Author: centic
Date: Wed Dec 6 19:50:02 2023
New Revision: 1914407
URL: http://svn.apache.org/viewvc?rev=1914407&view=rev
Log:
Make sure OLE2ScratchpadExtractorFacory is sorted first
Otherwise order of found extractors would depend on jar-loading order
and thus might have unexpected side-effects and missing features in
text-extraction.
Modified:
poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorFactory.java
poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorProvider.java
Modified:
poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorFactory.java
URL:
http://svn.apache.org/viewvc/poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorFactory.java?rev=1914407&r1=1914406&r2=1914407&view=diff
==============================================================================
--- poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorFactory.java
(original)
+++ poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorFactory.java
Wed Dec 6 19:50:02 2023
@@ -88,6 +88,21 @@ public final class ExtractorFactory {
private ExtractorFactory() {
ClassLoader cl = ExtractorFactory.class.getClassLoader();
ServiceLoader.load(ExtractorProvider.class, cl).forEach(provider::add);
+
+ // loading of service-files is non-deterministic as it depends on
order of loaded jars
+ // however we would like to "prefer" one Factory, so let's make sure
the more
+ // powerful "ScratchpadProvider" is sorted first
+ provider.sort((o1, o2) -> {
+ if (o1.getClass() != o2.getClass()) {
+ if
(o1.getClass().getSimpleName().equals("OLE2ScratchpadExtractorFactory")) {
+ return -1;
+ } else if
(o2.getClass().getSimpleName().equals("OLE2ScratchpadExtractorFactory")) {
+ return 1;
+ }
+ }
+
+ return o1.getClass().getName().compareTo(o2.getClass().getName());
+ });
}
/**
Modified:
poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorProvider.java
URL:
http://svn.apache.org/viewvc/poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorProvider.java?rev=1914407&r1=1914406&r2=1914407&view=diff
==============================================================================
--- poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorProvider.java
(original)
+++ poi/trunk/poi/src/main/java/org/apache/poi/extractor/ExtractorProvider.java
Wed Dec 6 19:50:02 2023
@@ -70,7 +70,8 @@ public interface ExtractorProvider {
* @param dirs a list to be filled with directory references holding
embedded
* @param nonPOIFS a list to be filled with streams which aren't based on
POIFS entries
*
- * @throws IOException when the format specific extraction fails because
of invalid entires
+ * @throws IOException when the format specific extraction fails because
of invalid entries
+ * @throws java.lang.IllegalArgumentException if implementations do not
overwrite this method
*/
default void identifyEmbeddedResources(POIOLE2TextExtractor ext,
List<Entry> dirs, List<InputStream> nonPOIFS) throws IOException {
throw new IllegalArgumentException("Error checking for Scratchpad
embedded resources");
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
