weasyprint 67.0__py3-none-any.whl → 68.1__py3-none-any.whl

weasyprint/__init__.py

@@ -5,7 +5,6 @@ importing sub-modules.
 
 """
 
-import contextlib
 from datetime import datetime
 from os.path import getctime, getmtime
 from pathlib import Path
@@ -15,9 +14,9 @@ import cssselect2
 import tinycss2
 import tinyhtml5
 
-VERSION = __version__ = '67.0'
+VERSION = __version__ = '68.1'
 
-#: Default values for command-line and Python API options. See
+#: Default values for command-line and Python API rendering options. See
 #: :func:`__main__.main` to learn more about specific options for
 #: command-line.
 #:
@@ -67,14 +66,15 @@ VERSION = __version__ = '67.0'
 #:     images are temporarily stored.
 DEFAULT_OPTIONS = {
     'stylesheets': None,
-    'media_type': 'print',
     'attachments': None,
+    'attachment_relationships': None,
     'pdf_identifier': None,
     'pdf_variant': None,
     'pdf_version': None,
     'pdf_forms': None,
     'pdf_tags': False,
     'uncompressed_pdf': False,
+    'xmp_metadata': None,
     'custom_metadata': False,
     'presentational_hints': False,
     'srgb': False,
@@ -92,8 +92,7 @@ __all__ = [
 
 
 # Import after setting the version, as the version is used in other modules
-from .urls import (  # noqa: I001, E402
-    fetch, default_url_fetcher, path2url, ensure_url, url_is_absolute)
+from .urls import URLFetcher, default_url_fetcher, select_source  # noqa: I001, E402
 from .logger import LOGGER, PROGRESS_LOGGER  # noqa: E402
 # Some imports are at the end of the file (after the CSS class)
 # to work around circular imports.
@@ -150,9 +149,7 @@ class HTML:
         :term:`file objects <file object>`.
     :type url_fetcher: :term:`callable`
     :param url_fetcher:
-        A function or other callable with the same signature as
-        :func:`default_url_fetcher` called to fetch external resources such as
-        stylesheets and images. (See :ref:`URL Fetchers`.)
+        An instance of :class:`urls.URLFetcher`. (See :ref:`URL Fetchers`.)
     :param str media_type:
         The media type to use for ``@media``. Defaults to ``'print'``.
         **Note:** In some cases like ``HTML(string=foo)`` relative URLs will be
@@ -161,25 +158,24 @@ class HTML:
     """
     def __init__(self, guess=None, filename=None, url=None, file_obj=None,
                  string=None, encoding=None, base_url=None,
-                 url_fetcher=default_url_fetcher, media_type='print'):
+                 url_fetcher=None, media_type='print'):
         PROGRESS_LOGGER.info(
             'Step 1 - Fetching and parsing HTML - %s',
             guess or filename or url or
             getattr(file_obj, 'name', 'HTML string'))
         if isinstance(base_url, Path):
             base_url = str(base_url)
-        result = _select_source(
+        if url_fetcher is None:
+            url_fetcher = URLFetcher()
+        result = select_source(
             guess, filename, url, file_obj, string, base_url, url_fetcher)
-        with result as (source_type, source, base_url, protocol_encoding):
-            if isinstance(source, str):
-                result = tinyhtml5.parse(source, namespace_html_elements=False)
-            else:
-                kwargs = {'namespace_html_elements': False}
-                if protocol_encoding is not None:
-                    kwargs['transport_encoding'] = protocol_encoding
-                if encoding is not None:
-                    kwargs['override_encoding'] = encoding
-                result = tinyhtml5.parse(source, **kwargs)
+        with result as (file_obj, base_url, protocol_encoding, _):
+            kwargs = {'namespace_html_elements': False}
+            if protocol_encoding is not None:
+                kwargs['transport_encoding'] = protocol_encoding
+            if encoding is not None:
+                kwargs['override_encoding'] = encoding
+            result = tinyhtml5.parse(file_obj, **kwargs)
         self.base_url = _find_base_url(result, base_url)
         self.url_fetcher = url_fetcher
         self.media_type = media_type
@@ -286,28 +282,26 @@ class CSS:
     of :class:`HTML` objects.
 
