(tika) branch main updated: follow on updates to new MarkdownParser (#2926)

[tallison]

This is an automated email from the ASF dual-hosted git repository.

tballison pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tika.git


The following commit(s) were added to refs/heads/main by this push:
     new 32ef64bc07 follow on updates to new MarkdownParser (#2926)
32ef64bc07 is described below

commit 32ef64bc0703971ce837b11ea639c07c42c38546
Author: Tim Allison <talli...@apache.org>
AuthorDate: Tue Jul 7 11:18:57 2026 -0400

    follow on updates to new MarkdownParser (#2926)
---
 .../java/org/apache/tika/parser/markdown/MarkdownParser.java | 12 ++++++++++--
 .../org/apache/tika/parser/markdown/MarkdownParserTest.java  |  8 ++++++++
 2 files changed, 18 insertions(+), 2 deletions(-)

diff --git 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/main/java/org/apache/tika/parser/markdown/MarkdownParser.java
 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/main/java/org/apache/tika/parser/markdown/MarkdownParser.java
index 1c3266a7b4..2a91c69061 100644
--- 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/main/java/org/apache/tika/parser/markdown/MarkdownParser.java
+++ 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/main/java/org/apache/tika/parser/markdown/MarkdownParser.java
@@ -97,8 +97,13 @@ public class MarkdownParser extends 
AbstractEncodingDetectorParser {
     private static final List<Extension> EXTENSIONS =
             List.of(TablesExtension.create(), StrikethroughExtension.create());
 
-    //immutable and thread-safe
-    private static final Parser COMMONMARK = 
Parser.builder().extensions(EXTENSIONS).build();
+    //immutable and thread-safe.
+    //maxOpenBlockParsers caps block nesting: deeper blocks are parsed as flat 
paragraph text
+    //rather than nested structure, so a pathologically deep block document 
still extracts.
+    //Kept below SecureContentHandler's 100-level element-nesting cap so the 
flattened output
+    //stays under that limit and is emitted rather than rejected as a 
suspected zip bomb.
+    private static final Parser COMMONMARK =
+            
Parser.builder().extensions(EXTENSIONS).maxOpenBlockParsers(64).build();
 
     public MarkdownParser() {
         super();
@@ -123,6 +128,9 @@ public class MarkdownParser extends 
AbstractEncodingDetectorParser {
             metadata.set(Metadata.CONTENT_TYPE, new MediaType(MARKDOWN, 
charset).toString());
             metadata.set(Metadata.CONTENT_ENCODING, charset.name());
             document = COMMONMARK.parseReader(reader);
+        } catch (StackOverflowError e) {
+            //for reasons
+            throw new TikaException("Markdown is too deeply nested to parse", 
e);
         }
 
         XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata, 
context);
diff --git 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/test/java/org/apache/tika/parser/markdown/MarkdownParserTest.java
 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/test/java/org/apache/tika/parser/markdown/MarkdownParserTest.java
index 88bcb14448..dfe8ffd5b8 100644
--- 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/test/java/org/apache/tika/parser/markdown/MarkdownParserTest.java
+++ 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-text-module/src/test/java/org/apache/tika/parser/markdown/MarkdownParserTest.java
@@ -115,6 +115,14 @@ public class MarkdownParserTest extends TikaTest {
                 
metadataList.get(1).get(TikaCoreProperties.EMBEDDED_RESOURCE_TYPE));
     }
 
+    @Test
+    public void testDeeplyNestedBlocksAreFlattenedNotFailed() throws Exception 
{
+        //maxOpenBlockParsers caps block nesting below SecureContentHandler's 
100-level limit,
+        //so a pathologically deep block document extracts (deeper nesting 
flattened to text)
+        //rather than being rejected as a suspected zip bomb or overflowing 
the stack.
+        assertContains("deep", parseString("> ".repeat(5000) + "deep\n").xml);
+    }
+
     @Test
     public void testRoundTripsBackToMarkdown() throws Exception {
         String markdown = "# Title\n\nSome **bold** and *italic* and 
~~struck~~ `inline`.\n";