davide fezzardi 

I don't know if the AI helped you as you left it marked as a verified answer, but just in case you want to persist with Excel I think you need to use Power Query, like this:

• Data → Get Data → From File → From XML
  Browse to your renamed file.

• Navigator dialog opens – you will usually see a single top‑level table (often called Table0 or xliff).
  Tick it and choose Transform Data (not Load yet).

• Power Query Editor opens.
  You will see one or more columns of type Table or Record.
  For each such column:

  • Click the ⤡/expand icon in the column header.

  • Untick “Use original column name as prefix” (tidier).

  • Tick Select All Columns.

    • Attributes appear with an @ prefix (e.g. @id, @source-language) – keep them all.

  • Press OK.

• Repeat the expand step as many times as necessary (XML is hierarchical, so you may have to drill down several levels until every column is plain text/number). Excel always retains attributes as columns, so nothing is lost.

• When everything you need is laid out in tabular form, press Home → Close & Load.
  Power Query writes the results into a new worksheet.

I got those instructions with AI, then tested them, and they are pretty good.  But it's a ridiculously labour intensive process, especially if you have lots of files.  So I asked the AI to find a quicker way with an Excel VB macro.  I liked it so much I put it here:

https://github.com/paulfilkin/VB-scripts/tree/main/ExtractTranslationAttributes

When you run it you are asked to select the sdlxliff, then it tells you how many translation units were imported and displays a nice clean solution specific to the translation like this:

I guess you could easily adapt it from this starting point if you needed to.

I also think Python would probably be a better solution in the end, but I was interested to see how to get it into Excel.

davide fezzardi 

I don't know if the AI helped you as you left it marked as a verified answer, but just in case you want to persist with Excel I think you need to use Power Query, like this:

• Data → Get Data → From File → From XML
  Browse to your renamed file.

• Navigator dialog opens – you will usually see a single top‑level table (often called Table0 or xliff).
  Tick it and choose Transform Data (not Load yet).

• Power Query Editor opens.
  You will see one or more columns of type Table or Record.
  For each such column:

  • Click the ⤡/expand icon in the column header.

  • Untick “Use original column name as prefix” (tidier).

  • Tick Select All Columns.

    • Attributes appear with an @ prefix (e.g. @id, @source-language) – keep them all.

  • Press OK.

• Repeat the expand step as many times as necessary (XML is hierarchical, so you may have to drill down several levels until every column is plain text/number). Excel always retains attributes as columns, so nothing is lost.

• When everything you need is laid out in tabular form, press Home → Close & Load.
  Power Query writes the results into a new worksheet.

I got those instructions with AI, then tested them, and they are pretty good.  But it's a ridiculously labour intensive process, especially if you have lots of files.  So I asked the AI to find a quicker way with an Excel VB macro.  I liked it so much I put it here:

https://github.com/paulfilkin/VB-scripts/tree/main/ExtractTranslationAttributes

When you run it you are asked to select the sdlxliff, then it tells you how many translation units were imported and displays a nice clean solution specific to the translation like this:

I guess you could easily adapt it from this starting point if you needed to.

I also think Python would probably be a better solution in the end, but I was interested to see how to get it into Excel.

 davide fezzardi 

In fact I realised how much this was missing and now I started it I had to make it somewhat more comprehensive.  So the new version - same link - will capture a lot more:

.