     """
-    def __init__(self, guess=None, filename=None, url=None, file_obj=None,
-                 string=None, encoding=None, base_url=None,
-                 url_fetcher=default_url_fetcher, _check_mime_type=False,
+    def __init__(self, guess=None, filename=None, url=None, file_obj=None, string=None,
+                 encoding=None, base_url=None, url_fetcher=None, _check_mime_type=False,
                  media_type='print', font_config=None, counter_style=None,
                  color_profiles=None, matcher=None, page_rules=None, layers=None,
                  layer=None):
         PROGRESS_LOGGER.info(
             'Step 2 - Fetching and parsing CSS - %s',
             filename or url or getattr(file_obj, 'name', 'CSS string'))
-        result = _select_source(
-            guess, filename, url, file_obj, string,
-            base_url=base_url, url_fetcher=url_fetcher,
-            check_css_mime_type=_check_mime_type)
-        with result as (source_type, source, base_url, protocol_encoding):
-            if source_type == 'file_obj':
-                source = source.read()
-            if isinstance(source, str):
-                # unicode, no encoding
-                stylesheet = tinycss2.parse_stylesheet(source)
+        if url_fetcher is None:
+            url_fetcher = URLFetcher()
+        result = select_source(
+            guess, filename, url, file_obj, string, base_url=base_url,
+            url_fetcher=url_fetcher, check_css_mime_type=_check_mime_type)
+        with result as (file_obj, base_url, protocol_encoding, mime_type):
+            css = file_obj.read()
+            if isinstance(css, str):
+                stylesheet = tinycss2.parse_stylesheet(css)
             else:
-                stylesheet, encoding = tinycss2.parse_stylesheet_bytes(
-                    source, environment_encoding=encoding,
+                stylesheet, _ = tinycss2.parse_stylesheet_bytes(
+                    css, environment_encoding=encoding,
                     protocol_encoding=protocol_encoding)
         self.base_url = base_url
         self.matcher = matcher or cssselect2.Matcher()
@@ -347,10 +341,12 @@ class Attachment:
 
     """
     def __init__(self, guess=None, filename=None, url=None, file_obj=None,
-                 string=None, base_url=None, url_fetcher=default_url_fetcher,
-                 name=None, description=None, created=None, modified=None,
+                 string=None, base_url=None, url_fetcher=None, name=None,
+                 description=None, created=None, modified=None,
                  relationship='Unspecified'):
-        self.source = _select_source(
+        if url_fetcher is None:
+            url_fetcher = URLFetcher()
+        self.source = select_source(
             guess, filename, url, file_obj, string, base_url=base_url,
             url_fetcher=url_fetcher)
         self.name = name
@@ -372,70 +368,6 @@ class Attachment:
         self.modified = modified
 
 
-@contextlib.contextmanager
-def _select_source(guess=None, filename=None, url=None, file_obj=None,
-                   string=None, base_url=None, url_fetcher=default_url_fetcher,
-                   check_css_mime_type=False):
-    """If only one input is given, return it with normalized ``base_url``."""
-    if base_url is not None:
-        base_url = ensure_url(base_url)
-
-    selected_params = [
-        param for param in (guess, filename, url, file_obj, string) if
-        param is not None]
-    if len(selected_params) != 1:
-        source = ', '.join(selected_params) or 'nothing'
-        raise TypeError(f'Expected exactly one source, got {source}')
-    elif guess is not None:
-        if hasattr(guess, 'read'):
-            type_ = 'file_obj'
-        elif isinstance(guess, Path):
-            type_ = 'filename'
-        elif url_is_absolute(guess):
-            type_ = 'url'
-        else:
-            type_ = 'filename'
-        result = _select_source(
-            base_url=base_url, url_fetcher=url_fetcher,
-            check_css_mime_type=check_css_mime_type,
-            **{type_: guess})
-        with result as result:
-            yield result
-    elif filename is not None:
-        if base_url is None:
-            base_url = path2url(filename)
-        with open(filename, 'rb') as file_obj:
-            yield 'file_obj', file_obj, base_url, None
-    elif url is not None:
-        with fetch(url_fetcher, url) as result:
-            if check_css_mime_type and result['mime_type'] != 'text/css':
-                LOGGER.error(
-                    'Unsupported stylesheet type %s for %s',
-                    result['mime_type'], result['redirected_url'])
-                yield 'string', '', base_url, None
-            else:
-                proto_encoding = result.get('encoding')
-                if base_url is None:
-                    base_url = result.get('redirected_url', url)
-                if 'string' in result:
-                    yield 'string', result['string'], base_url, proto_encoding
-                else:
-                    yield (
-                        'file_obj', result['file_obj'], base_url,
-                        proto_encoding)
-    elif file_obj is not None:
-        if base_url is None:
-            # filesystem file-like objects have a 'name' attribute.
-            name = getattr(file_obj, 'name', None)
-            # Some streams have a .name like '<stdin>', not a filename.
-            if name and not name.startswith('<'):
-                base_url = ensure_url(name)
-        yield 'file_obj', file_obj, base_url, None
-    else:
-        assert string is not None
-        yield 'string', string, base_url, None
-
-
 # Work around circular imports.
 from .css import preprocess_stylesheet  # noqa: I001, E402
 from .html import (  # noqa: E402

