I added an option to buildwatson for only indexing records which match the filter applied during the search.register process.
This option can be selected by ./manage.py buildwatson --slim
I see the prudence of indexing all records to prevent holes in the search index when modifying the register filter, so I left that behavior as the default. On the other hand, I have some tables that have millions of records, including legacy data, where I only need to search 1/10-2/3 of the data, so this should save a significant amount of time.
I added an option to
buildwatson
for only indexing records which match the filter applied during thesearch.register
process.This option can be selected by
./manage.py buildwatson --slim
I see the prudence of indexing all records to prevent holes in the search index when modifying the register filter, so I left that behavior as the default. On the other hand, I have some tables that have millions of records, including legacy data, where I only need to search 1/10-2/3 of the data, so this should save a significant amount of time.