Re: FW: Re: Apache::TestRequest + Cookies
Ok, here's my first crack at a quick n dirty cookies test. The new test is t/cookies.t, and it's supporting server file is t/htdocs/cgi-bin/cookies.pl. I figured the extra cgi-bin directory was good for sanity sake. I've also included the necessary patches to MANIFEST and extra.conf.in against A-T 1.16. -=Chris a-t-cookie-test.tar.gz Description: GNU Zip compressed data
Re: FW: Re: Apache::TestRequest + Cookies
Christopher H. Laco wrote: Thanks, Chris, committed with a few small adjustements. You can see the commits here: http://svn.apache.org/viewcvs?view=revrev=109277 http://svn.apache.org/viewcvs?view=revrev=109279 Thanks. Glad to contribute. It looks like the only change was naming the cookies.pl to cookies.pl.PL. That was needed to get the path to #!perl right What cnvention did I miss so I don't make that mistake again? The only missing thing was to check that CGI and CGI::Cookie were available in plan(). I moved htdocs/cgi-bin/ = cgi-bin/ and did a few minor syntax changes. You can see what was applied at the URLs quoted at the top of this email. Now looking forward for docs updates. Thanks again. -- __ Stas BekmanJAm_pH -- Just Another mod_perl Hacker http://stason.org/ mod_perl Guide --- http://perl.apache.org mailto:[EMAIL PROTECTED] http://use.perl.org http://apacheweek.com http://modperlbook.org http://apache.org http://ticketmaster.com
Re: FW: Re: Apache::TestRequest + Cookies
Stas Bekman wrote: Christopher H. Laco wrote: Thanks, Chris, committed with a few small adjustements. You can see the commits here: http://svn.apache.org/viewcvs?view=revrev=109277 http://svn.apache.org/viewcvs?view=revrev=109279 Thanks. Glad to contribute. It looks like the only change was naming the cookies.pl to cookies.pl.PL. That was needed to get the path to #!perl right What cnvention did I miss so I don't make that mistake again? The only missing thing was to check that CGI and CGI::Cookie were available in plan(). Duh. Silly me. I moved htdocs/cgi-bin/ = cgi-bin/ and did a few minor syntax changes. You can see what was applied at the URLs quoted at the top of this email. Now looking forward for docs updates. Thanks again.
Re: FW: Re: Apache::TestRequest + Cookies
Stas Bekman wrote: Christopher H. Laco wrote: Ok, here's my first crack at a quick n dirty cookies test. The new test is t/cookies.t, and it's supporting server file is t/htdocs/cgi-bin/cookies.pl. I figured the extra cgi-bin directory was good for sanity sake. I've also included the necessary patches to MANIFEST and extra.conf.in against A-T 1.16. Thanks, Chris, committed with a few small adjustements. You can see the commits here: http://svn.apache.org/viewcvs?view=revrev=109277 http://svn.apache.org/viewcvs?view=revrev=109279 You probably want to get the svn checkout to continue working with the latest version: svn checkout http://svn.apache.org/repos/asf/httpd/test/trunk/perl-framework/Apache-Test Apache-Test For docs (where you can find the latest testing.pod) it's: svn co http://svn.apache.org/repos/asf/perl/modperl/docs/trunk modperl-docs Thanks. Glad to contribute. It looks like the only change was naming the cookies.pl to cookies.pl.PL. What cnvention did I miss so I don't make that mistake again? -=Chris
Re: FW: Re: Apache::TestRequest + Cookies
CChristopher H. Laco wrote: Maybe relying on Apache::Session would cover more install scnerios instead. I was trying to avoid writing a custom handler just to accept and toss some cookies in the test environment, but maybe that's better. At least then everyone install should be able to run the A-T cookie test. I think the same_interp setup will do what you want. e.g. see in modperl-2.0 repository this test couple: t/response/TestModperl/sameinterp.pm t/modperl/sameinterp.t Yeah, I'll have to check that out. Thanks. (Yes, I just played the email switch-a-roo)
Re: FW: Re: Apache::TestRequest + Cookies
Christopher H. Laco wrote: Christopher H. Laco wrote: CChristopher H. Laco wrote: Maybe relying on Apache::Session would cover more install scnerios instead. I was trying to avoid writing a custom handler just to accept and toss some cookies in the test environment, but maybe that's better. At least then everyone install should be able to run the A-T cookie test. I think the same_interp setup will do what you want. e.g. see in modperl-2.0 repository this test couple: t/response/TestModperl/sameinterp.pm t/modperl/sameinterp.t Holy crap. That made my brain hurt. Would we use that directly, or just borrow the idea? Why don't you first post what you had on your mind. May be your idea doesn't need that at all. -- __ Stas BekmanJAm_pH -- Just Another mod_perl Hacker http://stason.org/ mod_perl Guide --- http://perl.apache.org mailto:[EMAIL PROTECTED] http://use.perl.org http://apacheweek.com http://modperlbook.org http://apache.org http://ticketmaster.com
Re: FW: Re: Apache::TestRequest + Cookies
Stas Bekman wrote:
[snip]
Why don't you first post what you had on your mind. May be your idea
doesn't need that at all.
Well, here's what I was thinking. First, write a quick and dirty
mod_perl handler that looked for a specific cookie; let's say 'ApacheTest'.
If it wasn't supplied by the client, send a new copy out. If it was sent
by the client, simply return the supplied cookie/value.
Shouldn't take more than a few lines of code, and it removes the
dependency of usertrack and Apache::Session.