weasyprint/__main__.py

@@ -4,14 +4,13 @@ import argparse
 import logging
 import platform
 import sys
-from functools import partial
 
 import pydyf
 
 from . import DEFAULT_OPTIONS, HTML, LOGGER, __version__
 from .pdf import VARIANTS
 from .text.ffi import pango
-from .urls import default_url_fetcher
+from .urls import URLFetcher
 
 
 class PrintInfo(argparse.Action):
@@ -33,117 +32,141 @@ class PrintInfo(argparse.Action):
 
 class Parser(argparse.ArgumentParser):
     def __init__(self, *args, **kwargs):
-        self._arguments = {}
+        self._groups = {None: {}}
         super().__init__(*args, **kwargs)
 
-    def add_argument(self, *args, **kwargs):
-        super().add_argument(*args, **kwargs)
+    def add_argument(self, *args, _group_name=None, **kwargs):
+        if _group_name is None:
+            super().add_argument(*args, **kwargs)
         key = args[-1].lstrip('-')
         kwargs['flags'] = args
         kwargs['positional'] = args[-1][0] != '-'
-        self._arguments[key] = kwargs
+        self._groups[_group_name][key] = kwargs
+
+    def add_argument_group(self, name, *args, **kwargs):
+        group = super().add_argument_group(name, *args, **kwargs)
+        self._groups[name] = {}
+        def add_argument(*args, **kwargs):
+            group._add_argument(*args, **kwargs)
+            self.add_argument(*args, _group_name=name, **kwargs)
+        group._add_argument = group.add_argument
+        group.add_argument = add_argument
+        return group
 
     @property
     def docstring(self):
-        self._arguments['help'] = self._arguments.pop('help')
+        self._groups[None].pop('help')
         data = []
-        for key, args in self._arguments.items():
-            data.append('.. option:: ')
-            action = args.get('action', 'store')
-            for flag in args['flags']:
-                data.append(flag)
-                if not args['positional'] and action in ('store', 'append'):
-                    data.append(f' <{key}>')
-                data.append(', ')
-            data[-1] = '\n\n'
-            data.append(f'  {args["help"][0].upper()}{args["help"][1:]}.\n\n')
-            if 'choices' in args:
-                choices = ", ".join(args['choices'])
-                data.append(f'  Possible choices: {choices}.\n\n')
-            if action == 'append':
-                data.append('  This option can be passed multiple times.\n\n')
+        for group, arguments in self._groups.items():
+            if not arguments:
+                continue
+            if group:
+                data.append(f'{group[0].title()}{group[1:]}\n')
+                data.append(f'{"~" * len(group)}\n\n')
+            for key, args in arguments.items():
+                data.append('.. option:: ')
+                action = args.get('action', 'store')
+                for flag in args['flags']:
+                    data.append(flag)
+                    if not args['positional'] and action in ('store', 'append'):
+                        data.append(f' <{key}>')
+                    data.append(', ')
+                data[-1] = '\n\n'
+                data.append(f'  {args["help"][0].upper()}{args["help"][1:]}.\n\n')
+                if 'choices' in args:
+                    choices = ", ".join(args['choices'])
+                    data.append(f'  Possible choices: {choices}.\n\n')
+                if action == 'append':
+                    data.append('  This option can be passed multiple times.\n\n')
         return ''.join(data)
 
