def print_candidates(self, chain):
"""Prints a XML file (mwetoolkit-candidates.dtd) from a temporary
candidates file generated by the treat_sentence callback function.
Repeated candidates are not printed several times: instead, each base
form has a joint frequency of the candidate in the corpus. Since the
new version of the "count.py" script, this initial frequency is only
printed if you explicitely ask to do it through the -f option.
@param filename: The file name of the corpus from which we generate the
candidates.
"""
global print_cand_freq, print_source
verbose("Outputting candidates file...")
for ngram_basestring, info in self.all_entities.iteritems():
cand = self.candidate_factory.make()
cand.from_string(ngram_basestring)
for corpus_name, (surface_dict, total_freq) in info.iteritems():
if print_cand_freq:
freq = Frequency(corpus_name, total_freq)
cand.add_frequency(freq)
for occur_string in surface_dict.keys():
occur_form = Ngram(None, None)
occur_form.from_string(occur_string)
sources = surface_dict[occur_string]
freq_value = len(sources)
freq = Frequency(corpus_name, freq_value)
occur_form.add_frequency(freq)
if print_source:
occur_form.add_sources(sources)
cand.add_occur(occur_form)
chain.handle_candidate(cand, info)
def print_candidates(self, chain):
"""Prints a XML file (mwetoolkit-candidates.dtd) from a temporary
candidates file generated by the treat_sentence callback function.
Repeated candidates are not printed several times: instead, each base
form has a joint frequency of the candidate in the corpus. Since the
new version of the "count.py" script, this initial frequency is only
printed if you explicitely ask to do it through the -f option.
@param filename: The file name of the corpus from which we generate the
candidates.
"""
global print_cand_freq, print_source
verbose("Outputting candidates file...")
for ngram_basestring, info in self.all_entities.iteritems() :
cand = self.candidate_factory.make()
cand.from_string(ngram_basestring)
for corpus_name, (surface_dict, total_freq) in info.iteritems():
if print_cand_freq :
freq = Frequency( corpus_name, total_freq )
cand.add_frequency( freq )
for occur_string in surface_dict.keys() :
occur_form = Ngram( None, None )
occur_form.from_string(occur_string)
sources = surface_dict[occur_string]
freq_value = len(sources)
freq = Frequency( corpus_name, freq_value )
occur_form.add_frequency( freq )
if print_source:
occur_form.add_sources(sources)
cand.add_occur( occur_form )
chain.handle_candidate(cand, info)
def finish(self, info={}):
"""After we read all input, we can finally be sure about which lines
need to be printed. Those correspond exactly to the unique lines added
to the buffer.
"""
global entity_buffer
verbose("Output the unified ngrams...")
for uniq_counter, (entity, info) in enumerate(entity_buffer.values()):
#entity.id_number = uniq_counter
if isinstance(entity, Candidate):
# WARNING: This is sort of specific for the VERBS 2010 paper. This
# whole script should actually be redefined and documented. But for
# the moment it's useful and I have no time to be a good programmer
# -Carlos
freq_sum = {}
for freq in entity.freqs:
freq_entry = freq_sum.get(freq.name, 0)
freq_entry += int(freq.value)
freq_sum[freq.name] = freq_entry
entity.freqs.clear()
for (name, value) in freq_sum.items():
entity.add_frequency(Frequency(name, value))
elif isinstance(entity, Entry):
pass
elif isinstance(entity, Sentence):
pass
self.chain.handle(entity, info)
self.chain.finish(info)
def finish(self, info={}):
"""After we read all input, we can finally be sure about which lines
need to be printed. Those correspond exactly to the unique lines added
to the buffer.
"""
global entity_buffer
verbose( "Output the unified ngrams..." )
for uniq_counter, (entity, info) in enumerate(entity_buffer.values()):
#entity.id_number = uniq_counter
if isinstance( entity, Candidate ) :
# WARNING: This is sort of specific for the VERBS 2010 paper. This
# whole script should actually be redefined and documented. But for
# the moment it's useful and I have no time to be a good programmer
# -Carlos
freq_sum = {}
for freq in entity.freqs :
freq_entry = freq_sum.get( freq.name, 0 )
freq_entry += int( freq.value )
freq_sum[ freq.name ] = freq_entry
entity.freqs.clear()
for ( name, value ) in freq_sum.items() :
entity.add_frequency( Frequency( name, value ) )
elif isinstance( entity, Entry ) :
pass
elif isinstance( entity, Sentence ) :
pass
self.chain.handle(entity, info)
self.chain.finish(info)
def treat_options( opts, arg, n_arg, usage_string ) :
"""Callback function that handles the command line options of this script.
@param opts The options parsed by getopts. Ignored.
@param arg The argument list parsed by getopts.
@param n_arg The number of arguments expected for this script.
"""
global filetype_corpus_ext
global filetype_candidates_ext
global output_filetype_ext
global action_annotate
global action_filter
treat_options_simplest(opts, arg, n_arg, usage_string)
detector_class = ContiguousLemmaDetector
candidates_fnames = []
n_gaps = None
for (o, a) in opts:
if o in ("-c", "--candidates"):
candidates_fnames.append(a)
elif o in ("-d", "--detector"):
detector_class = detectors.get(a,None)
if detector_class is None :
error("Unkown detector name: "+a)
elif o in ("-S", "--source"):
detector_class = SourceDetector
elif o in ("-g", "--gaps"):
n_gaps = int(a)
elif o == "--corpus-from":
filetype_corpus_ext = a
elif o == "--candidates-from":
filetype_candidates_ext = a
elif o == "--to":
output_filetype_ext = a
elif o == "--filter":
action_annotate = False
action_filter = True
elif o == "--filter-and-annot":
action_filter = True
else:
raise Exception("Bad arg: " + o)
if not candidates_fnames:
error("No candidates file given!")
