This little project has grown quite a bit over the past few weeks, both from my own code and patches from others. This latest release sees some major improvements over the initial release.

csvDB now supports, INSERT, UPDATE, DELETE, and ALTER TABLE queries. As well as sub-queries in WHERE:

SELECT ... FROM table1 WHERE column IN (SELECT ...)

WHERE comparisons now support more than just AND equals, with a full range of not equals, less than, greater than, etc. comparisons. As well as OR for matching any of multiple value.

SELECT can now output it's results INTO OUTFILE to complement LOAD DATA INFILE. And multiple tables can be selected from with implicit and explicit JOINs.

Column and table selection has been greatly enhanced, with in-query aliases, better name resolution, and table.column format support.

Queries have received much optimisation; tests have shown queries that 0.1 took upwards of 30 minutes to complete are finished in a fraction of a second with 0.4.

The shell has seen a change to readline rather than the previous raw tty code, and now includes table, column, and keyword auto-completion. Additionally the output format of results has been change to keep columns and rows neatly aligned.

The build system has also been converted to GNU autotools, this has lead to support for nonstandard platforms such as windows thanks to some additional patches from contributers.

All in all, csvDB is growing quite nicely.
The customary screenshot, showing some of the new query features, and result formatting: