txt2ebook 0.1.24__tar.gz → 0.1.26__tar.gz

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/CHANGELOG.md

@@ -7,6 +7,35 @@ and this project adheres to [0-based versioning](https://0ver.org/).
 
 ## [Unreleased]
 
+## v0.1.26 - 2023-03-12
+
+### Changed
+
+- Rename epub template names, `clean_no_indent` to `noindent`, and
+  `clean_no_paragraph_space` to `condense`
+- Raise warnings for invalid prepend length for `zh_words_to_numbers` function
+
+### Fixed
+
+- Fix missing translations
+- Fix missing EPUB template names not showing in help message
+
+## v0.1.25 - 2023-03-05
+
+### Added
+
+- Add new EPUB templates, `clean_no_indent` and `clean_no_paragraph_space`
+
+### Fixed
+
+- Add missing `zh-*` character for `zh_numeric`
+- Remove unused import
+- Fix incorrect dep added to wrong environment
+
+### Changed
+
+- Refactor ebook extensions variable into package
+
 ## v0.1.24 - 2023-02-26
 
 ### Changed

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/CONTRIBUTING.md

@@ -53,11 +53,17 @@ tox -e py37,py38,py39,py310,py311 -- tests/test_tokenizer.py
 For code lint, we're using `pre-commit`:
 
 ```bash
-pre-commit install
+pre-commit install # run once
 pre-commit clean
 pre-commit run --all-files
 ```
 
+Or specific hook:
+
+```bash
+pre-commit run pylint -a
+```
+
 We're using zero-based versioning.
 
 For patches or bug fixes:

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: txt2ebook
-Version: 0.1.24
+Version: 0.1.26
 Summary: CLI tool to convert txt file to ebook format
 Home-page: https://github.com/kianmeng/txt2ebook
 License: AGPL-3.0-or-later
@@ -31,7 +31,6 @@ Requires-Dist: CJKwrap (>=2.2,<3.0)
 Requires-Dist: EbookLib (>=0.17.1,<0.18.0)
 Requires-Dist: bs4 (>=0.0.1,<0.0.2)
 Requires-Dist: langdetect (>=1.0.9,<2.0.0)
-Requires-Dist: pytest-xdist (>=3.1.0,<4.0.0)
 Requires-Dist: regex (>=2021.11.10,<2022.0.0)
 Requires-Dist: tox-pyenv (>=1.1.0,<2.0.0)
 Project-URL: Repository, https://github.com/kianmeng/txt2ebook
@@ -67,8 +66,9 @@ txt2ebook --help
 usage: txt2ebook [-f {epub,txt}] [-t TITLE] [-l LANGUAGE] [-a AUTHOR]
                  [-c IMAGE_FILENAME] [-w WIDTH] [-ps SEPARATOR] [-rd REGEX]
                  [-rvc REGEX] [-rv REGEX] [-rc REGEX] [-rt REGEX] [-ra REGEX]
-                 [-rl REGEX] [-rr REGEX REGEX] [-et TEMPLATE] [-vp] [-tp]
-                 [-sp] [-hn] [-fw] [-rw] [-ob] [-ow] [-v] [-d] [-h] [-V]
+                 [-rl REGEX] [-rr REGEX REGEX] [-et {clean,noindent,condense}]
+                 [-vp] [-tp] [-sp] [-hn] [-fw] [-rw] [-ob] [-ow] [-v] [-d]
+                 [-h] [-V]
                  TXT_FILENAME [EBOOK_FILENAME]
 
 txt2ebook/tte is a cli tool to convert txt file to ebook format.
@@ -112,7 +112,7 @@ optional arguments:
       regex to delete whole line (default: '[]')
   -rr REGEX REGEX, --regex-replace REGEX REGEX
       regex to search and replace (default: '[]')
-  -et TEMPLATE, --epub-template TEMPLATE
+  -et {clean,noindent,condense}, --epub-template {clean,noindent,condense}
       CSS template for epub ebook (default: 'clean')
   -vp, --volume-page
       generate each volume as separate page (only 'epub' format)

