Make PyICU optional

Fix more locale issues
Update README.md

README.md

@@ -33,21 +33,13 @@ For details, screen-shots, and more, visit [the Wiki](https://github.com/comicta
 
 ## Installation
 
+It is reccomended to either use the Binaries or a package manager for the best experience
+
 ### Binaries
 
-Windows and macOS binaries are provided in the [Releases Page](https://github.com/comictagger/comictagger/releases).
+Windows, Linux and MacOS binaries are provided in the [Releases Page](https://github.com/comictagger/comictagger/releases).
 
-Just unzip the archive in any folder and run, no additional installation steps are required.
-
-### PIP installation
-
-A pip package is provided, you can install it with:
-
-```
- $ pip3 install comictagger[GUI]
-```
-
-There are two optional dependencies GUI and CBR. You can install the optional dependencies by specifying one or more of `GUI`,`CBR` or `all` in braces e.g. `comictagger[CBR,GUI]`
+Unzip the archive in any folder and run, no additional installation steps are required.
 
 ### Chocolatey installation (Windows only)
 
@@ -55,12 +47,26 @@ A [Chocolatey package](https://community.chocolatey.org/packages/comictagger), m
 ```powershell
 choco install comictagger
 ```
+
+### PyPi installation
+
+! Please note that pages may not be sorted correctly if ICU is not installed on Linux and MacOS.
+If you have issues installing ICU please see the PyICU documentation [here](https://pyicu.org)
+
+A PyPi package is provided, you can install it with:
+
+```
+ $ pip3 install comictagger[GUI,ICU]
+```
+
+There are several optional dependencies GUI, CBR, 7Z and ICU. You can install the optional dependencies by specifying one or more of `GUI`,`CBR` or `all` in braces e.g. `comictagger[CBR,GUI,ICU]`
+
 ### From source
 
  1. Ensure you have python 3.9 installed
  2. Clone this repository `git clone https://github.com/comictagger/comictagger.git`
  3. `pip3 install -r requirements_dev.txt`
- 7. `pip3 install .` or `pip3 install .[GUI]`
+ 7. `pip3 install .` or `pip3 install .[GUI,ICU]`
 
 
 ## Contributors

comicapi/comicarchive.py

@@ -22,7 +22,6 @@ import shutil
 import sys
 from typing import cast
 
-import natsort
 import wordninja
 
 from comicapi import filenamelexer, filenameparser, utils
@@ -280,7 +279,7 @@ class ComicArchive:
 
             # seems like some archive creators are on Windows, and don't know about case-sensitivity!
             if sort_list:
-                files = cast(list[str], natsort.os_sorted(files))
+                files = cast(list[str], utils.os_sorted(files))
 
             # make a sub-list of image files
             self.page_list = []

comicapi/utils.py

@@ -15,25 +15,69 @@
 from __future__ import annotations
 
 import json
+import locale
 import logging
 import os
 import pathlib
 import unicodedata
 from collections import defaultdict
-from collections.abc import Mapping
+from collections.abc import Iterable, Mapping
 from shutil import which  # noqa: F401
 from typing import Any
 
+import natsort
 import pycountry
 import rapidfuzz.fuzz
 
 import comicapi.data
 
+try:
+    import icu
+
+    del icu
+    icu_available = True
+except ImportError:
+    icu_available = False
+
 logger = logging.getLogger(__name__)
 
 
-class UtilsVars:
-    already_fixed_encoding = False
+def _custom_key(tup):
+    lst = []
+    for x in natsort.os_sort_keygen()(tup):
+        ret = x
+        if len(x) > 1 and isinstance(x[1], int) and isinstance(x[0], str) and x[0] == "":
+            ret = (str(x[1]), *x[1:])
+
+        lst.append(ret)
+    return tuple(lst)
+
+
+def os_sorted(lst: Iterable) -> Iterable:
+    if icu_available:
+        raise Exception("fuck off")
+        return natsort.os_sorted(lst)
+    return sorted(lst, key=_custom_key)
+
+
+def save_locale() -> dict[int, tuple[str | None, str | None]]:
+    locales: dict[int, tuple[str | None, str | None]] = {
+        locale.LC_ALL: (None, None),
+        locale.LC_COLLATE: (None, None),
+        locale.LC_CTYPE: (None, None),
+        locale.LC_MESSAGES: (None, None),
+        locale.LC_MONETARY: (None, None),
+        locale.LC_NUMERIC: (None, None),
+        locale.LC_TIME: (None, None),
+    }
+    for x in locales:
+        locales[x] = locale.getlocale(x)
+    return locales
+
+
+def set_locale(locales: dict[int, tuple[str | None, str | None]]) -> None:
+    for x, value in locales.items():
+        locale.setlocale(x, value)
 
 
 def combine_notes(existing_notes: str | None, new_notes: str | None, split: str) -> str:

comictaggerlib/gui.py

@@ -9,14 +9,20 @@ import types
 
 import settngs
 
+from comicapi.utils import save_locale, set_locale
 from comictaggerlib.graphics import graphics_path
 from comictalker.comictalker import ComicTalker
 
 logger = logging.getLogger("comictagger")
+
 try:
-    qt_available = True
+    loc = save_locale()
     from PyQt5 import QtCore, QtGui, QtWidgets
 