 
 PARSER = Parser(prog='weasyprint', description='Render web pages to PDF.')
+PARSER.add_argument('input', help='URL or filename of the HTML input, or - for stdin')
+PARSER.add_argument('output', help='filename where output is written, or - for stdout')
 PARSER.add_argument(
-    'input', help='URL or filename of the HTML input, or - for stdin')
-PARSER.add_argument(
-    'output', help='filename where output is written, or - for stdout')
-PARSER.add_argument(
-    '-e', '--encoding', help='force the input character encoding')
+    '-i', '--info', action=PrintInfo, nargs=0, help='print system information and exit')
 PARSER.add_argument(
+    '--version', action='version', version=f'WeasyPrint version {__version__}',
+    help='print WeasyPrint’s version number and exit')
+
+group = PARSER.add_argument_group('rendering options')
+group.add_argument(
     '-s', '--stylesheet', action='append', dest='stylesheets',
     help='URL or filename for a user CSS stylesheet')
-PARSER.add_argument(
-    '-m', '--media-type',
-    help='media type to use for @media, defaults to print')
-PARSER.add_argument(
-    '-u', '--base-url',
-    help='base for relative URLs in the HTML input, defaults to the '
-    'input’s own filename or URL or the current directory for stdin')
-PARSER.add_argument(
+group.add_argument(
     '-a', '--attachment', action='append', dest='attachments',
     help='URL or filename of a file to attach to the PDF document')
-PARSER.add_argument('--pdf-identifier', help='PDF file identifier')
-PARSER.add_argument(
-    '--pdf-variant', choices=VARIANTS, help='PDF variant to generate')
-PARSER.add_argument('--pdf-version', help='PDF version number')
-PARSER.add_argument(
-    '--pdf-forms', action='store_true', help='include PDF forms')
-PARSER.add_argument(
-    '--pdf-tags', action='store_true', help='tag PDF for accessibility')
-PARSER.add_argument(
+group.add_argument(
+    '--attachment-relationship', action='append', dest='attachment_relationships',
+    help='Relationship of the attachment file to attach to the PDF')
+group.add_argument('--pdf-identifier', help='PDF file identifier')
+group.add_argument('--pdf-variant', choices=VARIANTS, help='PDF variant to generate')
+group.add_argument('--pdf-version', help='PDF version number')
+group.add_argument('--pdf-forms', action='store_true', help='include PDF forms')
+group.add_argument('--pdf-tags', action='store_true', help='tag PDF for accessibility')
+group.add_argument(
     '--uncompressed-pdf', action='store_true',
     help='do not compress PDF content, mainly for debugging purpose')
-PARSER.add_argument(
+group.add_argument(
+    '--xmp-metadata', action='append',
+    help='URL or filename of a file to include into the XMP metadata')
+group.add_argument(
     '--custom-metadata', action='store_true',
     help='include custom HTML meta tags in PDF metadata')
-PARSER.add_argument(
+group.add_argument(
     '-p', '--presentational-hints', action='store_true',
     help='follow HTML presentational hints')
-PARSER.add_argument(
-    '--srgb', action='store_true',
-    help='include sRGB color profile')
-PARSER.add_argument(
+group.add_argument('--srgb', action='store_true', help='include sRGB color profile')
+group.add_argument(
     '--optimize-images', action='store_true',
     help='optimize size of embedded images with no quality loss')
-PARSER.add_argument(
+group.add_argument(
     '-j', '--jpeg-quality', type=int,
     help='JPEG quality between 0 (worst) to 95 (best)')
-PARSER.add_argument(
+group.add_argument(
+    '-D', '--dpi', type=int,
+    help='set maximum resolution of images embedded in the PDF')
+group.add_argument(
     '--full-fonts', action='store_true',
     help='embed unmodified font files when possible')
-PARSER.add_argument(
-    '--hinting', action='store_true',
-    help='keep hinting information in embedded fonts')
-PARSER.add_argument(
+group.add_argument(
+    '--hinting', action='store_true', help='keep hinting information in embedded fonts')
+group.add_argument(
     '-c', '--cache-folder', dest='cache',
     help='store cache on disk instead of memory, folder is '
     'created if needed and cleaned after the PDF is generated')
-PARSER.add_argument(
-    '-D', '--dpi', type=int,
-    help='set maximum resolution of images embedded in the PDF')
-PARSER.add_argument(
+
+group = PARSER.add_argument_group('HTML options')
+group.add_argument('-e', '--encoding', help='force the input character encoding')
+group.add_argument(
+    '-m', '--media-type', help='media type to use for @media, defaults to print',
+    default='print')
+group.add_argument(
+    '-u', '--base-url',
+    help='base for relative URLs in the HTML input, defaults to the '
+    'input’s own filename or URL or the current directory for stdin')
+
+group = PARSER.add_argument_group('URL fetcher options')
+group.add_argument(
+    '-t', '--timeout', type=int, help='set timeout in seconds for HTTP requests')
+group.add_argument(
+    '--allowed-protocols', dest='allowed_protocols',
+    help='only authorize comma-separated list of protocols for fetching URLs')
+group.add_argument(
+    '--no-http-redirects', action='store_true', help='do not follow HTTP redirects')
+group.add_argument(
+    '--fail-on-http-errors', action='store_true',
+    help='abort document rendering on any HTTP error')
+
+group = PARSER.add_argument_group('command-line logging options')
+group.add_argument(
     '-v', '--verbose', action='store_true',
     help='show warnings and information messages')
-PARSER.add_argument(
+group.add_argument(
     '-d', '--debug', action='store_true', help='show debugging messages')
-PARSER.add_argument(
-    '-q', '--quiet', action='store_true', help='hide logging messages')
-PARSER.add_argument(
-    '--version', action='version',
-    version=f'WeasyPrint version {__version__}',
-    help='print WeasyPrint’s version number and exit')
-PARSER.add_argument(
-    '-i', '--info', action=PrintInfo, nargs=0,
-    help='print system information and exit')
-PARSER.add_argument(
-    '-t', '--timeout', type=int,
-    help='set timeout in seconds for HTTP requests')
-PARSER.add_argument(
-    '--allowed-protocols', dest='allowed_protocols',
-    help='only authorize comma-separated list of protocols for fetching URLs')
+group.add_argument('-q', '--quiet', action='store_true', help='hide logging messages')
+
 PARSER.set_defaults(**DEFAULT_OPTIONS)
 
