XML 2 custom filetype: everything in bold, characters with diacritical marks not correctly processed

Question

Hello, community, 
 at my firm, we get XML files like the one attached for translation. For these files, I want to create a custom XML 2 filetype so we can see the RecordID attribute value in the DSI section in Studio. I managed to do this, but the filetype still has two problems: 
 
 it shows everything in bold 
 it does not correctly process characters with diacritical marks (even though entity conversion is enabled) 
 
 This behaviour changes when content processors are enabled: 
 
 Preview screenshot when NO content processor is enabled OR when it is enabled with the parser rule "//entry" OR when it is enabled defined by document structure information "sdl:cdata" (dsi information is shown, but everything is bold and characters with diacritical marks are not displayed correctly):

Preview screenshot when plain text processor is enabled inside CDATA element (text displayed correctly (not bold, characters with diacritical marks okay), but no dsi whatsoever):

What I'd like to have is the file being displayed as in the second screenshot, but with the structure information as in the first screenshot. Any help would be appreciated. 
 Thanks in advance. 
 Best regards, Katharina Studio version: Trados Studio 2022 - 17.0.6.14902 
 0245.Test.zip

Paul · Answer

Katharina Schuhmann 
 Katharina Schuhmann said: I did provide a sample in the .zip folder I attached to my post. 
 So you did... couldn't see it for looking! 
 Quite an interesting file and I went around the houses before we found the problem! I'll start with the problem... the file you are using has no XML declaration and it doesn't have a BOM (Byte Order Mark telling it the file is UTF-8). So when you open in Studio the file is actually seen as Baltic (Windows) and the incorrect encoding is used. The result... nothing you do will fix this problem of entities unless you fix in the file or change the encoding when you open it. 
 The last option is only available for single file projects and works by selecting UTF-8 when you open the file here: 
 
 You will then see this message: 
 
 Click yes and all is well :-) 
 
 We cannot change the bold I'm afraid as this is due to the text being in a CDATA section. It's only bold in the editor. 
 The next way to change it, seeing as you may have many files and working with a single file project is probably not optimal, is to pre-process the file and add an xml declaration to the start: 
 <? xml version="1.0" encoding="UTF-8" ?> 
 Easy enough to do and you could create a script to handle all the file in a project and then use a script to remove the declaration afterwards if you client cannot manage the files with this. 
 Another way is to add a BOM to your files. This may be even more preferable since you can handle multiple files in one go with this app and the change shouldn't be an issue for your customer: 
 https://appstore.rws.com/Plugin/58 
 You can find a short video to see how this works here: https://youtu.be/mJY212049WE 
 In the process of all of this I also tried to use the preview to workaround the issue and can do this: 
 
 It didn't help me with the workaround until I fixed the problem with the encoding, but I quite like it so here's the stylesheet in case you find it helpful too: recordID.zip 
 So I think this is solved apart from the bold bit, but I'm afraid there is no way around this since you cannot use the Document Structure for your rule, nor for the preview, if you use embedded content. But at least you seem to have a way forward.

Paul · Answer

Ana Benita Sanchez 
 ok - I can repro with your settings and if I recreate the filetype from scratch I cannot. It's very strange since these tags seem to be the line feeds coming from outside the element altogether. If you presss Ctrl+Shift+H they disappear... from view anyway, but they are still there. 
 Can you confirm how you created this filetype? Did you import settings from an old version? Have you tried recreating the filetype settings from scratch? I quickly rebuilt the file just to test and used the same embedded content processing method as you but I cannot cause this to happen. So I suggest you do the same and check the file at each stage of adding the rules back in case the problem is also caused by your regex rules somehow. 
 But do not try to take a shortcut and import a settings file or hack the settings file. Do it from scratch and let the software do the work using the latest filetype.

Paul · Answer

Katharina Schuhmann 
 Without a sample of your file we can't see why it may be all bold. But the entity conversion may be a problem we can't resolve as I think it requires work from the dev team to reslve this when working with embedded content filters. 
 But provide a SMALL sample file and we can take a look and see if we can resolve it.

Trados Studio > 1. Trados Studio

XML 2 custom filetype: everything in bold, characters with diacritical marks not correctly processed

Top Replies