def initialize():
chk_file = raw_input("What is the file to spellcheck? ")
field = raw_input("What FIELD do you want to spellcheck? ")
s_file = raw_input("What is name of final file? ")
checker = enchant.checker.SpellChecker("en_US")
cmdln = CmdLineChecker()
file_data = pd.read_csv(chk_file)
fields = list(file_data.apply(lambda x:'%s' % (x[field]),axis=1))
# maybe i don't even need this...
#fields = strip_html(fields)
corrected_text = []
for data_field in fields:
checker.set_text(str(data_field))
for err in checker:
print err.word
print err.suggest()
correct = raw_input("provide 0-index int of correct word or i to ignore, e to edit ")
if correct == 'i':
pass
elif correct == 'e':
suggest = raw_input("")
err.replace(suggest)
else:
correct = int(correct)
suggest = err.suggest()[correct]
err.replace(suggest)
corrected_text.append(checker.get_text())
saved_file = write_fixed_file(corrected_text, s_file)
def do_check(checker,to_check):
for text in to_check:
checker.set_text(text)
cmdline_checker = CmdLineChecker()
cmdline_checker.set_checker(checker)
cmdline_checker.run()
to_check[to_check.index(text)] = checker.get_text()
def spell_check_split_words(self, indices, doc, language, register=-1):
"""
Split misspelled words based on spell-checker suggestions.
Using this is usually not a good idea unless you have an insane
dictionary that contains all possible compound words in `language`.
Raise :exc:`enchant.Error` if dictionary instatiation fails.
"""
new_indices = []
new_texts = []
re_multispace = re.compile(r" +")
checker = self._get_enchant_checker(language)
indices = indices or self.get_all_indices()
for index in indices:
subtitle = self.subtitles[index]
text = subtitle.get_text(doc)
text = re_multispace.sub(" ", text)
checker.set_text(text)
while True:
try:
next(checker)
except StopIteration:
break
if checker.word.capitalize() == checker.word:
# Skip capitalized words, which are usually names
# and thus not always found in dictionaries.
continue
suggestions = []
for i, suggestion in enumerate(checker.suggest()):
if suggestion.find(" ") > 0:
if suggestion.replace(" ", "") == checker.word:
suggestions.append(suggestion)
# Split word only if only one two-word suggestion found that
# has all the same characters as the original unsplit word.
if len(suggestions) != 1: continue
text = checker.get_text()
a = checker.wordpos
z = checker.wordpos + len(checker.word)
checker.set_text(text[:a] + suggestions[0] + text[z:])
new_text = checker.get_text()
if new_text != text:
new_indices.append(index)
new_texts.append(new_text)
if not new_indices: return
self.replace_texts(new_indices, doc, new_texts, register=register)
description = _("Splitting words by spell-check suggestions")
self.set_action_description(register, description)
def spell_check_join_words(self, indices, doc, language, register=-1):
"""
Join misspelled words based on spell-checker suggestions.
Raise :exc:`enchant.Error` if dictionary instatiation fails.
"""
new_indices = []
new_texts = []
re_multispace = re.compile(r" +")
checker = self._get_enchant_checker(language)
seeker = self._get_enchant_checker(language)
for index in indices or self.get_all_indices():
subtitle = self.subtitles[index]
text = subtitle.get_text(doc)
text = re_multispace.sub(" ", text)
checker.set_text(text)
while True:
try:
next(checker)
except StopIteration:
break
text = checker.get_text()
a = checker.wordpos
z = checker.wordpos + len(checker.word)
ok_with_prev = ok_with_next = False
if checker.leading_context(1) == " ":
seeker.set_text(text[:a - 1] + text[a:])
poss = self._get_misspelled_indices(seeker)
ok_with_prev = not (a - 1) in poss
if checker.trailing_context(1) == " ":
seeker.set_text(text[:z] + text[z + 1:])
poss = self._get_misspelled_indices(seeker)
