def __init__(self,
trnkbsServerUrl,
sHttpProxy=None,
loggingLevel=logging.WARN):
TranskribusClient.__init__(self,
sServerUrl=self.sDefaultServerUrl,
proxies=sHttpProxy,
loggingLevel=loggingLevel)
def __init__(self,
trnkbsServerUrl,
sHttpProxy=None,
loggingLevel=logging.WARN):
TranskribusClient.__init__(self,
sServerUrl=self.sDefaultServerUrl,
proxies=sHttpProxy,
loggingLevel=loggingLevel)
self._trpMng = DoTranscript(self.sDefaultServerUrl,
sHttpProxy=sHttpProxy,
loggingLevel=loggingLevel)
self.percTest = 0.1
import logging
try: #to ease the use without proper Python installation
import TranskribusPyClient_version
except ImportError:
sys.path.append(
os.path.dirname(
(os.path.dirname(os.path.dirname(os.path.abspath(sys.argv[0]))))))
import TranskribusPyClient_version
from TranskribusPyClient.test import _colId_A, _docId_b
from TranskribusPyClient.client import TranskribusClient, getStoredCredentials
login, pwd = getStoredCredentials()
conn = TranskribusClient(proxies={'https': 'http://*****:*****@xrce.xerox.com", "trnjluc", sHttpsProxyUrl='http://cornillon:8000')
sXml = u"""<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<PcGts xmlns="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15 http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15/pagecontent.xsd">
<Metadata>
<Creator>TRP</Creator>
<Created>2016-08-18T13:35:08.767+07:00</Created>
<LastChange>2016-12-01T09:59:24.254+01:00</LastChange>
class TableProcessing(Component.Component):
usage = ""
version = "v.01"
description = "description: table layout analysis based on template"
sCOL = "col"
sMPXMLExtension = ".mpxml"
def __init__(self):
"""
Always call first the Component constructor.
"""
Component.Component.__init__(self, "TableProcessing", self.usage,
self.version, self.description)
self.colid = None
self.docid = None
self.bFullCol = False
# generate MPXML using Ext
self.useExtForMPXML = False
self.bRegenerateMPXML = False
self.sRowModelName = None
self.sRowModelDir = None
self.sHTRmodel = None
self.sDictName = None
def setParams(self, dParams):
"""
Always call first the Component setParams
Here, we set our internal attribute according to a possibly specified value (otherwise it stays at its default value)
"""
Component.Component.setParams(self, dParams)
if "coldir" in dParams:
self.coldir = dParams["coldir"]
if "colid" in dParams:
self.colid = dParams["colid"]
if "colid" in dParams:
self.docid = dParams["docid"]
if "useExt" in dParams:
self.useExtForMPXML = dParams["useExt"]
if 'regMPXML' in dParams:
self.bRegenerateMPXML = True
if "rowmodelname" in dParams:
self.sRowModelName = dParams["rowmodelname"]
if "rowmodeldir" in dParams:
self.sRowModelDir = dParams["rowmodeldir"]
if "htrmodel" in dParams:
self.sHTRmodel = dParams["htrmodel"]
if "dictname" in dParams:
self.sDictName = dParams["dictname"]
# Connection to Transkribus
self.myTrKCient = None
self.persist = False
self.loginInfo = False
if dParams.has_key("server"):
self.server = dParams["server"]
if dParams.has_key("persist"):
self.persist = dParams["persist"]
if dParams.has_key("login"):
self.loginInfo = dParams["login"]
def login(self, trnskrbs_client, trace=None, traceln=None):
"""
deal with the complicated login variants...
-trace and traceln are optional print methods
return True or raises an exception
"""
DEBUG = True
bOk = False
if self.persist:
#try getting some persistent session token
if DEBUG and trace:
trace(" ---login--- Try reusing persistent session ... ")
try:
bOk = trnskrbs_client.reusePersistentSession()
if DEBUG and traceln: traceln("OK!")
except:
if DEBUG and traceln: traceln("Failed")
if not bOk:
if self.loginInfo:
login, pwd = self.loginInfo, self.pwd
else:
if trace:
DEBUG and trace(
" ---login--- no login provided, looking for stored credentials... "
)
login, pwd = trnskrbs_client.getStoredCredentials(bAsk=False)
if DEBUG and traceln: traceln("OK")
if DEBUG and traceln:
trace(" ---login--- logging onto Transkribus as %s " % login)
trnskrbs_client.auth_login(login, pwd)
if DEBUG and traceln: traceln("OK")
bOk = True
return bOk
def downloadCollection(self,
colid,
destDir,
docid,
bNoImg=True,
bForce=False):
"""
download colID
replace destDir by '.' ?
"""
destDir = "."
# options.server, proxies, loggingLevel=logging.WARN)
#download
downloader = TranskribusDownloader(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
downloader.setSessionId(self.myTrKCient.getSessionId())
traceln("- Downloading collection %s to folder %s" %
(colid, os.path.abspath(destDir)))
# col_ts, colDir = downloader.downloadCollection(colid, destDir, bForce=options.bForce, bNoImage=options.bNoImage)
col_ts, colDir, ldocids, dFileListPerDoc = downloader.downloadCollection(
colid, destDir, bForce=bForce, bNoImage=bNoImg, sDocId=docid)
traceln("- Done")
with open(os.path.join(colDir, "config.txt"), "w") as fd:
fd.write("server=%s\nforce=%s\nstrict=%s\n" %
(self.server, True, False))
downloader.generateCollectionMultiPageXml(
os.path.join(colDir, TableProcessing.sCOL), dFileListPerDoc, False)
traceln('- Done, see in %s' % colDir)
return ldocids
def upLoadDocument(self,
colid,
coldir,
docid,
sNote="",
sTranscripExt='.mpxml'):
"""
download colID
"""
# options.server, proxies, loggingLevel=logging.WARN)
#download
# uploader = TranskribusTranscriptUploader(self.server,self.proxies)
uploader = TranskribusDUTranscriptUploader(
self.myTrKCient.getServerUrl(), self.myTrKCient.getProxies())
uploader.setSessionId(self.myTrKCient.getSessionId())
traceln("- uploading document %s to collection %s" % (docid, colid))
uploader.uploadDocumentTranscript(colid,
docid,
os.path.join(coldir, sCOL),
sNote,
'NLE Table',
sTranscripExt,
iVerbose=False)
traceln("- Done")
return
def applyLA_URO(self, colid, docid, nbpages):
"""
apply textline finder
"""
