travisgoodspeed / indextool Goto Github PK

It would be nice to quickly fetch a list of the least indexed articles, so that those can be given more coverage. The following query is pretty much what I want, except that it doesn't flag those articles with zero entries. (And maybe it shouldn't, if they are side pieces or otherwise not worth indexing.)

select filename, count(name) as c from indices group by filename order by c asc limit 20;

Warn when the same index work is used twice with different capitalization.

Indices are probably better for this.

select distinct a.name, b.name from
indices a inner join indices b on (lower(a.name)==lower(b.name) and a.name!=b.name);

Entries would be a stern warning.

select distinct a.name, b.name from
entries a inner join entries b on (lower(a.name)==lower(b.name) and a.name!=b.name);

Using the FTS4 table, we should be able to present text search results with a snippet near the relevant word. This might be more efficient for the poor writer than having to load each file individually in Emacs.