[tests] move utilities and static data into dedicated modules and remove `html5lib` (#12173)

Since #12168, HTML files are now XML compliant, hence ``html5lib`` is no more needed as a testing dependencies.

pyproject.toml

@@ -91,7 +91,6 @@ lint = [
 ]
 test = [
     "pytest>=6.0",
-    "html5lib",
     "defusedxml>=0.7.1", # for secure XML/HTML parsing
     "cython>=3.0",
     "setuptools>=67.0",  # for Cython compilation

sphinx/testing/util.py

@@ -7,12 +7,11 @@ __all__ = ('SphinxTestApp', 'SphinxTestAppWrapperForSkipBuilding')
 import contextlib
 import os
 import sys
-import warnings
 from io import StringIO
 from types import MappingProxyType
 from typing import TYPE_CHECKING
-from xml.etree import ElementTree
 
+from defusedxml.ElementTree import parse as xml_parse
 from docutils import nodes
 from docutils.parsers.rst import directives, roles
 
@@ -26,6 +25,7 @@ if TYPE_CHECKING:
     from collections.abc import Mapping
     from pathlib import Path
     from typing import Any, Final
+    from xml.etree.ElementTree import ElementTree
 
     from docutils.nodes import Node
 
@@ -70,10 +70,10 @@ def assert_node(node: Node, cls: Any = None, xpath: str = "", **kwargs: Any) ->
                 f'The node{xpath}[{key}] is not {value!r}: {node[key]!r}'
 
 
-def etree_parse(path: str) -> Any:
+# keep this to restrict the API usage and to have a correct return type
-    with warnings.catch_warnings(record=False):
+def etree_parse(path: str | os.PathLike[str]) -> ElementTree:
-        warnings.filterwarnings("ignore", category=DeprecationWarning)
+    """Parse a file into a (safe) XML element tree."""
-        return ElementTree.parse(path)  # NoQA: S314  # using known data in tests
+    return xml_parse(path)
 
 
 class SphinxTestApp(sphinx.application.Sphinx):

tests/test_builders/conftest.py

@@ -3,26 +3,26 @@ from __future__ import annotations
 from typing import TYPE_CHECKING
 
 import pytest
-from html5lib import HTMLParser
+
+from sphinx.testing.util import etree_parse
 
 if TYPE_CHECKING:
     from collections.abc import Callable, Generator
     from pathlib import Path
-    from xml.etree.ElementTree import Element
+    from xml.etree.ElementTree import ElementTree
 
-etree_cache: dict[Path, Element] = {}
+_etree_cache: dict[Path, ElementTree] = {}
 
 
-def _parse(fname: Path) -> Element:
+def _parse(path: Path) -> ElementTree:
-    if fname in etree_cache:
+    if path in _etree_cache:
-        return etree_cache[fname]
+        return _etree_cache[path]
-    with fname.open('rb') as fp:
+
-        etree = HTMLParser(namespaceHTMLElements=False).parse(fp)
+    _etree_cache[path] = tree = etree_parse(path)
-        etree_cache[fname] = etree
+    return tree
-        return etree
 
 
 @pytest.fixture(scope='package')
-def cached_etree_parse() -> Generator[Callable[[Path], Element], None, None]:
+def cached_etree_parse() -> Generator[Callable[[Path], ElementTree], None, None]:
     yield _parse
-    etree_cache.clear()
+    _etree_cache.clear()

tests/test_builders/test_build_html.py

@@ -12,43 +12,8 @@ from sphinx.errors import ConfigError
 from sphinx.util.console import strip_colors
 from sphinx.util.inventory import InventoryFile
 
-FIGURE_CAPTION = ".//figure/figcaption/p"
+from tests.test_builders.xpath_data import FIGURE_CAPTION
-
+from tests.test_builders.xpath_util import check_xpath
-
-def check_xpath(etree, fname, path, check, be_found=True):
-    nodes = list(etree.findall(path))
-    if check is None:
-        assert nodes == [], ('found any nodes matching xpath '
-                             f'{path!r} in file {fname}')
-        return
-    else:
-        assert nodes != [], ('did not find any node matching xpath '
-                             f'{path!r} in file {fname}')
-    if callable(check):
-        check(nodes)
-    elif not check:
-        # only check for node presence
-        pass
-    else:
-        def get_text(node):
-            if node.text is not None:
-                # the node has only one text
-                return node.text
-            else:
-                # the node has tags and text; gather texts just under the node
-                return ''.join(n.tail or '' for n in node)
-
-        rex = re.compile(check)
-        if be_found:
-            if any(rex.search(get_text(node)) for node in nodes):
-                return
-        else:
-            if all(not rex.search(get_text(node)) for node in nodes):
-                return
-
-        msg = (f'{check!r} not found in any node matching '
-               f'{path!r} in file {fname}: {[node.text for node in nodes]!r}')
-        raise AssertionError(msg)
 
