@ character in http login

[simonp]

I am trying to use the basic http://user:pa******@example.com
form for authentication inside a script.

My only problem is my username contains the @ character (its an
email address). As result, the url seems to be parsing
incorrectly.

Has anyone found a way around this problem?

Cheers
Simon

--
Stupendous Tales
www.stupendoustales.com
Speculative Fiction, Pulp Dreams

[Michael B. Trausch]

si****@nospam.com wrote in <e2**********@reader1.panix.com> on Sat April 22
2006 22:53:

I am trying to use the basic http://user:pa******@example.com
form for authentication inside a script.

My only problem is my username contains the @ character (its an
email address). As result, the url seems to be parsing
incorrectly.

Has anyone found a way around this problem?

Cheers
Simon

Well, the way it usually works is:

http://user:pass@server/

You may be able to try this:

http://user%40sample.com:pass@server

And see if it works. 0x40 is the hex character code for '@', and URL
encoding works by using % and then the hex char reference (i.e., %20 for a
space, %40 for '@', and so forth).

If that doesn't work, then I don't know what else you could do to make it
work, at least, when authenticating using basic auth like that.

- Mike

[Gordon Burditt]

>I am trying to use the basic http://user:pa******@example.com

form for authentication inside a script.

My only problem is my username contains the @ character (its an
email address). As result, the url seems to be parsing
incorrectly.

Has anyone found a way around this problem?

You may be able to use something like
http://username%40hotmail.com:pa******@example.com

where @ is converted to %40 . The same approach may apply if the
password contains odd characters like @, colon, space, %, etc.

Gordon L. Burditt

[Michael B. Trausch]

Gordon Burditt wrote in <12*************@corp.supernews.com> on Sat April 22
2006 23:18:

I am trying to use the basic http://user:pa******@example.com
form for authentication inside a script.

My only problem is my username contains the @ character (its an
email address). As result, the url seems to be parsing
incorrectly.

Has anyone found a way around this problem?

You may be able to use something like
http://username%40hotmail.com:pa******@example.com

where @ is converted to %40 . The same approach may apply if the
password contains odd characters like @, colon, space, %, etc.

Gordon L. Burditt

FYI, you may want to read to ensure that the question hasn't been answered
the same way you're about to answer it. It is bad form; the question was
answered twelve minutes prior to you posting the same thing.

Regards,
Mike

[Gordon Burditt]

>FYI, you may want to read to ensure that the question hasn't been answered

the same way you're about to answer it. It is bad form; the question was
answered twelve minutes prior to you posting the same thing.

I did. Remember that USENET transmission is not instantaneous.
Also, article composition takes time. It used to be that a feed
that was getting articles 3 days old from the far side of the net
(consisting largely of UUCP feeds) was operating unusually better
than normal.

Even now, I doubt that even all well-connected servers get a posted
article within 2 hours of when it is posted on one of them.

Gordon L. Burditt

[simonp]

Michael B. Trausch <mi*********************@comcast.net> wrote:

si****@nospam.com wrote in <e2**********@reader1.panix.com> on Sat April 22
2006 22:53:

I am trying to use the basic http://user:pa******@example.com
form for authentication inside a script.

My only problem is my username contains the @ character (its an
email address). As result, the url seems to be parsing
incorrectly.

Has anyone found a way around this problem?

Cheers
Simon

Well, the way it usually works is:

http://user:pass@server/

You may be able to try this:

http://user%40sample.com:pass@server

Thank you for the pointer, it pointed me to the right solution.

It looks like @ is the escape character that works for @,
rather than %40. On using the parse_url() function on the url,
the correct email username was returned for the user element in
the resulting array.

Cheers
Simon

--
Stupendous Tales
www.stupendoustales.com
Speculative Fiction, Pulp Dreams

[Michael B. Trausch]

si****@nospam.com wrote in <e2**********@reader1.panix.com> on Sun April 23
2006 01:51:

Thank you for the pointer, it pointed me to the right solution.