if detector_class == SourceDetector and n_gaps is not None:
error('Bad arguments: method "Source" with "--gaps"')
c = CandidatesHandler()
verbose("Reading MWE list from candidates file")
filetype.parse(candidates_fnames,
c, filetype_candidates_ext)
verbose("MWE list loaded in memory successfully")
global detector
detector = detector_class(c.info, n_gaps)
def after_file(self, fileobj, info={}):
global corpus_size_f
corpus_size_f = float(corpus_size)
verbose("Selecting ngrams through LocalMaxs...")
self.localmaxs()
verbose("Outputting candidates file...")
for ngram_key in selected_candidates:
if selected_candidates[ngram_key] and ngram_counts[ngram_key] >= min_frequency:
self.dump_ngram(ngram_key, None)
self.chain.after_file(fileobj, info)
def after_file(self, fileobj, info={}):
global corpus_size_f
corpus_size_f = float(corpus_size)
verbose("Selecting ngrams through LocalMaxs...")
self.localmaxs()
verbose("Outputting candidates file...")
for ngram_key in selected_candidates:
if selected_candidates[
ngram_key] and ngram_counts[ngram_key] >= min_frequency:
self.dump_ngram(ngram_key, None)
self.chain.after_file(fileobj, info)
def transform_format(rasp):
"""
Reads an input file and converts it into mwetoolkit corpus XML format,
printing the XML file to stdout.
@param rasp Is the input, file or piped.
"""
global morphg_folder
global work_path
if morphg_folder:
os.chdir(morphg_folder)
n_line = 0
l_empty = 2
first_line = True
phrase = {}
l = unicode(rasp.readline(), "utf-8")
#pdb.set_trace()
# THIS LOOP IS HORRIBLE. PLEASE REFACTOR ME, PLEEEEASE!!!! - CR 20140905
while l != "":
if l == "\n":
l_empty += 1
if l_empty == 1:
sorted_phrase = map(lambda x: x[1], sorted(phrase.items()))
write_entry(n_line, sorted_phrase)
phrase = {}
if n_line % 100 == 0:
verbose("Processing sentence number %d" % n_line)
n_line += 1
first_line = True
l = unicode(rasp.readline(), "utf-8")
continue
# too long sentences not parsed because -w word limit passed to parser
elif l.startswith("(X "):
while l != "\n":
l = unicode(rasp.readline(), "utf-8")
continue
if first_line:
if l_empty >= 1:
l_empty = 0
process_line(l, phrase)
#pdb.set_trace()
first_line = False
l = unicode(rasp.readline(), "utf-8") #ignore line
else:
l_empty = 0
first_line = True
else:
process_tree_branch(l, phrase)
l = unicode(rasp.readline(), "utf-8")
if l_empty != 1 and len(phrase) != 0: #save last entry
write_entry(n_line, map(lambda x: x[1], sorted(phrase.items())))
if morphg_folder:
os.chdir(work_path)
def treat_options(opts, arg, n_arg, usage_string):
"""Callback function that handles the command line options of this script.
@param opts The options parsed by getopts. Ignored.
@param arg The argument list parsed by getopts.
@param n_arg The number of arguments expected for this script.
"""
global filetype_corpus_ext
global filetype_candidates_ext
global output_filetype_ext
global action_annotate
global action_filter
treat_options_simplest(opts, arg, n_arg, usage_string)
detector_class = ContiguousLemmaDetector
candidates_fnames = []
n_gaps = None
for (o, a) in opts:
if o in ("-c", "--candidates"):
candidates_fnames.append(a)
elif o in ("-d", "--detector"):
detector_class = detectors.get(a, None)
if detector_class is None:
error("Unkown detector name: " + a)
elif o in ("-S", "--source"):
detector_class = SourceDetector
elif o in ("-g", "--gaps"):
n_gaps = int(a)
elif o == "--corpus-from":
filetype_corpus_ext = a
elif o == "--candidates-from":
filetype_candidates_ext = a
elif o == "--to":
output_filetype_ext = a
elif o == "--filter":
action_annotate = False
action_filter = True
elif o == "--filter-and-annot":
action_filter = True
else:
raise Exception("Bad arg: " + o)
if not candidates_fnames:
error("No candidates file given!")
if detector_class == SourceDetector and n_gaps is not None:
error('Bad arguments: method "Source" with "--gaps"')
c = CandidatesHandler()
verbose("Reading MWE list from candidates file")
filetype.parse(candidates_fnames, c, filetype_candidates_ext)
verbose("MWE list loaded in memory successfully")
global detector
detector = detector_class(c.info, n_gaps)
def transform_format(rasp):
"""
Reads an input file and converts it into mwetoolkit corpus XML format,
printing the XML file to stdout.
@param rasp Is the input, file or piped.
