OpenXLIFF Filters

An open source set of Java filters for creating, merging and validating XLIFF 1.2, 2.0 and 2.1 files.

With OpenXLIFF Filters you can create XLIFF files that don't use proprietary markup and are compatible with most CAT (Computer Asisted Translation) tools.

Features

Filters Configuration

XML and JSON filters are configurable

XML Filter Configuration
JSON Filter Configuration

Related Projects

XLIFF Manager implements an easy-to-use user interface for creating, merging, validating, and manipulating XLIFF files in a graphical environment.
Swordfish uses OpenXLIFF Filters to extract translatable text from supported formats.
RemoteTM uses OpenXLIFF Filters to handle all TMX processing.
Stingray uses OpenXLIFF Filters for extracting the text to align from supported monolingual documents.
TMXEditor relies on OpenXLIFF Filters XML support for processing TMX files.
XLIFF Validation web-based XLIFF Validation Service.
Fluenta a Translation Manager that uses OpenXLIFF Filters to generate XLIFF from DITA projects.
JavaPM a set of scripts for localizing Java .properties files using XLIFF.

Releases

Version	Comment	Release Date
3.22.0	Added option to ignore translatable SVG files when parsing DITA	September 26th, 2024
3.21.0	Fixed attributes parsing in HTML and improved JSON filter	August 29th, 2024
3.20.0	Improved region merging for Word documents	March 17th, 2024
3.19.0	Moved Language Tags handling to BCP47J	March 7th, 2024
3.18.0	Fixed scope builder for DITA filter; removed Machine Translation engines	February 5th, 2024
3.17.0	Improved extraction from Word text boxes; switched to Java 21	January 6th, 2024
3.16.0	Added catalog for XLIFF 2.2; adjusted ChatGPT models	October 29th, 2023
3.15.0	Added new options to JSON filter configuration	September 13th, 2023
3.14.0	Added option to generate XLIFF 2.1	September 1st, 2023
3.13.0	Added export as TMX scripts; allowed relative paths from command line	August 27th, 2023
3.12.0	Improved JSON support and localization	August 16th, 2023
3.11.0	Improved SVG handling and language sorting	July 31st, 2023
3.10.0	Fixed TMX exports	July 10th, 2023
3.9.0	Improved Machine Translation support and internal code	June 30th, 2023
3.8.0	Improved localization handling	June 2nd, 2023
3.7.0	Extracted "alt" text from Word images	May 15th, 2023
3.6.0	Improved segmentation of XLIFF 2.0	April 1st, 2023
3.5.0	Improved HTML filter	March 15th, 2023
3.4.0	Extracted localizable strings	February 27th, 2023
3.3.0	Detected loops while processing @keyref in DITA	February 20th, 2023
3.2.0	Improved file management	February 4th, 2023
3.1.0	Improved DITA merge	January 24th, 2023
3.0.0	Moved XML code to project XMLJava	January 9th, 2023
2.13.0	Ignored tracked changes from Oxygen XML Editor	December 22nd, 2022
2.12.0	Added "xmlfilter" parameter to conversion options	December 5th, 2022
2.11.0	Improved support for DITA from Astoria CMS	December 2nd, 2022
2.10.0	Fixed DITA segmentation	November 22nd, 2022
2.9.1	Fixed joining of XLIFF 2.0 files and improved PHP Array support	October 22nd, 2022
2.9.0	Added support for PHP Arrays	October 21st, 2022
2.8.0	Updated TXLF, JSON and DITA filters	October 8th, 2022
2.7.0	Fixed resegmenter for XLIFF 2.0	August 12th, 2022
2.6.0	Converted HTML fragments in Excel & Word files to tags	July 16th, 2022
2.5.0	Added configuration options to JSON filter; Added scripts to approve all segments; Updated language list	July 6th, 2022
2.4.2	Improved support for Trados Studio Packages	June 18th, 2022
2.4.1	Fixed conversion of third party XLIFF files	June 10th, 2022
2.4.0	Added remove all targets; added feedback for Fluenta on DITA filter	June 9th, 2022