 
 def test_html4_error(make_app, tmp_path):

tests/test_builders/test_build_html_5_output.py

@@ -4,7 +4,7 @@ import re
 
 import pytest
 
-from tests.test_builders.test_build_html import check_xpath
+from tests.test_builders.xpath_util import check_xpath
 
 
 def tail_check(check):
@@ -128,7 +128,7 @@ def tail_check(check):
     # ``seealso`` directive
     ('markup.html', ".//div/p[@class='admonition-title']", 'See also'),
     # a ``hlist`` directive
-    ('markup.html', ".//table[@class='hlist']/tbody/tr/td/ul/li/p", '^This$'),
+    ('markup.html', ".//table[@class='hlist']/tr/td/ul/li/p", '^This$'),
     # a ``centered`` directive
     ('markup.html', ".//p[@class='centered']/strong", 'LICENSE'),
     # a glossary

tests/test_builders/test_build_html_numfig.py

@@ -5,7 +5,8 @@ import re
 
 import pytest
 
-from tests.test_builders.test_build_html import FIGURE_CAPTION, check_xpath
+from tests.test_builders.xpath_data import FIGURE_CAPTION
+from tests.test_builders.xpath_util import check_xpath
 
 
 @pytest.mark.sphinx('html', testroot='numfig')

tests/test_builders/test_build_html_tocdepth.py

@@ -2,7 +2,7 @@
 
 import pytest
 
-from tests.test_builders.test_build_html import check_xpath
+from tests.test_builders.xpath_util import check_xpath
 
 
 @pytest.mark.parametrize(("fname", "path", "check", "be_found"), [

tests/test_builders/xpath_data.py

@@ -0,0 +1,8 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from typing import Final
+
+FIGURE_CAPTION: Final[str] = ".//figure/figcaption/p"

tests/test_builders/xpath_util.py

@@ -0,0 +1,79 @@
+from __future__ import annotations
+
+import re
+import textwrap
+from typing import TYPE_CHECKING
+from xml.etree.ElementTree import tostring
+
+if TYPE_CHECKING:
+    import os
+    from collections.abc import Callable, Iterable, Sequence
+    from xml.etree.ElementTree import Element, ElementTree
+
+
+def _get_text(node: Element) -> str:
+    if node.text is not None:
+        # the node has only one text
+        return node.text
+
+    # the node has tags and text; gather texts just under the node
+    return ''.join(n.tail or '' for n in node)
+
+
+def _prettify(nodes: Iterable[Element]) -> str:
+    def pformat(node: Element) -> str:
+        return tostring(node, encoding='unicode', method='html')
+
+    return ''.join(f'(i={index}) {pformat(node)}\n' for index, node in enumerate(nodes))
+
+
+def check_xpath(
+    etree: ElementTree,
+    filename: str | os.PathLike[str],
+    xpath: str,
+    check: str | re.Pattern[str] | Callable[[Sequence[Element]], None] | None,
+    be_found: bool = True,
+    *,
+    min_count: int = 1,
+) -> None:
+    """Check that one or more nodes satisfy a predicate.
+
+    :param etree: The element tree.
+    :param filename: The element tree source name (for errors only).
+    :param xpath: An XPath expression to use.
+    :param check: Optional regular expression or a predicate the nodes must validate.
+    :param be_found: If false, negate the predicate.
+    :param min_count: Minimum number of nodes expected to satisfy the predicate.
+
+    * If *check* is empty (``''``), only the minimum count is checked.
+    * If *check* is ``None``, no node should satisfy the XPath expression.
+    """
+    nodes = etree.findall(xpath)
+    assert isinstance(nodes, list)
+
+    if check is None:
+        # use == to have a nice pytest diff
+        assert nodes == [], f'found nodes matching xpath {xpath!r} in file {filename}'
+        return
+
+    assert len(nodes) >= min_count, (f'expecting at least {min_count} node(s) '
+                                     f'to satisfy {xpath!r} in file {filename}')
+
+    if check == '':
+        return
+
+    if callable(check):
+        check(nodes)
+        return
+
+    rex = re.compile(check)
+    if be_found:
+        if any(rex.search(_get_text(node)) for node in nodes):
+            return
+    else:
+        if all(not rex.search(_get_text(node)) for node in nodes):
+            return
+
+    ctx = textwrap.indent(_prettify(nodes), ' ' * 2)
+    msg = f'{check!r} not found in any node matching {xpath!r} in file {filename}:\n{ctx}'
+    raise AssertionError(msg)

