mirror of
https://github.com/arsenetar/dupeguru.git
synced 2024-11-18 13:09:02 +00:00
Andrew Senetar
7ba8aa3514
- Format all files with black - Update tox.ini flake8 arguments to be compatible - Add black to requirements-extra.txt - Reduce ignored flake8 rules and fix a few violations
451 lines
13 KiB
Python
451 lines
13 KiB
Python
# Created By: Virgil Dupras
|
|
# Created On: 2011-01-11
|
|
# Copyright 2015 Hardcoded Software (http://www.hardcoded.net)
|
|
#
|
|
# This software is licensed under the "GPLv3" License as described in the "LICENSE" file,
|
|
# which should be included with this package. The terms are also available at
|
|
# http://www.gnu.org/licenses/gpl-3.0.html
|
|
|
|
import sys
|
|
import os
|
|
import os.path as op
|
|
import re
|
|
from math import ceil
|
|
import glob
|
|
import shutil
|
|
from datetime import timedelta
|
|
|
|
from .path import Path, pathify, log_io_error
|
|
|
|
|
|
def nonone(value, replace_value):
|
|
"""Returns ``value`` if ``value`` is not ``None``. Returns ``replace_value`` otherwise.
|
|
"""
|
|
if value is None:
|
|
return replace_value
|
|
else:
|
|
return value
|
|
|
|
|
|
def tryint(value, default=0):
|
|
"""Tries to convert ``value`` to in ``int`` and returns ``default`` if it fails.
|
|
"""
|
|
try:
|
|
return int(value)
|
|
except (TypeError, ValueError):
|
|
return default
|
|
|
|
|
|
def minmax(value, min_value, max_value):
|
|
"""Returns `value` or one of the min/max bounds if `value` is not between them.
|
|
"""
|
|
return min(max(value, min_value), max_value)
|
|
|
|
|
|
# --- Sequence related
|
|
|
|
|
|
def dedupe(iterable):
|
|
"""Returns a list of elements in ``iterable`` with all dupes removed.
|
|
|
|
The order of the elements is preserved.
|
|
"""
|
|
result = []
|
|
seen = {}
|
|
for item in iterable:
|
|
if item in seen:
|
|
continue
|
|
seen[item] = 1
|
|
result.append(item)
|
|
return result
|
|
|
|
|
|
def flatten(iterables, start_with=None):
|
|
"""Takes a list of lists ``iterables`` and returns a list containing elements of every list.
|
|
|
|
If ``start_with`` is not ``None``, the result will start with ``start_with`` items, exactly as
|
|
if ``start_with`` would be the first item of lists.
|
|
"""
|
|
result = []
|
|
if start_with:
|
|
result.extend(start_with)
|
|
for iterable in iterables:
|
|
result.extend(iterable)
|
|
return result
|
|
|
|
|
|
def first(iterable):
|
|
"""Returns the first item of ``iterable``.
|
|
"""
|
|
try:
|
|
return next(iter(iterable))
|
|
except StopIteration:
|
|
return None
|
|
|
|
|
|
def stripfalse(seq):
|
|
"""Returns a sequence with all false elements stripped out of seq.
|
|
"""
|
|
return [x for x in seq if x]
|
|
|
|
|
|
def extract(predicate, iterable):
|
|
"""Separates the wheat from the shaft (`predicate` defines what's the wheat), and returns both.
|
|
"""
|
|
wheat = []
|
|
shaft = []
|
|
for item in iterable:
|
|
if predicate(item):
|
|
wheat.append(item)
|
|
else:
|
|
shaft.append(item)
|
|
return wheat, shaft
|
|
|
|
|
|
def allsame(iterable):
|
|
"""Returns whether all elements of 'iterable' are the same.
|
|
"""
|
|
it = iter(iterable)
|
|
try:
|
|
first_item = next(it)
|
|
except StopIteration:
|
|
raise ValueError("iterable cannot be empty")
|
|
return all(element == first_item for element in it)
|
|
|
|
|
|
def trailiter(iterable, skipfirst=False):
|
|
"""Yields (prev_element, element), starting with (None, first_element).
|
|
|
|
If skipfirst is True, there will be no (None, item1) element and we'll start
|
|
directly with (item1, item2).