# do the job...
# if options.trp_doc:
# trpdoc = json.load(codecs.open(options.trp_doc, "rb",'utf-8'))
# docId,sPageDesc = doer.buildDescription(colId,options.docid,trpdoc)
traceln('process %s pages...' % nbpages)
lretJobIDs = []
for i in range(1, nbpages + 1):
LA = DoLAbatch(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
LA._trpMng.setSessionId(self.myTrKCient.getSessionId())
LA.setSessionId(self.myTrKCient.getSessionId())
_, sPageDesc = LA.buildDescription(colid, "%s/%s" % (docid, i))
sPageDesc = LA.jsonToXMLDescription(sPageDesc)
_, lJobIDs = LA.run(colid, sPageDesc, "CITlabAdvancedLaJob", False)
traceln(lJobIDs)
lretJobIDs.extend(lJobIDs)
traceln("- LA running for page %d job:%s" % (i, lJobIDs))
return lretJobIDs
def applyHTRForRegions(self, colid, docid, nbpages, modelname, dictionary):
"""
apply an htr model at region level
"""
htrComp = DoHtrRnn(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
htrComp._trpMng.setSessionId(self.myTrKCient.getSessionId())
htrComp.setSessionId(self.myTrKCient.getSessionId())
_, sPageDesc = htrComp.buildDescription(colid,
"%s/%s" % (docid, nbpages))
sPages = "1-%d" % (nbpages)
sModelID = None
# get modelID
lColModels = self.myTrKCient.listRnns(colid)
for model in lColModels:
# print model['htrId'], type(model['htrId']), modelname,type(modelname)
if str(model['htrId']) == str(modelname):
sModelID = model['htrId']
traceln('model id = %s' % sModelID)
#some old? models do not have params field
# try: traceln("%s\t%s\t%s" % (model['htrId'],model['name'],model['params']))
# except KeyError: traceln("%s\t%s\tno params" % (model['htrId'],model['name']))
if sModelID == None:
raise Exception, "no model ID found for %s" % (modelname)
ret = htrComp.htrRnnDecode(colid,
sModelID,
dictionary,
docid,
sPageDesc,
bDictTemp=False)
traceln(ret)
return ret
def applyHTR(self, colid, docid, nbpages, modelname, dictionary):
"""
apply HTR on docid
htr id is needed: we have htrmodename
"""
htrComp = DoHtrRnn(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
htrComp._trpMng.setSessionId(self.myTrKCient.getSessionId())
htrComp.setSessionId(self.myTrKCient.getSessionId())
_, sPageDesc = htrComp.buildDescription(colid,
"%s/%s" % (docid, nbpages))
sPages = "1-%d" % (nbpages)
sModelID = None
# get modelID
lColModels = self.myTrKCient.listRnns(colid)
for model in lColModels:
# print model['htrId'], type(model['htrId']), modelname,type(modelname)
if str(model['htrId']) == str(modelname):
sModelID = model['htrId']
traceln('model id = %s' % sModelID)
#some old? models do not have params field
# try: traceln("%s\t%s\t%s" % (model['htrId'],model['name'],model['params']))
# except KeyError: traceln("%s\t%s\tno params" % (model['htrId'],model['name']))
if sModelID == None:
raise Exception, "no model ID found for %s" % (modelname)
ret = htrComp.htrRnnDecode(colid,
sModelID,
dictionary,
docid,
sPageDesc,
bDictTemp=False)
traceln(ret)
return ret
def extractFileNamesFromMPXML(self, mpxmldoc):
"""
to insure correct file order !
duplicated form performCVLLA.py
"""
xmlpath = os.path.abspath(os.path.join(self.coldir, sCOL, self.docid))
lNd = PageXml.getChildByName(mpxmldoc.getRootElement(), 'Page')
# for i in lNd:print i
return map(
lambda x: "%s%s%s.xml" %
(xmlpath, os.sep, x.prop('imageFilename')[:-4]), lNd)
def processDocument(self, coldir, colid, docid, dom=None):
"""
process a single document
1 python ../../src/xml_formats/PageXml.py trnskrbs_5400/col/17442 --ext=pxml
2 python ../../src/tasks/performCVLLA.py --coldir=trnskrbs_5400/ --docid=17442 -i trnskrbs_5400/col/17442.mpxml --bl --regTL --form
3 python ../../src/tasks/DU_ABPTable_T.py modelMultiType tableRow2 --run=trnskrbs_5400
4 python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442_du.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
5 python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml
6 python ../../../TranskribusPyClient/src/TranskribusCommands/TranskribusDU_transcriptUploader.py --nodu trnskrbs_5400 5400 17442
7 python ../../../TranskribusPyClient/src/TranskribusCommands/do_htrRnn.py <model-name> <dictionary-name> 5400 17442
wait
8 python ../../../TranskribusPyClient/src/TranskribusCommands/Transkribus_downloader.py 5400 --force
#covnert to ds
9 python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
10 python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml --doie --usetemplate
"""
#create Transkribus client
self.myTrKCient = TranskribusClient(sServerUrl=self.server,
proxies={},
loggingLevel=logging.WARN)
#login
_ = self.login(self.myTrKCient, trace=trace, traceln=traceln)
# self.downloadCollection(colid,coldir,docid,bNoImg=False,bForce=True)
## load dom
if dom is None:
self.inputFileName = os.path.abspath(
os.path.join(coldir, TableProcessing.sCOL,
docid + TableProcessing.sMPXMLExtension))
mpxml_doc = self.loadDom()
nbPages = MultiPageXml.getNBPages(mpxml_doc)
else:
# load provided mpxml
mpxml_doc = dom
nbPages = MultiPageXml.getNBPages(mpxml_doc)
# ### table registration: need to compute/select??? the template
# # perform LA separator, table registration, baseline with normalization
# #python ../../src/tasks/performCVLLA.py --coldir=trnskrbs_5400/ --docid=17442 -i trnskrbs_5400/col/17442.mpxml --bl --regTL --form
# tableregtool= LAProcessor()
# # latool.setParams(dParams)
# tableregtool.coldir = coldir
# tableregtool.docid = docid
# tableregtool.bTemplate, tableregtool.bSeparator , tableregtool.bBaseLine , tableregtool.bRegularTextLine = True,False,False,False
# # creates xml and a new mpxml
# mpxml_doc,nbPages = tableregtool.performLA(mpxml_doc)
#
#
# self.upLoadDocument(colid, coldir,docid,sNote='NLE workflow;table reg done')
lJobIDs = self.applyLA_URO(colid, docid, nbPages)
return
bWait = True
assert lJobIDs != []
jobid = lJobIDs[-1]
traceln("waiting for job %s" % jobid)
while bWait:
dInfo = self.myTrKCient.getJobStatus(jobid)
bWait = dInfo['state'] not in ['FINISHED', 'FAILED']
