def loadFrameFe(self, feNode):
"""
"""
goodNodes = getNoneTextChildNodes(feNode)
fe = {}
fe = loadXMLAttributes(fe, feNode.attributes)
fe['semtypes'] = {}
for gn in goodNodes:
if gn.nodeName == 'definition':
if not fe.has_key('definition'):
try:
fe['definition'] = gn.childNodes[0].nodeValue
except:
fe['definition'] = None
else:
print >> sys.stderr, 'Error , fe already have a definition:', fe['definition']
return None
elif gn.nodeName == 'semTypes':
goodSemTypeNodes = getNoneTextChildNodes(gn)
for gsn in goodSemTypeNodes:
semType = {}
loadXMLAttributes(semType, gsn.attributes)
fe['semtypes'][semType['ID']] = semType
else:
print >> sys.stderr, 'In loadFrameFe, found this node:', gn.nodeName
return None
return fe
def loadXML(self, fileName):
"""
"""
doc = xml.dom.minidom.parse(fileName)
relationTypeNodes = getNoneTextChildNodes(
doc.childNodes[1]) #the actual frame-relation-type nodes
for rtn in relationTypeNodes:
rt = {} #dictionary for frame-relation-type
loadXMLAttributes(rt, rtn.attributes)
relationNodes = getNoneTextChildNodes(
rtn) #the actual frame-relationS nodes
if len(relationNodes) != 1:
print >> sys.stderr, 'Got more than one frame-relations node in type:', rt[
'name']
return False
singleRelationNodes = getNoneTextChildNodes(
relationNodes[0]) # the actual frame-relation nodes
singleRelations = {}
i = 0
for srn in singleRelationNodes:
tmp = self.loadSingleRelation(srn)
if tmp == None:
print >> sys.stderr, 'Unable to load relation No.' + str(
i), 'for type', rt['name']
return False
singleRelations[tmp['ID']] = tmp
rt['frame-relations'] = singleRelations
self['relation-types'][rt['ID']] = rt
return True
def loadXML(self, fileName):
"""
"""
doc = xml.dom.minidom.parse(fileName)
relationTypeNodes = getNoneTextChildNodes(doc.childNodes[1]) #the actual frame-relation-type nodes
for rtn in relationTypeNodes:
rt = {} #dictionary for frame-relation-type
loadXMLAttributes(rt, rtn.attributes)
relationNodes = getNoneTextChildNodes(rtn) #the actual frame-relationS nodes
if len(relationNodes) != 1:
print >> sys.stderr, 'Got more than one frame-relations node in type:', rt['name']
return False
singleRelationNodes = getNoneTextChildNodes(relationNodes[0]) # the actual frame-relation nodes
singleRelations = {}
i = 0
for srn in singleRelationNodes:
tmp = self.loadSingleRelation(srn)
if tmp == None:
print >> sys.stderr, 'Unable to load relation No.' + str(i), 'for type', rt['name']
return False
singleRelations[tmp['ID']] = tmp
rt['frame-relations'] = singleRelations
self['relation-types'][rt['ID']] = rt
return True
def loadFrameFe(self, feNode):
"""
"""
goodNodes = getNoneTextChildNodes(feNode)
fe = {}
fe = loadXMLAttributes(fe, feNode.attributes)
fe['semtypes'] = {}
for gn in goodNodes:
if gn.nodeName == 'definition':
if not fe.has_key('definition'):
try:
fe['definition'] = gn.childNodes[0].nodeValue
except:
fe['definition'] = None
else:
print >> sys.stderr, 'Error , fe already have a definition:', fe[
'definition']
return None
elif gn.nodeName == 'semTypes':
goodSemTypeNodes = getNoneTextChildNodes(gn)
for gsn in goodSemTypeNodes:
semType = {}
loadXMLAttributes(semType, gsn.attributes)
fe['semtypes'][semType['ID']] = semType
else:
print >> sys.stderr, 'In loadFrameFe, found this node:', gn.nodeName
return None
return fe
def loadXMLNode(self, frameNode):
"""
"""
loadXMLAttributes(self, frameNode.attributes)
goodNodes = getNoneTextChildNodes(frameNode)
for node in goodNodes:
if node.nodeName == 'definition':
if not self.loadFrameDefinition(node):
print >> sys.stderr, 'Unable to read definition node for frame:', self[
'ID']
elif node.nodeName == 'fes':
if not self.loadFrameFes(node):
print >> sys.stderr, 'Unable to read a fes node for frame:', self[
'ID']
elif node.nodeName == 'lexunits':
if not self.loadFrameLexunits(node):
print >> sys.stderr, 'Unable to read a lexunits node for frame:', self[
'ID']
elif node.nodeName == 'semTypes':