|
|
"""
|
|
it = iter(iterable)
|
|
if skipfirst:
|
|
try:
|
|
prev = next(it)
|
|
except StopIteration:
|
|
return
|
|
else:
|
|
prev = None
|
|
for item in it:
|
|
yield prev, item
|
|
prev = item
|
|
|
|
|
|
def iterconsume(seq, reverse=True):
|
|
"""Iterate over ``seq`` and pops yielded objects.
|
|
|
|
Because we use the ``pop()`` method, we reverse ``seq`` before proceeding. If you don't need
|
|
to do that, set ``reverse`` to ``False``.
|
|
|
|
This is useful in tight memory situation where you are looping over a sequence of objects that
|
|
are going to be discarded afterwards. If you're creating other objects during that iteration
|
|
you might want to use this to avoid ``MemoryError``.
|
|
"""
|
|
if reverse:
|
|
seq.reverse()
|
|
while seq:
|
|
yield seq.pop()
|
|
|
|
|
|
# --- String related
|
|
|
|
|
|
def escape(s, to_escape, escape_with="\\"):
|
|
"""Returns ``s`` with characters in ``to_escape`` all prepended with ``escape_with``.
|
|
"""
|
|
return "".join((escape_with + c if c in to_escape else c) for c in s)
|
|
|
|
|
|
def get_file_ext(filename):
|
|
"""Returns the lowercase extension part of filename, without the dot.
|
|
"""
|
|
pos = filename.rfind(".")
|
|
if pos > -1:
|
|
return filename[pos + 1 :].lower()
|
|
else:
|
|
return ""
|
|
|
|
|
|
def rem_file_ext(filename):
|
|
"""Returns the filename without extension.
|
|
"""
|
|
pos = filename.rfind(".")
|
|
if pos > -1:
|
|
return filename[:pos]
|
|
else:
|
|
return filename
|
|
|
|
|
|
def pluralize(number, word, decimals=0, plural_word=None):
|
|
"""Returns a pluralized string with ``number`` in front of ``word``.
|
|
|
|
Adds a 's' to s if ``number`` > 1.
|
|
``number``: The number to go in front of s
|
|
``word``: The word to go after number
|
|
``decimals``: The number of digits after the dot
|
|
``plural_word``: If the plural rule for word is more complex than adding a 's', specify a plural
|
|
"""
|
|
number = round(number, decimals)
|
|
format = "%%1.%df %%s" % decimals
|
|
if number > 1:
|
|
if plural_word is None:
|
|
word += "s"
|
|
else:
|
|
word = plural_word
|
|
return format % (number, word)
|
|
|
|
|
|
def format_time(seconds, with_hours=True):
|
|
"""Transforms seconds in a hh:mm:ss string.
|
|
|
|
If ``with_hours`` if false, the format is mm:ss.
|
|
"""
|
|
minus = seconds < 0
|
|
if minus:
|
|
seconds *= -1
|
|
m, s = divmod(seconds, 60)
|
|
if with_hours:
|
|
h, m = divmod(m, 60)
|
|
r = "%02d:%02d:%02d" % (h, m, s)
|
|
else:
|
|
r = "%02d:%02d" % (m, s)
|
|
if minus:
|
|
return "-" + r
|
|
else:
|
|
return r
|
|
|
|
|
|
def format_time_decimal(seconds):
|
|
"""Transforms seconds in a strings like '3.4 minutes'.
|
|
"""
|
|
minus = seconds < 0
|
|
if minus:
|
|
seconds *= -1
|
|
if seconds < 60:
|
|
r = pluralize(seconds, "second", 1)
|
|
elif seconds < 3600:
|
|
r = pluralize(seconds / 60.0, "minute", 1)
|
|
elif seconds < 86400:
|
|
r = pluralize(seconds / 3600.0, "hour", 1)
|
|
else:
|
|
r = pluralize(seconds / 86400.0, "day", 1)
|
|
if minus:
|
|
return "-" + r
|
|
else:
|
|
return r
|
|
|
|
|
|
SIZE_DESC = ("B", "KB", "MB", "GB", "TB", "PB", "EB", "ZB", "YB")
|
|
SIZE_VALS = tuple(1024 ** i for i in range(1, 9))
|
|
|
|
|
|
def format_size(size, decimal=0, forcepower=-1, showdesc=True):
|
|
"""Transform a byte count in a formatted string (KB, MB etc..).