ok_with_next = not a in poss
# Join backwards or forwards if only one direction,
# but not both, produce a correctly spelled result.
if ok_with_prev and not ok_with_next:
checker.set_text(text[:a - 1] + text[a:])
if ok_with_next and not ok_with_prev:
checker.set_text(text[:z] + text[z + 1:])
new_text = checker.get_text()
if new_text != text:
new_indices.append(index)
new_texts.append(new_text)
if not new_indices: return
self.replace_texts(new_indices, doc, new_texts, register=register)
description = _("Joining words by spell-check suggestions")
self.set_action_description(register, description)
def spell_check_split_words(self, indices, doc, language, register=-1):
"""
Split misspelled words based on spell-checker suggestions.
Using this is usually not a good idea unless you have an insane
dictionary that contains all possible compound words in `language`.
Raise :exc:`enchant.Error` if dictionary instatiation fails.
"""
new_indices = []
new_texts = []
re_multispace = re.compile(r" +")
checker = self._get_enchant_checker(language)
indices = indices or self.get_all_indices()
for index in indices:
subtitle = self.subtitles[index]
text = subtitle.get_text(doc)
text = re_multispace.sub(" ", text)
checker.set_text(text)
while True:
try: next(checker)
except StopIteration: break
if checker.word.capitalize() == checker.word:
# Skip capitalized words, which are usually names
# and thus not always found in dictionaries.
continue
suggestions = []
for i, suggestion in enumerate(checker.suggest()):
if suggestion.find(" ") > 0:
if suggestion.replace(" ", "") == checker.word:
suggestions.append(suggestion)
# Split word only if only one two-word suggestion found that
# has all the same characters as the original unsplit word.
if len(suggestions) != 1: continue
text = checker.get_text()
a = checker.wordpos
z = checker.wordpos + len(checker.word)
checker.set_text(text[:a] + suggestions[0] + text[z:])
new_text = checker.get_text()
if new_text != text:
new_indices.append(index)
new_texts.append(new_text)
if not new_indices: return
self.replace_texts(new_indices, doc, new_texts, register=register)
description = _("Splitting words by spell-check suggestions")
self.set_action_description(register, description)
def spell_check_join_words(self, indices, doc, language, register=-1):
"""
Join misspelled words based on spell-checker suggestions.
Raise :exc:`enchant.Error` if dictionary instatiation fails.
"""
new_indices = []
new_texts = []
re_multispace = re.compile(r" +")
checker = self._get_enchant_checker(language)
seeker = self._get_enchant_checker(language)
for index in indices or self.get_all_indices():
subtitle = self.subtitles[index]
text = subtitle.get_text(doc)
text = re_multispace.sub(" ", text)
checker.set_text(text)
while True:
try: next(checker)
except StopIteration: break
text = checker.get_text()
a = checker.wordpos
z = checker.wordpos + len(checker.word)
ok_with_prev = ok_with_next = False
if checker.leading_context(1) == " ":
seeker.set_text(text[:a - 1] + text[a:])
poss = self._get_misspelled_indices(seeker)
ok_with_prev = not (a - 1) in poss
if checker.trailing_context(1) == " ":
seeker.set_text(text[:z] + text[z + 1:])
poss = self._get_misspelled_indices(seeker)
ok_with_next = not a in poss
# Join backwards or forwards if only one direction,
# but not both, produce a correctly spelled result.
if ok_with_prev and not ok_with_next:
checker.set_text(text[:a - 1] + text[a:])
if ok_with_next and not ok_with_prev:
checker.set_text(text[:z] + text[z + 1:])
new_text = checker.get_text()
if new_text != text:
new_indices.append(index)
new_texts.append(new_text)
if not new_indices: return
self.replace_texts(new_indices, doc, new_texts, register=register)
description = _("Joining words by spell-check suggestions")
self.set_action_description(register, description)
