[
https://issues.apache.org/jira/browse/DOXIA-690?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17682031#comment-17682031
]
ASF GitHub Bot commented on DOXIA-690:
--------------------------------------
kwin commented on code in PR #141:
URL: https://github.com/apache/maven-doxia/pull/141#discussion_r1090522065
##########
doxia-modules/doxia-module-markdown/src/main/java/org/apache/maven/doxia/module/markdown/MarkdownParser.java:
##########
@@ -156,6 +178,86 @@ public void parse(Reader source, Sink sink, String
reference) throws ParseExcept
}
}
+ private boolean processMetadataForHtml(StringBuilder html, StringBuilder
source) {
+ final Map<String, List<String>> metaData;
+ final int endOffset; // end of metadata within source
+ // support two types of metadata:
+ if (source.toString().startsWith("---")) {
+ // 1. YAML front matter
(https://github.com/vsch/flexmark-java/wiki/Extensions#yaml-front-matter)
+ Node documentRoot =
FLEXMARK_METADATA_PARSER.parse(source.toString());
+ YamlFrontMatterVisitor visitor = new YamlFrontMatterVisitor();
+ visitor.visit(documentRoot);
+ metaData = visitor.getData();
+ endOffset = visitor.getEndOffset();
+ } else {
+ // 2. Multimarkdown metadata
(https://fletcher.github.io/MultiMarkdown-5/metadata.html), not yet supported
+ // by Flexmark (https://github.com/vsch/flexmark-java/issues/550)
+ metaData = new LinkedHashMap<>();
+ Matcher metadataMatcher = METADATA_SECTION_PATTERN.matcher(source);
+ if (metadataMatcher.find()) {
+ String entry = metadataMatcher.group(0) + '\n';
+ Matcher entryMatcher = METADATA_ENTRY_PATTERN.matcher(entry);
+ while (entryMatcher.find()) {
+ String key = entryMatcher.group(1);
+ String value =
normalizeMultilineValue(entryMatcher.group(2));
+ metaData.put(key, Collections.singletonList(value));
+ }
+ endOffset = metadataMatcher.end(0);
+ } else {
+ endOffset = 0;
+ }
+ }
+ if (endOffset > 0) {
+ // Trim the metadata from the source
+ source.delete(0, endOffset);
+ }
+ return writeHtmlMetadata(html, metaData);
+ }
+
+ static String normalizeMultilineValue(String value) {
+ return value.trim().replaceAll("[ \\t]*[\\r\\n]+[ \\t]*", " ");
+ }
+
+ private boolean writeHtmlMetadata(StringBuilder html, Map<String,
List<String>> data) {
+ boolean containsTitle = false;
+ for (Entry<String, List<String>> entry : data.entrySet()) {
+ if (writeHtmlMetadata(html, entry.getKey(), entry.getValue())) {
+ containsTitle = true;
+ }
+ }
+ return containsTitle;
+ }
+
+ private boolean writeHtmlMetadata(StringBuilder html, String key,
List<String> values) {
+ if ("title".equalsIgnoreCase(key)) {
+ html.append("<title>");
+
html.append(HtmlTools.escapeHTML(values.stream().collect(Collectors.joining(",
")), false));
+ html.append("</title>");
+ return true;
+ } else {
+ if (key.equalsIgnoreCase("author") && values.size() > 1) {
+ // for multiple authors emit multiple meta tags
+ for (String value : values) {
+ writeHtmlMetadata(html, key,
Collections.singletonList(value));
+ }
+ } else {
+ // every other multivalue should just be concatenated and
emitted in a single meta tag
+ final String separator;
+ if (key.equalsIgnoreCase("keywords")) {
+ separator = ",";
+ } else {
+ separator = "\n";
Review Comment:
Fixed in
https://github.com/apache/maven-doxia/pull/141/commits/168071b3d8e0434ce73d668746e6c78f5f2051f2.
> Markdown Parser: Multiline metadata incorrectly rendered
> --------------------------------------------------------
>
> Key: DOXIA-690
> URL: https://issues.apache.org/jira/browse/DOXIA-690
> Project: Maven Doxia
> Issue Type: Bug
> Components: Module - Markdown
> Reporter: Konrad Windszus
> Assignee: Konrad Windszus
> Priority: Major
>
> Markdown uses the following metadata format in its sink:
> [http://fletcher.github.io/MultiMarkdown-5/metadata.html].
> In case a metadata key has multiple values on multiple lines it is emitted as
> following from the {{MarkdownSink}}:
> {code:java}
> title: Guide to creating a site
> author: Brett Porter
> Jason van Zyl
> date: 2015-07-18 {code}
> That leads to incorrect XHTML output as the second line for {{author}} is not
> detected correctly and instead of being emitted as metadata is emitted as
> regular paragraph in the HTML body. This is due to
> https://github.com/apache/maven-doxia/blob/7509feb03af4d4fb7d48b4f9ef38ff5c1a17a149/doxia-modules/doxia-module-markdown/src/main/java/org/apache/maven/doxia/module/markdown/MarkdownParser.java#L110
> only detecting single line metadata values.
> Although this might be a glitch of the underlying Markdown Parser
> implementation the metadata format explicitly recommends:
> {quote}
> To keep multiline metadata values from being confused with additional
> metadata, I recommend indenting each new line of metadata. If your metadata
> value includes a colon, it must be indented to keep it from being treated as
> a new key-value pair
> {quote}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