if not self.loadFrameSemtypes(node):
print >> sys.stderr, 'Unable to read a semtypes node for frame:', self[
'ID']
else:
print >> sys.stderr, 'Have no idea how to deal with node type:', node.nodeName
return False
return True
def loadSingleRelation(self, relationNode):
"""
"""
frRelation = {}
loadXMLAttributes(frRelation, relationNode.attributes)
feNodes = getNoneTextChildNodes(relationNode)
for fn in feNodes:
tmp = {}
loadXMLAttributes(tmp, fn.attributes)
frRelation[tmp['ID']] = tmp
return frRelation
def loadFrameSemtypes(self, node):
"""
"""
goodSemTypeNodes = getNoneTextChildNodes(node)
semTypes = {}
for gsn in goodSemTypeNodes:
semType = {}
loadXMLAttributes(semType, gsn.attributes)
semTypes[semType['ID']] = semType
self['semtypes'] = semTypes
return True
def loadSingleRelation(self, relationNode):
"""
"""
frRelation = {}
loadXMLAttributes(frRelation, relationNode.attributes)
feNodes = getNoneTextChildNodes(relationNode)
for fn in feNodes:
tmp = {}
loadXMLAttributes(tmp, fn.attributes)
frRelation[tmp['ID']] = tmp
return frRelation
def loadFrameSemtypes(self, node):
"""
"""
goodSemTypeNodes = getNoneTextChildNodes(node)
semTypes = {}
for gsn in goodSemTypeNodes:
semType = {}
loadXMLAttributes(semType, gsn.attributes)
semTypes[semType['ID']] = semType
self['semtypes'] = semTypes
return True
def loadSingleLayer(self, layerNode):
"""
Loads a single layer, <layer>
"""
layer = {'labels': {}}
try:
layer['ID'] = int(layerNode.attributes['ID'].value)
layer['name'] = str(layerNode.attributes['name'].value)
except:
print >> sys.stderr, '>> Unable to load layer id or name'
return None
labelsNode = filter(lambda x: x.nodeType != x.TEXT_NODE,
layerNode.childNodes)
if len(labelsNode) > 1:
print >> sys.stderr, '>> Got more than one labels node for a layer'
print >> sys.stderr, labelsNode
return None
if len(labelsNode) > 0:
labels = filter(lambda x: x.nodeType != x.TEXT_NODE,
labelsNode[0].childNodes)
for l in labels:
ll = {}
ll = loadXMLAttributes(ll, l.attributes)
layer['labels'][ll['ID']] = ll
self['layers'][layer['ID']] = layer
return layer
def loadSentence(self, sentNode, annotation):
"""
Loads the sentence
"""
goodNodes = filter(lambda x: x.nodeType != x.TEXT_NODE,
sentNode.childNodes)
if len(goodNodes) != 1:
print >> sys.stderr, '>> Sent node has:', len(
goodNodes), 'good nodes'
return False
sent = {}
try:
sent = loadXMLAttributes(sent, sentNode.attributes)
except:
print >> sys.stderr, '>> Unable to get one of ID or aPos'
return False
sent['text'] = goodNodes[0].childNodes[0].nodeValue
if not isinstance(sent['text'], unicode):
print >> sys.stderr, '>> Unable to get the sentence text from', goodNodes[
0].childNodes[0]
return False
try:
sent['text'] = str(sent['text'])
except:
pass
annotation['sentence'] = sent
return True
def loadSingleLayer(self, layerNode):
"""
Loads a single layer, <layer>
"""
layer = {"labels": {}}
try:
layer["ID"] = int(layerNode.attributes["ID"].value)
layer["name"] = str(layerNode.attributes["name"].value)
except:
print >>sys.stderr, ">> Unable to load layer id or name"
return None
labelsNode = filter(lambda x: x.nodeType != x.TEXT_NODE, layerNode.childNodes)
if len(labelsNode) > 1:
print >>sys.stderr, ">> Got more than one labels node for a layer"
print >>sys.stderr, labelsNode
return None
if len(labelsNode) > 0:
labels = filter(lambda x: x.nodeType != x.TEXT_NODE, labelsNode[0].childNodes)
for l in labels:
ll = {}
ll = loadXMLAttributes(ll, l.attributes)
layer["labels"][ll["ID"]] = ll
self["layers"][layer["ID"]] = layer
return layer
def loadSentence(self, sentNode, annotation):
"""
Loads the sentence
"""
goodNodes = filter(lambda x: x.nodeType != x.TEXT_NODE, sentNode.childNodes)
if len(goodNodes) != 1:
print >>sys.stderr, ">> Sent node has:", len(goodNodes), "good nodes"
return False
sent = {}
try:
sent = loadXMLAttributes(sent, sentNode.attributes)
except:
print >>sys.stderr, ">> Unable to get one of ID or aPos"
