Issue 58: DB reindexing improve

I have two seriously notices: 1. When free space on disk is end, recoll start index DB "since zero". With my library data=160 Gb, it is not nice, wait again several days. 2. If index process stop, recoll will reindexing from begin,too. Make pieces, please. Thanks and sorry for bad english

About 1: you can set a maximum disk occupation in the GUI: Preferences- >Indexing configuration- >max disk occupation. This will ensure that the indexing stops cleanly when the maximum is reached. Maybe I should change the current default value for the threshold (none). If Xapian does run out of space I imagine that the index can be unusable on restart.

About 2: unless you are certain, and have the log file to prove it :), recollindex does not restart from scratch when interrupted. It will walk the whole file system indeed (looking for updates), but only files which have changed since the last pass will be reindexed.

I’ll close this issue, don’t hesitate to reopen it if what you are seeing does not match what I wrote above, but I’d need more specifics then: system and recoll versions, and log files please.

Cheers,

jf