Make PyICU optional

Update README.md

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

@@ -18,22 +18,46 @@ import json
 import logging
 import os
 import pathlib
+import platform
 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 = ("a", *x[1:])
+
+        lst.append(ret)
+    return tuple(lst)
+
+
+def os_sorted(lst: Iterable) -> Iterable:
+    key = _custom_key
+    if icu_available or platform.system() == "Windows":
+        key = natsort.os_sort_keygen()
+    return sorted(lst, key=key)
 
 
 def combine_notes(existing_notes: str | None, new_notes: str | None, split: str) -> str:

comictagger.py

@@ -1,10 +0,0 @@
-#!/usr/bin/env python3
-from __future__ import annotations
-
-import localefix
-from comictaggerlib.main import App
-
-if __name__ == "__main__":
-    localefix.configure_locale()
-
-    App().run()

comictagger.spec

@@ -9,7 +9,7 @@ block_cipher = None
 
 
 a = Analysis(
-    ["comictagger.py"],
+    ["comictaggerlib/__main__.py"],
     pathex=[],
     binaries=[],
     datas=[],

comictaggerlib/__main__.py

@@ -0,0 +1,5 @@
+from __future__ import annotations
+
+from comictaggerlib.main import main
+
+main()

comictaggerlib/gui.py

@@ -73,12 +73,12 @@ try:
                 return True
             return super().event(event)
 
-except ImportError as e:
+except ImportError:
 
     def show_exception_box(log_msg: str) -> None:
         ...
 
-    logger.error(str(e))
+    logger.exception("Qt unavailable")
     qt_available = False
 
 

comictaggerlib/main.py

@@ -17,8 +17,12 @@ from __future__ import annotations
 
 import argparse
 import json
+import locale
+import logging
 import logging.handlers
+import os
 import signal
+import subprocess
 import sys
 
 import settngs
@@ -48,6 +52,49 @@ except Exception:
 logger.setLevel(logging.DEBUG)
 
 
+def _lang_code_mac() -> str:
+    """
+    stolen from https://github.com/mu-editor/mu
+    Returns the user's language preference as defined in the Language & Region
+    preference pane in macOS's System Preferences.
+    """
+
+    # Uses the shell command `defaults read -g AppleLocale` that prints out a
+    # language code to standard output. Assumptions about the command:
+    # - It exists and is in the shell's PATH.
+    # - It accepts those arguments.
+    # - It returns a usable language code.
+    #
+    # Reference documentation:
+    # - The man page for the `defaults` command on macOS.
+    # - The macOS underlying API:
+    #   https://developer.apple.com/documentation/foundation/nsuserdefaults.
+
+    lang_detect_command = "defaults read -g AppleLocale"
+
+    status, output = subprocess.getstatusoutput(lang_detect_command)
+    if status == 0:
+        # Command was successful.
+        lang_code = output
+    else:
+        logging.warning("Language detection command failed: %r", output)
+        lang_code = ""
+
+    return lang_code
+
+
+def configure_locale() -> None:
+    if sys.platform == "darwin" and "LANG" not in os.environ:
+        code = _lang_code_mac()
+        if code != "":
+            os.environ["LANG"] = f"{code}.utf-8"
+
+    locale.setlocale(locale.LC_ALL, "")
+    sys.stdout.reconfigure(encoding=sys.getdefaultencoding())  # type: ignore[attr-defined]
+    sys.stderr.reconfigure(encoding=sys.getdefaultencoding())  # type: ignore[attr-defined]
+    sys.stdin.reconfigure(encoding=sys.getdefaultencoding())  # type: ignore[attr-defined]
+
+
 def update_publishers(config: settngs.Config[settngs.Namespace]) -> None:
     json_file = config[0].runtime_config.user_config_dir / "publishers.json"
     if json_file.exists():
@@ -66,6 +113,7 @@ class App:
         self.config_load_success = False
 
     def run(self) -> None:
+        configure_locale()
         conf = self.initialize()
         self.initialize_dirs(conf.config)
         self.load_plugins(conf)

localefix.py

@@ -1,50 +0,0 @@
-from __future__ import annotations
-
-import locale
-import logging
-import os
-import subprocess
-import sys
-
-
-def _lang_code_mac() -> str:
-    """
-    stolen from https://github.com/mu-editor/mu
-    Returns the user's language preference as defined in the Language & Region
-    preference pane in macOS's System Preferences.
-    """
-
-    # Uses the shell command `defaults read -g AppleLocale` that prints out a
-    # language code to standard output. Assumptions about the command:
-    # - It exists and is in the shell's PATH.
-    # - It accepts those arguments.
-    # - It returns a usable language code.
-    #
-    # Reference documentation:
-    # - The man page for the `defaults` command on macOS.
-    # - The macOS underlying API:
-    #   https://developer.apple.com/documentation/foundation/nsuserdefaults.
-
-    lang_detect_command = "defaults read -g AppleLocale"
-
-    status, output = subprocess.getstatusoutput(lang_detect_command)
-    if status == 0:
-        # Command was successful.
-        lang_code = output
-    else:
-        logging.warning("Language detection command failed: %r", output)
-        lang_code = ""
-
-    return lang_code
-
-
-def configure_locale() -> None:
-    if sys.platform == "darwin" and "LANG" not in os.environ:
-        code = _lang_code_mac()
-        if code != "":
-            os.environ["LANG"] = f"{code}.utf-8"
-
-    locale.setlocale(locale.LC_ALL, "")
-    sys.stdout.reconfigure(encoding=sys.getdefaultencoding())  # type: ignore[attr-defined]
-    sys.stderr.reconfigure(encoding=sys.getdefaultencoding())  # type: ignore[attr-defined]
-    sys.stdin.reconfigure(encoding=sys.getdefaultencoding())  # type: ignore[attr-defined]

setup.py

@@ -36,7 +36,7 @@ install_requires = read("requirements.txt").splitlines()
 extras_require = {}
 extra_req_files = glob.glob("requirements-*.txt")
 for extra_req_file in extra_req_files:
-    name = os.path.splitext(extra_req_file)[0].replace("requirements-", "", 1)
+    name = os.path.splitext(extra_req_file)[0].removeprefix("requirements-")
     extras_require[name] = read(extra_req_file).splitlines()
 
 # If there are any extras, add a catch-all case that includes everything.

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")