Merge pull request #2257 from Kozea/html5

Use tinyhtml5 instead of html5lib

docs/first_steps.rst

@@ -13,7 +13,7 @@ WeasyPrint |version| depends on:
 * Pango_ ≥ 1.44.0
 * pydyf_ ≥ 0.10.0
 * CFFI_ ≥ 0.6
-* html5lib_ ≥ 1.1
+* tinyhtml5_ ≥ 2.0.0b1
 * tinycss2_ ≥ 1.3.0
 * cssselect2_ ≥ 0.1
 * Pyphen_ ≥ 0.9.1
@@ -23,8 +23,8 @@ WeasyPrint |version| depends on:
 .. _Python: https://www.python.org/
 .. _Pango: https://pango.gnome.org/
 .. _CFFI: https://cffi.readthedocs.io/
-.. _html5lib: https://html5lib.readthedocs.io/
 .. _pydyf: https://doc.courtbouillon.org/pydyf/
+.. _tinyhtml5: https://doc.courtbouillon.org/tinyhtml5/
 .. _tinycss2: https://doc.courtbouillon.org/tinycss2/
 .. _cssselect2: https://doc.courtbouillon.org/cssselect2/
 .. _Pyphen: https://pyphen.org/

pyproject.toml

@@ -14,7 +14,7 @@ license = {file = 'LICENSE'}
 dependencies = [
   'pydyf >=0.11.0,<0.12',
   'cffi >=0.6,<2',
-  'html5lib >=1.1,<2',
+  'tinyhtml5 >=2.0.0b1,<3',
   'tinycss2 >=1.3.0,<2',
   'cssselect2 >=0.1,<0.8',
   'Pyphen >=0.9.1,<0.16',

weasyprint/__init__.py

@@ -12,8 +12,8 @@
 from urllib.parse import urljoin
 
 import cssselect2
-import html5lib
 import tinycss2
+import tinyhtml5
 
 VERSION = __version__ = '62.3'
 
@@ -111,7 +111,7 @@ def _find_base_url(html_document, fallback_base_url):
 
 
 class HTML:
-    """HTML document parsed by html5lib.
+    """HTML document parsed by tinyhtml5.
 
     You can just create an instance with a positional argument:
     ``doc = HTML(something)``
@@ -169,14 +169,14 @@ def __init__(self, guess=None, filename=None, url=None, file_obj=None,
             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 = html5lib.parse(source, namespaceHTMLElements=False)
+                result = tinyhtml5.parse(source, namespace_html_elements=False)
             else:
-                kwargs = {'namespaceHTMLElements': False}
+                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 = html5lib.parse(source, **kwargs)
+                result = tinyhtml5.parse(source, **kwargs)
         self.base_url = _find_base_url(result, base_url)
         self.url_fetcher = url_fetcher
         self.media_type = media_type

weasyprint/urls.py

@@ -45,8 +45,8 @@ def close(self):
         GzipFile.close(self)
         self.fileobj_to_close.close()
 
-    # Inform html5lib to not rely on these:
-    seek = tell = None
+    def seekable(self):
+        return False
 
 
 def iri_to_uri(url):