Fiddle with the handling of markup inside of maybe autolinks

bikeshed/h/parser/parser.py

@@ -16,6 +16,7 @@
     SafeText,
     SelfClosedTag,
     StartTag,
+    escapeAttr,
     escapeHTML,
 )
 from .preds import charRefs
@@ -894,11 +895,11 @@ def parseCSSMaybe(s: Stream, start: int) -> Result[list[ParserNode]]:
     # also if it actually produces a link
     # (like ''width: <<length>>'' linking to 'width')
     # it'll be broken anyway.
-    # So we'll hack this in - << gets turned into <
+    # So we'll hack this in - << gets turned into &lt;
     # within a maybe.
     # No chance of a link, but won't misparse in weird ways.
     if "<<" in text:
-        text = re.sub(r"<<", "<", text)
+        text = re.sub(r"<<", "&lt;", text)
         text = re.sub(r">>", ">", text)
 
     # This syntax does double duty as both a linking syntax
@@ -925,19 +926,19 @@ def parseCSSMaybe(s: Stream, start: int) -> Result[list[ParserNode]]:
         startTag = StartTag(
             line=s.line(start),
             endLine=s.line(textStart),
-            tag="a",
+            tag="bs-link",
             attrs={
-                "bs-autolink-syntax": s[start:nodeEnd],
+                "bs-autolink-syntax": escapeAttr(s[start:nodeEnd]),
                 "class": "css",
                 "data-link-type": "propdesc",
                 "data-lt": propdescname,
             },
         )
         if for_:
-            startTag.attrs["data-link-for"] = for_
+            startTag.attrs["data-link-for"] = escapeAttr(for_)
             startTag.attrs["data-link-type"] = "descriptor"
         startTag.finalize()
-        tagMiddle = SafeText(
+        tagMiddle = RawText(
             line=s.line(textStart),
             endLine=s.line(textEnd),
             text=text,
@@ -988,16 +989,20 @@ def parseCSSMaybe(s: Stream, start: int) -> Result[list[ParserNode]]:
         startTag = StartTag(
             line=s.line(start),
             endLine=s.line(textStart),
-            tag="a",
+            tag="bs-link",
             attrs={
-                "bs-autolink-syntax": s[start:nodeEnd],
+                "bs-autolink-syntax": escapeAttr(s[start:nodeEnd]),
                 "class": "css",
                 "data-link-type": linkType,
-                "data-lt": valueName,
+                "data-lt": escapeAttr(valueName),
             },
         )
+        if "&lt;" in valueName:
+            m.die(f"The autolink {s[start:nodeEnd]} is using an HTML escape (or <<) in its value; you probably don't want to escape things there.", lineNum=s.line(start))
         if for_:
-            startTag.attrs["data-link-for"] = for_
+            if "&lt;" in for_:
+                m.die(f"The autolink {s[start:nodeEnd]} is using an HTML escape in its for value; you probably don't want to escape things there.", lineNum=s.line(start))
+            startTag.attrs["data-link-for"] = escapeAttr(for_)
         if (for_ is not None and not for_.endswith("<")) or match[3] is not None:
             tagMiddle = SafeText(
                 line=s.line(textStart),
@@ -1006,7 +1011,7 @@ def parseCSSMaybe(s: Stream, start: int) -> Result[list[ParserNode]]:
             )
         else:
             startTag.attrs["bs-replace-text-on-link-success"] = valueName
-            tagMiddle = SafeText(
+            tagMiddle = RawText(
                 line=s.line(textStart),
                 endLine=s.line(textEnd),
                 text=text,
@@ -1027,7 +1032,7 @@ def parseCSSMaybe(s: Stream, start: int) -> Result[list[ParserNode]]:
         endLine=s.line(textStart),
         tag="css",
     )
-    tagMiddle = SafeText(
+    tagMiddle = RawText(
         line=s.line(textStart),
         endLine=s.line(textEnd),
         text=text,