tests/test_domains/test_domain_std.py

@@ -5,7 +5,6 @@ from unittest import mock
 import pytest
 from docutils import nodes
 from docutils.nodes import definition, definition_list, definition_list_item, term
-from html5lib import HTMLParser
 
 from sphinx import addnodes
 from sphinx.addnodes import (
@@ -20,7 +19,7 @@ from sphinx.addnodes import (
 )
 from sphinx.domains.std import StandardDomain
 from sphinx.testing import restructuredtext
-from sphinx.testing.util import assert_node
+from sphinx.testing.util import assert_node, etree_parse
 
 
 def test_process_doc_handle_figure_caption():
@@ -375,9 +374,11 @@ def test_productionlist(app, status, warning):
     assert warnings[-1] == ''
     assert "Dup2.rst:4: WARNING: duplicate token description of Dup, other instance in Dup1" in warnings[0]
 
-    with (app.outdir / 'index.html').open('rb') as f:
+    etree = etree_parse(app.outdir / 'index.html')
-        etree = HTMLParser(namespaceHTMLElements=False).parse(f)
+    nodes = list(etree.iter('ul'))
-    ul = list(etree.iter('ul'))[1]
+    assert len(nodes) >= 2
+
+    ul = nodes[1]
     cases = []
     for li in list(ul):
         assert len(list(li)) == 1

tests/test_extensions/autodoc_util.py

@@ -0,0 +1,33 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+from unittest.mock import Mock
+
+# NEVER import those objects from sphinx.ext.autodoc directly
+from sphinx.ext.autodoc.directive import DocumenterBridge, process_documenter_options
+from sphinx.util.docutils import LoggingReporter
+
+if TYPE_CHECKING:
+    from typing import Any
+
+    from docutils.statemachine import StringList
+
+    from sphinx.application import Sphinx
+
+
+def do_autodoc(
+    app: Sphinx,
+    objtype: str,
+    name: str,
+    options: dict[str, Any] | None = None,
+) -> StringList:
+    options = {} if options is None else options.copy()
+    app.env.temp_data.setdefault('docname', 'index')  # set dummy docname
+    doccls = app.registry.documenters[objtype]
+    docoptions = process_documenter_options(doccls, app.config, options)
+    state = Mock()
+    state.document.settings.tab_width = 8
+    bridge = DocumenterBridge(app.env, LoggingReporter(''), docoptions, 1, state)
+    documenter = doccls(bridge, name)
+    documenter.generate()
+    return bridge.result

tests/test_extensions/test_ext_autodoc.py

@@ -20,8 +20,8 @@ from docutils.statemachine import ViewList
 
 from sphinx import addnodes
 from sphinx.ext.autodoc import ALL, ModuleLevelDocumenter, Options
-from sphinx.ext.autodoc.directive import DocumenterBridge, process_documenter_options
+
-from sphinx.util.docutils import LoggingReporter
+from tests.test_extensions.autodoc_util import do_autodoc
 
 try:
     # Enable pyximport to test cython module
@@ -34,20 +34,6 @@ if TYPE_CHECKING:
     from typing import Any
 
 
-def do_autodoc(app, objtype, name, options=None):
-    options = {} if options is None else options.copy()
-    app.env.temp_data.setdefault('docname', 'index')  # set dummy docname
-    doccls = app.registry.documenters[objtype]
-    docoptions = process_documenter_options(doccls, app.config, options)
-    state = Mock()
-    state.document.settings.tab_width = 8
-    bridge = DocumenterBridge(app.env, LoggingReporter(''), docoptions, 1, state)
-    documenter = doccls(bridge, name)
-    documenter.generate()
-
-    return bridge.result
-
-
 def make_directive_bridge(env):
     options = Options(
         inherited_members=False,
@@ -82,23 +68,6 @@ def make_directive_bridge(env):
 processed_signatures = []
 
