Hi all, Münt, Bernd pisze: >> Von: [email protected] [mailto:[email protected]] Im Auftrag > >> ok, I am late on that latest change. Please remove all the >> abort() statments then. > > I think it is not so simple when doing this on mosesserver. It was discussed > here last week that this might lead to memory leaks. > >> In general, it would be better to ensure that no faulty XML >> is generated in the first place. Graceful decay also leads to >> error that are hard to track down. > > I agree. But I sent phrases from a translation memory (TRADOS and TRANSIT) to > moses for machine aided translation. In 99% the XML is ok, but I found some > cases when it is not.
Wouldn't it be a better solution to simply pack the faulty xml in some additional tags (for example, pass it to XLIFF)? I know that it's not trivial to build XLIFF around non-valid XML as most XLIFF converters expect valid input but you can simply encode all < and > as > and < and you'll be all set. So I guess it's next to trivial, and in Trados translation memories you can also expect some additional non-valid XML garbage (like non-ASCII control characters). Regards Marcin _______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
