spaCy/spacy/orth.py
2014-09-10 18:27:44 +02:00

134 lines
3.2 KiB
Python

# -*- coding: utf8 -*-
from __future__ import unicode_literals
import unicodedata
import math
TAGS = 'adj adp adv conj det noun num pdt pos pron prt punct verb'.upper().split()
# Binary string features
def is_alpha(string, prob, case_stats, tag_stats):
return string.isalpha()
def is_digit(string, prob, case_stats, tag_stats):
return string.isdigit()
def is_punct(string, prob, case_stats, tag_stats):
for c in string:
if not unicodedata.category(c).startswith('P'):
return False
else:
return True
def is_space(string, prob, case_stats, tag_stats):
return string.isspace()
def is_ascii(string, prob, case_stats, tag_stats):
for c in string:
if ord(c) >= 128:
return False
else:
return True
def is_title(string, prob, case_stats, tag_stats):
return string.istitle()
def is_lower(string, prob, case_stats, tag_stats):
return string.islower()
def is_upper(string, prob, case_stats, tag_stats):
return string.isupper()
# Statistics features
def oft_case(name, thresh):
def wrapped(string, prob, case_stats, tag_stats):
return string
return wrapped
def can_tag(name, thresh=0.5):
def wrapped(string, prob, case_stats, tag_stats):
return string
return wrapped
# String features
def canon_case(string, prob, cluster, case_stats, tag_stats):
upper_pc = case_stats.get('upper', 0.0)
title_pc = case_stats.get('title', 0.0)
lower_pc = case_stats.get('lower', 0.0)
if upper_pc >= lower_pc and upper_pc >= title_pc:
return string.upper()
elif title_pc >= lower_pc:
return string.title()
else:
return string.lower()
def word_shape(string, *args):
length = len(string)
shape = ""
last = ""
shape_char = ""
seq = 0
for c in string:
if c.isalpha():
if c.isupper():
shape_char = "X"
else:
shape_char = "x"
elif c.isdigit():
shape_char = "d"
else:
shape_char = c
if shape_char == last:
seq += 1
else:
seq = 0
last = shape_char
if seq < 5:
shape += shape_char
return shape
def non_sparse(string, prob, cluster, case_stats, tag_stats):
if is_alpha(string, prob, case_stats, tag_stats):
return canon_case(string, prob, cluster, case_stats, tag_stats)
elif prob >= math.log(0.0001):
return string
else:
return word_shape(string, prob, cluster, case_stats, tag_stats)
def asciied(string, prob=0, cluster=0, case_stats=None, tag_stats=None):
'''"ASCIIfy" a Unicode string by stripping all umlauts, tildes, etc.'''
# Snippet from
# http://www.physic.ut.ee/~kkannike/english/prog/python/util/asciify/index.html
# TODO: Rewrite and improve this
lookup_table = {
u'': '"',
u'': '"'
}
temp = u''
for char in string:
if char in lookup_table:
temp += lookup_table[char]
else:
decomp = unicodedata.decomposition(char)
if decomp: # Not an empty string
temp += unichr(int(decomp.split()[0], 16))
else:
temp += char
return temp