Hi Tom Imhof , I have reviewed this issue today, but unable to reproduce the issue that you are reporting.

Prerequisites

• Trados Studio 2022 SR1 - 17.1.6.16252
• GS: 15.1.9.12529 - GS 2020 SR1 CU09

Steps

1. Create a server based tm from GroupShare
2. Open the server based TM in Stduio & add some custom fields
3. Add a few TU entries & assign a custom field value to one or two of the TU's
4. Commit changes so that they are updated on the server.
5. Create a standalone app to connect and export the server based TM to tmx
   1. Documentation: How to export serverbased TM to TMX

Actual result

The custom field values are exported correctly, visible from the TU's in the TMX that was exported from the serverbased TM

Follow up

Tom Imhof  Can you pls provide the GroupShare version + Trados Studio version you are using?

Dear Patrick, 

many thanks for your answer. My issue is about the *context matching information* - not about custom fields. Custom fields - I agree - do survive the export. The context matching information in the TM is stored in these fields:

<prop-type="x-IncludesContextContent">True</prop>

This turns the feature on, I guess.

Then on segment pair level, you will find:

<prop type="x-Context"> etc. as highlighted in the screenshot:

These elements - when you export via the API from a server TM - are *not* exported. I am quite sure you can repro it. I tested against an older GS 2020 SR 1 CU7 (gsdemo.gs.sdl.com) as well as my own VM running GS 2020 SR1 8.1. I will test against CU9 as well, but I am quite sure it is not a Groupshare issue - it is a Studio API issue. 

Hello Tom Imhof 

Can you please please tell us, please, what parameters you send in the TranslationMemoryQuery.

Also, here is an example of parameters that works: 

var input = new TranslationMemoryEntityQuery()
{
Size = 10,
Index = 1,
IsProject = true,
OwnerId = GuidOwnerId,
IncludeChildResourceGroups = true,
ResourceGroupPath = "Location"
};

Also send paratmeter IsProject as true, so you'll get all TMs.

Can you, please, try this an let us know how it goes?

Hello Tom Imhof 

How TranslationMemoryQuery looks?

Here you have an working example: 

var query = new Sdl.LanguagePlatform.TranslationMemoryApi.TranslationMemoryQuery() {
ResourceGroupPath = UserSettings.Organization,
IsProject = true,
IsMain = true,
IncludeChildResourceGroups = false
};

Also, set IsProject as true.

Can you, please, try this an let us know how it goes?

Thanks!

Hello Paul,

many thanks, I will try this and get back to you on the results. 

Kind regards

Tom

Hi Patrick,

In the meantime I have tested the following two use cases using an older GS-Server

Running export using (classic) TranslationMemoryExporter:

--> Original problem (context information missing in the TMX) can be reproduced.

versus:

Running export with the new method ScheduledServerTranslationMemoryExport:

--> Original problem cannot be reproduced.

So, for the original problem, I have a solution at least and can carry on with my work :-) Many thanks for the decisive tipp about using the scheduled export! However, the issue should also not occur using the TranslationMemoryExporter method, even if it is not recommended as it is slower than the scheduled one.

Many thanks,

Tom

Hi Tom Imhof , thank you for confirming this (y). I confirm that I can also reproduce the original issue you reported against GS server CU9 using TranslationMemoryExporter.  I overlooked this as I've always used ScheduledServerTranslationMemoryExport  for server based memories as it is the more optimal approach, especially over slower network.

You're absolutely correct, the content of the exported TMX should be the same irrespective of the users choice in using the scheduled export or not.  Daniel Brockmann ,  I'll escalate this issue to the dev team.

Dear Patrick,

many thanks for escalating the issue. 

Could you let me know how to unzip the archive that is downloaded using the ScheduledServerTranslationMemoryExport method programmatically?

I tried various unzip methods incl. sharpziplib and .net system.io.compression.unziptodirectory as well as gzipstream and others, but they all fail on this type of archive. Which format is it in the first place? 

Many thanks for any tip!

Hello Paul - this is to confirm that this approach works with CU9 of Groupshare! Many thanks for pointing me to the workaround!

Hi Tom Imhof , name the download file with the extension .tmx.gz; I'm using https://7-zip.org/. It supports GZip compressed files, such as: export.tmx.gz

Dear Patrick,

yes I do know 7-zip, but I would like to unzip the gz archive *programmatically* and without installing an additional desktop app. Is there any sample code available to do this in c#/.net?

Please try the following and let me know how it goes.

string sourceFilePath = "compressed.gz"; // Replace with your source GZip file path
string extractedFilePath = "extracted.txt"; // Replace with the desired output file path

try
{
    using (FileStream sourceFileStream = File.OpenRead(sourceFilePath))
    using (FileStream extractedFileStream = File.Create(extractedFilePath))
    using (GZipStream gzipStream = new GZipStream(sourceFileStream, CompressionMode.Decompress))
    {
        byte[] buffer = new byte[4096];
        int bytesRead;

        while ((bytesRead = gzipStream.Read(buffer, 0, buffer.Length)) > 0)
        {
            extractedFileStream.Write(buffer, 0, bytesRead);
        }

        Console.WriteLine("File extraction complete.");
    }
}
catch (Exception ex)
{
    //TODO
}

Please try the following and let me know how it goes.

string sourceFilePath = "compressed.gz"; // Replace with your source GZip file path
string extractedFilePath = "extracted.txt"; // Replace with the desired output file path

try
{
    using (FileStream sourceFileStream = File.OpenRead(sourceFilePath))
    using (FileStream extractedFileStream = File.Create(extractedFilePath))
    using (GZipStream gzipStream = new GZipStream(sourceFileStream, CompressionMode.Decompress))
    {
        byte[] buffer = new byte[4096];
        int bytesRead;

        while ((bytesRead = gzipStream.Read(buffer, 0, buffer.Length)) > 0)
        {
            extractedFileStream.Write(buffer, 0, bytesRead);
        }

        Console.WriteLine("File extraction complete.");
    }
}
catch (Exception ex)
{
    //TODO
}

Hi Patrick,

the GZipStream method indeed did the trick! Many thanks for your help!

Cheers

Tom