Fixes for python 3.14 (#17)

* Replace pkg_resources with importlib.resources for Python 3.14 compatibility

* fix self.XPATH() calls to OPF.XPath() instead since 3.14 made functools.partial a descriptor and thus injected self as the first argument

* updated install instruction to include lxml to avoid using non matching system version

* Fix regex in LRF output that stripped CThumbnail metadata. .*? matched until >\n and cut off all children on the same line

---------

Co-authored-by: wave1martian <wave1martian@buildsleeprepeat.com>

README.rst

@@ -123,7 +123,12 @@ managers), i.e:
    $ . venv/bin/activate
    (venv) $ git clone https://github.com/gryf/ebook-converter
    (venv) $ cd ebook-converter
-   (venv) $ pip install .
+   (venv) $ pip install --no-binary lxml .
+
+Note: the ``--no-binary lxml`` flag is required to ensure ``lxml`` is compiled
+against the same system ``libxml2`` as ``html5-parser``. Without it, pip may
+install a pre-built ``lxml`` wheel bundling a different ``libxml2`` version,
+which causes a ``RuntimeError``.
 
 Simple as that. And from now on, you can issue converter:
 

ebook_converter/__init__.py

@@ -1,6 +1,5 @@
 import mimetypes
-import pkg_resources
+from importlib.resources import files
 
 
-mimetypes.init([pkg_resources.
-                resource_filename('ebook_converter', 'data/mime.types')])
+mimetypes.init([str(files('ebook_converter').joinpath('data/mime.types'))])

ebook_converter/ebooks/conversion/cli.py

@@ -7,7 +7,7 @@ import mimetypes
 import numbers
 import optparse
 import os
-import pkg_resources
+from importlib.resources import files
 import re
 import sys
 
@@ -351,8 +351,7 @@ def read_sr_patterns(path, log=None):
 
 def main(args=sys.argv):
     log = logging.default_log
-    mimetypes.init([pkg_resources.resource_filename('ebook_converter',
-                                                    'data/mime.types')])
+    mimetypes.init([str(files('ebook_converter').joinpath('data/mime.types'))])
     parser, plumber = create_option_parser(args, log)
     opts, leftover_args = parser.parse_args(args)
     if len(leftover_args) > 3:

ebook_converter/ebooks/conversion/plugins/fb2_input.py

@@ -3,7 +3,7 @@ Convert .fb2 files to .lrf
 """
 import mimetypes
 import os
-import pkg_resources
+from importlib.resources import files
 import re
 
 from lxml import etree
@@ -85,8 +85,8 @@ class FB2Input(InputFormatPlugin):
             css = re.sub(r'name\s*=\s*', 'class=', css)
         self.extract_embedded_content(doc)
         log.debug('Converting XML to HTML...')
-        with open(pkg_resources.resource_filename('ebook_converter',
-                                                  'data/fb2.xsl')) as f:
+        with open(str(files('ebook_converter')
+                      .joinpath('data/fb2.xsl'))) as f:
             ss = f.read()
         ss = ss.replace("__FB_NS__", fb_ns)
         if options.no_inline_fb2_toc:

ebook_converter/ebooks/conversion/plugins/html_output.py

@@ -1,5 +1,5 @@
 import os
-import pkg_resources
+from importlib.resources import files
 import re
 import shutil
 
@@ -92,9 +92,8 @@ class HTMLOutput(OutputFormatPlugin):
             with open(opts.template_html_index, 'rb') as f:
                 template_html_index_data = f.read()
         else:
-            with open(pkg_resources.
-                      resource_filename('ebook_converter',
-                                        'data/html_export_default_index.tmpl')
+            with open(str(files('ebook_converter')
+                          .joinpath('data/html_export_default_index.tmpl'))
                      ) as fobj:
                 template_html_index_data = fobj.read().decode()
 
@@ -102,9 +101,8 @@ class HTMLOutput(OutputFormatPlugin):
             with open(opts.template_html, 'rb') as f:
                 template_html_data = f.read()
         else:
-            with open(pkg_resources.
-                      resource_filename('ebook_converter',
-                                        'data/html_export_default.tmpl')
+            with open(str(files('ebook_converter')
+                          .joinpath('data/html_export_default.tmpl'))
                      ) as fobj:
                 template_html_data = fobj.read().decode()
 
@@ -112,9 +110,8 @@ class HTMLOutput(OutputFormatPlugin):
             with open(opts.template_css, 'rb') as f:
                 template_css_data = f.read()
         else:
-            with open(pkg_resources.
-                      resource_filename('ebook_converter',
-                                        'data/html_export_default.css')
+            with open(str(files('ebook_converter')
+                          .joinpath('data/html_export_default.css'))
                      ) as fobj:
                 template_css_data = fobj.read().decode()
 

ebook_converter/ebooks/conversion/plugins/lrf_input.py

@@ -1,6 +1,6 @@
 import os
 import sys
-import pkg_resources
+from importlib.resources import files
 
 from lxml import etree
 
@@ -52,9 +52,8 @@ class LRFInput(InputFormatPlugin):
 
         self.log.info('Converting XML to HTML...')
 