2.3.0	Added copy source to target; Fixed DITA conversion and merge	May 25th, 2022
2.2.0	Added pseudo-translation	May 11th, 2022
2.1.0	Updated dependencies and improved validation of XLIFF 2.x	April 21st, 2022
2.0.0	Moved server code to XLIFF Manager project	March 29th, 2022
1.17.5	Updated DITA keyscope handling	March 18th, 2022
1.17.4	Fixed handling of nested untranslatables in DITA; Improved XLIFF 2.0 support	March 6th, 2022
1.17.2	Fixed support for FrameMaker MIF files	February 6th, 2022
1.17.1	Improved support for DITA	February 5th, 2022
1.17.0	Improved validation of XLIFF 2.0; Added SVG statistics for XLIFF 2.0	December 1st, 2021
1.16.0	Improved support for XLIFF 2.0; Switched to Java 17	November 23rd, 2021
1.15.2	MS Office and DITA fixes	November 13th, 2021
1.15.0	Initial support for RELAX NG gammars	November 8th, 2021
1.14.0	Improved segmentation for XLIFF 2.0	October 3rd, 2021
1.13.0	Improved DITA support	August 31st, 2021
1.12.7	Improved round trip 1.2 -> 2.0 -> 1.2; Ignored untranslatable SVG in DITA maps	July 4th, 2021
1.12.6	Improved validation; updated language management	June 19th, 2021
1.12.5	Improved support for bilingual files	February 3rd, 2021
1.12.4	Allowed concurrent access for XLIFF Validation	January 22nd, 2021
1.12.3	Improved support for Trados Studio packages	January 1st, 2021
1.12.2	Suppressed output for confirmed empty targets	December 7th, 2020
1.12.1	Improved conversion of XLIFF 1.2 files	December 3rd, 2020
1.12.0	Added support for Adobe InCopy ICML and SRT subtitles	November 23rd, 2020
1.11.1	Fixed JSON encoding; fixed import of XLIFF matches	November 1st, 2020
1.11.0	Added support for JSON files	September 25, 2020
1.10.1	Fixed handling of TXLF files and improved XML catalog	September 5th, 2020
1.10.0	Added conversion of 3rd party XLIFF; improved support for XLIFF 2.0; fixed issues with Trados Studio packages	August 25th, 2020
1.9.1	Added improvements required by Swordfish IV.	August 13th, 2020
1.9.0	Added 5 Machine Translation (MT) engines (Google, Microsoft Azure, DeepL, MyMemory & Yandex)	May 18th, 2020
1.8.4	Improved catalog and other minor fixes	April 25th, 2020
1.8.3	Fixed conversion of PO files	April 17th, 2020
1.8.2	Switched to synchronized classes in XML package	April 10th, 2020
1.8.1	Improved support for Trados Studio packages	April 3rd, 2020
1.8.0	Implemented support for Trados Studio packages	March 29th, 2020
1.7.0	Major code cleanup; Changed segmentation model for XLIFF 2.0	January 1st, 2020
1.6.0	Added support for XLIFF files from WPML WordPress Plugin	December 1st, 2019
1.5.2	Improved segmenter performance	October 29th, 2019
1.5.1	Fixed catalog on Windows	September 22nd, 2019
1.5.0	Improved support for empty <xref/> elements in DITA; improved support for XML catalogs	September 5th, 2019
1.4.2	Added option to join XLIFF files; Fixed merge errors in XLIFF 2.0; added tool info to XLIFF files; cleaned DITA attributes on merging	August 14th, 2019
1.4.1	Improved performance embedding skeleton; added Apache Ant building option; renamed module to 'openxliff'	July 25th, 2019
1.4.0	Improved report of task results	July 17th, 2019
1.3.3	Fixed merging of MS Office files from XLIFF 2.0	July 5th, 2019
1.3.2	Updated for Swordfish 3.4.3	June 30th, 2019
1.3.1	Updated for Swordfish 3.4-0	April 30th, 2019
1.3.0	Added option to export approved XLIFF segments as TMX	April 18th, 2019
1.2.1	Improved validation of XLIFF 2.0	April 6th, 2019
1.2.0	Added Translation Status Analysis	March 3rd, 2019
1.1.0	Incorporated XLIFFChecker code	November 20th, 2018
1.0.0	Initial Release	November 12th, 2018