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/README.md

@@ -28,8 +28,9 @@ txt2ebook --help
 usage: txt2ebook [-f {epub,txt}] [-t TITLE] [-l LANGUAGE] [-a AUTHOR]
                  [-c IMAGE_FILENAME] [-w WIDTH] [-ps SEPARATOR] [-rd REGEX]
                  [-rvc REGEX] [-rv REGEX] [-rc REGEX] [-rt REGEX] [-ra REGEX]
-                 [-rl REGEX] [-rr REGEX REGEX] [-et TEMPLATE] [-vp] [-tp]
-                 [-sp] [-hn] [-fw] [-rw] [-ob] [-ow] [-v] [-d] [-h] [-V]
+                 [-rl REGEX] [-rr REGEX REGEX] [-et {clean,noindent,condense}]
+                 [-vp] [-tp] [-sp] [-hn] [-fw] [-rw] [-ob] [-ow] [-v] [-d]
+                 [-h] [-V]
                  TXT_FILENAME [EBOOK_FILENAME]
 
 txt2ebook/tte is a cli tool to convert txt file to ebook format.
@@ -73,7 +74,7 @@ optional arguments:
       regex to delete whole line (default: '[]')
   -rr REGEX REGEX, --regex-replace REGEX REGEX
       regex to search and replace (default: '[]')
-  -et TEMPLATE, --epub-template TEMPLATE
+  -et {clean,noindent,condense}, --epub-template {clean,noindent,condense}
       CSS template for epub ebook (default: 'clean')
   -vp, --volume-page
       generate each volume as separate page (only 'epub' format)

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "txt2ebook"
-version = "0.1.24"
+version = "0.1.26"
 description = "CLI tool to convert txt file to ebook format"
 authors = ["Kian-Meng Ang <kianmeng@cpan.org>"]
 license = "AGPL-3.0-or-later"
@@ -37,7 +37,6 @@ langdetect = "^1.0.9"
 python = "^3.7"
 regex = "^2021.11.10"
 tox-pyenv = "^1.1.0"
-pytest-xdist = "^3.1.0"
 
 [tool.poetry.dev-dependencies]
 bandit = "^1.7.1"
@@ -56,6 +55,9 @@ pre-commit = "^2.20.0"
 txt2ebook = 'txt2ebook.txt2ebook:main'
 tte = 'txt2ebook.txt2ebook:main'
 
+[tool.poetry.group.dev.dependencies]
+pytest-xdist = "^3.2.0"
+
 [build-system]
 build-backend = "poetry.core.masonry.api"
 requires = [

