Debian Bug report logs - #278044
php4: mb_ereg fails to compile regex with \x80-\xF7 characters

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: Alexander Wirt <formorer@debian.org>

To: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: php4: mbstring.func_overload = 7 breaks nearly every string function

Date: Sun, 24 Oct 2004 15:29:04 +0200

Package: php4
Version: 4:4.3.9-1
Severity: grave
Justification: renders package unusable

I had mbstring.func_overload = 7 in my php.ini for a long time, after
switching to sarge I detected some strange behaviour in my php apps,
for example ilohamail created mails with wrong BASE64 encodings 
or drupal gave me many many error messages while the code of the apps
was correct.

I tracked down the problem to mbstring.func_overload, if its set
to 7 most grep und regexp function used wrong encodings or reported
strange errors, here some examples from drupal:

warning: mb_ereg(): mbregex compile err: invalid regular expression
warning: mb_strrpos(): Empty haystack in .... 

(ereg and strrpos get overload by multibye functions here)
And so on. 
This worked on older php releases and is now broken.

Sincerly Alex

-- System Information:
Debian Release: 3.1
  APT prefers testing
  APT policy: (500, 'testing')
Architecture: i386 (i686)
Kernel: Linux 2.6.8
Locale: LANG=de_DE@euro, LC_CTYPE=de_DE@euro

Versions of packages php4 depends on:
ii  libapache-mod-php4            4:4.3.9-1  server-side, HTML-embedded scripti
ii  php4-common                   4:4.3.9-1  Common files for packages built fr

-- debconf information excluded

Message #10 received at 278044-submitter@bugs.debian.org (full text, mbox, reply):

From: "Adam Conrad" <adconrad@trinitysoftware.com.au>

To: <278044-submitter@bugs.debian.org>

Cc: <control@bugs.debian.org>

Subject: mbstring.func_overload = 7 breaks nearly every string function

Date: Mon, 25 Oct 2004 13:03:56 +1000

severity 278044 important
thanks

Can you provide me with a small test case of mb_ereg() usage that worked as
expected in older versions but fails in 4.3.9?... Just having the error
message isn't terribly helpful.

... Adam

--
backup [n] (bak'up): The duplicate copy of crucial data that no one
                     bothered to make; used only in the abstract.

1024D/C6CEA0C9  C8B2 CB3E 3225 49BB 5ED2  0002 BE3C ED47 C6CE A0C9
 

Message #15 received at 278044-quiet@bugs.debian.org (full text, mbox, reply):

From: Alexander Wirt <formorer@formorer.de>

To: Adam Conrad <adconrad@trinitysoftware.com.au>, 278044-quiet@bugs.debian.org

Subject: Re: Bug#278044: mbstring.func_overload = 7 breaks nearly every string function

Date: Mon, 25 Oct 2004 07:30:58 +0200

Adam Conrad wrote:
> severity 278044 important
> thanks
> 
> Can you provide me with a small test case of mb_ereg() usage that worked as
> expected in older versions but fails in 4.3.9?... Just having the error
> message isn't terribly helpful.
I detected that this was also broken in 4.3.8, sorry for the 
desinformation :).
But the problem is still there, I tested this with a woody and with a 
sarge php4:

$name = "user/1/viewuser/1/edit";
if (ereg("[^\x80-\xF7 [:alnum:]@_.-]", $name)) print('The username 
contains an illegal character.');

With woody everything its fine (maybe the function didn't get 
overwritten by mb_ereg in this version) but with sarge:

Warning: mb_ereg(): mbregex compile err: invalid regular expression in 
/var/www/test.php on line 6

If overload is on and works with overload off.

Thanks for you work

Alexander

Message #20 received at 278044@bugs.debian.org (full text, mbox, reply):

From: Steve Langasek <vorlon@debian.org>

To: 278044-submitter@bugs.debian.org, 278044@bugs.debian.org

Subject: Re: php4: mbstring.func_overload = 7 breaks nearly every string function

Date: Tue, 19 Apr 2005 01:14:43 -0700

[Message part 1 (text/plain, inline)]

Hi Alexander,

At Adam's prompting, I've taken a look at this bug.  In your test case, the
trigger seems to be this part of the check:

  \x80-\xF7

We've noticed that splitting this into \x80-\xA0 and \xA1-\xF7 fixes the
problem.

\xA0, \xA1 is a boundary between character classes: \x80-\xA0 are control
characters, and \xA1-x\F7 are printable characters (latin1 punctuation and
letters).  I'm not sure if it's a feature for mb_ereg() to complain about
ranges that span character classes, but I'm inclined to think that you don't
actually want control characters to be in your usernames anyway.

Even with this change, though, the regexp doesn't seem to work so well with
actual multibyte UTF8 chars in the string being compared.

-- 
Steve Langasek
postmodern programmer

[signature.asc (application/pgp-signature, inline)]

Message #26 received at 278044-submitter@bugs.debian.org (full text, mbox, reply):

From: Ondrej Sury <ondrej@sury.org>

To: 278044-submitter@bugs.debian.org

Subject: Re: php4: mbstring.func_overload = 7 breaks nearly every string function

Date: Sat, 13 Aug 2005 13:17:34 +0200

[Message part 1 (text/plain, inline)]

forwarded 278044 http://bugs.php.net/bug.php?id=34119
retitle 278044 php4: mb_ereg fails to compile regex with \x80-\xF7 characters
thank you

I have forwarded your bug upstream.  It looks like it works ok in php5
(at least it doesn't print warning).

O.
-- 
Ondrej Sury <ondrej@sury.org>

[signature.asc (application/pgp-signature, inline)]

Message #37 received at 278044-done@bugs.debian.org (full text, mbox, reply):

From: Marco Rodrigues <gothicx@sapo.pt>

To: 278044-done@bugs.debian.org

Subject: php4 has been removed from Debian, closing #278044

Date: Fri, 23 May 2008 15:52:12 +0100

Version: 6:4.4.6-2+rm

The php4 package has been removed from Debian testing, unstable and
experimental, so I am now closing the bugs that were still opened
against it.

For more information about this package's removal, read
http://bugs.debian.org/428266 . That bug might give the reasons why
this package was removed, and suggestions of possible replacements.

Don't hesitate to reply to this mail if you have any question.

Thank you for your contribution to Debian.

-- 
Marco Rodrigues

http://Marco.Tondela.org

Send a report that this bug log contains spam.