Supported File Formats

OpenXLIFF Filters can generate XLIFF 1.2 and XLIFF 2.0 from these formats:

General Documentation
- Adobe InCopy ICML
- Adobe InDesign Interchange (INX)
- Adobe InDesign IDML CS4, CS5, CS6 & CC
- HTML
- Microsoft Office (2007 and newer)
- Microsoft Visio XML Drawings (2007 and newer)
- MIF (Maker Interchange Format)
- OpenOffice / LibreOffice / StarOffice
- PHP Arrays
- Plain Text
- SDLXLIFF (Trados Studio)
- SRT Subtitles
- Trados Studio Packages (*.sdlppx)
- TXML (GlobalLink/Wordfast PRO)
- WPML XLIFF (WordPress Multilingual Plugin)
- Wordfast/GlobalLink XLIFF (*.txlf)
- XLIFF from Other Tools (.mqxliff, .txlf, .xliff, etc.)
XML Formats
- XML (Generic)
- DITA 1.0, 1.1, 1.2 and 1.3
- DocBook 3.x, 4.x and 5.x
- SVG
- Word 2003 ML
- XHTML
Software Development
- JavaScript
- JSON
- Java Properties
- PHP Arrays
- PO (Portable Objects)
- RC (Windows C/C++ Resources)
- ResX (Windows .NET Resources)
- TS (Qt Linguist translation source)

Requirements

JDK 21 or newer is required for compiling and building. Pre-built binaries already include everything you need to run all options.
Apache Ant 1.10.13 or newer

Building

Checkout this repository.
Point your JAVA_HOME variable to JDK 21
Run ant to generate a binary distribution in ./dist

Steps for building

  git clone https://github.com/rmraya/OpenXLIFF.git
  cd OpenXLIFF
  ant

A binary distribution will be created in /dist folder.

Convert Document to XLIFF

You can use the library in your own Java code. Conversion to XLIFF is handled by the class com.maxprograms.converters.Convert.

If you use binaries from the command line, running .\convert.bat or ./convert.sh without parameters displays help for XLIFF generation.

Usage:

   convert.sh [-help] [-version] -file sourceFile -srcLang sourceLang
        [-tgtLang targetLang] [-skl skeletonFile] [-xliff xliffFile] [-type fileType]
        [-enc encoding] [-srx srxFile] [-catalog catalogFile] [-divatal ditaval]
        [-config configFile] [-embed] [-paragraph] [-xmlfilter folder][-2.0]
        [-ignoretc][-charsets]

Where:

   -help:      (optional) Display this help information and exit
   -version:   (optional) Display version & build information and exit
   -file:      source file to convert
   -srcLang:   source language code
   -tgtLang:   (optional) target language code
   -xliff:     (optional) XLIFF file to generate
   -skl:       (optional) skeleton file to generate
   -type:      (optional) document type
   -enc:       (optional) character set code for the source file
   -srx:       (optional) SRX file to use for segmentation
   -catalog:   (optional) XML catalog to use for processing
   -ditaval:   (optional) conditional processing file to use when converting DITA maps
   -config:    (optional) configuration file to use when converting JSON documents
   -embed:     (optional) store skeleton inside the XLIFF file
   -paragraph: (optional) use paragraph segmentation
   -xmlfilter: (optional) folder containing configuration files for the XML filter
   -ignoretc:  (optional) ignore tracked changes from Oxygen XML Editor in XML files
   -2.0:       (optional) generate XLIFF 2.0
   -charsets:  (optional) display a list of available character sets and exit

