On Tue, Jun 23, 2015 at 02:40:48PM -0400, Jean-Philippe Ouellet wrote: > On Sat, Jun 20, 2015 at 03:01:18PM +0200, Reyk Floeter wrote: > > there is some great interest in getting support for rewrites > > What do people think of something like our tftpd(8)'s -r > > -r socket > Issue filename rewrite requests to the specified UNIX domain > socket. tftpd will write lines in the format "IP OP filename", > terminated by a newline, where IP is the client's IP address, and > OP is one of "read" or "write". tftpd expects replies in the > format "filename" terminated by a newline. All rewrite requests > from the daemon must be answered (even if it is with the original > filename) before the TFTP request will continue. By default > tftpd does not use filename rewriting. > > I was working on a patch to bring it to httpd but ran out of free time. > Thought I'd pass the idea by you anyway. > > I think it's a sweet spot of a minimum incrase in complexity and maximum > incrase in flexibility. Then people could plug in whatever they wanted: > be it trivial string substitutions, guaranteed safe regexes with re2[1], > potentially unsafe regexes with pcre, or even database lookups or whatever. > > [1]: https://swtch.com/~rsc/regexp/regexp3.html >
I have no oppinion about the use in tftpd, but it would sound weird for httpd. The use of patterns.c was intended to have one reasonably simple implementation, not a button for a "maximum increase of flexibility". The real question is: what do people _have to_ do and can they also do it with the new patterns? > > Reyk: Sorry for the duplicate, meant to reply to the list. no problem, I have thousands of openbsd emails, this one won't hurt ;) Reyk