txt2ebook-0.1.26/setup.py

@@ -0,0 +1,59 @@
+# -*- coding: utf-8 -*-
+from setuptools import setup
+
+package_dir = \
+{'': 'src'}
+
+packages = \
+['txt2ebook',
+ 'txt2ebook.formats',
+ 'txt2ebook.formats.templates',
+ 'txt2ebook.formats.templates.epub',
+ 'txt2ebook.helpers',
+ 'txt2ebook.languages',
+ 'txt2ebook.models']
+
+package_data = \
+{'': ['*'],
+ 'txt2ebook': ['locales/*',
+               'locales/en/LC_MESSAGES/txt2ebook.mo',
+               'locales/en/LC_MESSAGES/txt2ebook.mo',
+               'locales/en/LC_MESSAGES/txt2ebook.po',
+               'locales/en/LC_MESSAGES/txt2ebook.po',
+               'locales/zh-cn/LC_MESSAGES/txt2ebook.mo',
+               'locales/zh-cn/LC_MESSAGES/txt2ebook.mo',
+               'locales/zh-cn/LC_MESSAGES/txt2ebook.po',
+               'locales/zh-cn/LC_MESSAGES/txt2ebook.po']}
+
+install_requires = \
+['CJKwrap>=2.2,<3.0',
+ 'EbookLib>=0.17.1,<0.18.0',
+ 'bs4>=0.0.1,<0.0.2',
+ 'langdetect>=1.0.9,<2.0.0',
+ 'regex>=2021.11.10,<2022.0.0',
+ 'tox-pyenv>=1.1.0,<2.0.0']
+
+entry_points = \
+{'console_scripts': ['tte = txt2ebook.txt2ebook:main',
+                     'txt2ebook = txt2ebook.txt2ebook:main']}
+
+setup_kwargs = {
+    'name': 'txt2ebook',
+    'version': '0.1.26',
+    'description': 'CLI tool to convert txt file to ebook format',
+    'long_description': "# txt2ebook\n\nConsole tool to convert txt file to different ebook format.\n\n## Installation\n\nStable version From PyPI:\n\n```console\npython3 -m pip install txt2ebook\n```\n\nLatest development version from GitHub:\n\n```console\npython3 -m pip install -e git+https://github.com/kianmeng/txt2ebook.git\n```\n\n## Usage\n\nShowing help message of command-line options:\n\n```console\ntxt2ebook --help\n```\n\n```console\nusage: txt2ebook [-f {epub,txt}] [-t TITLE] [-l LANGUAGE] [-a AUTHOR]\n                 [-c IMAGE_FILENAME] [-w WIDTH] [-ps SEPARATOR] [-rd REGEX]\n                 [-rvc REGEX] [-rv REGEX] [-rc REGEX] [-rt REGEX] [-ra REGEX]\n                 [-rl REGEX] [-rr REGEX REGEX] [-et {clean,noindent,condense}]\n                 [-vp] [-tp] [-sp] [-hn] [-fw] [-rw] [-ob] [-ow] [-v] [-d]\n                 [-h] [-V]\n                 TXT_FILENAME [EBOOK_FILENAME]\n\ntxt2ebook/tte is a cli tool to convert txt file to ebook format.\n  website: https://github.com/kianmeng/txt2ebook\n  issues: https://github.com/kianmeng/txt2ebook/issues\n\npositional arguments:\n  TXT_FILENAME\n      source text filename\n  EBOOK_FILENAME\n      converted ebook filename (default: 'TXT_FILENAME.{epub,txt}')\n\noptional arguments:\n  -f {epub,txt}, --format {epub,txt}\n      ebook format (default: '['epub', 'txt']')\n  -t TITLE, --title TITLE\n      title of the ebook (default: 'None')\n  -l LANGUAGE, --language LANGUAGE\n      language of the ebook (default: 'None')\n  -a AUTHOR, --author AUTHOR\n      author of the ebook (default: '[]')\n  -c IMAGE_FILENAME, --cover IMAGE_FILENAME\n      cover of the ebook\n  -w WIDTH, --width WIDTH\n      width for line wrapping\n  -ps SEPARATOR, --paragraph_separator SEPARATOR\n      paragraph separator (default: '\\n\\n')\n  -rd REGEX, --regex-delete REGEX\n      regex to delete word or phrase (default: '[]')\n  -rvc REGEX, --regex-volume-chapter REGEX\n      regex to parse volume and chapter header (default: by LANGUAGE)\n  -rv REGEX, --regex-volume REGEX\n      regex to parse volume header (default: by LANGUAGE)\n  -rc REGEX, --regex-chapter REGEX\n      regex to parse chapter header (default: by LANGUAGE)\n  -rt REGEX, --regex-title REGEX\n      regex to parse title of the book (default: by LANGUAGE)\n  -ra REGEX, --regex-author REGEX\n      regex to parse author of the book (default: by LANGUAGE)\n  -rl REGEX, --regex-delete-line REGEX\n      regex to delete whole line (default: '[]')\n  -rr REGEX REGEX, --regex-replace REGEX REGEX\n      regex to search and replace (default: '[]')\n  -et {clean,noindent,condense}, --epub-template {clean,noindent,condense}\n      CSS template for epub ebook (default: 'clean')\n  -vp, --volume-page\n      generate each volume as separate page (only 'epub' format)\n  -tp, --test-parsing\n      test parsing for volume/chapter header\n  -sp, --split-volume-and-chapter\n      Split volume or chapter into separate file (only 'txt' format).\n      Both --overwrite-backup and --overwrite option will be ignore.\n  -hn, --header-number\n      convert section header from words to numbers (only 'zh-cn/zh-tw' language)\n  -fw, --fullwidth\n      convert ASCII character from halfwidth to fullwidth (only 'zh-cn/zh-tw' language)\n  -rw, --raise-on-warning\n      raise exception and stop parsing upon warning\n  -ob, --overwrite-backup\n      overwrite massaged and backup original TXT_FILENAME\n  -ow, --overwrite\n      overwrite massaged TXT_FILENAME\n  -v, --verbose\n      show verbosity of debugging log, use -vv, -vvv for more details\n  -d, --debug\n      show debugging log and stacktrace\n  -h, --help\n      show this help message and exit\n  -V, --version\n      show program's version number and exit\n```\n\nConvert a txt file into epub:\n\n```console\ntxt2book ebook.txt\n```\n\n## Copyright and License\n\nCopyright (c) 2021,2022,2023 Kian-Meng Ang\n\nThis program is free software: you can redistribute it and/or modify it under\nthe terms of the GNU Affero General Public License as published by the Free\nSoftware Foundation, either version 3 of the License, or (at your option) any\nlater version.\n\nThis program is distributed in the hope that it will be useful, but WITHOUT ANY\nWARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A\nPARTICULAR PURPOSE. See the GNU Affero General Public License for more details.\n\nYou should have received a copy of the GNU Affero General Public License along\nwith this program. If not, see <https://www.gnu.org/licenses/>.\n",
+    'author': 'Kian-Meng Ang',
+    'author_email': 'kianmeng@cpan.org',
+    'maintainer': 'None',
+    'maintainer_email': 'None',
+    'url': 'https://github.com/kianmeng/txt2ebook',
+    'package_dir': package_dir,
+    'packages': packages,
+    'package_data': package_data,
+    'install_requires': install_requires,
+    'entry_points': entry_points,
+    'python_requires': '>=3.7,<4.0',
+}
+
+
+setup(**setup_kwargs)

