Hello !<p>I was thinking for a long time that sqlite3 and fossil are nice projects but
lacking a central repository that anyone can clone and try to extend/hack
then and share with others like github makes cooperation harder than it need
be. I saw several times people mention on mailing lists that they have some
extensions/modifications to sqlite and offering it to ones that ask for.<p>Then I found this https://github.com/mackyle/sqlite that follow
https://www.sqlite.org with less than a day lag, would be nice if the sqlite
author create and maintained a mirror on github to centralize third part
works/contributions, there was a section on sqlite/wiki for contributors but
it was closed long ago.<p>So here is my fork on github with the intention to implement "SQL PREPARE"
following the postgresql implementation
http://www.postgresql.org/docs/9.5/static/sql-prepare.html it has a simple
public user facing interface "PREPARE/EXECUTE/DEALLOCATE", I followed the
actual implementations of virtual tables and triggers to get the parser and
dummy functions to be called by the parser, so far it seems to accept and
parse something like this:<p>PREPARE astmt(id integer, name text) AS select 1 as id, 'dad' as name;<p>EXECUTE astmt(1, 'blue');<p>DEALLOCATE astmt;<p>Now I need to implement the empty functions to actually make it do the real
work.<p>I'm asking and inviting anyone interested to help implement this with the
same license as sqlite.<p>Here is the links to my fork, go there fork it and let's share the
implementation of it.<p>https://github.com/mingodad/sqlite/tree/sql-prepared<p>https://github.com/mingodad/sqlite/commit/3b53199a9d853f98946ffa6ea665fc7357fc4be7<p>I chose this as the first attempt because it has interesting characteristics
it will open the road for a kind of stored procedure because we can pass
arbitrary parameters to then and also they should return zero or more rows
with varying columns.<p>Cheers !