def create_patterns_file( ngram_range ) :
"""
Create an artificial list of MWE patterns in which all the parts of
the words are wildcards. Such artificial patterns match every ngram
of size n, which is exactly what we want to do with the option -n. This
may seem a weird way to extract ngrams, but it allows a single
transparent candidate extraction function, treat_sentence.
@param ngram_range String argument of the -n option.
"""
global patterns, usage_string, shortest_pattern, longest_pattern
result = interpret_ngram( ngram_range )
if result :
( shortest_pattern, longest_pattern ) = result
patterns.append(build_generic_pattern(shortest_pattern, longest_pattern))
else :
error("Invalid argument for -n.")
def create_patterns_file(ngram_range):
"""
Create an artificial list of MWE patterns in which all the parts of
the words are wildcards. Such artificial patterns match every ngram
of size n, which is exactly what we want to do with the option -n. This
may seem a weird way to extract ngrams, but it allows a single
transparent candidate extraction function, treat_sentence.
@param ngram_range String argument of the -n option.
"""
global patterns, usage_string, shortest_pattern, longest_pattern
result = interpret_ngram(ngram_range)
if result:
(shortest_pattern, longest_pattern) = result
patterns.append(
build_generic_pattern(shortest_pattern, longest_pattern))
else:
error("Invalid argument for -n.")
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 surface_instead_lemmas
global glue
global base_attr
global min_ngram
global max_ngram
global min_frequency
global ngram_counts
global selected_candidates
global use_shelve
global input_filetype_ext
treat_options_simplest( opts, arg, n_arg, usage_string )
mode = []
for ( o, a ) in opts:
if o in ("-s", "--surface") :
surface_instead_lemmas = True
base_attr = 'surface'
elif o in ("-f", "--freq") :
min_frequency = int(a)
elif o in ("-n", "--ngram") :
(min_ngram, max_ngram) = interpret_ngram(a)
elif o in ("-G", "--glue"):
if a == "scp":
glue = scp_glue
else:
error("Unknown glue function '%s'" % a)
elif o in ("-S", "--shelve"):
use_shelve = True
elif o == "--from":
input_filetype_ext = a
else:
raise Exception("Bad arg: " + o)
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 surface_instead_lemmas
global glue
global base_attr
global min_ngram
global max_ngram
global min_frequency
global ngram_counts
global selected_candidates
global use_shelve
global input_filetype_ext
treat_options_simplest(opts, arg, n_arg, usage_string)
mode = []
for (o, a) in opts:
if o in ("-s", "--surface"):
surface_instead_lemmas = True
base_attr = 'surface'
elif o in ("-f", "--freq"):
min_frequency = int(a)
elif o in ("-n", "--ngram"):
(min_ngram, max_ngram) = interpret_ngram(a)
elif o in ("-G", "--glue"):
if a == "scp":
glue = scp_glue
else:
error("Unknown glue function '%s'" % a)
elif o in ("-S", "--shelve"):
use_shelve = True
elif o == "--from":
input_filetype_ext = a
else:
raise Exception("Bad arg: " + o)
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 surface_instead_lemmas
global glue
global corpus_from_index
global base_attr
global min_ngram
global max_ngram
global min_frequency
treat_options_simplest( opts, arg, n_arg, usage_string )
mode = []
for ( o, a ) in opts:
if o in ("-s", "--surface") :
surface_instead_lemmas = True
base_attr = 'surface'
elif o in ("-f", "--freq") :
min_frequency = int(a)
elif o in ("-n", "--ngram") :
(min_ngram, max_ngram) = interpret_ngram(a)
elif o in ("-i", "--index") :
corpus_from_index = True
elif o in ("-G", "--glue"):
if a == "scp":
glue = scp_glue
else:
error("Unknown glue function '%s'" % a)
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 surface_instead_lemmas
global glue
global corpus_from_index
global base_attr
global min_ngram
global max_ngram
global min_frequency
treat_options_simplest(opts, arg, n_arg, usage_string)
mode = []
for (o, a) in opts:
if o in ("-s", "--surface"):
surface_instead_lemmas = True
base_attr = 'surface'
elif o in ("-f", "--freq"):
min_frequency = int(a)
elif o in ("-n", "--ngram"):
(min_ngram, max_ngram) = interpret_ngram(a)
elif o in ("-i", "--index"):
corpus_from_index = True
elif o in ("-G", "--glue"):
if a == "scp":
glue = scp_glue
else:
error("Unknown glue function '%s'" % a)
