def clean_text(self, address, **kwargs):
"""Basic clean-up."""
address = self.LINE_BREAKS.sub(', ', address)
address = self.COMMATA.sub(', ', address)
address = collapse_spaces(address)
if len(address):
return address
def normalize(text,
lowercase=True,
collapse=True,
latinize=False,
ascii=False,
decompose=False,
replace_categories=UNICODE_CATEGORIES):
"""The main normalization function for text.
This will take a string and apply a set of transformations to it so
that it can be processed more easily afterwards. Arguments:
* ``lowercase``: not very mysterious.
* ``collapse``: replace multiple whitespace-like characters with a
single whitespace. This is especially useful with category replacement
which can lead to a lot of whitespace.
* ``decompose``: apply a unicode normalization (NFKD) to separate
simple characters and their diacritics.
* ``replace_categories``: This will perform a replacement of whole
classes of unicode characters (e.g. symbols, marks, numbers) with a
given character. It is used to replace any non-text elements of the
input string.
"""
if not isinstance(text, six.string_types):
return
# TODO: Python 3?
if six.PY2 and not isinstance(text, six.text_type):
encoding = guess_encoding(text, 'utf-8')
text = text.decode(encoding)
if lowercase:
# Yeah I made a Python package for this.
text = text.lower()
if decompose:
text = decompose_nfkd(text)
if ascii:
# A stricter form of transliteration that leaves only ASCII
# characters.
text = ascii_text(text)
elif latinize:
# Perform unicode-based transliteration, e.g. of cyricllic
# or CJK scripts into latin.
text = latinize_text(text)
# Perform unicode category-based character replacement. This is
# used to filter out whole classes of characters, such as symbols,
# punctuation, or whitespace-like characters.
text = category_replace(text, replace_categories)
if collapse:
# Remove consecutive whitespace.
text = collapse_spaces(text)
return text
def clean_text(
self,
text: str,
fuzzy: bool = False,
format: Optional[str] = None,
proxy: Optional["EntityProxy"] = None,
) -> Optional[str]:
"""Basic clean-up."""
name = strip_quotes(text)
return collapse_spaces(name)
def normalize(text: Any,
lowercase: bool = True,
collapse: bool = True,
latinize: bool = False,
ascii: bool = False,
encoding_default: Encoding = DEFAULT_ENCODING,
encoding: Optional[str] = None,
replace_categories: Categories = UNICODE_CATEGORIES):
"""The main normalization function for text.
This will take a string and apply a set of transformations to it so
that it can be processed more easily afterwards. Arguments:
* ``lowercase``: not very mysterious.
* ``collapse``: replace multiple whitespace-like characters with a
single whitespace. This is especially useful with category replacement
which can lead to a lot of whitespace.
* ``decompose``: apply a unicode normalization (NFKD) to separate
simple characters and their diacritics.
* ``replace_categories``: This will perform a replacement of whole
classes of unicode characters (e.g. symbols, marks, numbers) with a
given character. It is used to replace any non-text elements of the
input string.
"""
text = stringify(text,
encoding_default=encoding_default,
encoding=encoding)
if text is None:
return
if lowercase:
# Yeah I made a Python package for this.
text = text.lower()
if ascii:
# A stricter form of transliteration that leaves only ASCII
# characters.
text = ascii_text(text)
elif latinize:
# Perform unicode-based transliteration, e.g. of cyricllic
# or CJK scripts into latin.
text = latinize_text(text)
if text is None:
return
# Perform unicode category-based character replacement. This is
# used to filter out whole classes of characters, such as symbols,
# punctuation, or whitespace-like characters.
text = category_replace(text, replace_categories)
if collapse:
# Remove consecutive whitespace.
text = collapse_spaces(text)
return text
def _safe_name(file_name: Optional[str], sep: str) -> Optional[str]:
"""Convert the file name to ASCII and normalize the string."""
file_name = stringify(file_name)
if file_name is None:
return None
file_name = ascii_text(file_name)
file_name = category_replace(file_name, UNICODE_CATEGORIES)
file_name = collapse_spaces(file_name)
if file_name is None or not len(file_name):
return None
return file_name.replace(WS, sep)
def clean_text(
self,
text: str,
fuzzy: bool = False,
format: Optional[str] = None,
proxy: Optional["EntityProxy"] = None,
) -> Optional[str]:
"""Basic clean-up."""
address = self.LINE_BREAKS.sub(", ", text)
address = self.COMMATA.sub(", ", address)
collapsed = collapse_spaces(address)
if collapsed is None or not len(collapsed):
return None
return collapsed
def slugify(value: Any, sep: str = "-") -> Optional[str]:
"""A simple slug generator. Slugs are pure ASCII lowercase strings
that can be used in URLs an other places where a name has to be
machine-safe."""
text = stringify(value)
if text is None:
return None
text = text.replace(sep, WS)
# run this first because it'll give better results on special
# characters.
text = category_replace(text, SLUG_CATEGORIES)
text = latinize_text(text, ascii=True)
if text is None:
return None
text = text.lower()
text = "".join([c for c in text if c in VALID_CHARS])
text = collapse_spaces(text)
if text is None or len(text) == 0:
return None
return text.replace(WS, sep)
def clean_text(self, name, **kwargs):
"""Basic clean-up."""
name = strip_quotes(name)
name = collapse_spaces(name)
return name