bikeshed/unsortedJunk.py

@@ -685,12 +685,14 @@ def processAutolinks(doc: t.SpecT) -> None:
     # An <a> without an href is an autolink.
     # <i> is a legacy syntax for term autolinks. If it links up, we change it into an <a>.
     # We exclude bibliographical links, as those are processed in `processBiblioLinks`.
-    query = "a:not([href]):not([data-link-type='biblio'])"
+    query = "a:not([href]):not([data-link-type='biblio']), bs-link"
     if doc.md.useIAutolinks:
         m.warn("Use <i> Autolinks is deprecated and will be removed. Please switch to using <a> elements.")
         query += ", i"
     autolinks = h.findAll(query, doc)
     for el in autolinks:
+        if el.tag == "bs-link":
+            el.tag = "a"
         # Explicitly empty linking text indicates this shouldn't be an autolink.
         if el.get("data-lt") == "":
             continue

tests/github/w3c/csswg-drafts/css-color-5/Overview.console.txt

@@ -1,5 +1,6 @@
 LINE 124: Shorthand <<color space>> does not match any recognized shorthand grammar.
 LINE 632: Shorthand <<color space>> does not match any recognized shorthand grammar.
+LINE 721: The autolink ''from <<color>>'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 79: The heading 'color spaces {#color spaces-section}' needs a manually-specified ID.
 LINE 47: Image doesn't exist, so I couldn't determine its width and height: 'images/LC-picker-scale.png'
 LINE 48: Image doesn't exist, so I couldn't determine its width and height: 'images/LC-picker-map2.png'

tests/github/w3c/csswg-drafts/css-fonts-4/Overview.console.txt

@@ -3,6 +3,7 @@ LINE 3908:7: Garbage after id=.
 LINE 3908:1: Garbage at 3908:4 in <p>.
 LINE 3943:21: Spurious / in <img>.
 LINE 3976:21: Spurious / in <img>.
+LINE 4641: The autolink ''<<integer [0,∞]>>+'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 4832:22: Spurious / in <img>.
 LINE 5687:2: Spurious / in <img>.
 LINE 5692:3: Spurious / in <img>.

tests/github/w3c/csswg-drafts/css-gcpm-3/Overview.console.txt

@@ -4,6 +4,8 @@ LINE 213:1: Spurious / in <img>.
 LINE 299:1: Spurious / in <img>.
 LINE 385:1: Spurious / in <img>.
 LINE 749:1: Spurious / in <img>.
+LINE 932: The autolink ''attr(&lt;identifier>)'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 991: The autolink ''attr(&lt;identifier>)'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINT: Your document appears to use spaces to indent, but line 23 starts with tabs.
 LINT: Your document appears to use spaces to indent, but line 46 starts with tabs.
 LINT: Your document appears to use spaces to indent, but line 50 starts with tabs.

tests/github/w3c/csswg-drafts/css-grid-1/Overview.console.txt

@@ -5,8 +5,10 @@ LINE 295:3: Spurious / in <img>.
 LINE 301:3: Spurious / in <img>.
 LINE 607:3: Spurious / in <img>.
 LINE 1250:4: Spurious / in <img>.
+LINE 1542: The autolink ''minmax(auto, <<flex>>)'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 1688:4: Spurious / in <img>.
 LINE 2488:5: Spurious / in <img>.
+LINE 3650: The autolink ''&lt;content-distribution>/stretch'' is using an HTML escape in its for value; you probably don't want to escape things there.
 LINT: Your document appears to use tabs to indent, but line 3681 starts with spaces.
 WARNING: There are 1136 WPT tests underneath your path prefix 'css/css-grid/' that aren't in your document and must be added. (Use a <wpt hidden> if you don't actually want them in your document.)
   css/css-grid/abspos/absolute-positioning-changing-containing-block-001.html

tests/github/w3c/csswg-drafts/css-grid-2/Overview.console.txt

@@ -5,8 +5,10 @@ LINE 281:3: Spurious / in <img>.
 LINE 287:3: Spurious / in <img>.
 LINE 593:3: Spurious / in <img>.
 LINE 1296:4: Spurious / in <img>.
+LINE 1620: The autolink ''minmax(auto, <<flex>>)'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 1766:4: Spurious / in <img>.
 LINE 2634:5: Spurious / in <img>.
+LINE 4188: The autolink ''&lt;content-distribution>/stretch'' is using an HTML escape in its for value; you probably don't want to escape things there.
 LINT: Your document appears to use tabs to indent, but line 4219 starts with spaces.
 LINE 103: Image doesn't exist, so I couldn't determine its width and height: 'images/flex-layout.png'
 LINE 113: Image doesn't exist, so I couldn't determine its width and height: 'images/grid-layout.png'