 
@@ -171,13 +194,17 @@ def main(argv=None, stdout=None, stdin=None, HTML=HTML):  # noqa: N803
     else:
         output = args.output
 
-    url_fetcher = default_url_fetcher
+    fetcher_args = {}
     if args.timeout is not None:
-        url_fetcher = partial(default_url_fetcher, timeout=args.timeout)
+        fetcher_args['timeout'] = args.timeout
     if args.allowed_protocols is not None:
-        protocols = {
+        fetcher_args['allowed_protocols'] = {
             protocol.strip().lower() for protocol in args.allowed_protocols.split(',')}
-        url_fetcher = partial(url_fetcher, allowed_protocols=protocols)
+    if args.no_http_redirects:
+        fetcher_args['allow_redirects'] = False
+    if args.fail_on_http_errors:
+        fetcher_args['fail_on_errors'] = True
+    url_fetcher = URLFetcher(**fetcher_args)
 
     options = {
         key: value for key, value in vars(args).items() if key in DEFAULT_OPTIONS}

weasyprint/css/__init__.py

@@ -14,7 +14,6 @@ on other functions in this module.
 
 import math
 from collections import namedtuple
-from io import BytesIO
 from itertools import groupby
 from logging import DEBUG, WARNING
 from math import inf
@@ -295,11 +294,10 @@ def find_stylesheets(wrapper_element, device_media_type, url_fetcher, base_url,
             if href is not None:
                 try:
                     yield CSS(
-                        url=href, url_fetcher=url_fetcher,
-                        _check_mime_type=True, media_type=device_media_type,
+                        url=href, url_fetcher=url_fetcher, media_type=device_media_type,
                         font_config=font_config, counter_style=counter_style,
                         color_profiles=color_profiles, page_rules=page_rules,
-                        layers=layers)
+                        layers=layers, _check_mime_type=True)
                 except URLFetchingError as exception:
                     LOGGER.error('Failed to load stylesheet at %s: %s', href, exception)
                     LOGGER.debug('Error while loading stylesheet:', exc_info=exception)
@@ -781,7 +779,6 @@ def resolve_math(token, computed=None, property_name=None, refer_to=None):
         return
 
     args = []
-    original_token = token
     function = Function(token)
     if function.name is None:
         return
@@ -797,7 +794,7 @@ def resolve_math(token, computed=None, property_name=None, refer_to=None):
     if function.name == 'calc':
         result = _resolve_calc_sum(computed, args[0], property_name, refer_to)
         if result is None:
-            return original_token
+            return
         else:
             return tokenize(result)
 
