tag:blogger.com,1999:blog-20155610.post5520775310020749355..comments2024-03-06T02:46:19.929+00:00Comments on Translation Tribulations: Complicated XML in memoQ: a filtering case exampleKevin Lossnerhttp://www.blogger.com/profile/14727800526216764023noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-20155610.post-88529785776877717502020-08-04T21:13:09.406+01:002020-08-04T21:13:09.406+01:00Thank you for this. I recently had this problem wi...Thank you for this. I recently had this problem with some files sent by one client. This was really helpfulAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-20155610.post-2705214308638772092018-04-04T14:13:51.418+01:002018-04-04T14:13:51.418+01:00Marek, I don't think it's luck in this cas...Marek, I don't think it's luck in this case. When I've seen bad files from CMS systems like this, it's typically a case of some IT worker who doesn't understand processes and standard interfaces very well and clumsily patches together his own hardly maintainable solution. At least this one wasn't using Excel as its exchange format. Converting to and from real XLIFF wouldn't offer me much in this case except an engineered process that I would have to pass on to a client who is just starting to come to grips with the technical aspects of memoQ filter customization, so inflicting Rainbow on them would not be kind. My personal geeky preference would be XSLT for the conversion, but that would be even less kind to the client. <br /><br />I agree that the Multilingual XML filter in memoQ might also ultimately be the route to go here, and the client was informed of such. However, that approach seemed overengineered for what I think was a first job with a new corporate client and no information about additional languages or other plans for the future. (In the meantime I have some info, and it sounds like an interesting grab bag of formats ahead.) My main objective in this case (the way I approached the filtering and the fact that I am blogging about it) is to start creating more public examples which include the other important aspect here: the need for multiple filters in sequence. These cases are ending up in my inbox with increasing frequency, and what seems to cause the most trouble in general for some of my agency clients these days is PMs or technical support staff getting their heads around this multiple filter problem. Individual translators have the same issue too, of course, but so far the cases with the most "layers" seem to hit agencies mostly. So I think this kind of case should be emphasized somewhere in the training workflows.......Kevin Lossnerhttps://www.blogger.com/profile/14727800526216764023noreply@blogger.comtag:blogger.com,1999:blog-20155610.post-55774059183218653322018-04-04T08:45:34.414+01:002018-04-04T08:45:34.414+01:00Good approach, but you were lucky there were no ac...Good approach, but you were lucky there were no actual translations in the target segments that you would have to retain. If you do have to deal with the real bilingual file, there are three approaches you can try:<br />1. Use memoQ's multilingual XML filter. It's decent and easy to modify to various languages once you have basic rules, plus you can import multiple languages at once if your XML contains more than two.<br />2. Use automated pre-processing with project template. Relatively simple script could add missing content and convert source file into real XLIFF (this trick can be used on multitude of bilingual source content).<br />3. Use Okapi-based Rainbow to generate XLIFF files - somewhat similar to approach 2, but with different tool and may be easier in some cases.<br />Wasatyhttps://www.blogger.com/profile/12415028316778775171noreply@blogger.com