+1 for the rename +1 for adding the two new functions On Mon, Aug 6, 2018 at 4:41 PM, Alan Carroll < solidwallofc...@oath.com.invalid> wrote:
> +1 > > On Mon, Aug 6, 2018 at 4:18 PM Leif Hedstrom <zw...@apache.org> wrote: > > > > > > > > On Aug 6, 2018, at 3:14 PM, Leif Hedstrom <zw...@apache.org> wrote: > > > > > > We have this API: > > > > > > void TSHttpTxnSetHttpRetStatus(TSHttpTxn txnp, TSHttpStatus > > http_retstatus); > > > > > > > > > Which on its own is a little strange, but looking at code, and talking > > to Alan, I believe this was added when we added remap plugins long ago, > to > > allow such plugins to modify status codes without having an HttpHdr > > response. > > > > > > So, I guess that is fine, and we use it fairly frequently, but looking > > at this, the naming is really non-standard. I’d like to propose that we > > change this API to > > > > > > void TSHttpTxnStatusSet(TSHttpTxn txnp, TSHttpStatus status); > > > > > > > > > > > > This seems much more inline with the other StatusSet that we have, > > TSHttpHdrStatusSet(). > > > > > > > > > If this is reasonable, I’d also like to add a > > TSHttpTxnStatusGet(TSHttpTxn txnp) call. > > > > > > > > > In addition, I’m contemplating adding two new APIs, on the same note: > > > > > > void TSHttpTxnStatusSet(TSHttpTxn txnp, const char* reason); > > > const char* TSHttpTxnStatusGet(TSHttpTxn txnp); > > > > > > > > Sigh, the last two should be > > > > void TSHttpTxnReasonSet(TSHttpTxn txnp, const char* reason); > > const char* TSHttpTxnReasonGet(TSHttpTxn txnp); > > > > > > The point being, I might want to set the transaction return code to a 403 > > in a remap plugin, but there’s no way to set a custom reason phrase right > > now. > > > > Cheers, > > > > — leif > > > > > > > > -- > *Beware the fisherman who's casting out his line in to a dried up > riverbed.* > *Oh don't try to tell him 'cause he won't believe. Throw some bread to the > ducks instead.* > *It's easier that way. *- Genesis : Duke : VI 25-28 > -- Derek
