapping -----")
# This applies the mapping function to each record in inputfile
map_xml(record_map, inputfile)
if targetcode == 'vddoc':
# For vddoc, also look for vddoc-la
target = 'vddoc-la'
map_xml(record_map, inputfile)
# Calculate the total time elapsed
tdiff = datetime.datetime.now() - tstart
# Close the output document
writer.close()
print(
f'Routine terminée en {tdiff.total_seconds()} secondes. Résultat enregistré dans {outputfile}'
)