Test : Add test to validate ParagraphPdfDocumentReader to skip Invalid Outline

Signed-off-by: WOONBE <[email protected]>

Author: WOONBE

auto-configurations/models/spring-ai-autoconfigure-model-openai/src/main/java/org/springframework/ai/model/openai/autoconfigure/OpenAiEmbeddingProperties.java

@@ -66,6 +66,7 @@ public void setEmbeddingsPath(String embeddingsPath) {
 	public int getMaxBatchChars() {
 		return this.maxBatchChars;
 	}
+
 	public void setMaxBatchChars(int maxBatchChars) {
 		this.maxBatchChars = maxBatchChars;
 	}

document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/ParagraphPdfDocumentReader.java

@@ -207,18 +207,22 @@ public String getTextBetweenParagraphs(Paragraph fromParagraph, Paragraph toPara
 				if (pageNumber == startPage && pageNumber == endPage) {
 					y = toPos;
 					h = fromPos - toPos;
-				} else if (pageNumber == startPage) {
+				}
+				else if (pageNumber == startPage) {
 					y = 0;
 					h = fromPos;
-				} else if (pageNumber == endPage) {
+				}
+				else if (pageNumber == endPage) {
 					y = toPos;
 					h = (int) pageHeight - toPos;
-				} else {
+				}
+				else {
 					y = 0;
 					h = (int) pageHeight;
 				}
 
-				if (h < 0) h = 0;
+				if (h < 0)
+					h = 0;
 
 				pdfTextStripper.addRegion("pdfPageRegion", new Rectangle(x, y, w, h));
 				pdfTextStripper.extractRegions(page);

document-readers/pdf-reader/src/test/java/org/springframework/ai/reader/pdf/ParagraphPdfDocumentReaderTests.java

@@ -16,7 +16,6 @@
 
 package org.springframework.ai.reader.pdf;
 
-
 import org.apache.pdfbox.Loader;
 import org.apache.pdfbox.io.RandomAccessRead;
 import org.apache.pdfbox.pdmodel.PDDocument;
@@ -78,7 +77,41 @@ public void testPdfWithoutToc() {
 
 	}
 
+	@Test
+	void shouldSkipInvalidOutline() throws IOException {
+
+		Resource basePdfResource = new ClassPathResource("sample3.pdf");
+
+		PDDocument documentToModify;
+		try (InputStream inputStream = basePdfResource.getInputStream()) {
+
+			byte[] pdfBytes = inputStream.readAllBytes();
 
+			documentToModify = Loader.loadPDF(pdfBytes);
+		}
+		PDDocumentOutline outline = documentToModify.getDocumentCatalog().getDocumentOutline();
+		if (outline != null && outline.getFirstChild() != null) {
+			PDOutlineItem chapter2OutlineItem = outline.getFirstChild().getNextSibling();
+			if (chapter2OutlineItem != null) {
 
+				chapter2OutlineItem.setDestination((PDDestination) null);
+			}
+		}
+		ByteArrayOutputStream baos = new ByteArrayOutputStream();
+		documentToModify.save(baos);
+		documentToModify.close();
+
+		Resource corruptedPdfResource = new ByteArrayResource(baos.toByteArray());
+
+		ParagraphPdfDocumentReader reader = new ParagraphPdfDocumentReader(corruptedPdfResource,
+				PdfDocumentReaderConfig.defaultConfig());
+
+		List<Document> documents = assertDoesNotThrow(() -> reader.get());
+
+		assertThat(documents).isNotNull();
+		assertThat(documents).hasSize(2);
+		assertThat(documents.get(0).getMetadata().get("title")).isEqualTo("Chapter 1");
+		assertThat(documents.get(1).getMetadata().get("title")).isEqualTo("Chapter 3");
+	}
 
 }