## coldir???
self.downloadCollection(colid, coldir, docid, bNoImg=True, bForce=True)
##STOP HERE FOR DAS newx testset:
return
# tag text for BIES cell
#python ../../src/tasks/DU_ABPTable_T.py modelMultiType tableRow2 --run=trnskrbs_5400
"""
needed : doer = DU_ABPTable_TypedCRF(sModelName, sModelDir,
"""
doer = DU_ABPTable_TypedCRF(self.sRowModelName, self.sRowModelDir)
doer.load()
## needed predict at file level, and do not store dom, but return it
rowpath = os.path.join(coldir, "col")
BIESFiles = doer.predict([rowpath], docid)
BIESDom = self.loadDom(BIESFiles[0])
# res= BIESDom.saveFormatFileEnc('test.mpxml', "UTF-8",True)
# MPXML2DS
#python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442_du.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
dsconv = primaAnalysis()
DSBIESdoc = dsconv.convert2DS(BIESDom, self.docid)
# create XMLDOC object
self.ODoc = XMLDSDocument()
self.ODoc.loadFromDom(
DSBIESdoc) #,listPages = range(self.firstPage,self.lastPage+1))
# create row
#python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml
rdc = RowDetection()
rdc.findRowsInDoc(self.ODoc)
#python ../../src/xml_formats/DS2PageXml.py -i trnskrbs_5400/out/17442.ds_xml --multi
# DS2MPXML
DS2MPXML = DS2PageXMLConvertor()
lPageXml = DS2MPXML.run(self.ODoc.getDom())
if lPageXml != []:
# if DS2MPXML.bMultiPages:
newDoc = MultiPageXml.makeMultiPageXmlMemory(
map(lambda xy: xy[0], lPageXml))
outputFileName = os.path.join(
self.coldir, sCOL,
self.docid + TableProcessing.sMPXMLExtension)
newDoc.write(outputFileName,
xml_declaration=True,
encoding="UTF-8",
pretty_print=True)
# else:
# DS2MPXML.storePageXmlSetofFiles(lPageXml)
return
#upload
# python ../../../TranskribusPyClient/src/TranskribusCommands/TranskribusDU_transcriptUploader.py --nodu trnskrbs_5400 5400 17442
self.upLoadDocument(colid,
coldir,
docid,
sNote='NLE workflow;table row done')
## apply HTR
## how to deal with specific dictionaries?
## here need to know the ontology and the template
nbPages = 1
jobid = self.applyHTR(colid, docid, nbPages, self.sHTRmodel,
self.sDictName)
bWait = True
traceln("waiting for job %s" % jobid)
while bWait:
dInfo = self.myTrKCient.getJobStatus(jobid)
bWait = dInfo['state'] not in ['FINISHED', 'FAILED', 'CANCELED']
# download where???
# python ../../../TranskribusPyClient/src/TranskribusCommands/Transkribus_downloader.py 5400 --force
# coldir is not right!! coldir must refer to the parent folder!
self.downloadCollection(colid, coldir, docid, bNoImg=True, bForce=True)
#done!!
# IE extr
## not here: specific to a usecas
#python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml --doie --usetemplate
def processCollection(self, coldir):
"""
process all files in a colelction
need mpxml files
"""
lsDocFilename = sorted(
glob.iglob(
os.path.join(coldir, "*" + TableProcessing.sMPXMLExtension)))
lDocId = []
for sDocFilename in lsDocFilename:
sDocId = os.path.basename(
sDocFilename)[:-len(TableProcessing.sMPXMLExtension)]
try:
docid = int(sDocId)
lDocId.append(docid)
except ValueError:
traceln("Warning: folder %s : %s invalid docid, IGNORING IT" %
(self.coldir, sDocId))
continue
# process each document
for docid in lDocId:
traceln("Processing %s : %s " % (self.coldir, sDocId))
self.processDocument(self.colid, docid)
traceln("\tProcessing done for %s " % (self.coldir, sDocId))
def processParameters(self):
"""
what to do with the parameters provided by the command line
"""
if self.colid is None:
print('collection id missing!')
sys.exit(1)
self.bFullCol = self.docid != None
if self.bRegenerateMPXML and self.docid is not None:
l = glob.glob(os.path.join(self.coldir, sCOL, self.docid,
"*.pxml"))
doc = MultiPageXml.makeMultiPageXml(l)
outputFileName = os.path.join(
self.coldir, sCOL,
self.docid + TableProcessing.sMPXMLExtension)
doc.write(outputFileName,
xml_declaration=True,
encoding="UTF-8",
pretty_print=True)
return doc
return None
def run(self):
"""
process at collection level or document level
"""
newMPXML = self.processParameters()
if self.bFullCol is None:
self.processCollection(self.colid)
else:
self.processDocument(self.coldir, self.colid, self.docid, newMPXML)
class TableProcessing(Component.Component):
usage = ""
version = "v.01"
description = "description: table layout analysis based on template"
sCOL = "col"
sMPXMLExtension = ".mpxml"
def __init__(self):
"""
Always call first the Component constructor.
"""
Component.Component.__init__(self, "TableProcessing", self.usage,
self.version, self.description)
self.colid = None
self.docid = None
self.bFullCol = False
# generate MPXML using Ext
self.useExtForMPXML = False
self.bRegenerateMPXML = False
self.sRowModelName = None
self.sRowModelDir = None
self.sHTRmodel = None
self.sDictName = None
def setParams(self, dParams):
"""
Always call first the Component setParams
Here, we set our internal attribute according to a possibly specified value (otherwise it stays at its default value)
"""
Component.Component.setParams(self, dParams)
if "coldir" in dParams:
self.coldir = dParams["coldir"]
if "colid" in dParams:
self.colid = dParams["colid"]
if "colid" in dParams:
self.docid = dParams["docid"]
if "useExt" in dParams:
self.useExtForMPXML = dParams["useExt"]
if 'mergeTLC' in dParams:
self.bUROCVLMerge = dParams["mergeTLC"]
if 'regMPXML' in dParams:
self.bRegenerateMPXML = dParams["regMPXML"]
if "rowmodelname" in dParams:
self.sRowModelName = dParams["rowmodelname"]
if "rowmodeldir" in dParams:
self.sRowModelDir = dParams["rowmodeldir"]
if "htrmodel" in dParams:
self.sHTRmodel = dParams["htrmodel"]
if "dictname" in dParams:
self.sDictName = dParams["dictname"]
# Connection to Transkribus
self.myTrKCient = None
self.persist = False
self.loginInfo = False
if "server" in dParams:
self.server = dParams["server"]
if "persist" in dParams:
self.persist = dParams["persist"]
if "login" in dParams:
self.loginInfo = dParams["login"]
def login(self, trnskrbs_client, trace=None, traceln=None):
"""
deal with the complicated login variants...