 
-def process_signature(app, what, name, obj, options, args, retann):
-    processed_signatures.append((what, name))
-    if name == 'bar':
-        return '42', None
-    return None
-
-
-def skip_member(app, what, name, obj, skip, options):
-    if name in ('__special1__', '__special2__'):
-        return skip
-    if name.startswith('__'):
-        return True
-    if name == 'skipmeth':
-        return True
-    return None
-
-
 def test_parse_name(app):
     def verify(objtype, name, result):
         inst = app.registry.documenters[objtype](directive, name)
@@ -139,6 +108,21 @@ def test_parse_name(app):
 
 
 def test_format_signature(app):
+    def process_signature(app, what, name, obj, options, args, retann):
+        processed_signatures.append((what, name))
+        if name == 'bar':
+            return '42', None
+        return None
+
+    def skip_member(app, what, name, obj, skip, options):
+        if name in ('__special1__', '__special2__'):
+            return skip
+        if name.startswith('__'):
+            return True
+        if name == 'skipmeth':
+            return True
+        return None
+
     app.connect('autodoc-process-signature', process_signature)
     app.connect('autodoc-skip-member', skip_member)
 

tests/test_extensions/test_ext_autodoc_autoattribute.py

@@ -6,7 +6,7 @@ source file translated by test_build.
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_autodoc_autoclass.py

@@ -11,7 +11,7 @@ from typing import Union
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_autodoc_autodata.py

@@ -6,7 +6,7 @@ source file translated by test_build.
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_autodoc_autofunction.py

@@ -6,7 +6,7 @@ source file translated by test_build.
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_autodoc_automodule.py

@@ -8,7 +8,7 @@ import sys
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_autodoc_autoproperty.py

@@ -6,7 +6,7 @@ source file translated by test_build.
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_autodoc_configs.py

@@ -8,7 +8,7 @@ import pytest
 
 from sphinx.testing import restructuredtext
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 IS_PYPY = platform.python_implementation() == 'PyPy'
 

tests/test_extensions/test_ext_autodoc_events.py

@@ -4,7 +4,7 @@ import pytest
 
 from sphinx.ext.autodoc import between, cut_lines
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_autodoc_preserve_defaults.py

@@ -2,7 +2,7 @@
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc',

tests/test_extensions/test_ext_autodoc_private_members.py

@@ -3,7 +3,7 @@
 
 import pytest
 
-from tests.test_extensions.test_ext_autodoc import do_autodoc
+from tests.test_extensions.autodoc_util import do_autodoc
 
 
 @pytest.mark.sphinx('html', testroot='ext-autodoc')

tests/test_extensions/test_ext_intersphinx.py

@@ -20,7 +20,7 @@ from sphinx.ext.intersphinx import (
 from sphinx.ext.intersphinx import setup as intersphinx_setup
 from sphinx.util.console import strip_colors
 
-from tests.test_util.test_util_inventory import inventory_v2, inventory_v2_not_having_version
+from tests.test_util.intersphinx_data import INVENTORY_V2, INVENTORY_V2_NO_VERSION
 from tests.utils import http_server
 
 
@@ -92,7 +92,7 @@ def test_fetch_inventory_redirection(_read_from_url, InventoryFile, app, status,
 
 def test_missing_reference(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {
         'https://docs.python.org/': str(inv_file),
         'py3k': ('https://docs.python.org/py3k/', str(inv_file)),
@@ -170,7 +170,7 @@ def test_missing_reference(tmp_path, app, status, warning):
 
 def test_missing_reference_pydomain(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {
         'https://docs.python.org/': str(inv_file),
     })
@@ -200,7 +200,7 @@ def test_missing_reference_pydomain(tmp_path, app, status, warning):
 
 def test_missing_reference_stddomain(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {
         'cmd': ('https://docs.python.org/', str(inv_file)),
     })
@@ -251,7 +251,7 @@ def test_missing_reference_stddomain(tmp_path, app, status, warning):
 @pytest.mark.sphinx('html', testroot='ext-intersphinx-cppdomain')
 def test_missing_reference_cppdomain(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {
         'https://docs.python.org/': str(inv_file),
     })
@@ -277,7 +277,7 @@ def test_missing_reference_cppdomain(tmp_path, app, status, warning):
 
 def test_missing_reference_jsdomain(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {
         'https://docs.python.org/': str(inv_file),
     })
@@ -301,7 +301,7 @@ def test_missing_reference_jsdomain(tmp_path, app, status, warning):
 
 def test_missing_reference_disabled_domain(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {
         'inv': ('https://docs.python.org/', str(inv_file)),
     })
@@ -363,7 +363,7 @@ def test_missing_reference_disabled_domain(tmp_path, app, status, warning):
 
