Re: trouble with my c/sql-code on SE 7.23

You cannot depend on the value in sqlca.sqlerrd[2] until the last row was returned from the server. I'm guessing that 92 is the number of rows that fits into a default sized (4K) communications buffer and so when the engine returns that first buffer to the ESQL/C library it already contains the last row and so the count has been updated. If there are more rows than fit into a single buffer, then the count in sqlca.sqlerrd[2] will not be updated until the buffer containing the last row is returned from the engine. You could expand the comms buffer size with the FET_BUF_SIZE environment variable or by setting the fetbusize global variable immediately after connecting to the database but before any DML is issued. You can set the buffer up to 32765 bytes which would hold over 1500 rows in a single physical fetch, but someday you might exceed that and then your reliance on this value will fail again. If you need the count before counting them yourself in the code you will need to execute an identically (or at least equivalently) filtered query that returns just the COUNT(*). You don't need the count to regulate the FETCH loop, you can do that using the "(sqlca.sqlcode == SQLNOTFOUND)" test or its inverse to manage the loop. For pre-sizing an array to fetch data into, you'll have to either go the COUNT(*) route, or dynamically resize the array as you go. I have many apps like that where I allocate an array big enough for most runs and if there is more data I reallocate the array with an additional block of N more rows and again if I've exhausted the extra space, etc. Art Art S. Kagel Advanced DataTools (www.advancedatatools.com) IIUG Board of Directors (art@iiug.org) See you at the 2010 IIUG Informix Conference April 25-28, 2010 Overland Park (Kansas City), KS www.iiug.org/conf Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Fri, Mar 5, 2010 at 10:22 AM, Peter Michel < proORGA-Leimen.without-this@web.de> wrote: > Hi, > the following code works great, > until the count of records is more than 92 8-(( > > $declare zeigerc cursor for select > chrst_charge, chrst_kurztext, chrst_status, chrst_ende, > chrst_benutzt, chrst_nr_intern > from chargestamm > where chrst_kategorie = $chrst_kategorie > and chrst_artikelnr = $chrst_artikelnr > and chrst_start <= TODAY and chrst_ende >= TODAY > and chrst_status in (" ", "X") > order by 5 desc, 4; > IFDBERROR("declare-c"); > > $open zeigerc; > IFDBERROR("open-c"); > > $fetch zeigerc into > $chrst_charge, $chrst_kurztext, $chrst_status, $chrst_ende, > $chrst_benutzt, $chrst_nr_intern; > IFSQLCODE IFNOTNOTFOUND("fetch-c"); > > anz = (short)sqlca.sqlerrd[2]; > > If the founded subset of records is <=92 the variable 'anz' has the > correct value (0 to 92) BUT is it >92 records the slqca.sqlerrd[2] is > zero and no error is shown (sqlcode is zero too) and - as expected - > there are all the wanted records present. > > Has anybody an idea? > > Cheers > Peter > > Für direkte email rot13 cebBETN-Yrvzra@jro.qr > oder entferne ".without-this" aus der Email-To. Danke! > _______________________________________________ > Informix-list mailing list > Informix-list@iiug.org > http://www.iiug.org/mailman/listinfo/informix-list >