Document Types

   INX = Adobe InDesign Interchange
   ICML = Adobe InCopy ICML
   IDML = Adobe InDesign IDML
   DITA = DITA Map
   HTML = HTML Page
   JS = JavaScript
   JSON = JSON
   JAVA = Java Properties
   MIF = MIF (Maker Interchange Format)
   OFF = Microsoft Office 2007 Document
   OO = OpenOffice Document
   PHPA = PHP Array
   PO = PO (Portable Objects)
   RC = RC (Windows C/C++ Resources)
   RESX = ResX (Windows .NET Resources)
   SDLPPX = Trados Studio Package
   SDLXLIFF = SDLXLIFF Document
   SRT = SRT Subtitle
   TEXT = Plain Text
   TS = TS (Qt Linguist translation source)
   TXLF = Wordfast/GlobalLink XLIFF
   TXML = TXML Document
   WPML = WPML XLIFF
   XLIFF = XLIFF Document
   XML = XML Document
   XMLG = XML (Generic)

Only two parameters are absolutely required: -file and -srcLang. The library tries to automatically detect format and encoding and exits with an error message if it can't guess them. If automatic detection doesn't work, add -type and -enc parameters.

Character sets vary with the operating system. Run the conversion script with -charsets to get a list of character sets available in your OS.

By default, XLIFF and skeleton are generated in the folder where the source document is located. Extensions used for XLIFF and Skeleton are .xlf and .skl.

The XML type handles multiple document formats, like XHTML, SVG or DocBook files.

Default XML catalog and SRX file are provided. You can also use custom versions if required.

Convert XLIFF to Original Format

You can convert XLIFF files created with OpenXLIFF Filters to original format using class com.maxprograms.converters.Merge in your Java code.

If you use binaries from the command line, running .\merge.bat or ./merge.sh without parameters will display the information you need to merge an XLIFF file.

Usage:

   merge.bat [-help] [-version] -xliff xliffFile -target targetFile [-catalog catalogFile] [-unapproved] [-export]

Where:

   -help:       (optional) Display this help information and exit
   -version:    (optional) Display version & build information and exit
   -xliff:      XLIFF file to merge
   -target:     translated file or folder where to store translated files
   -catalog:    (optional) XML catalog to use for processing
   -unapproved: (optional) accept translations from unapproved segments
   -export:     (optional) generate TMX file from approved segments

XLIFF Validation

The original XLIFFChecker code supports XLIFF 1.0, 1.1 and 1.2. The new version incorporated in OpenXLIFF Filters also supports XLIFF 2.0.

Standard XML Schema validation does not detect the use of duplicated 'id' attributes, wrong language codes and other constraints written in the different XLIFF specifications.

All XLIFF 2.0 modules are validated using XML Schema validation in a first pass. Extra validation is then performed using Java code for XLIFF 2.0 Core and for Metadata, Matches and Glossary modules.

You can validate XLIFF files using your own Java code. Validation of XLIFF files is handled by the class com.maxprograms.validation.XliffChecker.

If you use binaries from the command line, running .\xliffchecker.bat or ./xliffchecker.sh without parameters displays help for XLIFF validation.

Usage:

   xliffchecker.bat [-help] -file xliffFile [-catalog catalogFile]

Where:

   -help:      (optional) Display this help information and exit
   -file:      XLIFF file to validate
   -catalog:   (optional) XML catalog to use for processing

XLIFF Validation Service

You can validate your XLIFF files online at https://dev.maxprograms.com/Validation/

Translation Status Analysis

This library lets you produce an HTML file with word counts and segment status statistics from an XLIFF file.

If you use binaries from the command line, running .\analysis.bat or ./analysis.sh without parameters displays help for statistics generation.