 def test_inventory_not_having_version(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2_not_having_version)
+    inv_file.write_bytes(INVENTORY_V2_NO_VERSION)
     set_config(app, {
         'https://docs.python.org/': str(inv_file),
     })
@@ -385,7 +385,7 @@ def test_load_mappings_warnings(tmp_path, app, status, warning):
     identifiers are not string
     """
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {
         'https://docs.python.org/': str(inv_file),
         'py3k': ('https://docs.python.org/py3k/', str(inv_file)),
@@ -406,7 +406,7 @@ def test_load_mappings_warnings(tmp_path, app, status, warning):
 
 def test_load_mappings_fallback(tmp_path, app, status, warning):
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     set_config(app, {})
 
     # connect to invalid path
@@ -505,7 +505,7 @@ def test_inspect_main_noargs(capsys):
 def test_inspect_main_file(capsys, tmp_path):
     """inspect_main interface, with file argument"""
     inv_file = tmp_path / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
 
     inspect_main([str(inv_file)])
 
@@ -520,7 +520,7 @@ def test_inspect_main_url(capsys):
         def do_GET(self):
             self.send_response(200, "OK")
             self.end_headers()
-            self.wfile.write(inventory_v2)
+            self.wfile.write(INVENTORY_V2)
 
         def log_message(*args, **kwargs):
             # Silenced.
@@ -539,7 +539,7 @@ def test_inspect_main_url(capsys):
 @pytest.mark.sphinx('html', testroot='ext-intersphinx-role')
 def test_intersphinx_role(app, warning):
     inv_file = app.srcdir / 'inventory'
-    inv_file.write_bytes(inventory_v2)
+    inv_file.write_bytes(INVENTORY_V2)
     app.config.intersphinx_mapping = {
         'inv': ('https://example.org/', str(inv_file)),
     }

tests/test_extensions/test_ext_napoleon_docstring.py

@@ -9,7 +9,6 @@ from textwrap import dedent
 from unittest import mock
 
 import pytest
-from html5lib import HTMLParser
 
 from sphinx.ext.intersphinx import load_mappings, normalize_intersphinx_mapping
 from sphinx.ext.napoleon import Config
@@ -21,6 +20,7 @@ from sphinx.ext.napoleon.docstring import (
     _token_type,
     _tokenize_type_spec,
 )
+from sphinx.testing.util import etree_parse
 
 from tests.test_extensions.ext_napoleon_pep526_data_google import PEP526GoogleClass
 from tests.test_extensions.ext_napoleon_pep526_data_numpy import PEP526NumpyClass
@@ -2684,8 +2684,7 @@ int py:class 1 int.html -
 
     app.build(force_all=True)
 
-    buffer = (app.outdir / 'index.html').read_bytes()
+    etree = etree_parse(app.outdir / 'index.html')
-    etree = HTMLParser(namespaceHTMLElements=False).parse(buffer)
 
     for name, typename in product(('keyword', 'kwarg', 'kwparam'), ('paramtype', 'kwtype')):
         param = f'{name}_{typename}'

tests/test_markup/test_smartquotes.py

@@ -1,7 +1,8 @@
 """Test smart quotes."""
 
 import pytest
-from html5lib import HTMLParser
+
+from sphinx.testing.util import etree_parse
 
 
 @pytest.mark.sphinx(buildername='html', testroot='smartquotes', freshenv=True)
@@ -16,9 +17,7 @@ def test_basic(app, status, warning):
 def test_literals(app, status, warning):
     app.build()
 
-    with (app.outdir / 'literals.html').open(encoding='utf-8') as html_file:
+    etree = etree_parse(app.outdir / 'literals.html')
-        etree = HTMLParser(namespaceHTMLElements=False).parse(html_file)
-
     for code_element in etree.iter('code'):
         code_text = ''.join(code_element.itertext())
 