-trace and traceln are optional print methods
return True or raises an exception
"""
DEBUG = True
bOk = False
if self.persist:
#try getting some persistent session token
if DEBUG and trace:
trace(" ---login--- Try reusing persistent session ... ")
try:
bOk = trnskrbs_client.reusePersistentSession()
if DEBUG and traceln: traceln("OK!")
except:
if DEBUG and traceln: traceln("Failed")
if not bOk:
if self.loginInfo:
login, pwd = self.loginInfo, self.pwd
else:
if trace:
DEBUG and trace(
" ---login--- no login provided, looking for stored credentials... "
)
login, pwd = trnskrbs_client.getStoredCredentials(bAsk=False)
if DEBUG and traceln: traceln("OK")
if DEBUG and traceln:
trace(" ---login--- logging onto Transkribus as %s " % login)
trnskrbs_client.auth_login(login, pwd)
if DEBUG and traceln: traceln("OK")
bOk = True
return bOk
def downloadCollection(self,
colid,
destDir,
docid,
bNoImg=True,
bForce=False):
"""
download colID
replace destDir by '.' ?
"""
destDir = "."
# options.server, proxies, loggingLevel=logging.WARN)
#download
downloader = TranskribusDownloader(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
downloader.setSessionId(self.myTrKCient.getSessionId())
traceln("- Downloading collection %s to folder %s" %
(colid, os.path.abspath(destDir)))
# col_ts, colDir = downloader.downloadCollection(colid, destDir, bForce=options.bForce, bNoImage=options.bNoImage)
col_ts, colDir, ldocids, dFileListPerDoc = downloader.downloadCollection(
colid, destDir, bForce=bForce, bNoImage=bNoImg, sDocId=docid)
traceln("- Done")
with open(os.path.join(colDir, "config.txt"), "w") as fd:
fd.write("server=%s\nforce=%s\nstrict=%s\n" %
(self.server, True, False))
downloader.generateCollectionMultiPageXml(
os.path.join(colDir, TableProcessing.sCOL), dFileListPerDoc, False)
traceln('- Done, see in %s' % colDir)
return ldocids
def upLoadDocument(self,
colid,
coldir,
docid,
sNote="",
sTranscripExt='.mpxml'):
"""
download colID
"""
# options.server, proxies, loggingLevel=logging.WARN)
#download
# uploader = TranskribusTranscriptUploader(self.server,self.proxies)
uploader = TranskribusDUTranscriptUploader(
self.myTrKCient.getServerUrl(), self.myTrKCient.getProxies())
uploader.setSessionId(self.myTrKCient.getSessionId())
traceln("- uploading document %s to collection %s" % (docid, colid))
uploader.uploadDocumentTranscript(colid,
docid,
os.path.join(coldir, sCOL),
sNote,
'NLE Table',
sTranscripExt,
iVerbose=False)
traceln("- Done")
return
def applyLA_URO(self, colid, docid, nbpages):
"""
apply textline finder
"""
# do the job...
# if options.trp_doc:
# trpdoc = json.load(codecs.open(options.trp_doc, "rb",'utf-8'))
# docId,sPageDesc = doer.buildDescription(colId,options.docid,trpdoc)
traceln('process %s pages...' % nbpages)
lretJobIDs = []
for i in range(1, nbpages + 1):
LA = DoLAbatch(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
LA._trpMng.setSessionId(self.myTrKCient.getSessionId())
LA.setSessionId(self.myTrKCient.getSessionId())
_, sPageDesc = LA.buildDescription(colid, "%s/%s" % (docid, i))
sPageDesc = LA.jsonToXMLDescription(sPageDesc)
_, lJobIDs = LA.run(colid, sPageDesc, "CITlabAdvancedLaJob", False)
traceln(lJobIDs)
lretJobIDs.extend(lJobIDs)
traceln("- LA running for page %d job:%s" % (i, lJobIDs))
return lretJobIDs
def applyHTRForRegions(self, colid, docid, nbpages, modelname, dictionary):
"""
apply an htr model at region level
"""
htrComp = DoHtrRnn(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
htrComp._trpMng.setSessionId(self.myTrKCient.getSessionId())
htrComp.setSessionId(self.myTrKCient.getSessionId())
_, sPageDesc = htrComp.buildDescription(colid,
"%s/%s" % (docid, nbpages))
sPages = "1-%d" % (nbpages)
sModelID = None
# get modelID
lColModels = self.myTrKCient.listRnns(colid)
for model in lColModels:
# print model['htrId'], type(model['htrId']), modelname,type(modelname)
if str(model['htrId']) == str(modelname):
sModelID = model['htrId']
traceln('model id = %s' % sModelID)
#some old? models do not have params field
# try: traceln("%s\t%s\t%s" % (model['htrId'],model['name'],model['params']))
# except KeyError: traceln("%s\t%s\tno params" % (model['htrId'],model['name']))
if sModelID == None:
raise Exception("no model ID found for %s" % (modelname))
ret = htrComp.htrRnnDecode(colid,
sModelID,
dictionary,
docid,
sPageDesc,
bDictTemp=False)
traceln(ret)
return ret
def applyHTR(self, colid, docid, nbpages, modelname, dictionary):
"""
apply HTR on docid
htr id is needed: we have htrmodename
"""
htrComp = DoHtrRnn(self.myTrKCient.getServerUrl(),
self.myTrKCient.getProxies())
htrComp._trpMng.setSessionId(self.myTrKCient.getSessionId())
htrComp.setSessionId(self.myTrKCient.getSessionId())
_, sPageDesc = htrComp.buildDescription(colid,
"%s/%s" % (docid, nbpages))
sPages = "1-%d" % (nbpages)
sModelID = None
# get modelID
lColModels = self.myTrKCient.listRnns(colid)
for model in lColModels:
