Hi Anton, well the problem isn't really accessing the data as such, it's quite easy to use HDBC.ODBC for instance to read in the data in a very comfortable way. I was explicitly asking for Takusen because it make certain guarantees regarding resources, memory consumption and such. One of the files for instance has 298 k lines. A glance over sushi revealed that parsec is involved so I can only presume it is read into memory all at once. That would certainly be a problem. Thanks, Günther Am 31.08.2009, 19:47 Uhr, schrieb Anton van Straaten :

Günther Schmidt wrote:

...

My app then needs to read in dynamic data, a set of 4 CSV files, and I had been using Microsofts ODBC driver with HDBC.ODBC and that worked too. When I try to swap it for Takusen though I get an *** Exception: DBError ("HY","C00") 106 "[Microsoft][ODBC Text Driver] Optional Feature was not implemented Which is a bit of a shame, because using the ODBC text driver was much simpler than parsing the CSV files.

A simple but powerful way to access CSV files is via TxtSushi, which gives SQL access to CSV files, without needing ODBC:

Home page: http://keithsheppard.name/txt-sushi/ Hackage: http://hackage.haskell.org/package/txt-sushi

You'd have to use it directly, since afaik it won't work with Takusen, although a Takusen back end for it ought to be possible.

Anton