"""
global morphg_folder
global work_path
if morphg_folder :
os.chdir( morphg_folder )
n_line=0
l_empty=2
first_line=True
phrase = {}
l=unicode(rasp.readline(),"utf-8")
#pdb.set_trace()
# THIS LOOP IS HORRIBLE. PLEASE REFACTOR ME, PLEEEEASE!!!! - CR 20140905
while l != "":
if l=="\n":
l_empty+=1
if l_empty == 1:
sorted_phrase = map( lambda x: x[1], sorted( phrase.items() ) )
write_entry(n_line,sorted_phrase)
phrase = {}
if n_line % 100 == 0 :
verbose( "Processing sentence number %d" % n_line )
n_line+=1
first_line=True
l=unicode( rasp.readline(), "utf-8" )
continue
# too long sentences not parsed because -w word limit passed to parser
elif l.startswith( "(X " ) :
while l != "\n" :
l = unicode( rasp.readline(), "utf-8" )
continue
if first_line:
if l_empty>=1:
l_empty=0
process_line(l,phrase)
#pdb.set_trace()
first_line=False
l=unicode( rasp.readline(), "utf-8" ) #ignore line
else:
l_empty=0
first_line=True
else:
process_tree_branch(l,phrase)
l=unicode( rasp.readline(), "utf-8" )
if l_empty != 1 and len(phrase) != 0 : #save last entry
write_entry(n_line,map( lambda x: x[1], sorted( phrase.items() ) ))
if morphg_folder :
os.chdir( work_path )
def read_data( f_data ) :
"""
Reads the annotation data from a tab-separated file and generates a
matrix with Ni rows, one per item to annotate, and Nc columns, one per
rater. The content of the matrix are the categories from K assigned
by coder c to item i. However, each category is converted to an integer
unique ID from 0 to Nk-1, so that it is easier to sort the categories.
Also returns the total number of items, raters and categories.
@param f_data The input file from which the data is read
@return A tuple containing, in the first position, the matrix with one
subject per row, one rater per column, and the annotation category IDs
in the cells as integers. The second, third and fourth fields of the
tuple are the number of items Ni, of coders Nc and of categories Nk. The
fifth field is a list containing the names of the categories sorted by
their IDs (position 0 contains the name of category IDentified by 0, and
so on).
"""
global first_rater
global first_header
global separator
if first_header :
f_data.readline() # Ignores first row
annotations = []
all_categories = {}
Ni = 0
Nc = 0
for line in f_data.readlines() :
if len(line.strip()) > 0 : # Ignore blank lines
Ni = Ni + 1
annot_row = line.strip().split( separator )[first_rater:]
if Nc == 0 :
Nc = len( annot_row )
elif Nc != len( annot_row ) :
raise ValueError, "Row %d: the file must contain the same \
number of fields in all rows" % Ni
# improvement for cases where file was Space-Tab separated
clean_annot_row = [] # contains the annotation cleaned from spaces
for annotation in annot_row :
clean_annot = annotation.strip() # Remove spurious spaces
all_categories[ clean_annot ] = 1
clean_annot_row.append( clean_annot )
annotations.append( clean_annot_row )
(annotations,categ_names) = categories_to_ids( annotations, all_categories )
Nk = len( all_categories )
verbose( "\n%d items\n%d raters\n%d categories\n" % (Ni, Nc, Nk) )
return ( annotations, Ni, Nc, Nk, categ_names )
def read_data(f_data):
"""
Reads the annotation data from a tab-separated file and generates a
matrix with Ni rows, one per item to annotate, and Nc columns, one per
rater. The content of the matrix are the categories from K assigned
by coder c to item i. However, each category is converted to an integer
unique ID from 0 to Nk-1, so that it is easier to sort the categories.
Also returns the total number of items, raters and categories.
@param f_data The input file from which the data is read
@return A tuple containing, in the first position, the matrix with one
subject per row, one rater per column, and the annotation category IDs
in the cells as integers. The second, third and fourth fields of the
tuple are the number of items Ni, of coders Nc and of categories Nk. The
fifth field is a list containing the names of the categories sorted by
their IDs (position 0 contains the name of category IDentified by 0, and
so on).
"""
global first_rater
global first_header
global separator
if first_header:
f_data.readline() # Ignores first row
annotations = []
all_categories = {}
Ni = 0
Nc = 0
for line in f_data.readlines():
if len(line.strip()) > 0: # Ignore blank lines
Ni = Ni + 1
annot_row = line.strip().split(separator)[first_rater:]
if Nc == 0:
Nc = len(annot_row)
elif Nc != len(annot_row):
raise ValueError, "Row %d: the file must contain the same \
number of fields in all rows" % Ni
# improvement for cases where file was Space-Tab separated
clean_annot_row = [] # contains the annotation cleaned from spaces
for annotation in annot_row:
clean_annot = annotation.strip() # Remove spurious spaces
all_categories[clean_annot] = 1
clean_annot_row.append(clean_annot)
annotations.append(clean_annot_row)
(annotations, categ_names) = categories_to_ids(annotations, all_categories)
Nk = len(all_categories)
verbose("\n%d items\n%d raters\n%d categories\n" % (Ni, Nc, Nk))
return (annotations, Ni, Nc, Nk, categ_names)
def interpret_length(l, maxormin):
"""
Transform argument given to -a or -i options into integer + error checks.
@param l: A string passed as argument to -i or -a
@param maxormin: A string indicating whether this is "maximum" or "minimum"
@return: An integer corresponding to l
"""
try:
result = int(l)
if result < 0:
raise ValueError
verbose("%s length: %d" % (maxormin, result))
return result
except ValueError:
error("Argument of must be non-negative integer, got " + repr(l))
def interpret_length( l, maxormin ):
"""
Transform argument given to -a or -i options into integer + error checks.
@param l: A string passed as argument to -i or -a
@param maxormin: A string indicating whether this is "maximum" or "minimum"
@return: An integer corresponding to l
"""
try :
result = int( l )
if result < 0:
raise ValueError
verbose( "%s length: %d" % (maxormin, result) )
return result
except ValueError:
error("Argument of must be non-negative integer, got " + repr(l))
def open_index(prefix):
"""
Open the index files (valid index created by the `index.py` script).
@param prefix The string name of the index file.
"""
global freq_name, the_corpus_size
global index, suffix_array
assert prefix.endswith(".info")
prefix = prefix[:-len(".info")]
try:
verbose("Loading index files... this may take some time.")
index = Index(prefix)
index.load_metadata()
freq_name = re.sub(".*/", "", prefix)
#pdb.set_trace()
the_corpus_size = index.metadata["corpus_size"]
except IOError:
error("Error opening the index.\nTry again with another index filename")
except KeyError:
error("Error opening the index.\nTry again with another index filename")
def main():
"""
Main function.
