[sqlite] speeding up an FTS5 search with a JOIN

2020-03-07 Thread P Kishor
I asked this question on Stackoverflow with not much success, and a suggestion to ask it on the list. So here I am. I have two tables, t1(id, t1Id, … other cols …, fullText) and a FTS5 virtual table vt1(t1Id, fullText) ``` sqlite> EXPLAIN QUERY PLAN ...> SELECT Count(*) as num FROM t1 WHERE

Re: [sqlite] speeding up an FTS5 search with a JOIN

2020-03-07 Thread Dan Kennedy
On 7/3/63 14:58, P Kishor wrote: I asked this question on Stackoverflow with not much success, and a suggestion to ask it on the list. So here I am. I have two tables, t1(id, t1Id, … other cols …, fullText) and a FTS5 virtual table vt1(t1Id, fullText) ``` sqlite> EXPLAIN QUERY PLAN ...>

Re: [sqlite] Trigger name missing

2020-03-07 Thread Keith Medcalf
You mean like take a "boo" at the defined triggers? select name, tbl_name, sql from sqlite_master where type='trigger'; would pretty much make clear that the defined trigger is not what you thought it was ... create table x(x); create trigger after insert on x begin select 1; end; select

Re: [sqlite] Trigger name missing

2020-03-07 Thread John G
Would it be possible to create an SQL verification program, which just like 'sqlite3_analyzer' and 'sqldiff' could be run separately? It could *warn* about apparently incompletely defined triggers and other possible pitfalls. Then developers could use it before installing the next version of

Re: [sqlite] Bug in SQLite version 3.31.1 and 3.32?

2020-03-07 Thread Joe Mucchiello
I just want to point something out that might help the original poster. On Saturday, March 7, 2020, 7:00:21 AM EST, sqlite-users-requ...@mailinglists.sqlite.org wrote: > > 1.  NULL is NULL = Yes, True, > 2.  NULL is FALSE = Nope, False. > 3.  NULL is TRUE  = Nope, False. > 4.  NULL is

Re: [sqlite] Trigger name missing

2020-03-07 Thread Simon Slavin
On 7 Mar 2020, at 4:49pm, John G wrote: > Would it be possible to create an SQL verification program, which just like > 'sqlite3_analyzer' and 'sqldiff' could be run separately? > It could *warn* about apparently incompletely defined triggers and other > possible pitfalls. The shell tool has