You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tika.apache.org by ta...@apache.org on 2021/01/11 20:44:10 UTC

[tika] 01/02: trivial cleanup in AbstractOOXMLExtractor

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

tallison pushed a commit to branch TIKA-3266
in repository https://gitbox.apache.org/repos/asf/tika.git

commit da371381a5f9a7dbb5d620b643abc097042684a7
Author: tallison <ta...@apache.org>
AuthorDate: Mon Jan 11 15:24:14 2021 -0500

    trivial cleanup in AbstractOOXMLExtractor
---
 .../apache/tika/parser/microsoft/ooxml/AbstractOOXMLExtractor.java  | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/tika-parsers/tika-parsers-classic/tika-parsers-classic-modules/tika-parser-microsoft-module/src/main/java/org/apache/tika/parser/microsoft/ooxml/AbstractOOXMLExtractor.java b/tika-parsers/tika-parsers-classic/tika-parsers-classic-modules/tika-parser-microsoft-module/src/main/java/org/apache/tika/parser/microsoft/ooxml/AbstractOOXMLExtractor.java
index 6ab763c..5e3341c 100644
--- a/tika-parsers/tika-parsers-classic/tika-parsers-classic-modules/tika-parser-microsoft-module/src/main/java/org/apache/tika/parser/microsoft/ooxml/AbstractOOXMLExtractor.java
+++ b/tika-parsers/tika-parsers-classic/tika-parsers-classic-modules/tika-parser-microsoft-module/src/main/java/org/apache/tika/parser/microsoft/ooxml/AbstractOOXMLExtractor.java
@@ -140,10 +140,10 @@ public abstract class AbstractOOXMLExtractor implements OOXMLExtractor {
         buildXHTML(xhtml);
 
         // Now do any embedded parts
-        handleEmbeddedParts(handler, metadata);
+        handleEmbeddedParts(xhtml, metadata);
 
         // thumbnail
-        handleThumbnail(handler);
+        handleThumbnail(xhtml);
 
         xhtml.endDocument();
     }
@@ -180,6 +180,8 @@ public abstract class AbstractOOXMLExtractor implements OOXMLExtractor {
                 thumbnailMetadata.set(TikaCoreProperties.EMBEDDED_RELATIONSHIP_ID, thumbName);
                 thumbnailMetadata.set(Metadata.CONTENT_TYPE, tPart.getContentType());
                 thumbnailMetadata.set(TikaCoreProperties.TITLE, tPart.getPartName().getName());
+                thumbnailMetadata.set(TikaCoreProperties.EMBEDDED_RESOURCE_TYPE,
+                        TikaCoreProperties.EmbeddedResourceType.THUMBNAIL.name());
 
                 if (embeddedExtractor.shouldParseEmbedded(thumbnailMetadata)) {
                     embeddedExtractor.parseEmbedded(TikaInputStream.get(tStream), new EmbeddedContentHandler(handler), thumbnailMetadata, false);