"""
global corpus_size_f
if corpus_from_index:
index = Index(corpus_path)
index.load_main()
for sentence in index.iterate_sentences():
treat_sentence(sentence)
else:
input_file = open(corpus_path)
parser = xml.sax.make_parser()
parser.setContentHandler( CorpusXMLHandler( treat_sentence ) )
parser.parse( input_file )
input_file.close()
corpus_size_f = float(corpus_size)
localmaxs()
verbose("Outputting candidates file...")
print(XML_HEADER % { "category": "candidates", "ns": "" })
meta = Meta([CorpusSize("corpus", corpus_size)],
[MetaFeat("glue", "real")], [])
print(meta.to_xml().encode('utf-8'))
id = 0
for ngram in select:
if (len(ngram) >= min_ngram and len(ngram) <= max_ngram and
select[ngram] and ngram_counts[ngram] >= min_frequency):
dump_ngram(ngram, id)
id += 1
print(XML_FOOTER % { "category": "candidates" })
def open_index(prefix):
"""
Open the index files (valid index created by the `index.py` script).
@param prefix The string name of the index file.
"""
global freq_name, the_corpus_size
global index, suffix_array
assert prefix.endswith(".info")
prefix = prefix[:-len(".info")]
try:
verbose("Loading index files... this may take some time.")
index = Index(prefix)
index.load_metadata()
freq_name = re.sub(".*/", "", prefix)
#pdb.set_trace()
the_corpus_size = index.metadata["corpus_size"]
except IOError:
error(
"Error opening the index.\nTry again with another index filename")
except KeyError:
error(
"Error opening the index.\nTry again with another index filename")
def main():
"""
Main function.
"""
global corpus_size_f
if corpus_from_index:
index = Index(corpus_path)
index.load_main()
for sentence in index.iterate_sentences():
treat_sentence(sentence)
else:
input_file = open(corpus_path)
parser = xml.sax.make_parser()
parser.setContentHandler(CorpusXMLHandler(treat_sentence))
parser.parse(input_file)
input_file.close()
corpus_size_f = float(corpus_size)
localmaxs()
verbose("Outputting candidates file...")
print(XML_HEADER % {"category": "candidates", "ns": ""})
meta = Meta([CorpusSize("corpus", corpus_size)],
[MetaFeat("glue", "real")], [])
print(meta.to_xml().encode('utf-8'))
id = 0
for ngram in select:
if (len(ngram) >= min_ngram and len(ngram) <= max_ngram
and select[ngram] and ngram_counts[ngram] >= min_frequency):
dump_ngram(ngram, id)
id += 1
print(XML_FOOTER % {"category": "candidates"})
def main(corpus_paths):
"""
Main function.
"""
global use_shelve, ngram_counts, selected_candidates
# Dummy file initialization to avoid warnings in PyCharm
ngram_counts_tmpfile = selected_candidates_tmpfile = None
if use_shelve:
verbose("Making temporary file...")
(ngram_counts, ngram_counts_tmpfile) = make_shelve()
(selected_candidates, selected_candidates_tmpfile) = make_shelve()
verbose("Counting ngrams...")
filetype.parse(corpus_paths, NGramCounterHandler(), input_filetype_ext)
if use_shelve:
verbose("Removing temporary files...")
destroy_shelve(ngram_counts, ngram_counts_tmpfile)
destroy_shelve(selected_candidates, selected_candidates_tmpfile)
def main(corpus_paths):
"""
Main function.
"""
global use_shelve, ngram_counts, selected_candidates
# Dummy file initialization to avoid warnings in PyCharm
ngram_counts_tmpfile = selected_candidates_tmpfile = None
if use_shelve:
verbose("Making temporary file...")
(ngram_counts, ngram_counts_tmpfile) = make_shelve()
(selected_candidates, selected_candidates_tmpfile) = make_shelve()
verbose("Counting ngrams...")
filetype.parse(corpus_paths, NGramCounterHandler(), input_filetype_ext)
if use_shelve:
verbose("Removing temporary files...")
destroy_shelve(ngram_counts, ngram_counts_tmpfile)
destroy_shelve(selected_candidates, selected_candidates_tmpfile)
candidate.add_feat( Feature( "n", len( candidate ) ) )
case_classes = {}
#pdb.set_trace()
has_hyphen = False
for w in candidate :
case_class = w.get_case_class()
count_class = case_classes.get( case_class, 0 )
case_classes[ case_class ] = count_class + 1
has_hyphen = has_hyphen or "-" in w.lemma
argmax_case_class = max( zip( case_classes.values(),
case_classes.keys() ) )[ 1 ]
candidate.add_feat( Feature( "capitalized", argmax_case_class ) )
candidate.add_feat( Feature( "hyphen", str( has_hyphen ) ) )
self.chain.handle_candidate(candidate, info)
################################################################################
# MAIN SCRIPT
args = read_options( "", [], treat_options_simplest, 1, usage_string )
# Done in 2 passes, one to define the type of the feature and another to
# print the feature values for each candidate
verbose( "1st pass : recover all POS patterns for meta feature" )
# Will ignore meta information and simply recover all the possible patterns
filetype.parse(args, RecovererHandler())
# Second pass to print the metafeat header with all possible pattern values
verbose( "2nd pass : add the features" )
filetype.parse(args, FeatGeneratorHandler())
def treat_options( opts, arg, n_arg, usage_string ) :
"""
Callback function that handles the command line options of this script.
@param opts The options parsed by getopts.
@param arg The argument list parsed by getopts.
