FIX: Keep escapes and entities in image alt text

myst_parser/mdit_to_docutils/base.py

@@ -482,10 +482,11 @@ def renderInlineAsText(self, tokens: list[SyntaxTreeNode]) -> str:  # noqa: N802
         result = ""
 
         for token in tokens or []:
-            if token.type == "text":
+            if token.type in {"text", "text_special"}:
+                # An escape or entity is a text_special token with no children.
+                # Its content is the character the alt text must keep
+                # (executablebooks/MyST-Parser#1210).
                 result += token.content
-            # elif token.type == "image":
-            #     result += self.renderInlineAsText(token.children)
             else:
                 result += self.renderInlineAsText(token.children or [])
         return result

tests/test_image_alt.py

@@ -0,0 +1,30 @@
+"""Image alt text keeps CommonMark escapes and entities."""
+
+from docutils.core import publish_parts
+
+from myst_parser.parsers.docutils_ import Parser
+
+
+def _html(source: str) -> str:
+    return publish_parts(source, parser=Parser(), writer_name="html5")["body"].strip()
+
+
+def test_image_alt_keeps_escapes_and_entities():
+    """Escapes and entities inside an image alt are not dropped.
+
+    markdown-it-py represents them as ``text_special`` tokens with no children.
+    Reading only ``text`` tokens used to drop the character
+    (executablebooks/MyST-Parser#1210). The same source in a link already
+    rendered the character.
+    """
+    assert _html(r"![a \* b](x.png)") == '<p><img alt="a * b" src="x.png" /></p>'
+    assert _html("![a &amp; b](x.png)") == '<p><img alt="a &amp; b" src="x.png" /></p>'
+    assert (
+        _html(r"![C:\Python26](x.png)")
+        == '<p><img alt="C:\\Python26" src="x.png" /></p>'
+    )
+    assert (
+        _html(r"[a \* b &amp; c](https://example.com)")
+        == '<p><a class="reference external" href="https://example.com">'
+        "a * b &amp; c</a></p>"
+    )