def __init__(self,t,locationAnnotator):
"""
initialize a graph class, followed by converting a tree
@type t: Tree
@param tree: syntactic tree to be converted
@type id: int
@param id: a unique id for current Tree
@type gr: digraph
@var gr: the graph representing t
"""
if not t.id: # meaning this is the ROOT element
self.tree = t.children[0]
else:
self.tree = t
self.gr = GraphWrapper(t.get_original_sentence())
self.locationAnnotator = locationAnnotator
# maintain an appendix for easier browsing
self.types = appendix_types()
self.parse(self.tree)
def parseSentences(sent, HOME_DIR = BASE_PATH):
orig_Stdin = sys.stdin
strIn = StringIO(sent)
sys.stdin = strIn
strOut = StringIO()
parseInput(parser, opts, outputFile=strOut)
sys.stdin = orig_Stdin
# Now we can retrieve the output as a string:
result = strOut.getvalue()
#print >>sys.stderr, result
#print result
tmp_fn = "./tmp.tmp"
fout = open(tmp_fn,'w')
fout.write(result)
fout.close()
graphs = read_dep_graphs_file(tmp_fn,False,HOME_DIR)
ret = []
for graph in graphs:
g = convert(graph)
ret.append((g,g.tree_str))
if not graphs:#Berkley bug?
ret.append((GraphWrapper("",HOME_DIR),""))
strIn.close()
strOut.close()
return ret
def create_dep_graphs_from_stream(stream, HOME_DIR):
graphs = []
init = True
curGraph = GraphWrapper("", HOME_DIR)
nodesMap = {}
for line in stream:
line = line.strip()
# print line
if line:
init = False
m = pat.match(line)
rel, head, head_id, dep, dep_id = m.groups()
# head_id = int(head_id)
# dep_id = int(dep_id)
if head_id not in nodesMap:
nodesMap[head_id] = Node(
text=[Word(index=int(head_id.split("'")[0]), word=head)],
isPredicate=False,
features={},
gr=curGraph,
orderText=True)
if dep_id not in nodesMap:
nodesMap[dep_id] = Node(
text=[Word(index=int(dep_id.split("'")[0]), word=dep)],
isPredicate=False,
features={},
gr=curGraph,
orderText=True)
headNode = nodesMap[head_id]
depNode = nodesMap[dep_id]
if curGraph.has_edge((headNode, depNode)): # stanford bug
curGraph.del_edge((headNode, depNode))
curGraph.add_edge(edge=(nodesMap[head_id], nodesMap[dep_id]),
label=rel)
if (not line) and (not init):
init = True
graphs.append((curGraph, nodesMap))
curGraph = GraphWrapper("", HOME_DIR)
nodesMap = {}
return graphs
def load_prop_from_file(filename, HOME_DIR):
fin = open(filename)
flag = True
ret = []
for line in fin:
line = line.strip("\n")
if flag:
curSentence = line
flag = False
curGraph = GraphWrapper(curSentence, HOME_DIR)
parentsList = []
else:
if line:
uid, words, pos, isPredicate, isAsserted, parents = line.split(
"\t")
uid = int(uid)
isAsserted = bool(int(isAsserted))
text = [
Word(int(index), word) for index, word in
[ent.split(",") for ent in words.split(";")]
]
if isAsserted:
feats = {"top": isAsserted}
else:
feats = {}
if parents:
parentsList.extend([
((int(index), uid), rel) for rel, index in
[ent.split(",") for ent in parents.split(";")]
])
curNode = newNode.Node(text,
bool(int(isPredicate)),
feats,
curGraph,
uid=uid)
else:
for edge, rel in parentsList:
digraph.add_edge(curGraph, edge=edge, label=rel)
ret.append(curGraph)
flag = True
fin.close()
return ret
def tree_to_graph(tree):
"""
@type tree DepTree
"""
HOME_DIR = os.environ.get("PROPEXTRACTION_HOME_DIR") + "\\"
ret = GraphWrapper(tree[0].original_sentence, HOME_DIR)
graphNodes = {}
for t in tree.values():
if t.id:
if t.parent_relation != "erased":
graphNodes[t.id] = treeNode_to_graphNode(t, ret)
for t in tree.values():
if t.id:
curParent = t.parent.id
if curParent:
ret.add_edge(edge=(graphNodes[curParent], graphNodes[t.id]),
label=t.parent_relation)
return ret
def parseSentences(sent, HOME_DIR=BASE_PATH, stanford_json_sent=None):
if stanford_json_sent:
# Use Stanford json notation
result = sent['parse'].replace("\n", "") + "\n"
else:
# Use default berkeley parser
from BerkeleyInterface import parseInput
orig_Stdin = sys.stdin
strIn = StringIO(sent)
sys.stdin = strIn
strOut = StringIO()
parseInput(parser, opts, outputFile=strOut)
sys.stdin = orig_Stdin
result = strOut.getvalue()
tmp_fn = "./tmp.parse"
fout = open(tmp_fn, 'w')
fout.write(result)
fout.close()
graphs = read_dep_graphs_file(tmp_fn,
False,
HOME_DIR,
stanford_json_sent = sent \
if stanford_json_sent \
else False)
ret = []
for graph in graphs:
g = convert(graph)
ret.append((g, g.tree_str))
if not graphs: #Berkley bug?
ret.append((GraphWrapper("", HOME_DIR), ""))
if (not stanford_json_sent):
strIn.close()
strOut.close()
return ret