Hello,<div><br></div><div>I am working with the 0.7* code base, and I need to extract a sub-string from a location URL regex capture during the access phase. I'm working on an authentication scheme that matches a capture from a location with the payload of an encrypted cookie (similar to kerberos tickets). I could only find code that did this in the scripting engine codes, and I do not at this point require this level of dynamism at request-time; therefore, I would prefer not to instantiate a scripting engine within my module.</div>
<div><br></div><div>My question: Since "r->ncaptures" and "r->captures" represent the ovecsize and ovector respectively, Would it be advisable to use the extraction helper functions from the PCRE api ?</div>
<div><br></div><div>PCRE api documentation link: <a href="http://regexkit.sourceforge.net/Documentation/pcre/pcreapi.html#SEC15">http://regexkit.sourceforge.net/Documentation/pcre/pcreapi.html#SEC15</a></div><div><br></div>
<div>Thanks </div><div><br></div><div>Hassan</div><div><br></div><div>P.S., I would advise that at certain point in the code base (for example struct ngx_http_request_s { ..) the parallels between "r->ncaptures" and "r->captures" and the PCRE counterparts be made explicit with a comment -- Since I did not have any PCRE experience before Nginx, I assumed the ovector was a concept native to the Nginx code-base.</div>
<meta http-equiv="content-type" content="text/html; charset=utf-8">