handled sentences with multiple roots in pretty print

Author: P Sruthi

main.py

@@ -2,11 +2,20 @@
 from my_parser_eval import _write_input
 from my_parser_eval import _read_output
 from my_parser_eval import pretty_print
+import sys
+
+if (len(sys.argv) > 2):
+print("Please only call me with one parameter")
+sys.exit()
+
+sentence = sys.argv [1]
+print(sentence)
+print "====================="
 
 tagger = SyntaxNetProcess("brain_tagger")
 parser = SyntaxNetProcess("brain_parser")
 
-_write_input("Bob brought the pizza to Alice")
+_write_input(sentence)
 tagger.eval()
 print "TAGGER OUTPUT=================="
 print _read_output()

my_parser_eval.py

@@ -23,7 +23,7 @@
 from syntaxnet import sentence_pb2, structured_graph_builder
 from syntaxnet.ops import gen_parser_ops
 import syntaxnet.load_parser_ops
-from syntaxnet.conll2tree import to_dict
+# from syntaxnet.conll2tree import to_dict
 
 input_file_path = os.path.join(PROJECT_ROOT, "data", "input-file.txt")
 output_file_path = os.path.join(PROJECT_ROOT, "data", "output-file.txt")
@@ -49,6 +49,92 @@ def _write_input(sentence):
  input_file.flush()
 input_file.close()
 
+def as_asciitree(sentence):
+ import asciitree
+ from collections import defaultdict
+ children = defaultdict(list)
+ # since erased nodes may be missing, multiple tokens may have same
+ # index (CCprocessed), etc.
+ token_to_index = {}
+ roots = []
+ # for token in self:
+ for i in range(0, len(sentence.token)):
+ token = sentence.token[i]
+ print token.word, token.tag, token.label, token.head
+ children[token.head].append(token)
+ token_to_index[token.word+":"+token.label+":"+token.head] = i + 1
+ if token.head == -1:
+ roots.append(token)
+ 
+ assert roots, "Couldn't find root Token(s)"
+
+ if len(roots) > 1:
+ # multiple roots so we make a fake one to be their parent
+ root = Token(0, 'ROOT', 'ROOT-LEMMA', 'ROOT-CPOS', 'ROOT-POS',
+ None, None, 'ROOT-DEPREL', None, None, None)
+ token_to_index[root] = 0
+ children[0] = roots
+ else:
+ root = roots[0]
+
+ def child_func(token):
+ index = token_to_index[token.word+":"+token.label+":"+token.head]
+ return children[index]
+ if not str_func:
+ def str_func(token):
+ return ' %s [%s]' % (token.word, token.label)
+
+ return asciitree.draw_tree(root, child_func, str_func)
+
+def to_dict(sentence):
+ token_str = list()
+children = [[] for token in sentence.token]
+roots = []
+root = -1
+for i in range(0, len(sentence.token)):
+token = sentence.token[i]
+print "current token:", token
+token_str.append('%s %s %s @%d' %
+(token.word, token.tag, token.label, (i+1)))
+if token.head == -1:
+roots.append(i)
+root = i
+else:
+print "appending child:", i , token.word, " - to parent - ", token.head
+children[token.head].append(i)
+
+assert roots, "Couldnt find roots!!"
+
+if len(roots) > 1:
+# multiple roots so we make a fake one to be their parent
+# root = Token(0, 'ROOT', 'ROOT-LEMMA', 'ROOT-CPOS', 'ROOT-POS',
+#	None, None, 'ROOT-DEPREL', None, None, None)
+print ("========== FOUND > 1 ROOT ==========", roots)
+s = '%s %s %s @%d' %("","","",1)
+token_str.insert(0, s)
+children[0] = roots
+root = 0
+
+visited = []
+for i in range(len(children)):
+visited.append(0) 
+
+def _get_dict(i):
+d = collections.OrderedDict()
+for c in children[i]:
+print "CHILDREN:", c, token_str[c]
+if (visited[c] == 0):
+visited[c] = 1
+d[token_str[c]] = _get_dict(c)
+
+ # d[token_str[c]] = _get_dict(c)
+return d
+
+tree = collections.OrderedDict()
+print "ROOT:", root
+tree[token_str[root]] = _get_dict(root)
+return tree
+
 def pretty_print():
 _write_input(_read_output().strip())
 logging.set_verbosity(logging.INFO)
@@ -60,6 +146,9 @@ def pretty_print():
 while True:
 documents, finished = sess.run(src)
 logging.info('Read %d documents', len(documents))
+# for d in documents:
+#	sentence.ParseFromString(d)
+#	as_asciitree(sentence)
 for d in documents:
 sentence.ParseFromString(d)
 tr = asciitree.LeftAligned()