On Wednesday, August 31, 2011 06:33:32 am Thomas Ganss wrote: > John, > > some months ago I was hacking on my own on that front and > found Roger's stuff a few days afterwards > (typical, never do the really smart things first like search the lists...) > and changed to his code, working to read vfp dbf's > (after making sure they were vfp6 compliant.) > > IIRC I used some [quasi-standard ?] DBAPI calls to generate the schema. > Was only basic stuff involved/needed, no DBC in the picture > > The PyODBC cursor AFAIR was not made out of "dict-records" but > conformed to DBAPI. > > I got some results, but then came work from a totally different direction > and I never followed up to make it a rounded package was only needed as > a demo. > > HTH > > thomas > > John Fabiani schrieb: > > On Tuesday, August 30, 2011 03:07:13 pm Roger Gammans wrote: > >> On Tue, 2011-08-30 at 14:08 -0700, John Fabiani wrote: > >>> Hi, > >>> I have a little time available and I thought I would look into using > >>> pyodbc with Dabo. > >> > >> John, If I haven't mentioned it before it might be worth you looking at > >> http://hg.backslashat.org/repos/dabo-odbc/ which is contains > >> a set of patchs to an ancient dabo to do this. > >> > >>> The first issue is the connection string. The pyodbc does NOT > >>> > >>> follow the "host", "user", etc. that we have been using for the > >>> connection. For example to connect to an MS SQL database on windows you > >>> would use something like: > >>> DRIVER={SQL Server};SERVER=cloak;DATABASE=test;UID=user;PWD=password > >>> > >>> > >>> And of course the connection string would be different for other > >>> platforms and database engines. At first I thought I would just change > >>> CxnEditor to deal > >> > >> That's what I did locally and added a driver and dsn attribute. It was > >> good enough for what I needed so I haven't mantained it. But it would > >> probably help you in the right direction. > >> > >> The important patches are :- > >> http://hg.backslashat.org/repos/dabo-odbc/rev/af94151ed49a > >> http://hg.backslashat.org/repos/dabo-odbc/graph/aeb4a75ba49f > >> > >> and > >> http://hg.backslashat.org/repos/dabo-odbc/rev/eb192fd4cd54 > >> > >> But you could just the db/dbOdbc.py from the last one and hack from > >> there. > > > > Roger, > > > > thanks for the code - I'm sure I'll use it in the near future (I've > > actually haven't started yet). > > > > After reading some of the code I do have a few questions! > > Why did you find it necessary to create the "Cursor" class? It appears > > the pyodbc has pyodbc.Cursor class. Did you find it did not match the > > Dabo requirements - I ran into this with the old pymssql and was unable > > to get it to work - Ed finished the job. > > > > You did not use "INFORMATION_SCHEMA" to access table, field information. > > I was planned on using the INFORMATION_SCHEMA as a universal way of > > accessing the system info. Of course I'm under the impression that all > > the major database engines have it available. > > > > Last the code implies that you were ONLY connecting to MSSQL - is that > > correct? Did you try connecting to other database engines? > > > > Johnf > [excessive quoting removed by server]
_______________________________________________ Post Messages to: [email protected] Subscription Maintenance: http://leafe.com/mailman/listinfo/dabo-dev Searchable Archives: http://leafe.com/archives/search/dabo-dev This message: http://leafe.com/archives/byMID/[email protected]
