Automating InDesign with Regular Expressions

Chapter 8. Look Before You Leap

All the script illustrated earlier in this short cut assumed that once the regex had been matched against a text, all the matches could be processed. But this will not always be the case. Some compound words, for instance, need to be linked with an en dash rather than a hyphen. Examples are English–Polish dictionaries and the Keenan–Comrie hypothesis. The latter does not contain a double-barreled surname, but is a certain hypothesis proposed by two people, Keenan and Comrie. It would be wrong to assume that in any two words with initial capitals linked with a hyphen, this hyphen could be changed to an en dash. Not only would double-barreled names such as Robin Knox-Johnson be included but also combinations such as Afro-Asiatic and North-East.

To deal with situations like this we take two steps. The first step is to collect all instances and display them in a new document. We then inspect this list, discarding all hyphenated and init-capped words that should stay hyphenated. The remaining list is then processed by a second script. Here is the first script:

 doctext = documentContents( app.activeDocument ); found = doctext.match( /[A-Z]\w+-[A-Z]\w+/g ); if( found != null ) newDoc().contents = deleteDuplicates( found ); function deleteDuplicates( array ) { var str = array.sort().join('\r')+'\r'; str = str.replace(/([^\r]+\r)(\1)+/g,'$1'); return str.replace(/\r$/,''); } // start a new document, place a text frame on the first page, // and return that ...