Custom XML filetype adds unexpected spacing in target files

Question

Hi, 
 
 I created a custom XML filetype to manage a fileformat. The filetype works perfectly fine in input: 
 
 it correctly recognizes the file structure 
 it correctly extracts translatable content 
 it correctly converts entities into placeholder 
 it generates the target file (here comes the issue) 
 
 I imported it in the app the file came from to find that some of the text went missing. Comparing the source and target xmls I found out that Trados Studio add some spacing into closing tags that compromises the import. 
 Here you can see a comparison: 
 
 On the left you can see: 
 original closing marked in green boxes 
 source text in pink box 
 
 On the right: 
 The proper target text in pink box 
 Some weird additional spaces in red boxes 
 
 NOTE: the affected character are not extracted as translatable content! Even using the "show all content" filter they are not visible to the user 
 
 I manually removed them, and it worked! I checked the filetype settings, and apparently none of the settings is related with spacing but "Whitespaces" options, which I didn't edit: 
 
 I also checked the "Writer" options, since this is definitely something "written" in the target files:

For the time being, I'm batch finding and replacing this usign a tool, but it would be definitely better to have this fixed.

RWS Community AI · Accepted Answer

Hello, 
 The issue you're experiencing with unexpected spacing in target files when using a custom XML filetype in Trados Studio is likely due to the technology used for parsing and writing the file. Trados Studio uses Microsoft .NET XML libraries, which can sometimes add additional whitespace to self-closing tags, change empty tag pairs into self-closing tags, or alter the order of attributes. 
 While this behavior is inherent to the Microsoft .NET XML libraries and not directly controllable through Trados Studio, there are a few potential workarounds you might consider: 
 Step 1: You could try using a different XML parsing and writing library that doesn't add the additional whitespace. This would likely involve some custom development work. 
 Step 2: You could continue to use your current method of batch finding and replacing the additional spaces using a tool. While this is not ideal, it is a reliable way to ensure the spaces are removed. 
 Step 3: You could consider using a post-processing script to remove the additional spaces after the target file has been generated by Trados Studio. 
 I understand that this issue can be frustrating, and I hope one of these suggestions can help you find a suitable solution. 
 Best regards, 
 RWS Community AI

Paul · Answer

Julian Hamm

Yes... apologies for not coming back. I had an answer and forgot to tell you! This is the outcome:

We fully embrace semantic functionality of the XML - this means for example, that these sort of things will not be considered bugs:

reordering attribute values,
automatic normalization of the whitespaces between element attributes (does not apply to attribute values),
whitespace changes outside of the content,
changing empty tags to selfclosed (<a></a> to <a/>)

This problem you have described fits into the "whitespace changes outside of the content” category, so our development team will not do anything about this.

In general I am of the opinion that there is "being technically correct", and "being customer correct", and sometimes I think we can be a little too rigid when it comes to handling customer specific data where the solution a particular customer may use is not capable of properly handling valid XML. However, in this specific scenario I do think that the problem here is one that your customer should address. The XML specifications, and plenty of discussion topics on technical forums discussing this sort of issue, conclude that whitespace is irrelevant and all apps should parse the file irrespective of whether there is an extra whitespace there or not.

I think, if this is problem that your customer won't address then I would run a script and just add the whitespace back into the files. Probably trivial enough and given they seem to be able to handle the whitespace you could add this as a matter of course into every self-closing tag without having to do this manually for each file.

Paul Filkin | RWS Group

________________________
Design your own training!
You've done the courses and still need to go a little further, or still not clear?
Tell us what you need in our Community Solutions Hub

Paul · Answer

Luca Pesenti 
 Indeed I also cannot prevent these spaces being added. Even adding xml:space="preserve" to the xml seems to be ignored. I created a support ticket so support can take a look and if verified they will log a bug for resolution. 
 Case Details - 00830442 for reference.

Trados Studio > 1. Trados Studio

Custom XML filetype adds unexpected spacing in target files

Top Replies