From 433f820e8bd6e3df357b42bfb4cfc17a266c0112 Mon Sep 17 00:00:00 2001 From: Christian Tzolov Date: Fri, 23 Feb 2024 22:41:28 +0100 Subject: [PATCH] Improve metadata handling --- .../springframework/ai/reader/pdf/PagePdfDocumentReader.java | 5 ++--- .../ai/reader/pdf/ParagraphPdfDocumentReader.java | 5 ++--- .../ai/reader/pdf/PagePdfDocumentReaderTests.java | 4 ++-- 3 files changed, 6 insertions(+), 8 deletions(-) diff --git a/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/PagePdfDocumentReader.java b/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/PagePdfDocumentReader.java index 7163e2834..bf2daeda3 100644 --- a/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/PagePdfDocumentReader.java +++ b/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/PagePdfDocumentReader.java @@ -17,7 +17,6 @@ package org.springframework.ai.reader.pdf; import java.awt.Rectangle; -import java.io.File; import java.io.IOException; import java.util.ArrayList; import java.util.List; @@ -58,7 +57,7 @@ public class PagePdfDocumentReader implements DocumentReader { private PdfDocumentReaderConfig config; - private File resourceFileName; + private String resourceFileName; public PagePdfDocumentReader(String resourceUrl) { this(new DefaultResourceLoader().getResource(resourceUrl)); @@ -80,7 +79,7 @@ public class PagePdfDocumentReader implements DocumentReader { new org.apache.pdfbox.io.RandomAccessReadBuffer(pdfResource.getInputStream())); this.document = pdfParser.parse(); - this.resourceFileName = pdfResource.getFile(); + this.resourceFileName = pdfResource.getFilename(); this.config = config; } catch (Exception e) { diff --git a/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/ParagraphPdfDocumentReader.java b/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/ParagraphPdfDocumentReader.java index 1785c8325..27949ec8a 100644 --- a/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/ParagraphPdfDocumentReader.java +++ b/document-readers/pdf-reader/src/main/java/org/springframework/ai/reader/pdf/ParagraphPdfDocumentReader.java @@ -17,7 +17,6 @@ package org.springframework.ai.reader.pdf; import java.awt.Rectangle; -import java.io.File; import java.util.ArrayList; import java.util.Iterator; import java.util.List; @@ -65,7 +64,7 @@ public class ParagraphPdfDocumentReader implements DocumentReader { private PdfDocumentReaderConfig config; - private File resourceFileName; + private String resourceFileName; /** * Constructs a ParagraphPdfDocumentReader using a resource URL. @@ -108,7 +107,7 @@ public class ParagraphPdfDocumentReader implements DocumentReader { this.paragraphTextExtractor = new ParagraphManager(this.document); - this.resourceFileName = pdfResource.getFile(); + this.resourceFileName = pdfResource.getFilename(); } catch (IllegalArgumentException iae) { throw iae; diff --git a/document-readers/pdf-reader/src/test/java/org/springframework/ai/reader/pdf/PagePdfDocumentReaderTests.java b/document-readers/pdf-reader/src/test/java/org/springframework/ai/reader/pdf/PagePdfDocumentReaderTests.java index 305cd984e..8c31d4a73 100644 --- a/document-readers/pdf-reader/src/test/java/org/springframework/ai/reader/pdf/PagePdfDocumentReaderTests.java +++ b/document-readers/pdf-reader/src/test/java/org/springframework/ai/reader/pdf/PagePdfDocumentReaderTests.java @@ -1,5 +1,5 @@ /* - * Copyright 2023-2023 the original author or authors. + * Copyright 2023-2024 the original author or authors. * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. @@ -33,7 +33,7 @@ import static org.assertj.core.api.Assertions.assertThat; public class PagePdfDocumentReaderTests { @Test - public void test1() { + public void classpathRead() { PagePdfDocumentReader pdfReader = new PagePdfDocumentReader("classpath:/sample1.pdf", PdfDocumentReaderConfig.builder()