Really Strategies
RESOURCE CENTER
“Really Strategies provides us with the third-party expertise we need.”
USATODAY.com
Getting from Quark to XML: One Publisher's Path
May 2003Many publishers are faced with the question of how to get XML output from existing print production processes. There is no general right or wrong answer, but the approach depends on a number of factors, including in-house technical resources and specific workflow requirements.
Congressional Quarterly (CQ), a Washington D.C.-based publisher of government and politics information first addressed this issue in 1999. CQ needed to output XML for two of its print magazines, CQ Weekly and CQ Today—both managed in Quark Publishing System (QPS)—but did not find any tools on the market that could meet their requirements at that time. CQ wanted to introduce minimal changes in the current print workflow to allow journalists and editors to meet their print deadline with no extra interference. At the same time, they needed to move articles to the web via XML as soon as possible.
At the time, no tools on the market offered the ability to batch process articles based on status "behind the scenes" without interruption to production staff. Additionally, CQ was unhappy with the ability of third-party tools to map between Xpress tag markup (the text version of Quark XPress files) and XML. "The products out there were limited to mapping one Quark stylesheet to one XML element, without much ability to control the document's overall structure or to capture the richness of the printed story," said Peter Roybal, associate managing editor. "We also would have needed to assign a production person to export and batch convert the files, which didn't fit with our real-time news operation."
So CQ built its own process to do the job. Even in late 2001, when CQ upgraded QPS and was faced with the task of rewriting the processes, CQ choose to stick with a home built approach because the tools on the market didn't look promising and they didn't have time to go through a learning curve with a product that might not fit their needs.
The CQ process uses Callas Software's AutoPilot to automatically extract articles out of QPS when those articles are assigned a specific status indicating they are ready for the web. Scripts convert the Xpress tag mark up to XML in a two-part process. First, they create XML that directly mimics the Xpress tag markup, but using XML syntax. Another process then transforms this XML to the DTD needed by CQ's content management system.
Included in the process is a mapping file that matches the XML elements created from Xpress tags to XML elements required by the DTD. "The mapping file is easy enough for our production staff to update when we create new stylesheets in Quark, which happens frequently. It also lets us do some post-processing, such as stripping out the symbol we put at the end of stories in the magazine." Roybal said. "We can do all of this without touching the code."
"The bottom line is that the homegrown conversion process results in clean XML without burdening our editors or production staff," Roybal said. "It's fast, reliable and means we don't have to touch about 90 percent of the stories in XML - they just go straight from Quark to XML to the web."
But today, CQ says if they were just starting now, they might come to different decision on building a process or buying a package, primarily because the packages available on the market today do a much better job than those of only a year ago.
In an upcoming issue we will look at a publisher who took the other path, that of purchasing a third-party product to do the job.