It looks like @ is the escape character that works for @,
rather than %40. On using the parse_url() function on the url,
the correct email username was returned for the user element in
the resulting array.

Cheers
Simon

Very good... I didn't realize that the server would recognize it in that
fashion. You could probably use urldecode() for getting the %40 out of it,
did you try it that way?

If you do, I'd be curious to see the result. I don't have the ambition to
code that up for a test right now, since I've got other things keeping my
plate rather full at the moment... <G>

- Mike

[John Dunlop]

Gordon Burditt:

I did. Remember that USENET transmission is not instantaneous.
Also, article composition takes time. It used to be that a feed
that was getting articles 3 days old from the far side of the net
(consisting largely of UUCP feeds) was operating unusually better
than normal.

Even now, I doubt that even all well-connected servers get a posted
article within 2 hours of when it is posted on one of them.

And some articles simply don't arrive on some servers.

Besides, it seems the original poster somehow misconstrued Michael's
followup and has now confused the HTML notation @ with the
percent-encoding %40.

It is now necessary to point out that @ is a way of representing
<@> *in* *HTML*, not in URLs (it means something very different in
URLs), and that as far as the URL is concerned, <@> is still
"unescaped". %40 (for the third time :-) is the percent-encoding of
<@>.

--
Jock

[Jerry Stuckle]

Michael B. Trausch wrote:

Gordon Burditt wrote in <12*************@corp.supernews.com> on Sat April 22
2006 23:18:

I am trying to use the basic http://user:pa******@example.com
form for authentication inside a script.

My only problem is my username contains the @ character (its an
email address). As result, the url seems to be parsing
incorrectly.

Has anyone found a way around this problem?

You may be able to use something like
http://username%40hotmail.com:pa******@example.com

where @ is converted to %40 . The same approach may apply if the
password contains odd characters like @, colon, space, %, etc.

Gordon L. Burditt

FYI, you may want to read to ensure that the question hasn't been answered
the same way you're about to answer it. It is bad form; the question was
answered twelve minutes prior to you posting the same thing.

Regards,
Mike

And it's even worse form to criticize someone for repeating the content of YOUR
OWN post 12 minutes after you posted it.

This is not a "I got the right answer first" contest. This is about helping
each other.
--
==================
Remove the "x" from my email address
Jerry Stuckle
JDS Computer Training Corp.
js*******@attglobal.net
==================

[Michael B. Trausch]

Gordon Burditt wrote in <12*************@corp.supernews.com> on Sun April 23
2006 00:13:

FYI, you may want to read to ensure that the question hasn't been answered
the same way you're about to answer it. It is bad form; the question was
answered twelve minutes prior to you posting the same thing.

I did. Remember that USENET transmission is not instantaneous.
Also, article composition takes time. It used to be that a feed
that was getting articles 3 days old from the far side of the net
(consisting largely of UUCP feeds) was operating unusually better
than normal.

Even now, I doubt that even all well-connected servers get a posted
article within 2 hours of when it is posted on one of them.

My apologies. I didn't realize that there would've been that large of an
amount of lag between the servers.

- Mike

[simonp]

Michael B. Trausch <mi*********************@comcast.net> wrote:

Very good... I didn't realize that the server would recognize it in that
fashion. You could probably use urldecode() for getting the %40 out of it,
did you try it that way?

If you do, I'd be curious to see the result. I don't have the ambition to
code that up for a test right now, since I've got other things keeping my
plate rather full at the moment... <G>

- Mike

Another solution I've discovered, that is probably more flexible
inside a script, is to use the Snoopy php class:

http://sourceforge.net/projects/snoopy/

Snoopy acts like a web browser inside the script, so username and
password can be set as properties of the class, and sent to the
server requiring basic authentication.

The results are then returned as an array.