@param n_arg The number of arguments expected for this script.
@param usage_string The usage string printed if the arguments are wrong.
"""
global first_header
global first_rater
global calculate_pairwise
global calculate_confusion
global separator
global distances_matrix
global unknown
treat_options_simplest( opts, arg, n_arg, usage_string )
for ( o, a ) in opts:
if o in ("-r", "--raters") :
verbose( "First row in file ignored -> considered as rater labels")
first_header = True
if o in ("-i", "--items") :
verbose("First column in file ignored -> considered as item labels")
first_rater = 1
if o in ("-p", "--pairwise") :
verbose( "Computing pairwise coefficients" )
calculate_pairwise = True
if o in ("-u", "--unknown") :
verbose( "Unknown value - TODO: implement: " + a )
unknown = a
if o in ("-s", "--separator") :
verbose( "Field separator: " + a )
separator = a
if len( separator ) > 1 :
warn("Multi-char field separator!")
if o in ("-d", "--distance") :
verbose("Calculating weighted coefficients using distance file")
distances_matrix = read_distances( a )
if distances_matrix is None :
warn("Error in distance matrix! Weighted coefficients will use 1.0 as default distance")
if o in ("-c", "--confusion") :
verbose( "Calculating confusion matrices" )
calculate_confusion = True
def get_freq_web1t(surfaces, lemmas, pos):
"""
Gets the frequency (number of occurrences) of an ngram in Google's
Web 1T 5-gram Corpus.
"""
global build_entry, web1t_data_path
length = len(surfaces)
if length > 5:
warn("Cannot count the frequency of an n-gram, n>5!")
return 0
search_term = ' '.join(map(build_entry, surfaces, lemmas, pos))
# Find the file in which to look for the ngram.
if length == 1:
filename = web1t_data_path + "/1gms/vocab.gz"
else:
indexfile = web1t_data_path + "/%dgms/%dgm.idx" % (length, length)
filenames = [x.split("\t") for x in read_file(indexfile).split("\n")]
filename = None
for (name, first) in filenames:
# Assumes byte-value-based ordering!
if first > search_term:
break
else:
filename = name
if filename is None:
return 0
filename = "%s/%dgms/%s" % (web1t_data_path, length, filename)
verbose("WEB1T: Opening %s, looking for %s" % (filename, search_term))
# This has been absurdly slow in Python.
#file = gzip.open(filename, "rb")
#
#search_term += "\t"
#freq = 0
#
#for line in file:
# if line.startswith(search_term):
# freq = int(line.split("\t")[1])
# break
#
#print >>sys.stderr, "buenito: %d" % freq
#
#file.close()
file = subprocess.Popen(
["zgrep", "--", "^" + re.escape(search_term) + "\t", filename],
stdout=subprocess.PIPE).stdout
line = file.read()
file.close()
if line:
freq = int(line.split("\t")[1])
else:
freq = 0
verbose("freq =" + str(freq))
return freq
suffix_array = index.load("lemma+pos")
else: # Web search, entries are single surface or lemma forms
if surface_flag:
build_entry = lambda surface, lemma, pos: surface
else:
build_entry = lambda surface, lemma, pos: lemma
if len(mode) != 1:
error("Exactly one option -u, -w or -i, must be provided")
#elif text_input and web_freq is None:
# warn("-x option is recommended for web queries, not textual indices")
################################################################################
# MAIN SCRIPT
longopts = ["candidates-from=", "corpus-from=", "to=",
"yahoo", "google", "index=", "ignore-pos", "surface", "old",
"lower=", "upper=", "vars", "lang=", "no-joint", "bigrams",
"univ=", "web1t="]
args = read_options("ywi:gsoal:Jbu:T:", longopts,
treat_options, -1, usage_string)
try:
verbose("Counting ngrams in candidates file")
filetype.parse(args, CounterPrinter(), filetype_candidates_ext)
finally:
if web_freq:
web_freq.flush_cache() # VERY IMPORTANT!
def treat_options( opts, arg, n_arg, usage_string ) :
"""
Callback function that handles the command line options of this script.
@param opts The options parsed by getopts. Ignored.
@param arg The argument list parsed by getopts.
@param n_arg The number of arguments expected for this script.
"""
global thresh_source
global thresh_value
global equals_name
global equals_value
global reverse
global minlength
global maxlength
global min_mweoccurs
global max_mweoccurs
global input_filetype_ext
global output_filetype_ext
treat_options_simplest( opts, arg, n_arg, usage_string )
for ( o, a ) in opts:
if o in ( "-t", "--threshold" ) :
threshold = interpret_threshold( a )
if threshold :
(thresh_source, thresh_value) = threshold
else :
error( "The format of the -t argument must be <source>:"
"<value>\n<source> must be a valid corpus name and "
"<value> must be a non-negative integer")
elif o in ( "-e", "--equals" ) :
equals = interpret_equals( a )
if equals :
( equals_name, equals_value ) = equals
else :
error( "The format of the -e argument must be <name>:"
"<value>\n<name> must be a valid feat name and "
"<value> must be a non-empty string")
elif o in ("-p", "--patterns") :
verbose( "Reading patterns file" )
global patterns
patterns = filetype.parse_entities([a])
elif o in ("-r", "--reverse") :
reverse = True
verbose("Option REVERSE active")
elif o in ("-i", "--minlength") :
minlength = interpret_length( a, "minimum" )
elif o in ("-a", "--maxlength") :
maxlength = interpret_length( a, "maximum" )
elif o == "--min-mweoccurs":
min_mweoccurs = interpret_length(a, "minimum")
elif o == "--max-mweoccurs":
max_mweoccurs = interpret_length(a, "maximum")
elif o == "--from":
input_filetype_ext = a
elif o == "--to":
output_filetype_ext = a
else:
raise Exception("Bad arg: " + o)
if minlength > maxlength:
warn("minlength should be <= maxlength")
if min_mweoccurs > max_mweoccurs:
warn("min-mweoccurs should be <= max-mweoccurs")
treat_options_simplest( opts, arg, n_arg, usage_string )
for ( o, a ) in opts:
if o == "--from":
input_filetype_ext = a
elif o == "--to":
output_filetype_ext = a
elif o in ("-l","--lemmas" ) :
lower_attr = "lemma"
elif o in ("-a", "--algorithm"):
algoname = a.lower()
elif o in ("-m", "-x"):
error( "Deprecated options -x and -m. Run with -h for details" )
else:
raise Exception("Bad arg: " + o)
################################################################################
# MAIN SCRIPT
longopts = [ "from=", "to=", "algorithm=", "lemmas" ]
args = read_options( "a:xml", longopts, treat_options, 1, usage_string )
if algoname != "simple" :
verbose( "Pass 1: Reading vocabulary from file... please wait" )
filetype.parse(args, VocabReaderHandler(), input_filetype_ext)
verbose( "Pass 2: Lowercasing the words in the file" )
filetype.parse(args, LowercaserHandler(), input_filetype_ext)
def treat_options(opts, arg, n_arg, usage_string):
"""
Callback function that handles the command line options of this script.
