Package org.apache.tika.parser.html.charsetdetector

Class StandardHtmlEncodingDetector

java.lang.Object
org.apache.tika.parser.html.charsetdetector.StandardHtmlEncodingDetector
All Implemented Interfaces:
Serializable, org.apache.tika.detect.EncodingDetector
public final class StandardHtmlEncodingDetector extends Object implements org.apache.tika.detect.EncodingDetector
An encoding detector that tries to respect the spirit of the HTML spec part 12.2.3 "The input byte stream", or at least the part that is compatible with the implementation of tika.

https://html.spec.whatwg.org/multipage/parsing.html#the-input-byte-stream

If a resource was fetched over HTTP, then HTTP headers should be added to tika metadata when using detect(java.io.InputStream, org.apache.tika.metadata.Metadata), especially HttpHeaders.CONTENT_TYPE, as it may contain charset information.

This encoding detector may return null if no encoding is detected. It is meant to be used inside a CompositeEncodingDetector. For instance: 

 
     EncodingDetector detector = new CompositeEncodingDetector(
       Arrays.asList(
         new StandardHtmlEncodingDetector(),
         new Icu4jEncodingDetector()));

See Also:

  • Constructor Details

    • StandardHtmlEncodingDetector

      public StandardHtmlEncodingDetector()

  • Method Details

    • detect

      public Charset detect(InputStream input, org.apache.tika.metadata.Metadata metadata) throws IOException
      Specified by:
      detect in interface org.apache.tika.detect.EncodingDetector
      Throws:
      IOException

    • getMarkLimit

      public int getMarkLimit()

    • setMarkLimit

      @Field public void setMarkLimit(int markLimit)
      How far into the stream to read for charset detection. Default is 8192.