tests/test_util/intersphinx_data.py

@@ -0,0 +1,52 @@
+from __future__ import annotations
+
+import zlib
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from typing import Final
+
+INVENTORY_V1: Final[bytes] = b'''\
+# Sphinx inventory version 1
+# Project: foo
+# Version: 1.0
+module mod foo.html
+module.cls class foo.html
+'''
+
+INVENTORY_V2: Final[bytes] = b'''\
+# Sphinx inventory version 2
+# Project: foo
+# Version: 2.0
+# The remainder of this file is compressed with zlib.
+''' + zlib.compress(b'''\
+module1 py:module 0 foo.html#module-module1 Long Module desc
+module2 py:module 0 foo.html#module-$ -
+module1.func py:function 1 sub/foo.html#$ -
+module1.Foo.bar py:method 1 index.html#foo.Bar.baz -
+CFunc c:function 2 cfunc.html#CFunc -
+std cpp:type 1 index.html#std -
+std::uint8_t cpp:type 1 index.html#std_uint8_t -
+foo::Bar cpp:class 1 index.html#cpp_foo_bar -
+foo::Bar::baz cpp:function 1 index.html#cpp_foo_bar_baz -
+foons cpp:type 1 index.html#foons -
+foons::bartype cpp:type 1 index.html#foons_bartype -
+a term std:term -1 glossary.html#term-a-term -
+ls.-l std:cmdoption 1 index.html#cmdoption-ls-l -
+docname std:doc -1 docname.html -
+foo js:module 1 index.html#foo -
+foo.bar js:class 1 index.html#foo.bar -
+foo.bar.baz js:method 1 index.html#foo.bar.baz -
+foo.bar.qux js:data 1 index.html#foo.bar.qux -
+a term including:colon std:term -1 glossary.html#term-a-term-including-colon -
+The-Julia-Domain std:label -1 write_inventory/#$ The Julia Domain
+''')
+
+INVENTORY_V2_NO_VERSION: Final[bytes] = b'''\
+# Sphinx inventory version 2
+# Project: foo
+# Version:
+# The remainder of this file is compressed with zlib.
+''' + zlib.compress(b'''\
+module1 py:module 0 foo.html#module-module1 Long Module desc
+''')

tests/test_util/test_util_inspect.py

@@ -223,168 +223,140 @@ def test_signature_partialmethod():
 
 
 def test_signature_annotations():
-    from tests.test_util.typing_test_data import (
+    import tests.test_util.typing_test_data as mod
-        Node,
-        f0,
-        f1,
-        f2,
-        f3,
-        f4,
-        f5,
-        f6,
-        f7,
-        f8,
-        f9,
-        f10,
-        f11,
-        f12,
-        f13,
-        f14,
-        f15,
-        f16,
-        f17,
-        f18,
-        f19,
-        f20,
-        f21,
-        f22,
-        f23,
-        f24,
-        f25,
-    )
 
     # Class annotations
-    sig = inspect.signature(f0)
+    sig = inspect.signature(mod.f0)
     assert stringify_signature(sig) == '(x: int, y: numbers.Integral) -> None'
 
     # Generic types with concrete parameters
-    sig = inspect.signature(f1)
+    sig = inspect.signature(mod.f1)
     assert stringify_signature(sig) == '(x: list[int]) -> typing.List[int]'
 
     # TypeVars and generic types with TypeVars
-    sig = inspect.signature(f2)
+    sig = inspect.signature(mod.f2)
     assert stringify_signature(sig) == ('(x: typing.List[tests.test_util.typing_test_data.T],'
                                         ' y: typing.List[tests.test_util.typing_test_data.T_co],'
                                         ' z: tests.test_util.typing_test_data.T'
                                         ') -> typing.List[tests.test_util.typing_test_data.T_contra]')
 
     # Union types
-    sig = inspect.signature(f3)
+    sig = inspect.signature(mod.f3)
     assert stringify_signature(sig) == '(x: str | numbers.Integral) -> None'
 
     # Quoted annotations
-    sig = inspect.signature(f4)
+    sig = inspect.signature(mod.f4)
     assert stringify_signature(sig) == '(x: str, y: str) -> None'
 
     # Keyword-only arguments
-    sig = inspect.signature(f5)
+    sig = inspect.signature(mod.f5)
     assert stringify_signature(sig) == '(x: int, *, y: str, z: str) -> None'
 
     # Keyword-only arguments with varargs
-    sig = inspect.signature(f6)
+    sig = inspect.signature(mod.f6)
     assert stringify_signature(sig) == '(x: int, *args, y: str, z: str) -> None'
 
     # Space around '=' for defaults
-    sig = inspect.signature(f7)
+    sig = inspect.signature(mod.f7)
     if sys.version_info[:2] <= (3, 10):
         assert stringify_signature(sig) == '(x: int | None = None, y: dict = {}) -> None'
     else:
         assert stringify_signature(sig) == '(x: int = None, y: dict = {}) -> None'
 