-        with open(pkg_resources.
-                  resource_filename('ebook_converter',
-                                    'data/lrf.xsl')) as fobj:
+        with open(str(files('ebook_converter')
+                      .joinpath('data/lrf.xsl'))) as fobj:
             # TODO(gryf): change this nonsense to etree.parse() instead.
             styledoc = etree.fromstring(fobj.read())
         media_type = MediaType()

ebook_converter/ebooks/conversion/plugins/rtf_input.py

@@ -1,6 +1,6 @@
 import glob
 import os
-import pkg_resources
+from importlib.resources import files
 import re
 import textwrap
 
@@ -288,8 +288,8 @@ class RTFInput(InputFormatPlugin):
 
         self.log.info('Converting XML to HTML...')
         inline_class = InlineClass(self.log)
-        with open(pkg_resources.resource_filename('ebook_converter',
-                                                  'data/rtf.xsl')) as fobj:
+        with open(str(files('ebook_converter')
+                      .joinpath('data/rtf.xsl'))) as fobj:
             styledoc = etree.fromstring(fobj.read())
         extensions = {('calibre', 'inline-class'): inline_class}
         transform = etree.XSLT(styledoc, extensions=extensions)

ebook_converter/ebooks/lrf/pylrs/pylrs.py

@@ -670,7 +670,7 @@ class Info(Delegator):
         f = io.BytesIO()
         tree.write(f, encoding='utf-8', xml_declaration=True)
         xmlInfo = f.getvalue().decode('utf-8')
-        xmlInfo = re.sub(r"<CThumbnail.*?>\n", "", xmlInfo)
+        xmlInfo = re.sub(r"<CThumbnail[^>]*/>\s*", "", xmlInfo)
         xmlInfo = xmlInfo.replace("SumPage>", "Page>")
         lrfWriter.docInfoXml = xmlInfo
 

ebook_converter/ebooks/metadata/opf2.py

@@ -697,7 +697,7 @@ class OPF(object):  # {{{
     def find_toc(self):
         self.toc = None
         try:
-            spine = self.XPath('descendant::*[re:match(name(), "spine", '
+            spine = OPF.XPath('descendant::*[re:match(name(), "spine", '
                                '"i")]')(self.root)
             toc = None
             if spine:
@@ -801,7 +801,7 @@ class OPF(object):  # {{{
 
     def replace_spine_items_by_idref(self, idref, new_idrefs):
         items = list(map(self.create_spine_item, new_idrefs))
-        spine = self.XPath('/opf:package/*[re:match(name(), "spine", '
+        spine = OPF.XPath('/opf:package/*[re:match(name(), "spine", '
                            '"i")]')(self.root)[0]
         old = [i for i in self.iterspine() if i.get('idref', None) == idref]
         for x in old:
@@ -1005,7 +1005,7 @@ class OPF(object):  # {{{
 
     def get_identifiers(self):
         identifiers = {}
-        for x in self.XPath(
+        for x in OPF.XPath(
             'descendant::*[local-name() = "identifier" and text()]')(
                     self.metadata):
             found_scheme = False
@@ -1039,7 +1039,7 @@ class OPF(object):  # {{{
                 uuid_id = self.root.attrib[attr]
                 break
 
-        for x in self.XPath(
+        for x in OPF.XPath(
             'descendant::*[local-name() = "identifier"]')(
                     self.metadata):
             xid = x.get('id', None)
@@ -1175,7 +1175,7 @@ class OPF(object):  # {{{
 
     @property
     def page_progression_direction(self):
-        spine = self.XPath('descendant::*[re:match(name(), "spine", '
+        spine = OPF.XPath('descendant::*[re:match(name(), "spine", '
                            '"i")][1]')(self.root)
         if spine:
             for k, v in spine[0].attrib.items():
@@ -1185,7 +1185,7 @@ class OPF(object):  # {{{
 