+    qt_available = True
+
+    set_locale(loc)
+
     def show_exception_box(log_msg: str) -> None:
         """Checks if a QApplication instance is available and shows a messagebox with the exception message.
         If unavailable (non-console application), log an additional notice.

comictaggerlib/imagefetcher.py

@@ -25,11 +25,15 @@ import tempfile
 
 import requests
 
+from comicapi.utils import save_locale, set_locale
 from comictaggerlib import ctversion
 
 try:
+    loc = save_locale()
     from PyQt5 import QtCore, QtNetwork
 
+    set_locale(loc)
+
     qt_available = True
 except ImportError:
     qt_available = False

comictaggerlib/main.py

@@ -25,6 +25,7 @@ import settngs
 
 import comicapi
 import comictalker
+from comicapi.utils import save_locale, set_locale
 from comictaggerlib import cli, ctsettings
 from comictaggerlib.ctversion import version
 from comictaggerlib.log import setup_logging
@@ -36,9 +37,13 @@ else:
 
 logger = logging.getLogger("comictagger")
 
+
 try:
+    loc = save_locale()
     from comictaggerlib import gui
 
+    set_locale(loc)
+
     qt_available = gui.qt_available
 except Exception:
     logger.exception("Qt unavailable")

comictaggerlib/ui/qtutils.py

@@ -6,13 +6,17 @@ import io
 import logging
 import traceback
 
+from comicapi.utils import save_locale, set_locale
 from comictaggerlib.graphics import graphics_path
 
 logger = logging.getLogger(__name__)
 
 try:
+    loc = save_locale()
     from PyQt5 import QtGui, QtWidgets
 
+    set_locale(loc)
+
     qt_available = True
 except ImportError:
     qt_available = False

requirements-ICU.txt

@@ -0,0 +1 @@
+pyicu; sys_platform == 'linux' or sys_platform == 'darwin'

requirements.txt

@@ -5,7 +5,6 @@ natsort>=8.1.0
 pathvalidate
 pillow>=9.1.0, <10
 pycountry
-#pyicu; sys_platform == 'linux' or sys_platform == 'darwin'
 rapidfuzz>=2.12.0
 requests==2.*
 settngs==0.6.2

tests/comicarchive_test.py

@@ -12,7 +12,7 @@ from testing.filenames import datadir
 
 
 @pytest.mark.xfail(not comicapi.archivers.rar.rar_support, reason="rar support")
-def test_getPageNameList():
+def test_get_page_name_list():
     c = comicapi.comicarchive.ComicArchive(datadir / "fake_cbr.cbr")
     assert c.seems_to_be_a_comic_archive()
     pageNameList = c.get_page_name_list()

tests/utils_test.py

@@ -7,6 +7,46 @@ import pytest
 import comicapi.utils
 
 
+def test_os_sorted():
+    page_name_list = [
+        "cover.jpg",
+        "Page1.jpeg",
+        "!cover.jpg",
+        "page4.webp",
+        "test/!cover.tar.gz",
+        "!cover.tar.gz",
+        "00.jpg",
+        "ignored.txt",
+        "page0.jpg",
+        "test/00.tar.gz",
+        ".ignored.jpg",
+        "Page3.gif",
+        "!cover.tar.gz",
+        "Page2.png",
+        "page10.jpg",
+        "!cover",
+    ]
+
+    assert comicapi.utils.os_sorted(page_name_list) == [
+        "!cover",
+        "!cover.jpg",
+        "!cover.tar.gz",
+        "!cover.tar.gz",  # Depending on locale punctuation or numbers might come first (Linux, MacOS)
+        ".ignored.jpg",
+        "00.jpg",
+        "cover.jpg",
+        "ignored.txt",
+        "page0.jpg",
+        "Page1.jpeg",
+        "Page2.png",
+        "Page3.gif",
+        "page4.webp",
+        "page10.jpg",
+        "test/!cover.tar.gz",
+        "test/00.tar.gz",
+    ]
+
+
 def test_recursive_list_with_file(tmp_path) -> None:
     foo_png = tmp_path / "foo.png"
     foo_png.write_text("not a png")