Oh, Simon Michael! I hadn't noticed your username until now.
> I have not used the SQL output.
I've used it to process the register output in a pipeline.
Sometimes, my `hledger reg` queries result in multiple quantities on some transactions. The terminal output leaves the columns of the date, transaction description and account blank, for pretty-printing purposes, I suppose. Those blanks complicate adding up the quantities per transaction using a tool like awk. There's also the JSON output format, but the depth of the objects around quantities makes the CSV still simpler to process, even though I need to process the quantities' text in SQL so they can behave like quantities.
> I have not used the SQL output.
I've used it to process the register output in a pipeline.
Sometimes, my `hledger reg` queries result in multiple quantities on some transactions. The terminal output leaves the columns of the date, transaction description and account blank, for pretty-printing purposes, I suppose. Those blanks complicate adding up the quantities per transaction using a tool like awk. There's also the JSON output format, but the depth of the objects around quantities makes the CSV still simpler to process, even though I need to process the quantities' text in SQL so they can behave like quantities.
Just sharing how one user uses hledger.