Hi Jospeh, > Recently, I needed to parse some rather large ( 1+ GB) csv files and > found the performance of the csv library lacking. The attached patch > addresses the performance issues and simplifies the implementation. The > api remains the same but I have changed the library so that it preserves > leading and trailing whitespace for fields. The prior version stripped > these, but I now believe that the parser should not do so. I have also > updated the gflu example and recette so that they function correctly. Wonderful. I will apply it at once. BTW, I'm almost done migrating from Mercurial to Git. We will be soon able to use pull-requests for patching or adding new stuff.
Sincerely, -- Manuel
