Page 4 of 4   « previous page

Within the database for the CMS, the content was stored as segments of HTML. The Web pages were generated by assigning the HTML segments to placeholders on the page. Listing 5 shows an example of an HTML segment stored within the CMS.

We made use of in-line tags to preserve the formatting information from the HTML segment and to ensure that the file created was a valid, well-formed XLIFF file (see Listing 6).

The XLIFF files that were transported to and from the CMS were translated using a software translation tool. These tools use Translation Memory technology in order to reuse previous translations. The editor is designed so a translator can only edit those parts of the XLIFF that relate to translations. What was striking during this project was how stable XLIFF was as a localization format. At that time, we needed to localize problematic HTML files. Using XLIFF assured the technical quality of the end delivery.

A Brief History of XLIFF
XLIFF got its start as an informal group of like-minded localization and globalization professionals based in Dublin, Ireland who recognized that the lack of a standard localizable resource container added significant tools, development, and project management costs to the localization process. Present at the kickoff meeting were Novell, Oracle, Sun, and Berlitz (now Bowne Global Solutions). The group shortly thereafter welcomed Lotus/IBM, Moravia IT, RWS Group, and Lionbridge. Initial discussions did not define the core technology for the format, and the group was open to any ideas, including flat or binary data files. Once preliminary research was completed, the group concurred that XML solved our technology requirements and chose it for the core technology.

Our schedule was aggressive, and the 1.0 draft specification was completed within six months. Once the core deliverables were completed, the group began preparing the specification for public distribution throughout the industry for comment. We recognized that we lacked the organizational and intellectual property framework and procedural expertise that was needed to support the advancement of our work. So in the autumn of 2001, the XLIFF group commenced an investigation of consortia as potential homes for our work. Specifically, we researched OASIS, W3C, and LISA, and selected OASIS because of its XML focus, and because it provided us with a turnkey infrastructure within which to operate. A secondary but important factor was that many of our member companies were already members of OASIS. Over the life of the TC, corporate participation has included: Bowne Global Solutions, Convey Software, HP, IBM, Lionbridge, LRC, Microsoft, Moravia IT, Novell, Oracle, RWS Group/Enlaso, SAP, SDL International, Sun Microsystems, TRADOS, and Xerox. The educational institution Localisation Research Centre of the University of Limerick is also a member of the TC. Individual members have included Doug Domeny, Paul Gampe, Enda McDonnell, Florian Sachse, Bryan Schnabel, Shigemichi Yazawa, and Andrzej Zydron.

The XLIFF 1.0 Specification was submitted to the Technical Committee as the starting point for its work, and was reviewed and approved by the TC on April 15, 2002 as an OASIS Committee Specification. The TC decided not to submit the 1.0 specification as an official OASIS standard, choosing instead to focus on completing revision 1.1 to deliver key improvements, including customization and extensibility using private namespaces, embedded XLIFF, and improved data validation via XSD rather than 1.0's DTD. The XLIFF 1.1 Specification was approved as an OASIS committee specification on October 31, 2003.

The TC is currently authoring a number of profile specifications to help tools developers and implementers implement XLIFF 1.1 consistently. The HTML Profile has been completed and will be made available publicly, and the final draft of the PO/POT Files Profile shall be made public shortly. Profiles are being authored for Java Resource Bundles and other proprietary resources formats, and public drafts will be published within the next quarter. The TC is also finalizing guidelines for XLIFF's support of segmentation - the process of optimizing the reuse of previously translated discrete logical or linguistic units of translatable content. After the profiles and segmentation guidelines are published and publicly reviewed, the TC plans to submit the XLIFF 1.1 spec and related profiles to the OASIS standards review process for consideration as an official OASIS standard.

Conclusion
XLIFF is an emerging standard that provides a structured interchange format for localization data. It is increasingly used as a mechanism by many large companies to optimize enterprise localization workflow processes. The open source community is also developing tools and processes that use XLIFF. The XLIFF TC at OASIS is continuing to work on improving the standard. We welcome those with an interest and expertise in this area to either join and actively participate in the TC or to comment on our work.

References

• XLIFF TC Homepage: www.oasis-open.org/committees/ tc_home.php?wg_abbrev=xliff
• Comments Submission Form: www.oasis-open.org/committees/comments/ form.php?wg_abbrev=xliff
• XLIFF 1.1 Committee Specification: www.oasis-open.org/committees/xliff/ documents/cs-xliff-core-1.1-20031031.htm
• XLIFF 1.1 Schema: www.oasis-open.org/committees/xliff/ documents/xliff-core-1.1.xsd
• XLIFF TC Whitepaper: www.oasis-open.org/apps/group_public/ download.php/3110/XLIFF-core-whitepaper_1.1-cs.pdf
• XLIFF TC Frequently Asked Questions (FAQ): www.oasis-open.org/committees/xliff/faq.php
• XLIFF - The XML-based Open Standard for Localisable Content (PowerPoint): www.localisation.ie/publications/presentations/2004/ Conference/Presentations/TonyJewtushenko.ppt
• IBM Developer Works: "XML in localisation: Use XLIFF to translate documents": www-128.ibm.com/developerworks/library-combined/x-localis2/
• Sun's Java.Net open-language-tools Project: https://open-language-tools.dev.java.net/
• Bowne Global Solutions Whitepaper "An introduction to XLIFF": www.bowneglobal.com/english/media/WP_XLIFF_US_0505.pdf
• XLIFF Tools Open Source Project: http://xliff-tools.freedesktop.org/wiki/
• XLIFF Support in KBabel: http://developer.kde.org/summerofcode/xliff.html
• OASIS: www.oasis-open.org
• Translation Web Services Technical Committee: www.oasis-open.org/committees/tc_home.php?wg_abbrev=trans-ws
• Localisation Research Centre, University of Limerick: www.localisation.ie/
• Opentag: www.opentag.com/xliff.htm
• Localisation Industry Standards Association: www.lisa.org/
• W3C i18n Activity: www.w3.org/International/

Page 4 of 4   « previous page