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/__init__.py

@@ -21,7 +21,7 @@ import sys
 
 logger = logging.getLogger(__name__)
 
-__version__ = "0.1.24"
+__version__ = "0.1.26"
 
 
 def setup_logger(config: argparse.Namespace) -> None:

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/formats/__init__.py

@@ -19,11 +19,14 @@ import argparse
 from typing import Union
 
 import txt2ebook.models.book
+from txt2ebook.formats.epub import TEMPLATES as EPUB_TEMPLATES
 from txt2ebook.formats.epub import EpubWriter
 from txt2ebook.formats.txt import TxtWriter
 from txt2ebook.helpers import load_class, to_classname
 from txt2ebook.models import Book
 
+EBOOK_FORMATS = ["epub", "txt"]
+
 
 def create_format(
     book: txt2ebook.models.book.Book,
@@ -49,6 +52,8 @@ def create_format(
 
 __all__ = [
     "Book",
+    "EBOOK_FORMATS",
+    "EPUB_TEMPLATES",
     "EpubWriter",
     "TxtWriter",
 ]

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/formats/epub.py

@@ -19,7 +19,7 @@ import gettext
 import io
 import logging
 import uuid
-from importlib.resources import read_text
+from importlib.resources import contents, read_text
 from pathlib import Path
 from typing import Optional, Union
 
@@ -32,6 +32,12 @@ logger = logging.getLogger(__name__)
 
 SPACE = "\u0020"
 
+TEMPLATES = [
+    Path(file).stem
+    for file in list(contents(template_epub))
+    if Path(file).suffix == ".css"
+]
+
 
 class EpubWriter:
     """Module for writing ebook in epub format."""