# print model['htrId'], type(model['htrId']), modelname,type(modelname)
if str(model['htrId']) == str(modelname):
sModelID = model['htrId']
traceln('model id = %s' % sModelID)
#some old? models do not have params field
# try: traceln("%s\t%s\t%s" % (model['htrId'],model['name'],model['params']))
# except KeyError: traceln("%s\t%s\tno params" % (model['htrId'],model['name']))
if sModelID == None:
raise Exception("no model ID found for %s" % (modelname))
ret = htrComp.htrRnnDecode(colid,
sModelID,
dictionary,
docid,
sPageDesc,
bDictTemp=False)
traceln(ret)
return ret
def overlapX(self, zoneA, zoneB):
[x11, y11, x12, y12] = zoneA.getBoundingBox(
) #self.getX(),self.getY(),self.getHeight(),self.getWidth()
[x21, y21, x22, y22] = zoneB.getBoundingBox()
[a1, a2] = x11, x12
[b1, b2] = x21, x22 #zoneB.getX(),zoneB.getX()+ zoneB.getWidth()
return min(a2, b2) >= max(a1, b1)
def overlapY(self, zoneA, zoneB):
[x11, y11, x12, y12] = zoneA.getBoundingBox(
) #self.getX(),self.getY(),self.getHeight(),self.getWidth()
[x21, y21, x22, y22] = zoneB.getBoundingBox()
[a1, a2] = y11, y12
[b1, b2] = y22, y22 #zone.getY(),zone.getY() + zone.getHeight()
return min(a2, b2) >= max(a1, b1)
def signedOverlap(self, zoneA, zoneB):
"""
overlap self and zone
return surface of self in zone
"""
[x11, y11, x12, y12] = zoneA.getBoundingBox(
) #self.getX(),self.getY(),self.getHeight(),self.getWidth()
[x21, y21, x22, y22] = zoneB.getBoundingBox(
) #.getX(),zone.getY(),zone.getHeight(),zone.getWidth()
w1 = x12 - x11
h1 = y12 - y11
fOverlap = 0.0
if self.overlapX(zoneA, zoneB) and self.overlapY(zoneA, zoneB):
s1 = w1 * h1
# possible ?
if s1 == 0: s1 = 1.0
#intersection
nx1 = max(x11, x21)
nx2 = min(x12, x22)
ny1 = max(y11, y21)
ny2 = min(y12, y22)
h = abs(nx2 - nx1)
w = abs(ny2 - ny1)
inter = h * w
if inter > 0:
fOverlap = inter / s1
else:
# if overX and Y this is not possible !
fOverlap = 0.0
return fOverlap
def mergeBaselineCells(self, coldir, colid, docid):
"""
Take a file (pxml) with stuff processed on Transkribus
Tale the CVL template tool xml (xml)
merge them
regenerate a mpxml
"""
xmlpath = os.path.abspath(os.path.join(coldir, sCOL, docid))
# print (xmlpath)
mpxml = xmlpath + ".mpxml"
mpxmldoc = etree.parse(mpxml)
lxml = glob.glob(os.path.join(xmlpath, "*.xml"))
pxmldoc = MultiPageXml.makeMultiPageXml(lxml)
lhtrxml = glob.glob(os.path.join(xmlpath, "*.pxml"))
mpxmldoc = MultiPageXml.makeMultiPageXml(lhtrxml)
lPXMLPage = PageXml.getChildByName(mpxmldoc.getroot(), 'Page')
lXMLPage = PageXml.getChildByName(pxmldoc.getroot(), 'Page')
assert len(lXMLPage) == len(lPXMLPage)
for i, cvlpage in enumerate(lXMLPage):
## remove TextRegion from xcvlpage
lTextRegions = PageXml.getChildByName(cvlpage, 'TextRegion')
for tr in lTextRegions:
tr.getparent().remove(tr)
pxmlpage = lPXMLPage[i]
lTL = []
lTextRegions = PageXml.getChildByName(pxmlpage, 'TextRegion')
for x in lTextRegions:
lTL.extend(PageXml.getChildByName(x, 'TextLine'))
ltable = PageXml.getChildByName(cvlpage, 'TableRegion')
if len(ltable) == 0:
raise "NO TABLE"
lCells = PageXml.getChildByName(ltable[0], 'TableCell')
lC = [Polygon(PageXml.getPointList(c)) for c in lCells]
lT = [Polygon(PageXml.getPointList(t)) for t in lTL]
for i, tl in enumerate(lT):
## normalization
lCoordsPoints = PageXml.getChildByName(lTL[i], 'Coords')
lCoordsB = PageXml.getChildByName(lTL[i], 'Baseline')
coordB = lCoordsB[0]
coord = lCoordsPoints[0]
iHeight = 30 # in pixel
x1, y1, x2, y2 = Polygon(
PageXml.getPointList(coordB)).getBoundingBox()
if coord is not None:
coord.set(
'points', "%d,%d %d,%d %d,%d %d,%d" %
(x1, y1 - iHeight, x2, y1 - iHeight, x2, y2, x1, y2))
tl = Polygon(PageXml.getPointList(coordB))
lOverlap = []
for _, c in enumerate(lC):
# print (lCells[j].get('row'),lCells[j].get('col'), self.signedOverlap(c,tl),tl.getBoundingBox(),c.getBoundingBox())
lOverlap.append(self.signedOverlap(
c, tl)) #.getBoundingBox()))
## region of the same size as the textline
# print (j,max(lOverlap),lOverlap.index(max(lOverlap)))
if max(lOverlap) == 0:
region = PageXml.createPageXmlNode('TextRegion')
cvlpage.append(region)
region.append(lTL[i])
else:
cell = lCells[lOverlap.index(max(lOverlap))]
cell.append(lTL[i])
# print (cell.get('row'),cell.get('col'),''.join(lTL[i].itertext()))
pxmldoc.write(mpxml)
"""
lOverlap=[]
for region in lRegions:
lOverlap.append(self.signedRatioOverlap(region))
if max(lOverlap) == 0: return None
return lRegions[lOverlap.index(max(lOverlap))]
"""
"""
fOverlap = 0.0
if self.overlapX(zone) and self.overlapY(zone):
[x11,y11,x12,y12] = [x1,y1,x1+w1,y1+h1]
[x21,y21,x22,y22] = [x2,y2,x2+w2,y2+h2]
s1 = w1 * h1
# possible ?
if s1 == 0: s1 = 1.0
#intersection
nx1 = max(x11,x21)
nx2 = min(x12,x22)
ny1 = max(y11,y21)
ny2 = min(y12,y22)
h = abs(nx2 - nx1)
w = abs(ny2 - ny1)
inter = h * w
if inter > 0 :
fOverlap = inter/s1
else:
# if overX and Y this is not possible !
fOverlap = 0.0
return fOverlap
"""
def extractFileNamesFromMPXML(self, mpxmldoc):
"""
to insure correct file order !
