def rebuilt2xmi(ci, output_dir, typesystem_path):
"""
Converts a rebuilt ContentItem into Apache UIMA/XMI format.
The resulting file will be named after the content item's ID, adding
the `.xmi` extension.
:param ci: the content item to be converted
:type ci: `impresso_commons.classes.ContentItem`
:param output_dir: the path to the output directory
:type output_dir: str
:param typesystem_path: TypeSystem file containing defitions of annotation
layers.
:type typesystem_path: str
"""
tsf = TypeSystemFactory()
tsf = tsf.readTypeSystem(typesystem_path)
cas = CAS(tsf)
cas.documentText = ci.fulltext
cas.sofaMimeType = 'text'
sentType = 'de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence'
imgLinkType = 'webanno.custom.ImpressoImages'
# create sentence-level annotations
start_offset = 0
for break_offset in ci.lines:
start = start_offset
end = break_offset
start_offset = break_offset
sntc = cas.createAnnotation(sentType, {'begin': start, 'end': end})
cas.addToIndex(sntc)
iiif_links = compute_image_links(ci)
for iiif_link, start, end in iiif_links:
imglink = cas.createAnnotation(imgLinkType, {
'begin': start,
'end': end,
'link': iiif_link
})
cas.addToIndex(imglink)
writer = XmiWriter()
outfile_path = os.path.join(output_dir, f'{ci.id}.xmi')
writer.write(cas, outfile_path)
def buildCAS(self,xmifilepath,typefilepath):
#create type ystem object
typesystem = TypeSystemFactory.readTypeSystem(self, typefilepath)
#create a CAS object
cas = CAS(typesystem)
#create cas xmi perser object to fetch elements from xmi file
casXmiParser = CasXmiParser()
casXmiParser.setXmiAsFile(xmifilepath)
return self.__build(cas, casXmiParser)
def buildCASfromStrings(self, xmistring, typesysstemString):
# create type ystem object
typesystem = TypeSystemFactory.readTypeSystemString(self, typesysstemString)
# create a CAS object
cas = CAS(typesystem)
# create cas xmi perser object to fetch elements from xmi file
casXmiParser = CasXmiParser()
casXmiParser.setXmiAsString(xmistring)
return self.__build(cas, casXmiParser)
def test_Cas(self):
typeSystemFilePath = 'typesystem.xml'
typesystem = TypeSystemFactory.readTypeSystem(self, typeSystemFilePath)
cas = CAS(typesystem)
cas.documentText = 'These steps install the basis system requirements'
cas.sofaMimeType = 'text'
sentenceType = 'de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence'
tokenType = 'de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token'
posType = 'de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS'
tagDescType = 'de.tudarmstadt.ukp.dkpro.core.api.metadata.type.TagDescription'
tagSetDescType = 'de.tudarmstadt.ukp.dkpro.core.api.metadata.type.TagsetDescription'
fsSentence1 = cas.createFS(sentenceType)
cas.addToIndex(fsSentence1)
#check empty FS
self.assertEqual(fsSentence1.getFeatureValsAsDictList(), [])
#throws error, 1, 'as FSid is already occupied' , assigned to auto generated Sofa FS
self.assertEqual('', '')
with self.assertRaises(ValueError):
fsSentence2 = cas.createFS(sentenceType, {
'begin': 10,
'end': 20
}, 1)
fsPOS = cas.createFS(posType, {'PosValue': 'NN'}, 3)
cas.addToIndex(fsPOS)
fsPOS.PosValue = 'Noun'
#throws error, need to provide feature as dictionary as second argument
self.assertEqual('', '')
with self.assertRaises(TypeError):
fsPOSx = cas.createAnnotation(posType, 2)
#throws error, as 3 is already set as FS id , lower than 3 can not be set by user
self.assertEqual('', '')
with self.assertRaises(ValueError):
fsPOSx = cas.createAnnotation(posType, {'PosValue': 'NN'}, 2)
#throws error, needs both begin and end
self.assertEqual('', '')
with self.assertRaises(ValueError):
fsPOSx = cas.createAnnotation(posType, {
'begin': 10,
'PosValue': 'NN'
}, 4)
#create a valid annotation FS
fsPOS = cas.createAnnotation(posType, {
'begin': 0,
'end': 5,
'PosValue': 'NN'
})
#add it to index
cas.addToIndex(fsPOS)
fsPOS1 = cas.createAnnotation(posType, {
'begin': 0,
'end': 5
})
fsPOS1.PosValue = 'NN'
cas.addToIndex(fsPOS1)
fsToken1 = cas.createAnnotation(tokenType, {
'begin': 0,
'end': 5,
'pos': fsPOS
})
cas.addToIndex(fsToken1)
tdlist = []
fstagDesc1 = cas.createFS(tagDescType, {'name': '#'})
tdlist.append(fstagDesc1)
fstagDesc2 = cas.createFS(tagDescType, {'name': '$'})
tdlist.append(fstagDesc2)
fstagDesc3 = cas.createFS(tagDescType, {'name': '-LRB-'})
tdlist.append(fstagDesc3)
fstagSetDesc = cas.createAnnotation(
tagSetDescType, {
'begin': 0,
'end': 152,
'layer':
'de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS',
'name': 'ptb',
'tags': tdlist
})
cas.addToIndex(fstagSetDesc)
#get all FS in index
for e in cas.getAnnotationIndex():
#check FSid
if (e.FStype.name == "uima.cas.Sofa"):
self.assertEqual(e.FSid, 1)
if (e.FStype.name ==
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence"
):
self.assertEqual(e.FSid, 2)
if (e.FStype.name ==
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token"
):
self.assertEqual(e.FSid, 7)
#check FS type
if (e.FSid == 1):
self.assertEqual(e.FStype.name, "uima.cas.Sofa")
if (e.FSid == 2):
self.assertEqual(
e.FStype.name,
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence"
)
if (e.FSid == 7):
self.assertEqual(
e.FStype.name,
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token"
)
#check sofa
self.assertEqual(cas.sofaMimeType, cas.sofaFS.mimeType)
self.assertEqual(1, cas.sofaFS.sofaNum)
self.assertEqual('_InitialView', cas.sofaFS.sofaID)
self.assertEqual(cas.documentText, cas.sofaFS.sofaString)
tokens = cas.getAnnotation(
'de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token')
self.assertEqual(tokens[0].getCoveredText(), "These ")