I'm crawling "lots" (100K+) of excel files using spreadsheetgear, however i found out that when Spreadsheetgear hits a chart with lots of points, everything breaks loose: memory isn't released & it takes up a lot of memory as well as being very slow.
For example i have a 63MB excel file, containing 35 charts with 96K points each, getting the bitmap for each of those is taking up 100MB+ per and does not get released (at least not untill i close the worksheet). If i let my program run on all charts i end up with >9GB commited! And i definately can't ask my customers to upgrade to the 32GB+ required to support all files.
Are there any alternatives to doing this? I'd be fine with skipping files that would be slow & tried to do this (checking the count on Points in seriescollection) however it seems that Points access is the issue so this didn't help (just checking points.count seems to load all the points as well as their associated data!).
Any help is most appreciated, either to be able to grab the graph in an alternate way (with spreadsheetgear OR with a library that doesn't require anything installed AND supports all excel file formats) or a way to check for graphs that have too many points without using points.count.
It's a bit of a wild shot asking something so specific but i'm really lost there.