     # Callable types
-    sig = inspect.signature(f8)
+    sig = inspect.signature(mod.f8)
     assert stringify_signature(sig) == '(x: typing.Callable[[int, str], int]) -> None'
 
-    sig = inspect.signature(f9)
+    sig = inspect.signature(mod.f9)
     assert stringify_signature(sig) == '(x: typing.Callable) -> None'
 
     # Tuple types
-    sig = inspect.signature(f10)
+    sig = inspect.signature(mod.f10)
     assert stringify_signature(sig) == '(x: typing.Tuple[int, str], y: typing.Tuple[int, ...]) -> None'
 
     # Instance annotations
-    sig = inspect.signature(f11)
+    sig = inspect.signature(mod.f11)
     assert stringify_signature(sig) == '(x: CustomAnnotation, y: 123) -> None'
 
     # tuple with more than two items
-    sig = inspect.signature(f12)
+    sig = inspect.signature(mod.f12)
     assert stringify_signature(sig) == '() -> typing.Tuple[int, str, int]'
 
     # optional
-    sig = inspect.signature(f13)
+    sig = inspect.signature(mod.f13)
     assert stringify_signature(sig) == '() -> str | None'
 
     # optional union
-    sig = inspect.signature(f20)
+    sig = inspect.signature(mod.f20)
     assert stringify_signature(sig) in ('() -> int | str | None',
                                         '() -> str | int | None')
 
     # Any
-    sig = inspect.signature(f14)
+    sig = inspect.signature(mod.f14)
     assert stringify_signature(sig) == '() -> typing.Any'
 
     # ForwardRef
-    sig = inspect.signature(f15)
+    sig = inspect.signature(mod.f15)
     assert stringify_signature(sig) == '(x: Unknown, y: int) -> typing.Any'
 
     # keyword only arguments (1)
-    sig = inspect.signature(f16)
+    sig = inspect.signature(mod.f16)
     assert stringify_signature(sig) == '(arg1, arg2, *, arg3=None, arg4=None)'
 
     # keyword only arguments (2)
-    sig = inspect.signature(f17)
+    sig = inspect.signature(mod.f17)
     assert stringify_signature(sig) == '(*, arg3, arg4)'
 
-    sig = inspect.signature(f18)
+    sig = inspect.signature(mod.f18)
     assert stringify_signature(sig) == ('(self, arg1: int | typing.Tuple = 10) -> '
                                         'typing.List[typing.Dict]')
 
     # annotations for variadic and keyword parameters
-    sig = inspect.signature(f19)
+    sig = inspect.signature(mod.f19)
     assert stringify_signature(sig) == '(*args: int, **kwargs: str)'
 
     # default value is inspect.Signature.empty
-    sig = inspect.signature(f21)
+    sig = inspect.signature(mod.f21)
     assert stringify_signature(sig) == "(arg1='whatever', arg2)"
 
     # type hints by string
-    sig = inspect.signature(Node.children)
+    sig = inspect.signature(mod.Node.children)
     assert stringify_signature(sig) == '(self) -> typing.List[tests.test_util.typing_test_data.Node]'
 
-    sig = inspect.signature(Node.__init__)
+    sig = inspect.signature(mod.Node.__init__)
     assert stringify_signature(sig) == '(self, parent: tests.test_util.typing_test_data.Node | None) -> None'
 
     # show_annotation is False
-    sig = inspect.signature(f7)
+    sig = inspect.signature(mod.f7)
     assert stringify_signature(sig, show_annotation=False) == '(x=None, y={})'
 
     # show_return_annotation is False
-    sig = inspect.signature(f7)
+    sig = inspect.signature(mod.f7)
     if sys.version_info[:2] <= (3, 10):
         assert stringify_signature(sig, show_return_annotation=False) == '(x: int | None = None, y: dict = {})'
     else:
         assert stringify_signature(sig, show_return_annotation=False) == '(x: int = None, y: dict = {})'
 
     # unqualified_typehints is True
-    sig = inspect.signature(f7)
+    sig = inspect.signature(mod.f7)
     if sys.version_info[:2] <= (3, 10):
         assert stringify_signature(sig, unqualified_typehints=True) == '(x: int | None = None, y: dict = {}) -> None'
     else:
         assert stringify_signature(sig, unqualified_typehints=True) == '(x: int = None, y: dict = {}) -> None'
 
     # case: separator at head
-    sig = inspect.signature(f22)
+    sig = inspect.signature(mod.f22)
     assert stringify_signature(sig) == '(*, a, b)'
 
