This series is an alternative approach to the patches Rich posted to ticket 3024. At first I thought the patches looked okay, but after thinking about it I wasn't happy with the wasted cycles tokenizing and iterating through to_chars for every index entry every time.
This series instead adds a a new datatype for the to_chars and status_chars options, that pre-tokenizes them. It is hooked into the unset, reset, and set functions in init.c, so the tokenizing takes places only once (or as the options are changed). The first patch creates the new type and structure. The second patch converts to_chars and status_chars over. I'm posting this for early review. I'm tired and need to take a closer look at the patches when I'm fresh tomorrow, but welcome feedback! -- Kevin J. McCarthy GPG Fingerprint: 8975 A9B3 3AA3 7910 385C 5308 ADEF 7684 8031 6BDA
signature.asc
Description: PGP signature
