Change query IDs in CSV results

[nck-mlcnv]

The query summary of a task stores the query IDs in the format <querysource-hashcode>;<query-id-inside-of-querysource>. which
may working with the file a bit more difficult. The hashcode should be instead written in a separate column.

Example:

queryID     |
------------|
1218510187:0|

should be instead:

queryID     | query-source |
------------|--------------|
          0 | 1218510187   |

[bigerl]

Do you have a example where it is specifically problematic? Because, on the other hand, this makes queryID a unique key for queries even if you have results from different sources in one CSV fil.

[nck-mlcnv]

Not a specific case, but it just seemed to make more sense to me to split up the query id and the hash of the query-source (also in terms of normalising the table). You could still use both values as the key, as opposed to solely using the queryID row. It was just some random thing that I noticed, I don't know if it is really an issue.

[bigerl]

query-index	source-hash	query-hash
0	1218510187	8263967382

I would suggest to have these columns then.

queryID should generally be replaced with query-index together with, where necessary source-hash. query-hash is an additional information for easier matching of queries within a source or across sources.