Add support for recovering from missing top level header

Signed-off-by: Jarno Elovirta <jarno@elovirta.com>
jelovirt · May 11, 2024 · 98ee818 · 98ee818
1 parent 53ca716
commit 98ee818
Show file tree

Hide file tree

Showing 8 changed files with 90 additions and 18 deletions.
diff --git a/src/main/java/com/elovirta/dita/markdown/MarkdownParserImpl.java b/src/main/java/com/elovirta/dita/markdown/MarkdownParserImpl.java
@@ -130,25 +130,43 @@ protected void validate(Document root) {
    * If document doesn't start with H1, generate H1 from YAML metadata or file name.
    */
   protected Document preprocess(Document root, URI input) throws SAXException {
-    if (DitaRenderer.WIKI.get(options) && isWiki(root)) {
-      generateRootHeading(root, input);
-    }
-    if (DitaRenderer.FIX_ROOT_HEADING.get(options) && isWiki(root)) {
-      if (errorHandler != null) {
-        errorHandler.warning(
-          new SAXParseException(MESSAGES.getString("error.missing_title"), null, input.toString(), 1, 1)
-        );
+    if (isWiki(root)) {
+      if (DitaRenderer.WIKI.get(options)) {
+        generateRootHeading(root, input);
+      } else if (DitaRenderer.FIX_ROOT_HEADING.get(options)) {
+        if (errorHandler != null) {
+          errorHandler.warning(
+            new SAXParseException(MESSAGES.getString("error.missing_title"), null, input.toString(), 1, 1)
+          );
+        }
+        generateRootHeading(root, input);
+      } else if (MarkdownReader.PROCESSING_MODE.get(options)) {
+        if (errorHandler != null) {
+          errorHandler.error(
+            new SAXParseException(MESSAGES.getString("error.missing_title"), null, input.toString(), 1, 1)
+          );
+        }
+      } else {
+        if (errorHandler != null) {
+          errorHandler.warning(
+            new SAXParseException(MESSAGES.getString("error.missing_title"), null, input.toString(), 1, 1)
+          );
+        }
+        generateRootHeading(root, null);
       }
-      generateRootHeading(root, input);
     }
     return root;
   }
 
+  /**
+   * Generate root heading using YAML header or input file name. If neither YAML or file name is not available, create empty heading.
+   *
+   */
   private void generateRootHeading(Document root, URI input) {
     final YamlFrontMatterBlock yaml = root.getFirstChild() instanceof YamlFrontMatterBlock
       ? (YamlFrontMatterBlock) root.getFirstChild()
       : null;
-    String title = getTextFromFile(input);
+    String title = input != null ? getTextFromFile(input) : null;
     final Heading heading = new Heading();
     if (yaml != null) {
       final AbstractYamlFrontMatterVisitor v = new AbstractYamlFrontMatterVisitor();
@@ -170,9 +188,11 @@ private void generateRootHeading(Document root, URI input) {
       }
     }
     heading.setLevel(1);
-    final AnchorLink anchorLink = new AnchorLink();
-    anchorLink.appendChild(new Text(title));
-    heading.appendChild(anchorLink);
+    if (title != null) {
+      final AnchorLink anchorLink = new AnchorLink();
+      anchorLink.appendChild(new Text(title));
+      heading.appendChild(anchorLink);
+    }
     root.prependChild(heading);
   }
 

diff --git a/src/main/java/com/elovirta/dita/markdown/MarkdownReader.java b/src/main/java/com/elovirta/dita/markdown/MarkdownReader.java
@@ -45,6 +45,7 @@ public class MarkdownReader implements XMLReader {
   private static final ServiceLoader<SchemaProvider> schemaLoader = ServiceLoader.load(SchemaProvider.class);
 
   public static final DataKey<Collection<String>> FORMATS = new DataKey<>("FORMATS", List.of("markdown", "md"));
+  /** Strict processing mode. */
   public static final DataKey<Boolean> PROCESSING_MODE = new DataKey<>("PROCESSING_MODE", false);
 
   /**

diff --git a/src/test/java/com/elovirta/dita/markdown/MarkdownReaderTest.java b/src/test/java/com/elovirta/dita/markdown/MarkdownReaderTest.java
@@ -17,12 +17,20 @@
 import java.nio.charset.StandardCharsets;
 import java.util.Arrays;
 import java.util.Map;
+import javax.xml.transform.Transformer;
+import javax.xml.transform.dom.DOMResult;
+import javax.xml.transform.sax.SAXSource;
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.params.ParameterizedTest;
 import org.junit.jupiter.params.provider.CsvSource;
 import org.junit.jupiter.params.provider.ValueSource;
+import org.opentest4j.AssertionFailedError;
+import org.w3c.dom.Document;
 import org.xml.sax.*;
 import org.xml.sax.helpers.XMLFilterImpl;
+import org.xmlunit.builder.DiffBuilder;
+import org.xmlunit.diff.DefaultComparisonFormatter;
+import org.xmlunit.diff.Diff;
 
 public class MarkdownReaderTest extends AbstractReaderTest {
 
@@ -140,6 +148,25 @@ public void test_missingHeader(String file) throws Exception {
     final TestErrorHandler errorHandler = new TestErrorHandler();
     reader.setErrorHandler(errorHandler);
 
