Fix XML context reporting (#144)

facelessuser · web-flow · commit e64fed948b16 · 2021-08-05T09:07:32.000-06:00
* Fix XML context reporting

This affects HTML and other plugins derived from XML

* Add test
diff --git a/docs/src/markdown/about/changelog.md b/docs/src/markdown/about/changelog.md
@@ -1,5 +1,9 @@
 # Changelog
 
+## 2.7.3
+
+- **FIX**: Fix context reporting in the XML, HTML, and other filters derived from XML the filter.
+
 ## 2.7.2
 
 - **FIX**: Add note in `--help` option about `--source` behavior.
diff --git a/pyspelling/filters/xml.py b/pyspelling/filters/xml.py
@@ -198,13 +198,18 @@ def extract_string(self, node, is_comments):
                 self._block_text[self._current_block].append(string)
                 self._block_text[self._current_block].append(' ')
 
-    def set_block(self, node, force=False):
-        """Set the current block."""
+    def pop_block(self, node, force=False):
+        """Pop block if the current blocks sibling is found."""
 
-        if not force and node is self._block_stack[-1][1]:
+        while self._block_stack and node is self._block_stack[-1][1]:
             self._block_stack.pop(-1)
             self._current_block = self._block_stack[-1][0]
 
+    def set_block(self, node, force=False):
+        """Set the current block."""
+
+        self.pop_block(node, force)
+
         if force or self.is_break_tag(node):
             self._block_stack.append((node, self.get_last_descendant(node)))
             self._block_text[node] = []
@@ -265,6 +270,8 @@ def to_text(self, root):
                         if next_good is None:
                             break
                 else:
+                    self.pop_block(node)
+
                     # Handle test nodes: normal text and comments
                     is_comments = isinstance(node, bs4.Comment)
                     if (self.comments and is_comments) or (not is_comments and not isinstance(node, NON_CONTENT)):
diff --git a/tests/filters/test_html.py b/tests/filters/test_html.py
@@ -1498,3 +1498,62 @@ def test_html_after_text(self):
 
         self.mktemp('test.txt', template, 'utf-8')
         self.assert_spellcheck('.html.yml', bad_words)
+
+
+class TestHTMLContext(util.PluginTestCase):
+    """Test HTML plugin with default `lxml` (HTML) mode."""
+
+    def setup_fs(self):
+        """Setup file system."""
+
+        config = self.dedent(
+            """
+            matrix:
+            - name: html
+              sources:
+              - '{}/**/*.txt'
+              aspell:
+                lang: en
+                d: en_US
+              hunspell:
+                d: en_US
+              pipeline:
+              - pyspelling.filters.html:
+                  attributes:
+                  - alt
+                  ignores:
+                  - ':is(code, pre)'
+                  - 'span:matches(.some-class, #some-id)'
+            """
+        ).format(self.tempdir)
+        self.mktemp('.html.yml', config, 'utf-8')
+
+    def test_html_context(self):
+        """Test HTML context."""
+
+        template = self.dedent(
+            """
+            <html>
+            <head>
+            </head>
+            <body>
+            <ol>
+            <li>
+            <p>Some <code>dkdd</code> text
+                flga</p>
+                <div><pre><code>kdjk</code></pre></div>
+                helo
+            </li>
+            </ol>
+            </html>
+            """
+        )
+
+        self.mktemp('test.txt', template, 'utf-8')
+        self.assert_context(
+            '.html.yml',
+            [
+                'test.txt: html>body>ol>li',
+                'test.txt: html>body>ol>li>p'
+            ]
+        )
diff --git a/tests/util.py b/tests/util.py
@@ -135,6 +135,54 @@ def assert_spell_required(self, running):
                 "'{}', is not compatible with '{}'".format(running, required)
             )
 
+    def assert_context(self, config_file, expected, names=None, groups=None, sources=None, verbose=4):
+        """Check context."""
+
+        hunspell_location = which(HUNSPELL)
+        aspell_location = which(ASPELL)
+        if hunspell_location and aspell_location:
+            running = "both"
+        elif hunspell_location:
+            running = "hunspell"
+        elif aspell_location:
+            running = "aspell"
+        else:
+            running = "none"
+        self.assert_spell_required(running)
+
+        if hunspell_location:
+            context = []
+            for results in spellcheck(
+                os.path.join(self.tempdir, config_file),
+                names=names,
+                groups=groups,
+                sources=sources,
+                checker='hunspell',
+                binary=hunspell_location,
+                debug=True,
+                verbose=verbose
+            ):
+                if results.error:
+                    print(results.error)
+                context.append(results.context)
+            self.assertEqual([('{}/{}'.format(self.tempdir, i) if i else i) for i in expected[:]], context)
+        if aspell_location:
+            context = []
+            for results in spellcheck(
+                os.path.join(self.tempdir, config_file),
+                names=names,
+                groups=groups,
+                sources=sources,
+                checker='aspell',
+                binary=aspell_location,
+                debug=True,
+                verbose=verbose
+            ):
+                if results.error:
+                    print(results.error)
+                context.append(results.context)
+            self.assertEqual([('{}/{}'.format(self.tempdir, i) if i else i) for i in expected[:]], context)
+
     def assert_spellcheck(self, config_file, expected, names=None, groups=None, sources=None, verbose=4):
         """Spell check."""
 