duplicated form performCVLLA.py
"""
xmlpath = os.path.abspath(os.path.join(self.coldir, sCOL, self.docid))
lNd = PageXml.getChildByName(mpxmldoc.getRootElement(), 'Page')
# for i in lNd:print i
return map(
lambda x: "%s%s%s.xml" %
(xmlpath, os.sep, x.prop('imageFilename')[:-4]), lNd)
def processDocument(self, coldir, colid, docid, dom=None):
"""
process a single document
1 python ../../src/xml_formats/PageXml.py trnskrbs_5400/col/17442 --ext=pxml
2 python ../../src/tasks/performCVLLA.py --coldir=trnskrbs_5400/ --docid=17442 -i trnskrbs_5400/col/17442.mpxml --bl --regTL --form
3 python ../../src/tasks/DU_ABPTable_T.py modelMultiType tableRow2 --run=trnskrbs_5400
4 python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442_du.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
5 python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml
6 python ../../../TranskribusPyClient/src/TranskribusCommands/TranskribusDU_transcriptUploader.py --nodu trnskrbs_5400 5400 17442
7 python ../../../TranskribusPyClient/src/TranskribusCommands/do_htrRnn.py <model-name> <dictionary-name> 5400 17442
wait
8 python ../../../TranskribusPyClient/src/TranskribusCommands/Transkribus_downloader.py 5400 --force
#covnert to ds
9 python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
10 python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml --doie --usetemplate
"""
#create Transkribus client
self.myTrKCient = TranskribusClient(sServerUrl=self.server,
proxies={},
loggingLevel=logging.WARN)
#login
_ = self.login(self.myTrKCient, trace=trace, traceln=traceln)
# self.downloadCollection(colid,coldir,docid,bNoImg=False,bForce=True)
## load dom
if dom is None:
self.inputFileName = os.path.abspath(
os.path.join(coldir, TableProcessing.sCOL,
docid + TableProcessing.sMPXMLExtension))
mpxml_doc = self.loadDom()
nbPages = MultiPageXml.getNBPages(mpxml_doc)
else:
# load provided mpxml
mpxml_doc = dom
nbPages = MultiPageXml.getNBPages(mpxml_doc)
# ### table registration: need to compute/select??? the template
# # perform LA separator, table registration, baseline with normalization
# #python ../../src/tasks/performCVLLA.py --coldir=trnskrbs_5400/ --docid=17442 -i trnskrbs_5400/col/17442.mpxml --bl --regTL --form
# tableregtool= LAProcessor()
# # latool.setParams(dParams)
# tableregtool.coldir = coldir
# tableregtool.docid = docid
# tableregtool.bTemplate, tableregtool.bSeparator , tableregtool.bBaseLine , tableregtool.bRegularTextLine = True,False,False,False
# # creates xml and a new mpxml
# mpxml_doc,nbPages = tableregtool.performLA(mpxml_doc)
#
#
# self.upLoadDocument(colid, coldir,docid,sNote='NLE workflow;table reg done')
lJobIDs = self.apply_URO(colid, docid, nbPages)
return
bWait = True
assert lJobIDs != []
jobid = lJobIDs[-1]
traceln("waiting for job %s" % jobid)
while bWait:
dInfo = self.myTrKCient.getJobStatus(jobid)
bWait = dInfo['state'] not in ['FINISHED', 'FAILED']
## coldir???
self.downloadCollection(colid, coldir, docid, bNoImg=True, bForce=True)
##STOP HERE FOR DAS newx testset:
return
# tag text for BIES cell
#python ../../src/tasks/DU_ABPTable_T.py modelMultiType tableRow2 --run=trnskrbs_5400
"""
needed : doer = DU_ABPTable_TypedCRF(sModelName, sModelDir,
"""
doer = DU_ABPTable_TypedCRF(self.sRowModelName, self.sRowModelDir)
doer.load()
## needed predict at file level, and do not store dom, but return it
rowpath = os.path.join(coldir, "col")
BIESFiles = doer.predict([rowpath], docid)
BIESDom = self.loadDom(BIESFiles[0])
# res= BIESDom.saveFormatFileEnc('test.mpxml', "UTF-8",True)
# MPXML2DS
#python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442_du.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
dsconv = primaAnalysis()
DSBIESdoc = dsconv.convert2DS(BIESDom, self.docid)
# create XMLDOC object
self.ODoc = XMLDSDocument()
self.ODoc.loadFromDom(
DSBIESdoc) #,listPages = range(self.firstPage,self.lastPage+1))
# create row
#python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml
rdc = RowDetection()
rdc.findRowsInDoc(self.ODoc)
#python ../../src/xml_formats/DS2PageXml.py -i trnskrbs_5400/out/17442.ds_xml --multi
# DS2MPXML
DS2MPXML = DS2PageXMLConvertor()
lPageXml = DS2MPXML.run(self.ODoc.getDom())
if lPageXml != []:
# if DS2MPXML.bMultiPages:
newDoc = MultiPageXml.makeMultiPageXmlMemory(
map(lambda xy: xy[0], lPageXml))
outputFileName = os.path.join(
self.coldir, sCOL,
self.docid + TableProcessing.sMPXMLExtension)
newDoc.write(outputFileName,
xml_declaration=True,
encoding="UTF-8",
pretty_print=True)
# else:
# DS2MPXML.storePageXmlSetofFiles(lPageXml)
return
#upload
# python ../../../TranskribusPyClient/src/TranskribusCommands/TranskribusDU_transcriptUploader.py --nodu trnskrbs_5400 5400 17442
self.upLoadDocument(colid,
coldir,
docid,
sNote='NLE workflow;table row done')
## apply HTR
## how to deal with specific dictionaries?
## here need to know the ontology and the template
## OPTION: put it after LA on server (just one download needed )
nbPages = 1
jobid = self.applyHTR(colid, docid, nbPages, self.sHTRmodel,
self.sDictName)
bWait = True
traceln("waiting for job %s" % jobid)
while bWait:
dInfo = self.myTrKCient.getJobStatus(jobid)
bWait = dInfo['state'] not in ['FINISHED', 'FAILED', 'CANCELED']
# download where???
# python ../../../TranskribusPyClient/src/TranskribusCommands/Transkribus_downloader.py 5400 --force
# coldir is not right!! coldir must refer to the parent folder!
self.downloadCollection(colid, coldir, docid, bNoImg=True, bForce=True)
#done!!