txt2ebook-0.1.26/src/txt2ebook/formats/templates/epub/condense.css

@@ -0,0 +1,42 @@
+/* cover page */
+div#cover {
+  text-align: center;
+  width: 100%;
+}
+
+/* navigation page - toc */
+ol {
+  list-style-type: none;
+  padding-left: 0;
+}
+
+ol li {
+  padding-top: 0.2em;
+}
+
+ol li ol li:last-of-type {
+  padding-bottom: 1em;
+}
+
+ol li a {
+  text-decoration: none;
+}
+
+/* chapters & section */
+div.volume {
+  display: table;
+  width: 100%;
+  height: 100%;
+}
+
+h1.volume {
+  display: table-cell;
+  text-align: center;
+  vertical-align: middle;
+}
+
+h2 {
+}
+p {
+  margin: 0em;
+}

txt2ebook-0.1.26/src/txt2ebook/formats/templates/epub/noindent.css

@@ -0,0 +1,42 @@
+/* cover page */
+div#cover {
+  text-align: center;
+  width: 100%;
+}
+
+/* navigation page - toc */
+ol {
+  list-style-type: none;
+  padding-left: 0;
+}
+
+ol li {
+  padding-top: 0.2em;
+}
+
+ol li ol li:last-of-type {
+  padding-bottom: 1em;
+}
+
+ol li a {
+  text-decoration: none;
+}
+
+/* chapters & section */
+div.volume {
+  display: table;
+  width: 100%;
+  height: 100%;
+}
+
+h1.volume {
+  display: table-cell;
+  text-align: center;
+  vertical-align: middle;
+}
+
+h2 {
+}
+p {
+  text-indent: -2em;
+}

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/locales/en/LC_MESSAGES/txt2ebook.po

@@ -5,7 +5,7 @@
 msgid ""
 msgstr ""
 "Project-Id-Version: PACKAGE VERSION\n"
-"POT-Creation-Date: 2023-02-04 22:03+0800\n"
+"POT-Creation-Date: 2023-03-06 04:02+0800\n"
 "PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
 "Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
 "Language-Team: LANGUAGE <LL@li.org>\n"
@@ -15,22 +15,22 @@ msgstr ""
 "Content-Transfer-Encoding: 8bit\n"
 "Generated-By: pygettext.py 1.5\n"
 
-#: src/txt2ebook/formats/epub.py:193
+#: src/txt2ebook/formats/epub.py:199
 msgid "cover"
 msgstr "Cover"
 
-#: src/txt2ebook/formats/txt.py:88
+#: src/txt2ebook/formats/txt.py:84
 msgid "metadata"
 msgstr "metadata"
 
-#: src/txt2ebook/formats/txt.py:158
+#: src/txt2ebook/formats/txt.py:154
 msgid "title:"
 msgstr "Title:"
 
-#: src/txt2ebook/formats/txt.py:159
+#: src/txt2ebook/formats/txt.py:155
 msgid "author:"
 msgstr "Author:"
 
-#: src/txt2ebook/formats/txt.py:160
+#: src/txt2ebook/formats/txt.py:156
 msgid "tag:"
 msgstr "Tag:"

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/locales/txt2ebook.pot

@@ -5,7 +5,7 @@
 msgid ""
 msgstr ""
 "Project-Id-Version: PACKAGE VERSION\n"
-"POT-Creation-Date: 2023-02-05 10:12+0800\n"
+"POT-Creation-Date: 2023-03-06 04:02+0800\n"
 "PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
 "Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
 "Language-Team: LANGUAGE <LL@li.org>\n"
@@ -15,23 +15,23 @@ msgstr ""
 "Generated-By: pygettext.py 1.5\n"
 
 
-#: src/txt2ebook/formats/epub.py:193
+#: src/txt2ebook/formats/epub.py:199
 msgid "cover"
 msgstr ""
 
-#: src/txt2ebook/formats/txt.py:88
+#: src/txt2ebook/formats/txt.py:84
 msgid "metadata"
 msgstr ""
 
-#: src/txt2ebook/formats/txt.py:158
+#: src/txt2ebook/formats/txt.py:154
 msgid "title:"
 msgstr ""
 
-#: src/txt2ebook/formats/txt.py:159
+#: src/txt2ebook/formats/txt.py:155
 msgid "author:"
 msgstr ""
 
-#: src/txt2ebook/formats/txt.py:160
+#: src/txt2ebook/formats/txt.py:156
 msgid "tag:"
 msgstr ""
 

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/locales/zh-cn/LC_MESSAGES/txt2ebook.po

