Modify default encoding detector

Replace HtmlEncodingDetector to StandardHtmlEncodingDetector Adjust some test case
apache · Aug 13, 2020 · 2a1b6a5 · 2a1b6a5
1 parent 5021f34
commit 2a1b6a5
Show file tree

Hide file tree

Showing 3 changed files with 5 additions and 4 deletions.
diff --git a/tika-parsers/src/main/resources/META-INF/services/org.apache.tika.detect.EncodingDetector b/tika-parsers/src/main/resources/META-INF/services/org.apache.tika.detect.EncodingDetector
@@ -13,6 +13,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-org.apache.tika.parser.html.HtmlEncodingDetector
+org.apache.tika.parser.html.charsetdetector.StandardHtmlEncodingDetector
 org.apache.tika.parser.txt.UniversalEncodingDetector
 org.apache.tika.parser.txt.Icu4jEncodingDetector
diff --git a/tika-parsers/src/test/java/org/apache/tika/config/TikaEncodingDetectorTest.java b/tika-parsers/src/test/java/org/apache/tika/config/TikaEncodingDetectorTest.java
@@ -39,6 +39,7 @@
 import org.apache.tika.parser.Parser;
 import org.apache.tika.parser.ParserDecorator;
 import org.apache.tika.parser.html.HtmlEncodingDetector;
+import org.apache.tika.parser.html.charsetdetector.StandardHtmlEncodingDetector;
 import org.apache.tika.parser.txt.Icu4jEncodingDetector;
 import org.apache.tika.parser.txt.TXTParser;
 import org.apache.tika.parser.txt.UniversalEncodingDetector;
@@ -52,7 +53,7 @@ public void testDefault() {
  assertTrue(detector instanceof CompositeEncodingDetector);
  List<EncodingDetector> detectors = ((CompositeEncodingDetector) detector).getDetectors();
  assertEquals(3, detectors.size());
- assertTrue(detectors.get(0) instanceof HtmlEncodingDetector);
+ assertTrue(detectors.get(0) instanceof StandardHtmlEncodingDetector);
  assertTrue(detectors.get(1) instanceof UniversalEncodingDetector);
  assertTrue(detectors.get(2) instanceof Icu4jEncodingDetector);
  }

diff --git a/...src/test/resources/org/apache/tika/config/TIKA-2273-exclude-encoding-detector-default.xml b/...src/test/resources/org/apache/tika/config/TIKA-2273-exclude-encoding-detector-default.xml
@@ -19,9 +19,9 @@
  <!-- Explicitly request default parsers -->
  <parsers/>
  <encodingDetectors>
- <!-- All detectors except HtmlEncodingDetector -->
+ <!-- All detectors except StandardHtmlEncodingDetector -->
  <encodingDetector class="org.apache.tika.detect.DefaultEncodingDetector">
- <encodingDetector-exclude class="org.apache.tika.parser.html.HtmlEncodingDetector"/>
+ <encodingDetector-exclude class="org.apache.tika.parser.html.charsetdetector.StandardHtmlEncodingDetector"/>
  </encodingDetector>
  <!-- One other detector, to check ordering -->
  <encodingDetector class="org.apache.tika.detect.NonDetectingEncodingDetector"/>