def get_language(samples, is_html=False, hint_language: str = None) -> LocaleInfo:
"""Return the language of this document.
By default we use a heuristic based on Google's CLD2.
"""
try:
(success, _, ((_, lang, _, _), *_)) = detect_language(
utf8Bytes=samples, isPlainText=not is_html, hintLanguage=hint_language
)
except CLD2Error as e:
log.error(f"Failed to recognize document language", exc_info=True)
success = False
if success:
return LocaleInfo(lang)
return LocaleInfo("en")
def as_document_info(self) -> DocumentInfo:
kwargs = self.metadata.copy()
kwargs["uri"] = self.uri
kwargs["cover_image"] = self.cover_image
kwargs["language"] = LocaleInfo(kwargs["language"])
kwargs.setdefault("data", {}).update(database_id=self.get_id())
return DocumentInfo(**kwargs)
def get_recognition_languages(cls) -> t.List[LocaleInfo]:
langs = []
for lng in pytesseract.get_languages():
try:
langs.append(LocaleInfo.from_three_letter_code(lng))
except ValueError:
continue
return langs
def language(self):
if (epub_lang := self.epub_metadata.get("language")) is not None:
try:
return LocaleInfo(epub_lang)
except:
log.exception(
"Failed to parse epub language `{epub_lang}`", exc_info=True
)
def get_voices(self):
rv = []
for voice in self.synth.GetVoices():
rv.append(
VoiceInfo(
id=voice.Id,
name=voice.Name,
desc=voice.Description,
language=LocaleInfo(voice.Language),
data={"voice_obj": voice},
))
return rv
def populate_list(self, set_focus=True):
installed_languages = [(True, info) for info in sorted(
TesseractOcrEngine.get_recognition_languages(),
key=lambda l: l.english_name,
)]
if self.is_offline:
languages = installed_languages
else:
languages = []
added_locale_infos = set(l[1] for l in installed_languages)
for lang in self.online_languages:
loc_info = LocaleInfo.from_three_letter_code(lang)
if loc_info in added_locale_infos:
continue
else:
languages.append((False, loc_info))
added_locale_infos.add(loc_info)
column_defn = [
ColumnDefn(
# Translators: the title of a column in the Tesseract language list
_("Language"),
"left",
450,
lambda lang: lang[1].description,
),
ColumnDefn(
# Translators: the title of a column in the Tesseract language list
_("Installed"),
"center",
100,
lambda lang: _("Yes") if lang[0] else _("No"),
),
]
self.tesseractLanguageList.set_columns(column_defn)
if set_focus:
self.tesseractLanguageList.set_objects(languages, focus_item=0)
else:
self.tesseractLanguageList.set_objects(languages, set_focus=False)
# Maintain the state of the list
if not any(languages):
if not self.is_offline:
self.addBestButton.Enable(False)
self.addFastButton.Enable(False)
else:
self.removeButton.Enable(False)
self.btnPanel.Enable(False)
def get_voices(self, language=None):
rv = []
for voice in self.synth.GetInstalledVoices():
if not voice.Enabled:
continue
info = voice.VoiceInfo
if (voice_culture := info.Culture) is not None:
voice_language = LocaleInfo(voice_culture.IetfLanguageTag)
else:
log.exception(
f"Failed to obtain culture information for voice {info.Name}"
)
continue
rv.append(
VoiceInfo(
id=info.Id,
name=info.Name,
desc=info.Description,
language=voice_language,
gender=info.Gender,
age=info.Age,
))
def language(self):
if lang := self.presentation.core_properties.language:
try:
return LocaleInfo(lang)
except ValueError:
pass
def get_recognition_languages(cls) -> t.List[LocaleInfo]:
return [
LocaleInfo(lang)
for lang in Win10DocrEngine.get_supported_languages()
]
def get_user_locale():
return LocaleInfo(CultureInfo.CurrentUICulture.IetfLanguageTag)
def get_user_locale():
return LocaleInfo.from_babel_locale(Locale.default())
author=author.removeprefix("By ").strip(),
description=desc,
publication_year=publish_date,
publisher=info.get("publisher", ""),
)
@cached_property
def language(self):
if (epub_lang := self.epub_metadata.get("language")) is not None:
try:
return LocaleInfo(epub_lang)
except:
log.exception(
"Failed to parse epub language `{epub_lang}`", exc_info=True
)
return self.get_language(self.html_content, is_html=True) or LocaleInfo("en")
def get_content(self):
return self.structure.get_text()
def get_document_semantic_structure(self):
return self.structure.semantic_elements
def get_document_style_info(self):
return self.structure.styled_elements
def resolve_link(self, link_range) -> LinkTarget:
href = urllib_parse.unquote(self.structure.link_targets[link_range])
if is_external_url(href):
return LinkTarget(url=href, is_external=True)
else: