Currently in the code, the only indexes in ES that can be used with 'scan' are: 647 our %scan_field_convert = ( 648 'ti' => 'title', 649 'au' => 'author', 650 'su' => 'subject', 651 'se' => 'title-series', 652 'pb' => 'publisher', 653 ); This is not apparent to the end user, and causes confusion. We need to broaden support, or limit the displayed list of options for 'scan indexes'
Limiting the list when Elasticsearch is used makes sense to me. Should we hardcode the list or would it be possible to look-up which of the indexes are configured to work? I would support both fixes, we have the most important ones covered I think.
I agree that the use of the scan feature is somehow obscure now. I couldn't find a clear explanation in the doc about it. In the code it is more or less clear, but for the user -- not. How about sending a user to the page like: /cgi-bin/koha/catalogue/search.pl?advsearch=1&idx=au&q=a&scan=1&sort_by=relevance after checking the "Scan indexes" checkbox? Currently this page is available only as a result page. And the list of the indexes with the scan/browse feature is there. Or even better: how about if we made a separate "Scan [browse] indexes" search page, as a third search page (Advanced search, Item search, Browse indexes)?