@@ -5,7 +5,7 @@
 msgid ""
 msgstr ""
 "Project-Id-Version: \n"
-"POT-Creation-Date: 2023-02-04 22:03+0800\n"
+"POT-Creation-Date: 2023-03-06 04:02+0800\n"
 "PO-Revision-Date: 2023-01-14 18:07+0800\n"
 "Last-Translator: \n"
 "Language-Team: \n"
@@ -16,22 +16,22 @@ msgstr ""
 "Generated-By: pygettext.py 1.5\n"
 "X-Generator: Poedit 3.1.1\n"
 
-#: src/txt2ebook/formats/epub.py:193
+#: src/txt2ebook/formats/epub.py:199
 msgid "cover"
 msgstr "封面"
 
-#: src/txt2ebook/formats/txt.py:88
+#: src/txt2ebook/formats/txt.py:84
 msgid "metadata"
 msgstr "元数据"
 
-#: src/txt2ebook/formats/txt.py:158
+#: src/txt2ebook/formats/txt.py:154
 msgid "title:"
 msgstr "书名："
 
-#: src/txt2ebook/formats/txt.py:159
+#: src/txt2ebook/formats/txt.py:155
 msgid "author:"
 msgstr "作者："
 
-#: src/txt2ebook/formats/txt.py:160
+#: src/txt2ebook/formats/txt.py:156
 msgid "tag:"
 msgstr "票签："

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/parser.py

@@ -26,7 +26,7 @@ import regex as re
 
 from txt2ebook.models import Book, Chapter, Volume
 from txt2ebook.tokenizer import Tokenizer
-from txt2ebook.zh_utils import zh_halfwidth_to_fullwidth, zh_numeric, zh_words_to_numbers
+from txt2ebook.zh_utils import zh_halfwidth_to_fullwidth, zh_words_to_numbers
 
 logger = logging.getLogger(__name__)
 
@@ -135,7 +135,6 @@ class Parser:
         )
         return replaced_words
 
-
     def parse_tokens(self, tokenizer: Tokenizer) -> Tuple:
         """Parse the tokens and organize into book structure."""
         toc: List[Union[Volume, Chapter]] = []

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/txt2ebook.py

@@ -26,11 +26,9 @@ from langdetect import detect
 
 from txt2ebook import __version__, setup_logger
 from txt2ebook.exceptions import EmptyFileError
-from txt2ebook.formats import create_format
+from txt2ebook.formats import EBOOK_FORMATS, EPUB_TEMPLATES, create_format
 from txt2ebook.parser import Parser
 
-EBOOK_EXTS = ["epub", "txt"]
-
 logger = logging.getLogger(__name__)
 
 
@@ -104,7 +102,7 @@ def build_parser(
         default=None,
         help=(
             "converted ebook filename "
-            "(default: 'TXT_FILENAME.{" + ",".join(EBOOK_EXTS) + "}')"
+            "(default: 'TXT_FILENAME.{" + ",".join(EBOOK_FORMATS) + "}')"
         ),
         metavar="EBOOK_FILENAME",
     )
@@ -114,7 +112,7 @@ def build_parser(
         "--format",
         dest="format",
         default=["epub", "txt"],
-        choices=EBOOK_EXTS,
+        choices=EBOOK_FORMATS,
         action="append",
         help="ebook format (default: '%(default)s')",
     )
@@ -264,9 +262,9 @@ def build_parser(
         "-et",
         "--epub-template",
         default="clean",
+        choices=EPUB_TEMPLATES,
         dest="epub_template",
         help="CSS template for epub ebook (default: '%(default)s')",
-        metavar="TEMPLATE",
     )
 
     parser.add_argument(

{txt2ebook-0.1.24 → txt2ebook-0.1.26}/src/txt2ebook/zh_utils.py

@@ -16,10 +16,12 @@
 """String helper functions for handling zh related text."""
 
 import re
+import warnings
 from typing import Any
 from unicodedata import numeric
 
