Thank you!

Thanks for landing this! My name is Jakob Kummerow, email: jakob.kummerow@gmail.com

ahiemstra: Thanks! I don't have commit access, so I'd appreciate it if you could land this for me.

Minor update: size_t instead of int.

OK, here's a version that scales the buffer size as needed.

cfeck: Sure, an even better long-term fix would be to stop depending on a hard-coded buffer size. Interleaved reading and parsing would be one way to achieve that; another way would be a dynamic buffer size: start small, and if it's too small then allocate a new buffer that's twice as large and try again.