You can generate statistics using your own Java code. Statistics generation is handled by the class com.maxprograms.stats.RepetitionAnalysis.

Usage:

   analysis.sh [-help] -file xliffFile [-catalog catalogFile]

Where:

   -help:      (optional) Display this help information and exit
   -file:      XLIFF file to analyze
   -catalog:   (optional) XML catalog to use for processing

The HTML report is generated in the folder where the XLIFF file is located and its name is the name of the XLIFF plus .log.html.

Join multiple XLIFF files

You can combine several XLIFF files into a larger one using class com.maxprograms.converters.Join from your Java code or using the provided scripts.

Running .\join.bat or ./join.sh without parameters displays help for joining files.

Usage:

   join.sh [-help] -target targetFile -files file1,file2,file3...

 Where:

   -help:     (optional) Display this help information and exit
   -target:   combined output XLIFF file

The merge process automatically splits the files when converting back to original format.

Pseudo-translate XLIFF file

You can pseudo-translate all untranslated segments using class com.maxprograms.converters.PseudoTranslation from your Java code or using the provided scripts.

Running .\pseudotranslate.bat or ./pseudotranslate.sh without parameters displays help for pseudo-translating an XLIFF file.

Usage:

   pseudotranslate.bat [-help] -xliff xliffFile [-catalog catalogFile]

Where:

   -help:      (optional) Display this help information and exit
   -xliff:     XLIFF file to pseudo-translate
   -catalog:   (optional) XML catalog to use for processing

Copy Source to Target

You can copy the content of <source> elements to new <target> elements for all untranslated segments using class com.maxprograms.converters.CopySources from your Java code or using the provided scripts.

Running .\copysources.bat or ./copysources.sh without parameters displays help for copying source to target in an XLIFF file.

Usage:

   copysources.bat [-help] -xliff xliffFile [-catalog catalogFile]

Where:

   -help:      (optional) Display this help information and exit
   -xliff:     XLIFF file to process
   -catalog:   (optional) XML catalog to use for processing

Approve All Segments

You can set all <trans-unit> or <segment> elements as approved or final if they contain target text using class com.maxprograms.converters.ApproveAll from your Java code or using the provided scripts.

Running .\approveall.bat or ./approveall.sh without parameters displays help for approving or confirming all segments in an XLIFF file.

Usage:

   approveall.bat [-help] -xliff xliffFile [-catalog catalogFile]

Where:

   -help:      (optional) Display this help information and exit
   -xliff:     XLIFF file to process
   -catalog:   (optional) XML catalog to use for processing

Remove All Targets

You can remove<target> elements from all <segment> or <trans-unit> elements using class com.maxprograms.converters.RemoveTargets from your Java code or using the provided scripts.

Running .\removetargets.bat or ./removetargets.sh without parameters displays help for removing targets from an XLIFF file.

Usage:

   removetargets.bat [-help] -xliff xliffFile [-catalog catalogFile]

Where:

   -help:      (optional) Display this help information and exit
   -xliff:     XLIFF file to process
   -catalog:   (optional) XML catalog to use for processing

Export Approved Segments as TMX

You can export all aproved segments from an XLIFF file as TMX using class com.maxprograms.converters.TmxExporter from your Java code or using the provided scripts.

Running .\exporttmx.bat or ./exporttmx.sh without parameters displays help for exporting approved segments from an XLIFF file.

Usage:

exporttmx.sh [-help] -xliff xliffFile [-tmx tmxFile] [-catalog catalogFile]

Where:

    -help:      (optional) Display this help information and exit
    -xliff:     XLIFF file to process
    -tmx:       (optional) TMX file to generate
    -catalog:   (optional) XML catalog to use for processing

If the optional -tmx parameter is not provided, the TMX file will be generated in the same folder as the XLIFF file and its name will be the same as the XLIFF file plus .tmx.