@param opts The options parsed by getopts. Ignored.
@param arg The argument list parsed by getopts.
@param n_arg The number of arguments expected for this script.
"""
global thresh_source
global thresh_value
global equals_name
global equals_value
global reverse
global minlength
global maxlength
global min_mweoccurs
global max_mweoccurs
global input_filetype_ext
global output_filetype_ext
treat_options_simplest(opts, arg, n_arg, usage_string)
for (o, a) in opts:
if o in ("-t", "--threshold"):
threshold = interpret_threshold(a)
if threshold:
(thresh_source, thresh_value) = threshold
else:
error("The format of the -t argument must be <source>:"
"<value>\n<source> must be a valid corpus name and "
"<value> must be a non-negative integer")
elif o in ("-e", "--equals"):
equals = interpret_equals(a)
if equals:
(equals_name, equals_value) = equals
else:
error("The format of the -e argument must be <name>:"
"<value>\n<name> must be a valid feat name and "
"<value> must be a non-empty string")
elif o in ("-p", "--patterns"):
verbose("Reading patterns file")
global patterns
patterns = filetype.parse_entities([a])
elif o in ("-r", "--reverse"):
reverse = True
verbose("Option REVERSE active")
elif o in ("-i", "--minlength"):
minlength = interpret_length(a, "minimum")
elif o in ("-a", "--maxlength"):
maxlength = interpret_length(a, "maximum")
elif o == "--min-mweoccurs":
min_mweoccurs = interpret_length(a, "minimum")
elif o == "--max-mweoccurs":
max_mweoccurs = interpret_length(a, "maximum")
elif o == "--from":
input_filetype_ext = a
elif o == "--to":
output_filetype_ext = a
else:
raise Exception("Bad arg: " + o)
if minlength > maxlength:
warn("minlength should be <= maxlength")
if min_mweoccurs > max_mweoccurs:
warn("min-mweoccurs should be <= max-mweoccurs")
for ( o, a ) in opts:
if o in ( "-m", "--measures" ) :
try :
measures = []
measures = interpret_measures( a )
except ValueError as message :
error( str(message)+"\nargument must be list separated by "
"\":\" and containing the names: "+
str( supported_measures ))
elif o in ( "-o", "--original" ) :
main_freq_name = a
elif o in ( "-a", "--all" ) :
join_all_contrastive = True
if not main_freq_name :
error( "Option -o is mandatory")
################################################################################
# MAIN SCRIPT
longopts = ["measures=", "original=", "all"]
args = read_options( "m:o:a", longopts, treat_options, 1, usage_string )
for a in args :
verbose( "Pass 1 for " + a )
filetype.parse([a], TotalCalculatorHandler())
# First calculate Nc for each contrastive corpus
verbose( "Pass 2 for " + a )
filetype.parse([a], MeasureCalculatorHandler())
def treat_options(opts, arg, n_arg, usage_string):
"""
Callback function that handles the command line options of this script.
@param opts The options parsed by getopts.
@param arg The argument list parsed by getopts.
@param n_arg The number of arguments expected for this script.
@param usage_string The usage string printed if the arguments are wrong.
"""
global first_header
global first_rater
global calculate_pairwise
global calculate_confusion
global separator
global distances_matrix
global unknown
treat_options_simplest(opts, arg, n_arg, usage_string)
for (o, a) in opts:
if o in ("-r", "--raters"):
verbose("First row in file ignored -> considered as rater labels")
first_header = True
if o in ("-i", "--items"):
verbose(
"First column in file ignored -> considered as item labels")
first_rater = 1
if o in ("-p", "--pairwise"):
verbose("Computing pairwise coefficients")
calculate_pairwise = True
if o in ("-u", "--unknown"):
verbose("Unknown value - TODO: implement: " + a)
unknown = a
if o in ("-s", "--separator"):
verbose("Field separator: " + a)
separator = a
if len(separator) > 1:
warn("Multi-char field separator!")