tests/github/w3c/csswg-drafts/css-inline-3/Overview.console.txt

@@ -1,4 +1,5 @@
 LINE 2526:27: Garbage after the tagname in </h4>.
+LINE 2928: The autolink ''[&lt;baseline-keyword> <<percentage>>]+'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINT: Your document appears to use tabs to indent, but line 2 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 3 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 2019 starts with spaces.

tests/github/w3c/csswg-drafts/css-pseudo-4/Overview.console.txt

@@ -1,3 +1,5 @@
+LINE 963: The autolink ''<button>'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 964: The autolink ''<input>'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 317: Image doesn't exist, so I couldn't determine its width and height: 'images/first-letter2.gif'
 LINE ~181: No 'property' refs found for 'color' with spec 'css-color-3'.
 'color'

tests/github/w3c/csswg-drafts/css-syntax-3/Overview.console.txt

@@ -1,4 +1,5 @@
 LINE 172:42: Character reference '&B' didn't end in ;.
+LINE 3099: The autolink ''u <<number>>'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 2386: The var 'top-level flag' (in global scope) is only used once.
 If this is not a typo, please add an ignore='' attribute to the <var>.
 LINE 3368: Multiple elements have the same ID 'at-rule'.

tests/github/w3c/csswg-drafts/css-text-4/Overview.console.txt

@@ -1,3 +1,5 @@
+LINE 320: The autolink ''word-boundary-detection/auto(<<lang>>)'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 1236: The autolink ''match-indent | <<length>> | <<percentage>>'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINT: Your document appears to use tabs to indent, but line 1909 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 1911 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 1913 starts with spaces.

tests/github/w3c/csswg-drafts/css-text-4/Overview.html

@@ -1496,7 +1496,7 @@ <h4 class="heading settled" data-level="2.2.1" id="word-boundary-detection"><spa
 	language-specific content analysis
 	in response to user preferences.
 	User agents with this behavior must do this
-	by setting the <a data-link-type="dfn" href="https://drafts.csswg.org/css-cascade-5/#declared-value" id="ref-for-declared-value">declared value</a> of <a class="property css" data-link-type="property" href="#propdef-word-boundary-detection" id="ref-for-propdef-word-boundary-detection①">word-boundary-detection</a> to <span class="css">auto(&lt;&lt;lang>>)</span> in the <a data-link-type="dfn" href="https://drafts.csswg.org/css-cascade-5/#cascade-origin-user" id="ref-for-cascade-origin-user">User Origin</a>.
+	by setting the <a data-link-type="dfn" href="https://drafts.csswg.org/css-cascade-5/#declared-value" id="ref-for-declared-value">declared value</a> of <a class="property css" data-link-type="property" href="#propdef-word-boundary-detection" id="ref-for-propdef-word-boundary-detection①">word-boundary-detection</a> to <span class="css">auto(&amp;lt;lang>)</span> in the <a data-link-type="dfn" href="https://drafts.csswg.org/css-cascade-5/#cascade-origin-user" id="ref-for-cascade-origin-user">User Origin</a>.
 	User agents that do not support the <span id="ref-for-cascade-origin-user①">User Origin</span> may use the <a data-link-type="dfn" href="https://drafts.csswg.org/css-cascade-5/#cascade-origin-ua" id="ref-for-cascade-origin-ua">User-Agent Origin</a> instead.</p>
    <div class="advisement">
      Manual analysis of the content can be more reliable than UA heuristics.

tests/github/w3c/csswg-drafts/css-transforms-2/Overview.console.txt

@@ -1,3 +1,7 @@
+LINE 978: The autolink ''rotate3d(1, 0, 0, <angle>)'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 980: The autolink ''rotate3d(0, 1, 0, <angle>)'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 982: The autolink ''rotate3d(0, 0, 1, <angle>)'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 982: The autolink ''rotate(<angle>)'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINT: Your document appears to use tabs to indent, but line 52 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 53 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 54 starts with spaces.