# IE extr
## not here: specific to a usecas
#python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml --doie --usetemplate
def processCollection(self, coldir):
"""
process all files in a colelction
need mpxml files
"""
lsDocFilename = sorted(
glob.iglob(
os.path.join(coldir, "*" + TableProcessing.sMPXMLExtension)))
lDocId = []
for sDocFilename in lsDocFilename:
sDocId = os.path.basename(
sDocFilename)[:-len(TableProcessing.sMPXMLExtension)]
try:
docid = int(sDocId)
lDocId.append(docid)
except ValueError:
traceln("Warning: folder %s : %s invalid docid, IGNORING IT" %
(self.coldir, sDocId))
continue
# process each document
for docid in lDocId:
traceln("Processing %s : %s " % (self.coldir, sDocId))
self.processDocument(self.colid, docid)
traceln("\tProcessing done for %s " % (self.coldir, sDocId))
def processParameters(self):
"""
what to do with the parameters provided by the command line
"""
if self.colid is None:
print('collection id missing!')
sys.exit(1)
self.bFullCol = self.docid != None
if self.bRegenerateMPXML and self.docid is not None:
l = glob.glob(os.path.join(self.coldir, sCOL, self.docid,
"*.pxml"))
doc = MultiPageXml.makeMultiPageXml(l)
outputFileName = os.path.join(
self.coldir, sCOL,
self.docid + TableProcessing.sMPXMLExtension)
doc.write(outputFileName,
xml_declaration=True,
encoding="UTF-8",
pretty_print=True)
return doc
return None
def run(self):
"""
process at collection level or document level
"""
newMPXML = self.processParameters()
if self.bFullCol is None:
self.processCollection(self.colid)
else:
if self.bUROCVLMerge:
self.mergeBaselineCells(self.coldir, self.colid, self.docid)
return
self.processDocument(self.coldir, self.colid, self.docid, newMPXML)
def processDocument(self, coldir, colid, docid, dom=None):
"""
process a single document
1 python ../../src/xml_formats/PageXml.py trnskrbs_5400/col/17442 --ext=pxml
2 python ../../src/tasks/performCVLLA.py --coldir=trnskrbs_5400/ --docid=17442 -i trnskrbs_5400/col/17442.mpxml --bl --regTL --form
3 python ../../src/tasks/DU_ABPTable_T.py modelMultiType tableRow2 --run=trnskrbs_5400
4 python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442_du.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
5 python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml
6 python ../../../TranskribusPyClient/src/TranskribusCommands/TranskribusDU_transcriptUploader.py --nodu trnskrbs_5400 5400 17442
7 python ../../../TranskribusPyClient/src/TranskribusCommands/do_htrRnn.py <model-name> <dictionary-name> 5400 17442
wait
8 python ../../../TranskribusPyClient/src/TranskribusCommands/Transkribus_downloader.py 5400 --force
#covnert to ds
9 python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
10 python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml --doie --usetemplate
"""
#create Transkribus client
self.myTrKCient = TranskribusClient(sServerUrl=self.server,
proxies={},
loggingLevel=logging.WARN)
#login
_ = self.login(self.myTrKCient, trace=trace, traceln=traceln)
# self.downloadCollection(colid,coldir,docid,bNoImg=False,bForce=True)
## load dom
if dom is None:
self.inputFileName = os.path.abspath(
os.path.join(coldir, TableProcessing.sCOL,
docid + TableProcessing.sMPXMLExtension))
mpxml_doc = self.loadDom()
nbPages = MultiPageXml.getNBPages(mpxml_doc)
else:
# load provided mpxml
mpxml_doc = dom
nbPages = MultiPageXml.getNBPages(mpxml_doc)
# ### table registration: need to compute/select??? the template
# # perform LA separator, table registration, baseline with normalization
# #python ../../src/tasks/performCVLLA.py --coldir=trnskrbs_5400/ --docid=17442 -i trnskrbs_5400/col/17442.mpxml --bl --regTL --form
# tableregtool= LAProcessor()
# # latool.setParams(dParams)
# tableregtool.coldir = coldir
# tableregtool.docid = docid
# tableregtool.bTemplate, tableregtool.bSeparator , tableregtool.bBaseLine , tableregtool.bRegularTextLine = True,False,False,False
# # creates xml and a new mpxml
# mpxml_doc,nbPages = tableregtool.performLA(mpxml_doc)
#
#
# self.upLoadDocument(colid, coldir,docid,sNote='NLE workflow;table reg done')
lJobIDs = self.applyLA_URO(colid, docid, nbPages)
return
bWait = True
assert lJobIDs != []
jobid = lJobIDs[-1]
traceln("waiting for job %s" % jobid)
while bWait:
dInfo = self.myTrKCient.getJobStatus(jobid)
bWait = dInfo['state'] not in ['FINISHED', 'FAILED']