@@ -1196,10 +1193,10 @@ class ComputedStyle(dict):
                 try:
                     token = resolve_math(function, self, key)
                 except PercentageInMath:
-                    token = None
-                if token is None:
                     solved_tokens.append(function)
                 else:
+                    if token is None:
+                        raise Exception
                     solved_tokens.append(token)
                 original_key = key.replace('_', '-')
                 value = validate_non_shorthand(solved_tokens, original_key)[0][1]
@@ -1645,13 +1642,9 @@ def preprocess_stylesheet(device_media_type, base_url, stylesheet_rules, url_fet
                     rule.source_column)
                 continue
 
-            with fetch(url_fetcher, descriptors['src'][1]) as result:
-                if 'string' in result:
-                    file_object = BytesIO(result['string'])
-                else:
-                    file_object = result['file_obj']
+            with fetch(url_fetcher, descriptors['src'][1]) as response:
                 try:
-                    color_profile = ColorProfile(file_object, descriptors)
+                    color_profile = ColorProfile(response, descriptors)
                 except BaseException:
                     LOGGER.warning(
                         'Invalid profile file for profile named %r, the whole '

weasyprint/css/functions.py

@@ -68,6 +68,11 @@ def check_attr(token, allowed_type=None):
         name_and_type, fallback = parts[0], ''
     elif len(parts) == 2:
         name_and_type, fallback = parts
+        # TODO: support fallbacks with multiple tokens and follow type.
+        if len(fallback) >= 1 and fallback[0].type == 'string':
+            fallback = fallback[0].value
+        else:
+            fallback = ''
     else:
         return
 

weasyprint/css/html5_ua.css

@@ -124,7 +124,7 @@ dir, menu, ul { list-style-type: disc }
 :is(dir, menu, ol, ul) ul { list-style-type: circle }
 :is(dir, menu, ol, ul) :is(dir, menu, ol, ul) ul { list-style-type: square }
 
-::marker { font-variant-numeric: tabular-nums }
+::marker { font-variant-numeric: tabular-nums; white-space: pre; text-transform: none }
 
 [dir=ltr i] { direction: ltr }
 [dir=rtl i] { direction: rtl }

weasyprint/css/tokens.py

@@ -432,7 +432,10 @@ def get_angle(token):
         token = resolve_math(token) or token
     except (PercentageInMath, FontUnitInMath):
         return
-    if token.type == 'dimension':
+    if token.type == 'number' and token.value == 0:
+        # Legacy syntax: https://drafts.csswg.org/css-values-4/#angles.
+        return 0
+    elif token.type == 'dimension':
         factor = ANGLE_TO_RADIANS.get(token.unit.lower())
         if factor is not None:
             return token.value * factor

weasyprint/css/validation/properties.py

@@ -137,7 +137,7 @@ def other_colors(token):
 def outline_color(token):
     if get_keyword(token) == 'invert':
         return 'currentcolor'
-    else:
+    elif parse_color(token):
         return token
 
 
@@ -161,7 +161,7 @@ def color(token):
     result = parse_color(token)
     if result == 'currentcolor':
         return 'inherit'
-    else:
+    elif result:
         return token
 
 
@@ -849,7 +849,7 @@ def font_feature_settings(tokens):
             tokens, token = tokens[:-1], tokens[-1]
             if token.type == 'ident':
                 value = {'on': 1, 'off': 0}.get(token.value)
-            elif number := get_number(token, negative=False):
+            elif number := get_number(token, negative=False, integer=True):
                 value = number.value
         elif len(tokens) == 1:
             value = 1
@@ -1211,7 +1211,7 @@ def text_decoration_thickness(token):
     """``text-decoration-thickness`` property validation."""
     if length := get_length(token, percentage=True):
         return length
-    elif keyword := get_keyword(token) in ('auto', 'from-font'):
+    elif (keyword := get_keyword(token)) in ('auto', 'from-font'):
         return keyword