tests/github/w3c/csswg-drafts/css-values-3/Overview.console.txt

@@ -1,3 +1,9 @@
+LINE 251: The autolink ''<length>#{1,4}'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 1552: The autolink ''<resolution>/dpi'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 1552: The autolink ''<resolution>/dpcm'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 1552: The autolink ''<resolution>/dppx'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 1574: The autolink ''<resolution>/dppx'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 1674: The autolink ''<<position>> <<length>>?'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 423: Couldn't find WPT test 'css/css-values/initial-background-color.html' - did you misspell something?
 LINE 423: Couldn't find WPT test 'css/css-values/unset-value-storage.html' - did you misspell something?
 LINE 1808: Couldn't find WPT test 'css/vendor-imports/mozilla/mozilla-central-reftests/values3/calc-background-image-gradient-1.html' - did you misspell something?

tests/github/w3c/csswg-drafts/css-values-4/Overview.console.txt

@@ -1,3 +1,9 @@
+LINE 247: The autolink ''<length>#{1,4}'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 2122: The autolink ''<resolution>/dpi'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 2122: The autolink ''<resolution>/dpcm'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 2122: The autolink ''<resolution>/dppx'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 2145: The autolink ''<resolution>/dppx'' is using an HTML escape in its for value; you probably don't want to escape things there.
+LINE 2276: The autolink ''<<position>> <<length>>?'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 545: Couldn't find WPT test 'css/css-values/initial-background-color.html' - did you misspell something?
 LINE 545: Couldn't find WPT test 'css/css-values/unset-value-storage.html' - did you misspell something?
 LINE 1509: Image doesn't exist, so I couldn't determine its width and height: 'images/Typography_Line_Terms.svg'

tests/github/w3c/csswg-drafts/css-writing-modes-3/Overview.console.txt

@@ -1,3 +1,5 @@
+LINE 1097: The autolink ''<length>'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 1098: The autolink ''<percentage>'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINT: Line 1914's indent contains tabs after spaces.
 LINT: Line 1915's indent contains tabs after spaces.
 LINT: Your document appears to use spaces to indent, but line 2510 starts with tabs.

tests/github/w3c/csswg-drafts/css-writing-modes-4/Overview.console.txt

@@ -1,3 +1,5 @@
+LINE 1175: The autolink ''<length>'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 1176: The autolink ''<percentage>'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINT: Line 2010's indent contains tabs after spaces.
 LINT: Line 2011's indent contains tabs after spaces.
 LINT: Your document appears to use spaces to indent, but line 2788 starts with tabs.

tests/github/w3c/csswg-drafts/css2/Overview.console.txt

@@ -1,3 +1,4 @@
+LINE 15173: The autolink ''<<border-style>>/none'' is using an HTML escape in its for value; you probably don't want to escape things there.
 LINT: Your document appears to use spaces to indent, but line 559 starts with tabs.
 LINT: Your document appears to use spaces to indent, but line 560 starts with tabs.
 LINT: Your document appears to use spaces to indent, but line 561 starts with tabs.

tests/github/w3c/csswg-drafts/mediaqueries-4/Overview.console.txt

@@ -1,3 +1,5 @@
+LINE 624: The autolink ''(600px < height)'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 633: The autolink ''(400px < width < 1000px)'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 978:11: Character reference '&test' didn't end in ;.
 LINT: Your document appears to use tabs to indent, but line 2260 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 2261 starts with spaces.

tests/github/w3c/csswg-drafts/mediaqueries-5/Overview.console.txt

@@ -1,3 +1,5 @@
+LINE 630: The autolink ''(600px < height)'' is using an HTML escape in its value; you probably don't want to escape things there.
+LINE 639: The autolink ''(400px < width < 1000px)'' is using an HTML escape in its value; you probably don't want to escape things there.
 LINE 984:11: Character reference '&test' didn't end in ;.
 LINT: Your document appears to use tabs to indent, but line 1463 starts with spaces.
 LINT: Your document appears to use tabs to indent, but line 1464 starts with spaces.