|
|
|
|
``size`` is the number of bytes to format.
|
|
``decimal`` is the number digits after the dot.
|
|
``forcepower`` is the desired suffix. 0 is B, 1 is KB, 2 is MB etc.. if kept at -1, the suffix
|
|
will be automatically chosen (so the resulting number is always below 1024).
|
|
if ``showdesc`` is ``True``, the suffix will be shown after the number.
|
|
Usage example::
|
|
|
|
>>> format_size(1234, decimal=2, showdesc=True)
|
|
'1.21 KB'
|
|
"""
|
|
if forcepower < 0:
|
|
i = 0
|
|
while size >= SIZE_VALS[i]:
|
|
i += 1
|
|
else:
|
|
i = forcepower
|
|
if i > 0:
|
|
div = SIZE_VALS[i - 1]
|
|
else:
|
|
div = 1
|
|
format = "%%%d.%df" % (decimal, decimal)
|
|
negative = size < 0
|
|
divided_size = (0.0 + abs(size)) / div
|
|
if decimal == 0:
|
|
divided_size = ceil(divided_size)
|
|
else:
|
|
divided_size = ceil(divided_size * (10 ** decimal)) / (10 ** decimal)
|
|
if negative:
|
|
divided_size *= -1
|
|
result = format % divided_size
|
|
if showdesc:
|
|
result += " " + SIZE_DESC[i]
|
|
return result
|
|
|
|
|
|
_valid_xml_range = "\x09\x0A\x0D\x20-\uD7FF\uE000-\uFFFD"
|
|
if sys.maxunicode > 0x10000:
|
|
_valid_xml_range += "%s-%s" % (chr(0x10000), chr(min(sys.maxunicode, 0x10FFFF)))
|
|
RE_INVALID_XML_SUB = re.compile("[^%s]" % _valid_xml_range, re.U).sub
|
|
|
|
|
|
def remove_invalid_xml(s, replace_with=" "):
|
|
return RE_INVALID_XML_SUB(replace_with, s)
|
|
|
|
|
|
def multi_replace(s, replace_from, replace_to=""):
|
|
"""A function like str.replace() with multiple replacements.
|
|
|
|
``replace_from`` is a list of things you want to replace. Ex: ['a','bc','d']
|
|
``replace_to`` is a list of what you want to replace to.
|
|
If ``replace_to`` is a list and has the same length as ``replace_from``, ``replace_from``
|
|
items will be translated to corresponding ``replace_to``. A ``replace_to`` list must
|
|
have the same length as ``replace_from``
|
|
If ``replace_to`` is a string, all ``replace_from`` occurence will be replaced
|
|
by that string.
|
|
``replace_from`` can also be a str. If it is, every char in it will be translated
|
|
as if ``replace_from`` would be a list of chars. If ``replace_to`` is a str and has
|
|
the same length as ``replace_from``, it will be transformed into a list.
|
|
"""
|
|
if isinstance(replace_to, str) and (len(replace_from) != len(replace_to)):
|
|
replace_to = [replace_to for r in replace_from]
|
|
if len(replace_from) != len(replace_to):
|
|
raise ValueError("len(replace_from) must be equal to len(replace_to)")
|
|
replace = list(zip(replace_from, replace_to))
|
|
for r_from, r_to in [r for r in replace if r[0] in s]:
|
|
s = s.replace(r_from, r_to)
|
|
return s
|
|
|
|
|
|
# --- Date related
|
|
|
|
# It might seem like needless namespace pollution, but the speedup gained by this constant is
|
|
# significant, so it stays.
|
|
ONE_DAY = timedelta(1)
|
|
|
|
|
|
def iterdaterange(start, end):
|
|
"""Yields every day between ``start`` and ``end``.
|
|
"""
|
|
date = start
|
|
while date <= end:
|
|
yield date
|
|
date += ONE_DAY
|
|
|
|
|
|
# --- Files related
|
|
|
|
|
|
@pathify
|
|
def modified_after(first_path: Path, second_path: Path):
|
|
"""Returns ``True`` if first_path's mtime is higher than second_path's mtime.