 WORD_NUMERIC_MAP = {
+    "两": 2.0,
     "圩": 50.0,
     "圓": 60.0,
     "進": 70.0,
@@ -130,9 +132,18 @@ def zh_words_to_numbers(words: str, length: int = 0) -> str:
             header_nums += int(zh_numeric(word_grp))
 
     padded_header_nums = str(header_nums)
-    # todo: padded_header_nums length must smaller than length
+
     if length > 0:
-        padded_header_nums = padded_header_nums.rjust(length, "0")
+        word_length = len(padded_header_nums)
+        if word_length < length:
+            padded_header_nums = padded_header_nums.rjust(length, "0")
+        else:
+            warnings.warn(
+                (
+                    "prepend zero length less than word length, "
+                    f"word length: {word_length}, prepend length: {length}"
+                )
+            )
 
     replaced_words = words.replace(header_words, padded_header_nums)
     return replaced_words

txt2ebook-0.1.24/setup.py

@@ -1,60 +0,0 @@
-# -*- coding: utf-8 -*-
-from setuptools import setup
-
-package_dir = \
-{'': 'src'}
-
-packages = \
-['txt2ebook',
- 'txt2ebook.formats',
- 'txt2ebook.formats.templates',
- 'txt2ebook.formats.templates.epub',
- 'txt2ebook.helpers',
- 'txt2ebook.languages',
- 'txt2ebook.models']
-
-package_data = \
-{'': ['*'],
- 'txt2ebook': ['locales/*',
-               'locales/en/LC_MESSAGES/txt2ebook.mo',
-               'locales/en/LC_MESSAGES/txt2ebook.mo',
-               'locales/en/LC_MESSAGES/txt2ebook.po',
-               'locales/en/LC_MESSAGES/txt2ebook.po',
-               'locales/zh-cn/LC_MESSAGES/txt2ebook.mo',
-               'locales/zh-cn/LC_MESSAGES/txt2ebook.mo',
-               'locales/zh-cn/LC_MESSAGES/txt2ebook.po',
-               'locales/zh-cn/LC_MESSAGES/txt2ebook.po']}
-
-install_requires = \
-['CJKwrap>=2.2,<3.0',
- 'EbookLib>=0.17.1,<0.18.0',
- 'bs4>=0.0.1,<0.0.2',
- 'langdetect>=1.0.9,<2.0.0',
- 'pytest-xdist>=3.1.0,<4.0.0',
- 'regex>=2021.11.10,<2022.0.0',
- 'tox-pyenv>=1.1.0,<2.0.0']
-
-entry_points = \
-{'console_scripts': ['tte = txt2ebook.txt2ebook:main',
-                     'txt2ebook = txt2ebook.txt2ebook:main']}
-
-setup_kwargs = {
-    'name': 'txt2ebook',
-    'version': '0.1.24',
-    'description': 'CLI tool to convert txt file to ebook format',
-    'long_description': "# txt2ebook\n\nConsole tool to convert txt file to different ebook format.\n\n## Installation\n\nStable version From PyPI:\n\n```console\npython3 -m pip install txt2ebook\n```\n\nLatest development version from GitHub:\n\n```console\npython3 -m pip install -e git+https://github.com/kianmeng/txt2ebook.git\n```\n\n## Usage\n\nShowing help message of command-line options:\n\n```console\ntxt2ebook --help\n```\n\n```console\nusage: txt2ebook [-f {epub,txt}] [-t TITLE] [-l LANGUAGE] [-a AUTHOR]\n                 [-c IMAGE_FILENAME] [-w WIDTH] [-ps SEPARATOR] [-rd REGEX]\n                 [-rvc REGEX] [-rv REGEX] [-rc REGEX] [-rt REGEX] [-ra REGEX]\n                 [-rl REGEX] [-rr REGEX REGEX] [-et TEMPLATE] [-vp] [-tp]\n                 [-sp] [-hn] [-fw] [-rw] [-ob] [-ow] [-v] [-d] [-h] [-V]\n                 TXT_FILENAME [EBOOK_FILENAME]\n\ntxt2ebook/tte is a cli tool to convert txt file to ebook format.\n  website: https://github.com/kianmeng/txt2ebook\n  issues: https://github.com/kianmeng/txt2ebook/issues\n\npositional arguments:\n  TXT_FILENAME\n      source text filename\n  EBOOK_FILENAME\n      converted ebook filename (default: 'TXT_FILENAME.