def load(self,args):
fh = None
try:
fh = args['filehandle']
except:
filename = args['filename']
fh = open(filename, 'r')
fh = codecs.getreader('utf8')(fh)
nodes = []
comment = ''
for line in fh:
if re.search('^#',line):
comment = comment + line
elif re.search('^\d+\-',line): # HACK: multiword tokens temporarily avoided
pass
elif line.strip():
if not nodes:
bundle = Bundle()
self.bundles.append(bundle)
root = Root() # TODO: nahradit bundle.create_tree, az bude odladene
root._aux['comment'] = comment # TODO: ulozit nekam poradne
nodes = [root]
bundle.trees.append(root)
columns = line.strip().split('\t')
node = Node()
nodes.append(node)
for index in xrange(0,len(Document.attrnames)):
setattr( node, Document.attrnames[index], columns[index] )
try: # TODO: kde se v tomhle sloupecku berou podtrzitka
node.head = int(node.head)
except ValueError:
node.head = 0
try: # TODO: poresit multitokeny
node.ord = int(node.ord)
except ValueError:
node.ord = 0
else: # an empty line is guaranteed even after the last sentence in a conll-u file
nodes[0]._aux['descendants'] = nodes[1:]
for node in nodes[1:]:
node.set_parent( nodes[node.head] )
nodes = []
comment = ''
class LinkedList:
def __init__(self,value = None):
# if no value is passed in
if(value == None):
self._size = 0
self._headNode = None
self._tailNode = None
else:
self._headNode = Node(value,None)
self._tailNode = self._headNode
self._size = 1
def get(self,index):
self._verifyIndex(index)
if(index == 0):
return self._headNode.head()
elif(index == self._size - 1):
return self._tailNode.head()
else:
i = 0
walk = self._headNode
while(i < index):
walk = walk.tail()
i += 1
return walk.head()
def set(self,index,value):
self._verifyIndex(index)
i = 0
walk = self._headNode
while(i < index):
walk = walk.tail()
i += 1
walk.setHead(value)
def frontadd(self,value):
self._headNode = Node(value,self._headNode)
# if the headNode has tail None (is the last node)
# then it IS tailNode
if(self._headNode.tail() == None):
self._tailNode = self._headNode
self._size += 1
def backadd(self,value):
if(self._headNode == None):
self.frontadd(value)
return
self._tailNode.setTail(Node(value,None))
self._tailNode = self._tailNode.tail()
self._size += 1
def indexadd(self,index,value):
if(index == 0):
self.frontadd(value)
return
if(index == self._size):
self.backadd(value)
return
if(index > self._size):
raise IndexError
i = 0
walk = self._headNode
while(i < index-1):
walk = walk.tail()
i += 1
walk.setTail(Node(value,walk.tail()))
self._size += 1
def indexremove(self,index):
if(index == 0):
self.frontremove()
return
if(index == self._size - 1):
rval = self.backremove()
return rval
self._verifyIndex(index)
i = 0
walk = self._headNode
while(i < index-1):
walk = walk.tail()
i += 1
rval = walk.tail().head()
walk.setTail(walk.tail().tail())
self._size -= 1
return rval
def frontremove(self):
rval = self._headNode.head()
self._headNode = self._headNode.tail()
self._size -= 1
return rval
def backremove(self):
rval = self._tailNode.head()
walk = self._headNode
while(walk.tail() != self._tailNode):
walk = walk.tail()
self._tailNode = walk
self._tailNode.setTail(None)
self._size -= 1
return rval
def extract(self):
if(self._headNode == None):
return []
rlist = []
walk = self._headNode
while(walk.tail() != None):
rlist.append(walk.head())
walk = walk.tail()
rlist.append(walk.head())
return rlist
def size(self):
return self._size
def _verifyIndex(self,index):
if(index > self._size-1):
raise IndexError("index out of bounds")
else: return