|
|
|
|
If one of the files doesn't exist or is ``None``, it is considered "never modified".
|
|
"""
|
|
try:
|
|
first_mtime = first_path.stat().st_mtime
|
|
except (EnvironmentError, AttributeError):
|
|
return False
|
|
try:
|
|
second_mtime = second_path.stat().st_mtime
|
|
except (EnvironmentError, AttributeError):
|
|
return True
|
|
return first_mtime > second_mtime
|
|
|
|
|
|
def find_in_path(name, paths=None):
|
|
"""Search for `name` in all directories of `paths` and return the absolute path of the first
|
|
occurrence. If `paths` is None, $PATH is used.
|
|
"""
|
|
if paths is None:
|
|
paths = os.environ["PATH"]
|
|
if isinstance(paths, str): # if it's not a string, it's already a list
|
|
paths = paths.split(os.pathsep)
|
|
for path in paths:
|
|
if op.exists(op.join(path, name)):
|
|
return op.join(path, name)
|
|
return None
|
|
|
|
|
|
@log_io_error
|
|
@pathify
|
|
def delete_if_empty(path: Path, files_to_delete=[]):
|
|
"""Deletes the directory at 'path' if it is empty or if it only contains files_to_delete.
|
|
"""
|
|
if not path.exists() or not path.isdir():
|
|
return
|
|
contents = path.listdir()
|
|
if any(p for p in contents if (p.name not in files_to_delete) or p.isdir()):
|
|
return False
|
|
for p in contents:
|
|
p.remove()
|
|
path.rmdir()
|
|
return True
|
|
|
|
|
|
def open_if_filename(infile, mode="rb"):
|
|
"""If ``infile`` is a string, it opens and returns it. If it's already a file object, it simply returns it.
|
|
|
|
This function returns ``(file, should_close_flag)``. The should_close_flag is True is a file has
|
|
effectively been opened (if we already pass a file object, we assume that the responsibility for
|
|
closing the file has already been taken). Example usage::
|
|
|
|
fp, shouldclose = open_if_filename(infile)
|
|
dostuff()
|
|
if shouldclose:
|
|
fp.close()
|
|
"""
|
|
if isinstance(infile, Path):
|
|
return (infile.open(mode), True)
|
|
if isinstance(infile, str):
|
|
return (open(infile, mode), True)
|
|
else:
|
|
return (infile, False)
|
|
|
|
|
|
def ensure_folder(path):
|
|
"Create `path` as a folder if it doesn't exist."
|
|
if not op.exists(path):
|
|
os.makedirs(path)
|
|
|
|
|
|
def ensure_file(path):
|
|
"Create `path` as an empty file if it doesn't exist."
|
|
if not op.exists(path):
|
|
open(path, "w").close()
|
|
|
|
|
|
def delete_files_with_pattern(folder_path, pattern, recursive=True):
|
|
"""Delete all files (or folders) in `folder_path` that match the glob `pattern`.
|
|
"""
|
|
to_delete = glob.glob(op.join(folder_path, pattern))
|
|
for fn in to_delete:
|
|
if op.isdir(fn):
|
|
shutil.rmtree(fn)
|
|
else:
|
|
os.remove(fn)
|
|
if recursive:
|
|
subpaths = [op.join(folder_path, fn) for fn in os.listdir(folder_path)]
|
|
subfolders = [p for p in subpaths if op.isdir(p)]
|
|
for p in subfolders:
|
|
delete_files_with_pattern(p, pattern, True)
|
|
|
|
|
|
class FileOrPath:
|
|
"""Does the same as :func:`open_if_filename`, but it can be used with a ``with`` statement.
|
|
|
|
Example::
|
|
|
|
with FileOrPath(infile):
|
|
dostuff()
|
|
"""
|
|
|
|
def __init__(self, file_or_path, mode="rb"):
|
|
self.file_or_path = file_or_path
|
|
self.mode = mode
|
|
self.mustclose = False
|
|
self.fp = None
|
|
|
|
def __enter__(self):
|
|
self.fp, self.mustclose = open_if_filename(self.file_or_path, self.mode)
|
|
return self.fp
|
|
|
|
def __exit__(self, exc_type, exc_value, traceback):
|
|
if self.fp and self.mustclose:
|
|
self.fp.close()
|