Then, simply write a test that makes a series of requests to that
handler. The first response should return a new ApacheTest cookie/value,
and the second test should pass back the same ApacheTest cookie/value if
Apache::TestRequest::user_agent( cookie_jar = {});
and it's surroundings are woroking fine.
-=Chris
Re: FW: Re: Apache::TestRequest + Cookies
Christopher H. Laco wrote:
Stas Bekman wrote:
[snip]
Why don't you first post what you had on your mind. May be your idea
doesn't need that at all.
Well, here's what I was thinking. First, write a quick and dirty
mod_perl handler that looked for a specific cookie; let's say 'ApacheTest'.
If it wasn't supplied by the client, send a new copy out. If it was sent
by the client, simply return the supplied cookie/value.
Shouldn't take more than a few lines of code, and it removes the
dependency of usertrack and Apache::Session.
Then, simply write a test that makes a series of requests to that
handler. The first response should return a new ApacheTest cookie/value,
and the second test should pass back the same ApacheTest cookie/value if
Apache::TestRequest::user_agent( cookie_jar = {});
and it's surroundings are woroking fine.
Sounds good. Though it'll probably need to leave in the modperl2's test
suite and not Apache-Test's one. but let's see the code first and then see
where it's the best to apply it. You can use the
http://perl.apache.org/~geoff/Apache-Test-skeleton-mp2.tar.gz
to make it easy to write the test.
--
__
Stas BekmanJAm_pH -- Just Another mod_perl Hacker
http://stason.org/ mod_perl Guide --- http://perl.apache.org
mailto:[EMAIL PROTECTED] http://use.perl.org http://apacheweek.com
http://modperlbook.org http://apache.org http://ticketmaster.com
Re: FW: Re: Apache::TestRequest + Cookies
Stas Bekman wrote: Sounds good. Though it'll probably need to leave in the modperl2's test suite and not Apache-Test's one. but let's see the code first and then see where it's the best to apply it. You can use the http://perl.apache.org/~geoff/Apache-Test-skeleton-mp2.tar.gz to make it easy to write the test. I guess that's where I'm getting confused. Why would this be in mp2 rather than in A-T, since I'm really trying to confirm a scenerio under A-T. Obviously there's something I'm missing. -=Chris
Re: FW: Re: Apache::TestRequest + Cookies
Christopher H. Laco wrote: Stas Bekman wrote: Sounds good. Though it'll probably need to leave in the modperl2's test suite and not Apache-Test's one. but let's see the code first and then see where it's the best to apply it. You can use the http://perl.apache.org/~geoff/Apache-Test-skeleton-mp2.tar.gz to make it easy to write the test. I guess that's where I'm getting confused. Why would this be in mp2 rather than in A-T, since I'm really trying to confirm a scenerio under A-T. Obviously there's something I'm missing. It should be in A-T, but since your test requires modperl it'd be better be in modperl, since chances are that A-T won't have modperl available and then it'll be skipped. So may be your original mod_usertrack idea was better, as it doesn't depend on mod_perl's availability (but at the same time depends on mod_usertrack). The reason I'm OK with adding A-T tests to modperl is that we run modperl2 tests all the time, much more often than A-T tests, so it's probably a second best place to put those cookie tests. In fact you can already find some cookie tests in the modperl2 test suite (against CGI.pm). -- __ Stas BekmanJAm_pH -- Just Another mod_perl Hacker http://stason.org/ mod_perl Guide --- http://perl.apache.org mailto:[EMAIL PROTECTED] http://use.perl.org http://apacheweek.com http://modperlbook.org http://apache.org http://ticketmaster.com
Re: FW: Re: Apache::TestRequest + Cookies
Stas Bekman wrote: Christopher H. Laco wrote: Stas Bekman wrote: Sounds good. Though it'll probably need to leave in the modperl2's test suite and not Apache-Test's one. but let's see the code first and then see where it's the best to apply it. You can use the http://perl.apache.org/~geoff/Apache-Test-skeleton-mp2.tar.gz to make it easy to write the test. I guess that's where I'm getting confused. Why would this be in mp2 rather than in A-T, since I'm really trying to confirm a scenerio under A-T. Obviously there's something I'm missing. It should be in A-T, but since your test requires modperl it'd be better be in modperl, since chances are that A-T won't have modperl available and then it'll be skipped. So may be your original mod_usertrack idea was better, as it doesn't depend on mod_perl's availability (but at the same time depends on mod_usertrack). Can't the test be done with mod_cgi and not mod_perl? This should be available more often than mod_usertrack I would think. -- Michael Peters Developer Plus Three, LP
Re: FW: Re: Apache::TestRequest + Cookies
Stas Bekman wrote: Michael Peters wrote: Stas Bekman wrote: Christopher H. Laco wrote: Stas Bekman wrote: Sounds good. Though it'll probably need to leave in the modperl2's test suite and not Apache-Test's one. but let's see the code first and then see where it's the best to apply it. You can use the http://perl.apache.org/~geoff/Apache-Test-skeleton-mp2.tar.gz to make it easy to write the test. I guess that's where I'm getting confused. Why would this be in mp2 rather than in A-T, since I'm really trying to confirm a scenerio under A-T. Obviously there's something I'm missing. It should be in A-T, but since your test requires modperl it'd be better be in modperl, since chances are that A-T won't have modperl available and then it'll be skipped. So may be your original mod_usertrack idea was better, as it doesn't depend on mod_perl's availability (but at the same time depends on mod_usertrack). Can't the test be done with mod_cgi and not mod_perl? This should be available more often than mod_usertrack I would think. Yes, I was thinking the same. Good idea Michael. And it's enabled in the default shipping httpd.conf too! Great idea. I'll pursue it that method for now. -=Chris