     # case: separator in the middle
-    sig = inspect.signature(f23)
+    sig = inspect.signature(mod.f23)
     assert stringify_signature(sig) == '(a, b, /, c, d)'
 
-    sig = inspect.signature(f24)
+    sig = inspect.signature(mod.f24)
     assert stringify_signature(sig) == '(a, /, *, b)'
 
     # case: separator at tail
-    sig = inspect.signature(f25)
+    sig = inspect.signature(mod.f25)
     assert stringify_signature(sig) == '(a, b, /)'
 
 

tests/test_util/test_util_inventory.py

@@ -1,61 +1,21 @@
 """Test inventory util functions."""
 import os
 import posixpath
-import zlib
 from io import BytesIO
 
 import sphinx.locale
 from sphinx.testing.util import SphinxTestApp
 from sphinx.util.inventory import InventoryFile
 
-inventory_v1 = b'''\
+from tests.test_util.intersphinx_data import (
-# Sphinx inventory version 1
+    INVENTORY_V1,
-# Project: foo
+    INVENTORY_V2,
-# Version: 1.0
+    INVENTORY_V2_NO_VERSION,
-module mod foo.html
+)
-module.cls class foo.html
-'''
-
-inventory_v2 = b'''\
-# Sphinx inventory version 2
-# Project: foo
-# Version: 2.0
-# The remainder of this file is compressed with zlib.
-''' + zlib.compress(b'''\
-module1 py:module 0 foo.html#module-module1 Long Module desc
-module2 py:module 0 foo.html#module-$ -
-module1.func py:function 1 sub/foo.html#$ -
-module1.Foo.bar py:method 1 index.html#foo.Bar.baz -
-CFunc c:function 2 cfunc.html#CFunc -
-std cpp:type 1 index.html#std -
-std::uint8_t cpp:type 1 index.html#std_uint8_t -
-foo::Bar cpp:class 1 index.html#cpp_foo_bar -
-foo::Bar::baz cpp:function 1 index.html#cpp_foo_bar_baz -
-foons cpp:type 1 index.html#foons -
-foons::bartype cpp:type 1 index.html#foons_bartype -
-a term std:term -1 glossary.html#term-a-term -
-ls.-l std:cmdoption 1 index.html#cmdoption-ls-l -
-docname std:doc -1 docname.html -
-foo js:module 1 index.html#foo -
-foo.bar js:class 1 index.html#foo.bar -
-foo.bar.baz js:method 1 index.html#foo.bar.baz -
-foo.bar.qux js:data 1 index.html#foo.bar.qux -
-a term including:colon std:term -1 glossary.html#term-a-term-including-colon -
-The-Julia-Domain std:label -1 write_inventory/#$ The Julia Domain
-''')
-
-inventory_v2_not_having_version = b'''\
-# Sphinx inventory version 2
-# Project: foo
-# Version:
-# The remainder of this file is compressed with zlib.
-''' + zlib.compress(b'''\
-module1 py:module 0 foo.html#module-module1 Long Module desc
-''')
 
 
 def test_read_inventory_v1():
-    f = BytesIO(inventory_v1)
+    f = BytesIO(INVENTORY_V1)
     invdata = InventoryFile.load(f, '/util', posixpath.join)
     assert invdata['py:module']['module'] == \
         ('foo', '1.0', '/util/foo.html#module-module', '-')
@@ -64,7 +24,7 @@ def test_read_inventory_v1():
 
 
 def test_read_inventory_v2():
-    f = BytesIO(inventory_v2)
+    f = BytesIO(INVENTORY_V2)
     invdata = InventoryFile.load(f, '/util', posixpath.join)
 
     assert len(invdata['py:module']) == 2
@@ -82,7 +42,7 @@ def test_read_inventory_v2():
 
 
 def test_read_inventory_v2_not_having_version():
-    f = BytesIO(inventory_v2_not_having_version)
+    f = BytesIO(INVENTORY_V2_NO_VERSION)
     invdata = InventoryFile.load(f, '/util', posixpath.join)
     assert invdata['py:module']['module1'] == \
         ('foo', '', '/util/foo.html#module-module1', 'Long Module desc')
@@ -102,7 +62,7 @@ def _build_inventory(srcdir):
     app = SphinxTestApp(srcdir=srcdir)
     app.build()
     sphinx.locale.translators.clear()
-    return (app.outdir / 'objects.inv')
+    return app.outdir / 'objects.inv'
 
 
 def test_inventory_localization(tmp_path):