if o in ("-d", "--distance"):
verbose("Calculating weighted coefficients using distance file")
distances_matrix = read_distances(a)
if distances_matrix is None:
warn(
"Error in distance matrix! Weighted coefficients will use 1.0 as default distance"
)
if o in ("-c", "--confusion"):
verbose("Calculating confusion matrices")
calculate_confusion = True
if o in ("-m", "--measures"):
try:
measures = []
measures = interpret_measures(a)
except ValueError as message:
error(
str(message) + "\nargument must be list separated by "
"\":\" and containing the names: " +
str(supported_measures))
elif o in ("-o", "--original"):
main_freq_name = a
elif o in ("-a", "--all"):
join_all_contrastive = True
if not main_freq_name:
error("Option -o is mandatory")
################################################################################
# MAIN SCRIPT
longopts = ["measures=", "original=", "all"]
args = read_options("m:o:a", longopts, treat_options, 1, usage_string)
for a in args:
verbose("Pass 1 for " + a)
filetype.parse([a], TotalCalculatorHandler())
# First calculate Nc for each contrastive corpus
verbose("Pass 2 for " + a)
filetype.parse([a], MeasureCalculatorHandler())
def before_file(self, fileobj, info={}):
if not self.chain:
self.chain = self.make_printer(info, output_filetype_ext)
self.chain.before_file(fileobj, info)
verbose("Annotating corpus with MWEs found in list")
def get_freq_web1t(surfaces, lemmas, pos):
"""
Gets the frequency (number of occurrences) of an ngram in Google's
Web 1T 5-gram Corpus.
"""
global build_entry, web1t_data_path
length = len(surfaces)
if length > 5:
warn("Cannot count the frequency of an n-gram, n>5!")
return 0
search_term = ' '.join(map(build_entry, surfaces, lemmas, pos))
# Find the file in which to look for the ngram.
if length == 1:
filename = web1t_data_path + "/1gms/vocab.gz"
else:
indexfile = web1t_data_path + "/%dgms/%dgm.idx" % (length, length)
filenames = [x.split("\t") for x in read_file(indexfile).split("\n")]
filename = None
for (name, first) in filenames:
# Assumes byte-value-based ordering!
if first > search_term:
break
else:
filename = name
if filename is None:
return 0
filename = "%s/%dgms/%s" % (web1t_data_path, length, filename)
verbose("WEB1T: Opening %s, looking for %s" % (filename, search_term))
# This has been absurdly slow in Python.
#file = gzip.open(filename, "rb")
#
#search_term += "\t"
#freq = 0
#
#for line in file:
# if line.startswith(search_term):
# freq = int(line.split("\t")[1])
# break
#
#print >>sys.stderr, "buenito: %d" % freq
#
#file.close()
file = subprocess.Popen(
["zgrep", "--", "^" + re.escape(search_term) + "\t", filename],
stdout=subprocess.PIPE).stdout
line = file.read()
file.close()
if line:
freq = int(line.split("\t")[1])
else:
freq = 0
verbose("freq =" + str(freq))
return freq
cache_out[ k1 ] = cache1[ k1 ]
# Update entries in cache_out if corresponding entry in cache_2 is newer
for k2 in cache2.keys() :
( freq2, date2 ) = cache2[ k2 ]
( freq_out, date_out ) = cache_out.get( k2, ( -1, None ) )
if date_out is None :
cache_out[ k2 ] = ( freq2, date2 )
elif date2 < date_out :
cache_out[ k2 ] = ( freq2, date2 )
################################################################################
# MAIN SCRIPT
longopts = []
arg = read_options( "", longopts, treat_options_simplest, 3, usage_string )
verbose( "Opening files and checking consistency" )
cache1_desc = open( arg[ 0 ], "r" )
cache2_desc = open( arg[ 1 ], "r" )
cache_out_desc = open( arg[ 2 ], "w" )
cache1 = cPickle.load( cache1_desc )
cache2 = cPickle.load( cache2_desc )
cache_out = {}
verbose( "Combining cache files..." )
combine_caches( cache1, cache2, cache_out )
verbose( "Writing new cache file..." )
cPickle.dump( cache_out, cache_out_desc )
verbose( "{c} had {n} entries".format(c=arg[ 0 ], n=len(cache1)) )
verbose( "{c} had {n} entries".format(c=arg[ 1 ], n=len(cache2)) )
verbose( "Result has {n} entries".format(n=len(cache_out)) )
# Update entries in cache_out if corresponding entry in cache_2 is newer
for k2 in cache2.keys():
(freq2, date2) = cache2[k2]
(freq_out, date_out) = cache_out.get(k2, (-1, None))
if date_out is None:
cache_out[k2] = (freq2, date2)
elif date2 < date_out:
cache_out[k2] = (freq2, date2)
################################################################################
# MAIN SCRIPT
longopts = []
arg = read_options("", longopts, treat_options_simplest, 3, usage_string)
verbose("Opening files and checking consistency")
cache1_desc = open(arg[0], "r")
cache2_desc = open(arg[1], "r")
cache_out_desc = open(arg[2], "w")
cache1 = cPickle.load(cache1_desc)
cache2 = cPickle.load(cache2_desc)
cache_out = {}
verbose("Combining cache files...")
combine_caches(cache1, cache2, cache_out)
verbose("Writing new cache file...")
cPickle.dump(cache_out, cache_out_desc)
verbose("{c} had {n} entries".format(c=arg[0], n=len(cache1)))
verbose("{c} had {n} entries".format(c=arg[1], n=len(cache2)))
verbose("Result has {n} entries".format(n=len(cache_out)))
else: # Web search, entries are single surface or lemma forms
if surface_flag:
build_entry = lambda surface, lemma, pos: surface
else:
build_entry = lambda surface, lemma, pos: lemma
if len(mode) != 1:
error("Exactly one option -u, -w or -i, must be provided")
#elif text_input and web_freq is None:
# warn("-x option is recommended for web queries, not textual indices")
################################################################################
# MAIN SCRIPT
longopts = [
"candidates-from=", "corpus-from=", "to=", "yahoo", "google", "index=",
"ignore-pos", "surface", "old", "lower=", "upper=", "vars", "lang=",
"no-joint", "bigrams", "univ=", "web1t="
]
args = read_options("ywi:gsoal:Jbu:T:", longopts, treat_options, -1,
usage_string)
try:
verbose("Counting ngrams in candidates file")
filetype.parse(args, CounterPrinter(), filetype_candidates_ext)
finally:
if web_freq:
web_freq.flush_cache() # VERY IMPORTANT!