+    run(getSrc() + file, getExp() + "wiki/" + file.replaceAll("\\.md$", ".dita"));
+
+    assertEquals(1, errorHandler.warnings.size());
+    final SAXParseException e = errorHandler.warnings.get(0);
+    assertEquals("Document content doesn't start with heading", e.getMessage());
+    assertEquals("classpath:/markdown/" + file, e.getSystemId());
+    assertNull(e.getPublicId());
+    assertEquals(1, e.getLineNumber());
+    assertEquals(1, e.getColumnNumber());
+  }
+
+  @ParameterizedTest
+  @ValueSource(strings = { "missing_root_header.md", "missing_root_header_with_yaml.md" })
+  public void test_emptyHeader(String file) throws Exception {
+    reader =
+      new MarkdownReader(new MutableDataSet().set(Parser.EXTENSIONS, singletonList(YamlFrontMatterExtension.create())));
+    final TestErrorHandler errorHandler = new TestErrorHandler();
+    reader.setErrorHandler(errorHandler);
+
     run(file);
 
     assertEquals(1, errorHandler.warnings.size());

diff --git a/src/test/java/com/elovirta/dita/utils/AbstractReaderTest.java b/src/test/java/com/elovirta/dita/utils/AbstractReaderTest.java
@@ -65,7 +65,7 @@ public void run(final String input) throws Exception {
     run(getSrc() + input, getExp() + input.replaceAll("\\.(md|html)$", ".dita"));
   }
 
-  void run(final String input, final String expFile) throws Exception {
+  protected void run(final String input, final String expFile) throws Exception {
     final Document act;
     try (final InputStream in = getClass().getResourceAsStream("/" + input)) {
       act = db.newDocument();

diff --git a/src/test/resources/dita/missing_root_header.dita b/src/test/resources/dita/missing_root_header.dita
@@ -1,7 +1,7 @@
 <topic xmlns:ditaarch="http://dita.oasis-open.org/architecture/2005/" class="- topic/topic "
   specializations="@props/audience @props/deliveryTarget @props/otherprops @props/platform @props/product"
-  id="missing-root-header" ditaarch:DITAArchVersion="2.0">
-  <title class="- topic/title ">missing root header</title>
+  id="" ditaarch:DITAArchVersion="2.0">
+  <title class="- topic/title "></title>
   <body class="- topic/body ">
     <p class="- topic/p ">Root topic content.</p>
   </body>

diff --git a/src/test/resources/dita/missing_root_header_with_yaml.dita b/src/test/resources/dita/missing_root_header_with_yaml.dita
@@ -1,7 +1,7 @@
 <topic xmlns:ditaarch="http://dita.oasis-open.org/architecture/2005/" class="- topic/topic "
   specializations="@props/audience @props/deliveryTarget @props/otherprops @props/platform @props/product"
-  id="yaml-title" ditaarch:DITAArchVersion="2.0">
-  <title class="- topic/title ">YAML Title</title>
+  id="" ditaarch:DITAArchVersion="2.0">
+  <title class="- topic/title "></title>
   <body class="- topic/body ">
     <p class="- topic/p ">Root topic content.</p>
   </body>

diff --git a/src/test/resources/dita/wiki/missing_root_header.dita b/src/test/resources/dita/wiki/missing_root_header.dita
@@ -0,0 +1,16 @@
+<topic xmlns:ditaarch="http://dita.oasis-open.org/architecture/2005/" class="- topic/topic "
+  specializations="@props/audience @props/deliveryTarget @props/otherprops @props/platform @props/product"
+  id="missing-root-header" ditaarch:DITAArchVersion="2.0">
+  <title class="- topic/title ">missing root header</title>
+  <body class="- topic/body ">
+    <p class="- topic/p ">Root topic content.</p>
+  </body>
+  <topic class="- topic/topic "
+    specializations="@props/audience @props/deliveryTarget @props/otherprops @props/platform @props/product"
+    id="nested-topic" ditaarch:DITAArchVersion="2.0">
+    <title class="- topic/title ">Nested Topic</title>
+    <body class="- topic/body ">
+      <p class="- topic/p ">Nested topic content.</p>
+    </body>
+  </topic>
+</topic>
diff --git a/src/test/resources/dita/wiki/missing_root_header_with_yaml.dita b/src/test/resources/dita/wiki/missing_root_header_with_yaml.dita
@@ -0,0 +1,8 @@
+<topic xmlns:ditaarch="http://dita.oasis-open.org/architecture/2005/" class="- topic/topic "
+  specializations="@props/audience @props/deliveryTarget @props/otherprops @props/platform @props/product"
+  id="yaml-title" ditaarch:DITAArchVersion="2.0">
+  <title class="- topic/title ">YAML Title</title>
+  <body class="- topic/body ">
+    <p class="- topic/p ">Root topic content.</p>
+  </body>
+</topic>