GNOME Bugzilla – Bug 94298
Implicit iteration for operators
Last modified: 2004-12-22 21:47:04 UTC
I have a formula which works in MS Excel, but does not work in Gnumeric. It's a pretty ugly formula, but it extracts a number from a string. (i.e. 240 from PP240X) The forumula is: =MID(A1,MATCH(FALSE,ISERROR(1*MID(A1,ROW(INDIRECT("1:10")),1)),0),10-SUM (1*ISERROR(1*MID(A1,ROW(INDIRECT("1:10")),1))))*1 *Note that it is an array formula and should be entered with Ctrl+Shift & Enter When I do this in Gnumeric I get #VALUE. Thankfully, I didn't come up with this forumla myself, but found it at: http://www.geocities.com/davemcritchie/excel/join.htm If you either have another forumla which performs the same functionality and works in both environments, that would be great. However, I also wanted to make sure that you are aware of this possible bug. Cheers, Tommy.
This is hitting several limitations of 1.0.x 1) we did not support 1:10 as a reference to several rows Fixed in CVS 2) We did not support implicit iteration for array formulas Fixed in CVS patch will be in 1.1.9 3) indirect. hmm, this this is still kinda dodgey even in 1.1.x Hopefully will be solid for 1.2
Hmm, even cvs head still returns #VALUE! Can you submit a sample XL sheet with a working usage ? Thanks
Created attachment 11275 [details] Excel file example which works in MS Excel
I am confused about the comments listed. Is CVS another Spreadsheet applications? Also, I am currently unable to install Gnumeric 1.1.x due to my gcc compiler version. Is there a workaround or something else available? Thanks! Tommy.
Sorry for being unclear. CVS is a version control system. By saying it is fixed in CVS I mean that there is a publicly available patch in the development version. I di not reccomend using a 1.1.x version because it is under development and is not meant to be stable until the next even numbered release 1.2.0 loosely targeted at Dec. You can manually remove the gcc version check things will mostly work. However, some of the statistical routines end up with incorrect results as a result of some versions of that compiler. The problem will be gone as 3.x releases become readily available.
Well, we're getting closer. I can't close this in good conscience until we support implicit iteration for operators too.
Thanks for your persistence and dedication to getting this resolved. That's why I have faith in Open Source. :)
121440 is a bug report also reminding us of missing implict iteration for for operators. *** This bug has been marked as a duplicate of 121440 ***