if algoname == "simple" : # Redundant, kept for clarity
sent_handler = LowercaserHandler.handle_sentence_simple
elif algoname == "complex" :
sent_handler = LowercaserHandler.handle_sentence_complex
elif algoname == "aggressive" : # Redundant, kept for clarity
sent_handler = LowercaserHandler.handle_sentence_aggressive
else :
ctxinfo.error("Bad algorithm name `{name}`", name=algoname)
elif o == "-m":
ctxinfo.error("Deprecated option. Use --from=Moses instead" )
elif o == "-x":
ctxinfo.error("Deprecated option. " \
"Use --from=PlainCorpus instead")
else:
raise Exception("Bad arg: " + o)
################################################################################
# MAIN SCRIPT
longopts = [ "from=", "to=", "algorithm=", "lemmas" ]
args = util.read_options( "a:xml", longopts, treat_options, 1, usage_string )
if sent_handler != LowercaserHandler.handle_sentence_simple :
util.verbose( "Pass 1: Reading vocabulary from file... please wait" )
filetype.parse(args, VocabReaderHandler(), input_filetype_ext)
util.verbose( "Pass 2: Lowercasing the words in the file" )
filetype.parse(args, LowercaserHandler(), input_filetype_ext)
def _fallback_entity(self, entity, info={}) :
"""For each candidate, verifies whether its number of occurrences in a
given source corpus is superior or equal to the threshold. If no source
corpus was provided (thresh_source is None), then all the corpora will
be considered when verifying the threshold constraint. A candidate is
printed to stdout only if it occurrs thres_value times or more in the
corpus names thresh_source.
@param entity: The `Ngram` that is being read from the XML file.
"""
global thresh_source
global thresh_value
global equals_name
global equals_value
global reverse
global patterns
global maxlength
global minlength
print_it = True
ngram_to_print = entity
# Threshold test
if entity.freqs :
for freq in entity.freqs :
if thresh_source :
if ( thresh_source == freq.name or
thresh_source == freq.name + ".xml" ) and \
freq.value < thresh_value :
print_it = False
else :
if freq.value < thresh_value :
print_it = False
# Equality test
if print_it and equals_name :
print_it = False
for feat in entity.features :
if feat.name == equals_name and feat.value == equals_value :
print_it = True
# NOTE: Different patterns may match the same ngram, with different
# results, when the 'ignore' pattern attribute is involved. Currently,
# we are only printing the first such match.
if print_it and patterns :
print_it = False
words = entity
for pattern in patterns :
for (match_ngram, wordnums) in pattern.matches(words,
anchored_begin=True, anchored_end=True):
print_it = True
ngram_to_print = match_ngram
break
if print_it :
break
# Filter out too long or too short elements
lenentity = len(entity)
if lenentity < minlength or lenentity > maxlength :
print_it = False
verbose("Filtered out: %d tokens" % lenentity)
# Filter out sentences with too few/too many MWE candidates
if info["kind"] == "sentence":
n = len(entity.mweoccurs)
if not (min_mweoccurs <= n <= max_mweoccurs):
print_it = False
if reverse :
print_it = not print_it
if print_it :
self.chain.handle(ngram_to_print, info)
def before_file(self, fileobj, info={}):
if not self.chain:
self.chain = self.make_printer(info, output_filetype_ext)
self.chain.before_file(fileobj, info)
verbose("Annotating corpus with MWEs found in list")
def _fallback_entity(self, entity, info={}):
"""For each candidate, verifies whether its number of occurrences in a
given source corpus is superior or equal to the threshold. If no source
corpus was provided (thresh_source is None), then all the corpora will
be considered when verifying the threshold constraint. A candidate is
printed to stdout only if it occurrs thres_value times or more in the
corpus names thresh_source.
@param entity: The `Ngram` that is being read from the XML file.
"""
global thresh_source
global thresh_value
global equals_name
global equals_value
global reverse
global patterns
global maxlength
global minlength
print_it = True
ngram_to_print = entity
# Threshold test
if entity.freqs:
for freq in entity.freqs:
if thresh_source:
if ( thresh_source == freq.name or
thresh_source == freq.name + ".xml" ) and \
freq.value < thresh_value :
print_it = False
else:
if freq.value < thresh_value:
print_it = False
# Equality test
if print_it and equals_name:
print_it = False
for feat in entity.features:
if feat.name == equals_name and feat.value == equals_value:
print_it = True
# NOTE: Different patterns may match the same ngram, with different
# results, when the 'ignore' pattern attribute is involved. Currently,
# we are only printing the first such match.
if print_it and patterns:
print_it = False
words = entity
for pattern in patterns:
for (match_ngram,
wordnums) in pattern.matches(words,
anchored_begin=True,
anchored_end=True):
print_it = True
ngram_to_print = match_ngram
break
if print_it:
break
# Filter out too long or too short elements
lenentity = len(entity)
if lenentity < minlength or lenentity > maxlength:
print_it = False
verbose("Filtered out: %d tokens" % lenentity)
# Filter out sentences with too few/too many MWE candidates
if info["kind"] == "sentence":
n = len(entity.mweoccurs)
if not (min_mweoccurs <= n <= max_mweoccurs):
print_it = False
if reverse:
print_it = not print_it
if print_it:
self.chain.handle(ngram_to_print, info)