Cheers
Simon

--
Stupendous Tales
www.stupendoustales.com
Speculative Fiction, Pulp Dreams

[Michael B. Trausch]

Jerry Stuckle wrote in <Kt******************************@comcast.com> on Sun
April 23 2006 11:50:

And it's even worse form to criticize someone for repeating the content of
YOUR OWN post 12 minutes after you posted it.

This is not a "I got the right answer first" contest. This is about
helping each other.

I agree with you on that. Now, I didn't realize that the lag was 12 whole
minutes -- that's atypical for many networks these days, even Usenet.
Often, I can post something to Usenet and find it on Google Groups within
five minutes of my having posted it. Thus, it didn't seem unreasonable to
me to prompt Gordon to check prior to posting. It would seem that he is
not to blame for not checking; rather, his NNTP spooling service is a bit
behind on getting the feeds from Usenet.

Looking at their web site, it could also be a slowdown caused by their
Usenet spam filtration software or something similar; either way, that's a
great deal of lag. :( It isn't the 1980s anymore. <g>

- Mike

[Jerry Stuckle]

Michael B. Trausch wrote:

Jerry Stuckle wrote in <Kt******************************@comcast.com> on Sun
April 23 2006 11:50:

And it's even worse form to criticize someone for repeating the content of
YOUR OWN post 12 minutes after you posted it.

This is not a "I got the right answer first" contest. This is about
helping each other.

I agree with you on that. Now, I didn't realize that the lag was 12 whole
minutes -- that's atypical for many networks these days, even Usenet.
Often, I can post something to Usenet and find it on Google Groups within
five minutes of my having posted it. Thus, it didn't seem unreasonable to
me to prompt Gordon to check prior to posting. It would seem that he is
not to blame for not checking; rather, his NNTP spooling service is a bit
behind on getting the feeds from Usenet.

Looking at their web site, it could also be a slowdown caused by their
Usenet spam filtration software or something similar; either way, that's a
great deal of lag. :( It isn't the 1980s anymore. <g>

- Mike

Not atypical at all, Mike. It all depends on how often usenet servers sync with
each other, how many servers are between the posting one and the receiving one,
any network delays which can creep in, and even if a router is temporarily down
for maintenance.

12 minutes is nothing. I, too, have regularly seen delays of 2 hours or more.

--
==================
Remove the "x" from my email address
Jerry Stuckle
JDS Computer Training Corp.
js*******@attglobal.net
==================

[Joshie Surber]

Jerry Stuckle wrote:

Michael B. Trausch wrote:

Jerry Stuckle wrote in <Kt******************************@comcast.com> on Sun
April 23 2006 11:50:

And it's even worse form to criticize someone for repeating the content of
YOUR OWN post 12 minutes after you posted it.

This is not a "I got the right answer first" contest. This is about
helping each other.

I agree with you on that. Now, I didn't realize that the lag was 12 whole
minutes -- that's atypical for many networks these days, even Usenet.
Often, I can post something to Usenet and find it on Google Groups within
five minutes of my having posted it. Thus, it didn't seem unreasonable to
me to prompt Gordon to check prior to posting. It would seem that he is
not to blame for not checking; rather, his NNTP spooling service is a bit
behind on getting the feeds from Usenet.

Looking at their web site, it could also be a slowdown caused by their
Usenet spam filtration software or something similar; either way, that's a
great deal of lag. :( It isn't the 1980s anymore. <g>

- Mike

Not atypical at all, Mike. It all depends on how often usenet servers sync with
each other, how many servers are between the posting one and the receiving one,
any network delays which can creep in, and even if a router is temporarily down
for maintenance.

12 minutes is nothing. I, too, have regularly seen delays of 2 hours or more.

--
==================
Remove the "x" from my email address
Jerry Stuckle
JDS Computer Training Corp.
js*******@attglobal.net
==================

Just blame slow typing and everyone get along with your lives.