Repository navigation
Restore the straight quotes the FAQ conversion turned typographic - #3425
Merged
Merged
Conversation
Follow-up to #3423, which converted the FAQ from FML to Markdown and is already merged. Two defects in that page, one visible to readers. flexmark enables smart-punctuation, so a straight double quote in Markdown prose is rendered as a typographic quote. FML went through XdocParser, which does no such rewriting, so the page had straight quotes for years. Twenty-two of them changed appearance when the page became Markdown: the forked-VM message, "extraordinary", "jar", "classesDirectory", "${project.build.outputDirectory}", the corrupted-channel warning and both occurrences of "enableProcessChecker". They are written as " here, which is the convention the already-converted maven-site-plugin pages use for the same reason. Quotes inside code spans and fenced blocks are untouched; smart-punctuation never applied to those. The anchors also move from <a name="..."> to <a id="...">. This module is safe either way today, because the root pom pins maven-site-plugin 3.22.0 explicitly rather than inheriting it from maven-parent 47, and 3.22.0 preserves the name attribute. 3.21.0 strips it, which would delete every anchor on the page while the build stayed green. Since the pin is the only thing standing between this page and that outcome, and since Xhtml5BaseParser reads Attribute.ID first and only falls back to NAME - and name on <a> is obsolete in HTML5 - id is the form to be holding. Verified by building maven-surefire-plugin's site and comparing the rendered page against the FML original from before #3423. All ten anchors plus #top still served, <head> byte-identical, no duplicate-id warnings, and no typographic quotes left in the visible text. Note for anyone rebuilding this locally: the reactor does not build on JDK 26. sisu-maven-plugin 0.9.0.M4 fails on maven-surefire-common with "Unsupported class file major version 70". JDK 21 works. Generated-by: Claude Opus 5 (1M context)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Follow-up to #3423 (merged). Two defects in the converted FAQ page, one of which readers can see.
The visible one: smart quotes
flexmark enables smart-punctuation, so a straight
"in Markdown prose renders as a typographic quote. FML went throughXdocParser, which does no such rewriting β so this page served straight quotes for years and now does not. Twenty-two of them changed appearance:Written as
"here β the convention the already-convertedmaven-site-pluginpages use for exactly this reason. Quotes inside code spans and fenced blocks are untouched; smart-punctuation never applied to those.This is the class of defect the anchor test cannot catch β it is a text change, not an anchor change.
The latent one:
<a name>β<a id>maven-site-plugin 3.21.0 strips the
nameattribute from inline HTML; 3.22.0 keeps it. Stripping it would delete every anchor on this page while the build stayed green.This module is safe today β but only because the root pom pins
maven-site-plugin3.22.0 explicitly, rather than inheriting frommaven-parent47. That pin is the only thing standing between this page and silent anchor loss.Xhtml5BaseParserreadsAttribute.IDfirst and falls back toNAME, soidis the primary path rather than a workaround, andnameon<a>is obsolete in HTML5 anyway.Each anchor stays on its own line β folding one into heading text suppresses the section's own generated id.
Verification
Built
maven-surefire-plugin's site and compared against the FML original from before #3423, deletingtarget/sitefirst so a page that failed to re-render could not be silently compared against itself:#topstill served β#dumpfilesmatters beyond the live site,examples/shutdown.apt.vmlinks to../faq.html#dumpfiles<head>byte-identicalAnchor name β¦ used more than oncewarningsRebuilding locally needs JDK 21 β the reactor does not build on JDK 26,
sisu-maven-plugin0.9.0.M4 fails onmaven-surefire-commonwith "Unsupported class file major version 70". Unrelated to this change.Drafted with Claude β please verify