my ancient open sourced csv parser has now been updated to NOT read directly from a binary stream but to use an internal buffer
this should make it perform better
I’ve tested it agains several csv’s I generated
a small one of 368 K
a medium one of 25Mb
and one of 1.24Gb
As well I factored out the buffered binary input stream into a separate github project
Hard to believe I wrote this nearly 15 years ago
Now there are piles for JS, C# etc
There werent when I originally wrote it