return False
sent["text"] = goodNodes[0].childNodes[0].nodeValue
if not isinstance(sent["text"], unicode):
print >>sys.stderr, ">> Unable to get the sentence text from", goodNodes[0].childNodes[0]
return False
try:
sent["text"] = str(sent["text"])
except:
pass
annotation["sentence"] = sent
return True
def loadXMLNode(self, frameNode):
"""
"""
loadXMLAttributes(self, frameNode.attributes)
goodNodes = getNoneTextChildNodes(frameNode)
for node in goodNodes:
if node.nodeName == 'definition':
if not self.loadFrameDefinition(node):
print >> sys.stderr, 'Unable to read definition node for frame:', self['ID']
elif node.nodeName == 'fes':
if not self.loadFrameFes(node):
print >> sys.stderr, 'Unable to read a fes node for frame:', self['ID']
elif node.nodeName == 'lexunits':
if not self.loadFrameLexunits(node):
print >> sys.stderr, 'Unable to read a lexunits node for frame:', self['ID']
elif node.nodeName == 'semTypes':
if not self.loadFrameSemtypes(node):
print >> sys.stderr, 'Unable to read a semtypes node for frame:', self['ID']
else:
print >> sys.stderr, 'Have no idea how to deal with node type:', node.nodeName
return False
return True
def loadFrameLexicalUnit(self, lexunitNode):
"""
"""
lexunit = {}
loadXMLAttributes(lexunit, lexunitNode.attributes)
goodNodes = getNoneTextChildNodes(lexunitNode)
for gn in goodNodes:
if gn.nodeName == 'definition':
try:
lexunit['definition'] = gn.childNodes[0].nodeValue
except:
lexunit['definition'] = None
elif gn.nodeName == 'annotation':
annoNodes = getNoneTextChildNodes(gn)
anno = {}
for an in annoNodes:
try:
anno[str(an.nodeName)] = an.childNodes[0].nodeValue
try:
n = int(anno[str(an.nodeName)])
anno[str(an.nodeName)] = n
except:
pass
except:
anno[str(an.nodeName)] = None
print >> sys.stderr, 'Warning!! unable to retrieve', an.nodeName, 'for annotation'
lexunit['annotation'] = anno
elif gn.nodeName == 'lexemes':
goodSemTypeNodes = getNoneTextChildNodes(gn)
lexemes = {}
for gsn in goodSemTypeNodes:
lexeme = {}
loadXMLAttributes(lexeme, gsn.attributes)
# store the actual word
lexeme['lexeme'] = gsn.childNodes[0].nodeValue
lexemes[lexeme['ID']] = lexeme
lexunit['lexeme'] = lexemes
elif gn.nodeName == 'semTypes':
goodSemTypeNodes = getNoneTextChildNodes(gn)
semTypes = {}
for gsn in goodSemTypeNodes:
semType = {}
loadXMLAttributes(semType, gsn.attributes)
semTypes[semType['ID']] = semType
lexunit['semtypes'] = semTypes
else:
print >> sys.stderr, 'Error, encounted the node:', gn.nodeName, 'in', lexunitNode.nodeName, 'lexunit'
return None
return lexunit
def loadFrameLexicalUnit(self, lexunitNode):
"""
"""
lexunit = {}
loadXMLAttributes(lexunit, lexunitNode.attributes)
goodNodes = getNoneTextChildNodes(lexunitNode)
for gn in goodNodes:
if gn.nodeName == 'definition':
try:
lexunit['definition'] = gn.childNodes[0].nodeValue
except:
lexunit['definition'] = None
elif gn.nodeName == 'annotation':
annoNodes = getNoneTextChildNodes(gn)
anno = {}
for an in annoNodes:
try:
anno[str(an.nodeName)] = an.childNodes[0].nodeValue
try:
n = int(anno[str(an.nodeName)])
anno[str(an.nodeName)] = n
except:
pass
except:
anno[str(an.nodeName)] = None
print >> sys.stderr, 'Warning!! unable to retrieve', an.nodeName, 'for annotation'
lexunit['annotation'] = anno
elif gn.nodeName == 'lexemes':
goodSemTypeNodes = getNoneTextChildNodes(gn)
lexemes = {}
for gsn in goodSemTypeNodes:
lexeme = {}
loadXMLAttributes(lexeme, gsn.attributes)
# store the actual word
lexeme['lexeme'] = gsn.childNodes[0].nodeValue
lexemes[lexeme['ID']] = lexeme
lexunit['lexeme'] = lexemes
elif gn.nodeName == 'semTypes':
goodSemTypeNodes = getNoneTextChildNodes(gn)
semTypes = {}
for gsn in goodSemTypeNodes:
semType = {}
loadXMLAttributes(semType, gsn.attributes)
semTypes[semType['ID']] = semType
lexunit['semtypes'] = semTypes
else:
print >> sys.stderr, 'Error, encounted the node:', gn.nodeName, 'in', lexunitNode.nodeName, 'lexunit'
return None
return lexunit