     @property
     def primary_writing_mode(self):
-        for m in self.XPath('//*[local-name()="meta" and @name="primary-'
+        for m in OPF.XPath('//*[local-name()="meta" and @name="primary-'
                             'writing-mode" and @content]')(self.root):
             return m.get('content')
 

ebook_converter/ebooks/oeb/polish/toc.py

@@ -1,7 +1,7 @@
 import collections
 import functools
 import operator
-import pkg_resources
+from importlib.resources import files
 import re
 import urllib.parse
 
@@ -719,9 +719,8 @@ def commit_nav_toc(container, toc, lang=None, landmarks=None,
         if previous_nav is not None:
             root = previous_nav[1]
         else:
-            with open(pkg_resources.
-                      resource_filename('ebook_converter',
-                                        'data/new_nav.html')) as fobj:
+            with open(str(files('ebook_converter')
+                          .joinpath('data/new_nav.html'))) as fobj:
                 root = container.parse_xhtml(fobj.read())
         container.replace(tocname, root)
     else:
@@ -874,8 +873,7 @@ def toc_to_html(toc, container, toc_name, title, lang=None):
 
     E = ElementMaker(namespace=const.XHTML_NS, nsmap={None: const.XHTML_NS})
     # TODO(gryf): revisit lack of css.
-    css_f = pkg_resources.resource_filename('ebook_converter',
-                                            'data/inline_toc_styles.css')
+    css_f = str(files('ebook_converter').joinpath('data/inline_toc_styles.css'))
     html = E.html(E.head(E.title(title),
                          E.style(css_f, type='text/css')),
                   E.body(E.h2(title), E.ul(),

ebook_converter/ebooks/oeb/stylizer.py

@@ -2,7 +2,7 @@
 CSS property propagation class.
 """
 import os, re, logging, copy, unicodedata, numbers
-import pkg_resources
+from importlib.resources import files
 from operator import itemgetter
 from weakref import WeakKeyDictionary
 from xml.dom import SyntaxErr as CSSSyntaxError
@@ -28,8 +28,7 @@ _html_css_stylesheet = None
 def html_css_stylesheet():
     global _html_css_stylesheet
     if _html_css_stylesheet is None:
-        with open(pkg_resources.resource_filename('ebook_converter',
-                                                  'data/html.css'), 'rb') as f:
+        with open(str(files('ebook_converter').joinpath('data/html.css')), 'rb') as f:
             html_css = f.read().decode('utf-8')
         _html_css_stylesheet = parseString(html_css, validate=False)
     return _html_css_stylesheet

ebook_converter/spell/__init__.py

@@ -1,6 +1,6 @@
 from collections import namedtuple
 import json
-import pkg_resources
+from importlib.resources import files
 
 from ebook_converter.utils.localization import canonicalize_lang
 
@@ -16,8 +16,7 @@ ccodes, ccodemap, country_names = None, None, None
 def get_codes():
     global ccodes, ccodemap, country_names
     if ccodes is None:
-        src = pkg_resources.resource_filename('ebook_converter',
-                                              'data/iso_3166-1.json')
+        src = str(files('ebook_converter').joinpath('data/iso_3166-1.json'))
         with open(src, 'rb') as f:
             db = json.load(f)
         codes = set()

ebook_converter/utils/config_base.py

@@ -10,7 +10,7 @@ import pickle
 import re
 import traceback
 
-import pkg_resources
+from importlib.resources import files as _resource_files
 
 from ebook_converter.constants_old import config_dir
 from ebook_converter.constants_old import filesystem_encoding
@@ -591,8 +591,7 @@ def exec_tweaks(path):
 
 
 def default_tweaks_raw():
-    return pkg_resources.resource_filename('ebook_converter',
-                                           'data/default_tweaks.py')
+    return str(_resource_files('ebook_converter').joinpath('data/default_tweaks.py'))
 
 
 def read_tweaks():

ebook_converter/utils/localization.py

@@ -1,5 +1,5 @@
 import json
-import pkg_resources
+from importlib.resources import files
 
 
 def get_lang():
@@ -39,8 +39,7 @@ def _load_iso639():
     # excerpt form Calibre transform code which is executed during Calibre
     # build).
     if _iso639 is None:
-        src = pkg_resources.resource_filename('ebook_converter',
-                                              'data/iso_639-3.json')
+        src = str(files('ebook_converter').joinpath('data/iso_639-3.json'))
 
         with open(src, 'rb') as f:
             root = json.load(f)