This is an automated email from the ASF dual-hosted git repository.
Croway pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/camel.git
The following commit(s) were added to refs/heads/main by this push:
new 90c44ee31ae4 chore: align Docling integration expectations
90c44ee31ae4 is described below
commit 90c44ee31ae4a4cc2b6f6a8be3ae74be24674dc0
Author: Croway <[email protected]>
AuthorDate: Fri Aug 28 16:03:23 2026 +0200
chore: align Docling integration expectations
---
.../org/apache/camel/catalog/docs/docling-component.adoc | 8 ++++----
.../camel-docling/src/main/docs/docling-component.adoc | 8 ++++----
.../component/docling/integration/MetadataExtractionIT.java | 11 ++++++-----
.../camel/component/docling/integration/OcrExtractionIT.java | 12 +++++++-----
4 files changed, 21 insertions(+), 18 deletions(-)
diff --git
a/catalog/camel-catalog/src/generated/resources/org/apache/camel/catalog/docs/docling-component.adoc
b/catalog/camel-catalog/src/generated/resources/org/apache/camel/catalog/docs/docling-component.adoc
index 0aef255d2be8..0fedaa6098d5 100644
---
a/catalog/camel-catalog/src/generated/resources/org/apache/camel/catalog/docs/docling-component.adoc
+++
b/catalog/camel-catalog/src/generated/resources/org/apache/camel/catalog/docs/docling-component.adoc
@@ -350,9 +350,10 @@ YAML::
Docling recognizes page headers and footers (page numbers, copyright lines,
running titles, footnotes, and similar
content) during OCR, classifying them as page _furniture_ in docling's
`FURNITURE` content layer.
-Since docling v1.30.0, page furniture (headers and footers) is included in the
default body export (Markdown, text
-and HTML). Earlier versions excluded the `FURNITURE` layer, so header and
footer text was omitted from the converted
-output even though the OCR engine read it correctly.
+Docling's Markdown, text and HTML exports include only the `BODY` layer by
default, so header and footer text is
+omitted from the converted output even though the OCR engine read it
correctly. When using `useDoclingServe=true`,
+the component cannot currently include the `FURNITURE` layer because
docling-serve does not expose a content-layer
+selection option. This is tracked in
https://github.com/docling-project/docling-serve/issues/271[docling-serve#271].
=== Using headers to control processing
@@ -1327,4 +1328,3 @@ YAML::
3. **Set appropriate timeouts**: Adjust `asyncTimeout` based on your largest
expected document size. The default 5 minutes is suitable for most documents.
4. **Monitor with logging**: Use DEBUG level logging to monitor API calls and
async task status.
-
diff --git
a/components/camel-ai/camel-docling/src/main/docs/docling-component.adoc
b/components/camel-ai/camel-docling/src/main/docs/docling-component.adoc
index 0aef255d2be8..0fedaa6098d5 100644
--- a/components/camel-ai/camel-docling/src/main/docs/docling-component.adoc
+++ b/components/camel-ai/camel-docling/src/main/docs/docling-component.adoc
@@ -350,9 +350,10 @@ YAML::
Docling recognizes page headers and footers (page numbers, copyright lines,
running titles, footnotes, and similar
content) during OCR, classifying them as page _furniture_ in docling's
`FURNITURE` content layer.
-Since docling v1.30.0, page furniture (headers and footers) is included in the
default body export (Markdown, text
-and HTML). Earlier versions excluded the `FURNITURE` layer, so header and
footer text was omitted from the converted
-output even though the OCR engine read it correctly.
+Docling's Markdown, text and HTML exports include only the `BODY` layer by
default, so header and footer text is
+omitted from the converted output even though the OCR engine read it
correctly. When using `useDoclingServe=true`,
+the component cannot currently include the `FURNITURE` layer because
docling-serve does not expose a content-layer
+selection option. This is tracked in
https://github.com/docling-project/docling-serve/issues/271[docling-serve#271].
=== Using headers to control processing
@@ -1327,4 +1328,3 @@ YAML::
3. **Set appropriate timeouts**: Adjust `asyncTimeout` based on your largest
expected document size. The default 5 minutes is suitable for most documents.
4. **Monitor with logging**: Use DEBUG level logging to monitor API calls and
async task status.
-
diff --git
a/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/MetadataExtractionIT.java
b/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/MetadataExtractionIT.java
index f7150eb705ac..26fde44f26f9 100644
---
a/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/MetadataExtractionIT.java
+++
b/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/MetadataExtractionIT.java
@@ -65,13 +65,14 @@ class MetadataExtractionIT extends DoclingITestSupport {
DocumentMetadata metadata =
template.requestBody("direct:extract-metadata",
testFile.toString(), DocumentMetadata.class);
- assertNotNull(metadata, "Metadata should not be null");
- assertNotNull(metadata.getFileName(), "File name should be extracted");
- assertTrue(metadata.getFileSizeBytes() > 0, "File size should be
greater than 0");
- assertNotNull(metadata.getFilePath(), "File path should be set");
+ assertThat(metadata).isNotNull();
+ assertThat(metadata.getFileName()).isNotNull();
+ assertThat(metadata.getFileSizeBytes()).isPositive();
+ assertThat(metadata.getFilePath()).isNotNull();
assertThat(metadata.getPageCount()).isEqualTo(5);
assertThat(metadata.getFormat()).isEqualTo("application/pdf");
- assertThat(metadata.getTitle()).isEqualTo("The Evolution of the Word
Processor");
+ // Docling classifies the leading heading in this fixture as a section
header, not a document title.
+ assertThat(metadata.getTitle()).isNull();
assertThat(metadata.getDocumentType()).isEqualTo("PDF");
LOG.info("Successfully extracted metadata: {}", metadata);
diff --git
a/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/OcrExtractionIT.java
b/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/OcrExtractionIT.java
index c3f40981129a..9690ad8efef6 100644
---
a/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/OcrExtractionIT.java
+++
b/components/camel-ai/camel-docling/src/test/java/org/apache/camel/component/docling/integration/OcrExtractionIT.java
@@ -175,13 +175,15 @@ class OcrExtractionIT extends CamelTestSupport {
boolean foundFirst = resultLower.contains("first");
boolean foundSecond = resultLower.contains("second");
- assertTrue(foundFirst && foundSecond,
- "OCR should extract at least some of the expected text. Got: "
+ result);
+ assertThat(foundFirst && foundSecond)
+ .as("OCR should extract at least some of the expected text.
Got: %s", result)
+ .isTrue();
- // Since docling v1.30.0, page furniture (headers/footers) is included
in the default body export.
+ // Docling recognizes the footer but classifies it as page furniture,
which the default body export excludes.
boolean foundFooter = resultLower.contains("footer");
- assertTrue(foundFooter,
- "Footer text (page furniture) should be included in the
docling body export since v1.30.0. Got: " + result);
+ assertThat(foundFooter)
+ .as("Footer text should be excluded from the default body
export. Got: %s", result)
+ .isFalse();
LOG.info("OCR extraction with multiple text blocks result:\n{}",
result);
LOG.info("Successfully extracted text from image with multiple text
blocks");
