We have a request to drop phonetic strings from xlsx as the default in Tika.  
I'm not familiar enough with Japanese to know if users would generally expect 
to be able to search on these as well as the original.  The current practice is 
to include them.
  Any recommendations?  Thank you!



-----Original Message-----
From: Takahiro Ochi (JIRA) [] 
Sent: Tuesday, August 8, 2017 2:28 AM
Subject: [jira] [Created] (TIKA-2440) Phonetic strings handling for 
multilingual environments.

Takahiro Ochi created TIKA-2440:

             Summary: Phonetic strings handling for multilingual environments.
                 Key: TIKA-2440
             Project: Tika
          Issue Type: Improvement
          Components: parser
            Reporter: Takahiro Ochi
            Priority: Minor

Hi there,

I would like to propose an idea to improve phonetic strings handling for 
multilingual environments. I believe Tika should not concatenate phonetic 
strings because text with phonetic strings is recognized as noisy text in most 
situations of natural language processing.

Excel files include phonetic strings in some languages such as Japanese, 
Chinese and so on. Apache POI concatenates phonetic strings onto the shared 
strings when Tika extract text from Excel files.

Recent Apache POI has an switch flag for phonetic strings concatination as 

Tika should set the 2nd argument "includePhoneticRuns" as false. Here is the 
simple patch for my idea.

$ diff -ru 
---    2017-06-10 19:13:33.355412625 +0900
 2017-06-10 19:14:30.452411830 +0900
@@ -130,7 +130,7 @@
             styles = xssfReader.getStylesTable();

             iter = (XSSFReader.SheetIterator) xssfReader.getSheetsData();
-            strings = new ReadOnlySharedStringsTable(container);
+            strings = new ReadOnlySharedStringsTable(container,false);
         } catch (InvalidFormatException e) {
             throw new XmlException(e);
         } catch (OpenXML4JException oe) {


This message was sent by Atlassian JIRA

Reply via email to