## coldir???
self.downloadCollection(colid, coldir, docid, bNoImg=True, bForce=True)
##STOP HERE FOR DAS newx testset:
return
# tag text for BIES cell
#python ../../src/tasks/DU_ABPTable_T.py modelMultiType tableRow2 --run=trnskrbs_5400
"""
needed : doer = DU_ABPTable_TypedCRF(sModelName, sModelDir,
"""
doer = DU_ABPTable_TypedCRF(self.sRowModelName, self.sRowModelDir)
doer.load()
## needed predict at file level, and do not store dom, but return it
rowpath = os.path.join(coldir, "col")
BIESFiles = doer.predict([rowpath], docid)
BIESDom = self.loadDom(BIESFiles[0])
# res= BIESDom.saveFormatFileEnc('test.mpxml', "UTF-8",True)
# MPXML2DS
#python ../../src/xml_formats/Page2DS.py --pattern=trnskrbs_5400/col/17442_du.mpxml -o trnskrbs_5400/xml/17442.ds_xml --docid=17442
dsconv = primaAnalysis()
DSBIESdoc = dsconv.convert2DS(BIESDom, self.docid)
# create XMLDOC object
self.ODoc = XMLDSDocument()
self.ODoc.loadFromDom(
DSBIESdoc) #,listPages = range(self.firstPage,self.lastPage+1))
# create row
#python src/IE_test.py -i trnskrbs_5400/xml/17442.ds_xml -o trnskrbs_5400/out/17442.ds_xml
rdc = RowDetection()
rdc.findRowsInDoc(self.ODoc)
#python ../../src/xml_formats/DS2PageXml.py -i trnskrbs_5400/out/17442.ds_xml --multi
# DS2MPXML
DS2MPXML = DS2PageXMLConvertor()
lPageXml = DS2MPXML.run(self.ODoc.getDom())
if lPageXml != []:
# if DS2MPXML.bMultiPages:
newDoc = MultiPageXml.makeMultiPageXmlMemory(
map(lambda xy: xy[0], lPageXml))
outputFileName = os.path.join(
self.coldir, sCOL,
self.docid + TableProcessing.sMPXMLExtension)
newDoc.write(outputFileName,
xml_declaration=True,
encoding="UTF-8",
pretty_print=True)
# else:
# DS2MPXML.storePageXmlSetofFiles(lPageXml)
return
#upload
# python ../../../TranskribusPyClient/src/TranskribusCommands/TranskribusDU_transcriptUploader.py --nodu trnskrbs_5400 5400 17442
self.upLoadDocument(colid,
coldir,
docid,
sNote='NLE workflow;table row done')
## apply HTR
## how to deal with specific dictionaries?
## here need to know the ontology and the template
nbPages = 1
jobid = self.applyHTR(colid, docid, nbPages, self.sHTRmodel,
self.sDictName)
bWait = True
traceln("waiting for job %s" % jobid)
while bWait:
dInfo = self.myTrKCient.getJobStatus(jobid)
bWait = dInfo['state'] not in ['FINISHED', 'FAILED', 'CANCELED']
# download where???
# python ../../../TranskribusPyClient/src/TranskribusCommands/Transkribus_downloader.py 5400 --force
# coldir is not right!! coldir must refer to the parent folder!
self.downloadCollection(colid, coldir, docid, bNoImg=True, bForce=True)
import sys, os
import logging
try: #to ease the use without proper Python installation
import TranskribusPyClient_version
except ImportError:
sys.path.append( os.path.dirname(os.path.dirname( os.path.abspath(sys.argv[0]) )) )
import TranskribusPyClient_version
from TranskribusPyClient.test import _colId_A
from TranskribusPyClient.client import TranskribusClient, getStoredCredentials
login, pwd = getStoredCredentials()
conn = TranskribusClient(proxies={'https':'http://cornillon:8000'}
, loggingLevel=logging.INFO)
sessionID = conn.auth_login(login, pwd)
doc = conn.listEditDeclFeatures(_colId_A)
doc.saveFormatFileEnc("-", "UTF-8", True)
conn.xmlFreeDoc(doc)
print conn.auth_logout()
"""
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<edFeatures>
<edFeature>
<featureId>1</featureId>
<title>Long S</title>
<description>Source uses long "s"</description>
# -*- coding: utf-8 -*-
#optional: useful if you want to choose the logging level to something else than logging.WARN
import sys, os
import logging
try: #to ease the use without proper Python installation
import TranskribusPyClient_version
except ImportError:
sys.path.append(
os.path.dirname(os.path.dirname(os.path.abspath(sys.argv[0]))))
import TranskribusPyClient_version
from TranskribusPyClient.test import _coldId_Sandbox, _docId_a
from TranskribusPyClient.client import TranskribusClient, getStoredCredentials
login, pwd = getStoredCredentials()
conn = TranskribusClient(proxies={'https': 'http://cornillon:8000'},
loggingLevel=logging.INFO)
sessionID = conn.auth_login(login, pwd)
data = conn.addDocToCollection(_coldId_Sandbox, _docId_a)
"""
True or Exception
"""
print conn.auth_logout()
# -*- coding: utf-8 -*-
#optional: useful if you want to choose the logging level to something else than logging.WARN
import sys, os
import logging
try: #to ease the use without proper Python installation
import TranskribusPyClient_version
except ImportError:
sys.path.append(
os.path.dirname(os.path.dirname(os.path.abspath(sys.argv[0]))))
import TranskribusPyClient_version
from TranskribusPyClient.test import _colId_A, _coldId_Sandbox, _docId_c, _docId_d
from TranskribusPyClient.client import TranskribusClient, getStoredCredentials
login, pwd = getStoredCredentials()
conn = TranskribusClient(proxies={'https': 'http://cornillon:8000'},
loggingLevel=logging.INFO)
sessionID = conn.auth_login(login, pwd)
data = conn.duplicateDoc(_colId_A, _docId_c, _coldId_Sandbox, "named_by_JL")
data = conn.duplicateDoc(_colId_A, _docId_d, _coldId_Sandbox)
"""
True or Exception
"""
print conn.auth_logout()
import sys, os
import logging
try: #to ease the use without proper Python installation
import TranskribusPyClient_version
except ImportError:
sys.path.append( os.path.dirname(os.path.dirname( os.path.abspath(sys.argv[0]) )) )
import TranskribusPyClient_version
from TranskribusPyClient.test import _colId_A, _docId_a
from TranskribusPyClient.client import TranskribusClient, getStoredCredentials
login, pwd = getStoredCredentials()
conn = TranskribusClient(proxies={'https':'http://*****:*****@xrce.xerox.com'}]},
u'createdFromTimestamp': 33175290,
u'createdToTimestamp': 33175290,
import sys, os
import logging
try: #to ease the use without proper Python installation
import TranskribusPyClient_version
except ImportError:
sys.path.append(
os.path.dirname(os.path.dirname(os.path.abspath(sys.argv[0]))))
import TranskribusPyClient_version
from TranskribusPyClient.test import _colId_A, _docId_a
from TranskribusPyClient.client import TranskribusClient, getStoredCredentials
login, pwd = getStoredCredentials()
conn = TranskribusClient(proxies={'https': 'http://*****:*****@xrce.xerox.com", "trnjluc", sHttpsProxyUrl='http://cornillon:8000')
# ret = conn.getDocumentFromServer(colid, docid)
#ret = conn.getDocumentFromServer("3571", "7750")
data = conn.getDocByIdAsXml(_colId_A, str(_docId_a)) #str just to stress-test
#data = conn.getDocByIdAsXml(3571, "7750")
print data
"""
