def convert_file(self, input_file, output_file):
text = ""
actual_word = ""
new_para = False
new_word = False
for line in input_file:
if ("</para>" in line):
new_para = True
elif ("<para>" in line and new_para):
output_file.write('\n\n')
new_para = False
elif ("<w " in line):
new_word = True
elif ("</w>" in line and new_word):
output_file.write(actual_word)
actual_word = ""
new_word = False
elif ("<token>" in line):
token = get_interstring(line, '>', '<') + " "
actual_word += token
elif ("<no_space_after>" in line):
value = get_interstring(line, '>', '<')
if (value == "1"):
actual_word = actual_word[:-1]
output_file.write(text)
def process_coref( self, first_line, actual_info, perspron): # -> Coreference_record
"""
reads coreferent ID and creates a new coreference record
"""
( actual_ID, actual_type ) = actual_info
coref_ID = ""
if ( "<coref_gram" in first_line ):
line = self.pdt_t_input.readline()
coref_ID = get_interstring( line, '>', '<')
elif ( "<coref_text" in first_line ):
self.pdt_t_input.readline() # <LM>
line = self.pdt_t_input.readline()
coref_ID = get_interstring( line, '>', '<')
#elif ( "<coref_special" in first_line ):
# pass # segments or exophorae - they don't refere to any word in file
coref_type = self.ID_type( coref_ID) # string "word", "dropped", "sentence", "other"
#actual_ID = actual_info[0]
#actual_dropped = actual_info[1]
if ( coref_type == "dropped" ): # replace a reference to dropped pronoun with a reference to its non-dropped supernode
coref_ID = self.get_dropped( coref_ID)
if ( actual_ID != None and coref_ID != None ):
record = Coreference_record( actual_type == "dropped", actual_ID, coref_type == "dropped", coref_ID, perspron)
return record
return None
def read_infos( self, first_line): # -> (string, Node_type)
"""
obtains information about a node from its first line
return a pair of id string (substring of the first line containing node id) and enum
"""
id_string = get_interstring( first_line, '"', '"') # gots substring between first two quotes
type = self.get_node_type( id_string) # recognizes node type
return ( id_string, type)
def get_paragraph_ID( self, id_string): # -> int
"""
gets the number of the actual paragraph from the id of the actual node
id_string ... of a sentence, e.g. t-lnd94103-052-p1s11
"""
last = id_string.split( '-')[-1]
id = get_interstring( last, 'p', 's')
return int( id)
def next_pdt_word(self):
pdt_line = self.pdt_w_input.readline()
while (not "</doc>" in pdt_line):
#print(self.para_ID, self.sent_ID )
#print(pdt_line)
if ("<para" in pdt_line):
self.para_ID += 1
self.sent_ID = 1
elif ("<w id" in pdt_line):
pdt_ID = get_interstring(pdt_line, '"', '"')
token_line = self.pdt_w_input.readline()
token = get_interstring(token_line, '>', '<')
return (pdt_ID, token)
elif (pdt_line == ""):
return ("", "")
pdt_line = self.pdt_w_input.readline()
return ("", "")
def process_coref( self, first_line, actual_info): # -> Coreference_record
"""
reads coreferent ID and creates a new coreference record
"""
( actual_ID, actual_type ) = actual_info # id_string and Node_type (word or dropped)
coref_ID = ""
if ( "<coref_gram" in first_line ):
line = self.pdt_t_input.readline()
coref_ID = get_interstring( line, '>', '<')
elif ( "<coref_text" in first_line ):
self.pdt_t_input.readline() # <LM>
line = self.pdt_t_input.readline()
coref_ID = get_interstring( line, '>', '<')
#elif ( "<coref_special" in first_line ):
# pass # segments or exophorae - they don't refere to any word in file
coref_type = self.get_node_type( coref_ID) #
if ( coref_type == Node_type.Dropped ): # replace a reference to dropped pronoun with a reference to its non-dropped supernode
coref_ID = self.get_dropped( coref_ID)
if ( actual_ID != None and coref_ID != None ): # creating a new coreference record
record = Coreference_record( actual_type == Node_type.Dropped, actual_ID, coref_type == Node_type.Dropped, coref_ID)
return record
def paragraph_ID( self, string):
last = string.split( '-')[-1]
para_string = get_interstring( last, 'p', 's')
return int( para_string)
def read_infos( self, string):
id_string = get_interstring( string, '"', '"')
type = self.ID_type( id_string)
return ( id_string, type)