{epub,txt}')\n\noptional arguments:\n  -f {epub,txt}, --format {epub,txt}\n      ebook format (default: '['epub', 'txt']')\n  -t TITLE, --title TITLE\n      title of the ebook (default: 'None')\n  -l LANGUAGE, --language LANGUAGE\n      language of the ebook (default: 'None')\n  -a AUTHOR, --author AUTHOR\n      author of the ebook (default: '[]')\n  -c IMAGE_FILENAME, --cover IMAGE_FILENAME\n      cover of the ebook\n  -w WIDTH, --width WIDTH\n      width for line wrapping\n  -ps SEPARATOR, --paragraph_separator SEPARATOR\n      paragraph separator (default: '\\n\\n')\n  -rd REGEX, --regex-delete REGEX\n      regex to delete word or phrase (default: '[]')\n  -rvc REGEX, --regex-volume-chapter REGEX\n      regex to parse volume and chapter header (default: by LANGUAGE)\n  -rv REGEX, --regex-volume REGEX\n      regex to parse volume header (default: by LANGUAGE)\n  -rc REGEX, --regex-chapter REGEX\n      regex to parse chapter header (default: by LANGUAGE)\n  -rt REGEX, --regex-title REGEX\n      regex to parse title of the book (default: by LANGUAGE)\n  -ra REGEX, --regex-author REGEX\n      regex to parse author of the book (default: by LANGUAGE)\n  -rl REGEX, --regex-delete-line REGEX\n      regex to delete whole line (default: '[]')\n  -rr REGEX REGEX, --regex-replace REGEX REGEX\n      regex to search and replace (default: '[]')\n  -et TEMPLATE, --epub-template TEMPLATE\n      CSS template for epub ebook (default: 'clean')\n  -vp, --volume-page\n      generate each volume as separate page (only 'epub' format)\n  -tp, --test-parsing\n      test parsing for volume/chapter header\n  -sp, --split-volume-and-chapter\n      Split volume or chapter into separate file (only 'txt' format).\n      Both --overwrite-backup and --overwrite option will be ignore.\n  -hn, --header-number\n      convert section header from words to numbers (only 'zh-cn/zh-tw' language)\n  -fw, --fullwidth\n      convert ASCII character from halfwidth to fullwidth (only 'zh-cn/zh-tw' language)\n  -rw, --raise-on-warning\n      raise exception and stop parsing upon warning\n  -ob, --overwrite-backup\n      overwrite massaged and backup original TXT_FILENAME\n  -ow, --overwrite\n      overwrite massaged TXT_FILENAME\n  -v, --verbose\n      show verbosity of debugging log, use -vv, -vvv for more details\n  -d, --debug\n      show debugging log and stacktrace\n  -h, --help\n      show this help message and exit\n  -V, --version\n      show program's version number and exit\n```\n\nConvert a txt file into epub:\n\n```console\ntxt2book ebook.txt\n```\n\n## Copyright and License\n\nCopyright (c) 2021,2022,2023 Kian-Meng Ang\n\nThis program is free software: you can redistribute it and/or modify it under\nthe terms of the GNU Affero General Public License as published by the Free\nSoftware Foundation, either version 3 of the License, or (at your option) any\nlater version.\n\nThis program is distributed in the hope that it will be useful, but WITHOUT ANY\nWARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A\nPARTICULAR PURPOSE. See the GNU Affero General Public License for more details.\n\nYou should have received a copy of the GNU Affero General Public License along\nwith this program. If not, see <https://www.gnu.org/licenses/>.\n",
-    'author': 'Kian-Meng Ang',
-    'author_email': 'kianmeng@cpan.org',
-    'maintainer': 'None',
-    'maintainer_email': 'None',
-    'url': 'https://github.com/kianmeng/txt2ebook',
-    'package_dir': package_dir,
-    'packages': packages,
-    'package_data': package_data,
-    'install_requires': install_requires,
-    'entry_points': entry_points,
-    'python_requires': '>=3.7,<4.0',
-}
-
-
-setup(**setup_kwargs)