Re: Fixing busted citext function declarations

On May 11, 2015, at 5:01 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:

> Me too.  Something fell through the cracks rather badly there :-(.
> Would you check your commit history to see if anything else got missed?

Let’s see…

In https://github.com/theory/citext/commit/4030b4e1ad9fd9f994a6cdca1126a903682acae4 I copied your use of specifying the
fullpath to pg_catalog function, which is still in core. 

In https://github.com/theory/citext/commit/c24132c098a822f5a8669ed522e747e01e1c0835, I made some tweaks based on you
changeyou made to some version of my patch. Most are minor, or just for functions needed for 8.4 and not later
versions.

In https://github.com/theory/citext/commit/2c7e997fd60e2b708d06c128e5fd2db51c7a9f33, I added a cast to bpchar, which is
incore. 

In https://github.com/theory/citext/commit/cf988024d18a6ddd9a8146ab8cabfe6e0167ba26 and
https://github.com/theory/citext/commit/22f91a0d50003a0c1c27d1fbf0bb5c0a1e3a3cadI switched from VARSIZE_ANY_EXHDR() to
strlen()at your suggestion. Also still there. 


Anyway, those are all from 2008 and pretty much just copy changes you made to core. The return value of
regexp_matches()is the only significant change since then. So I think we’re good. 

Best,

David\