# PCRE error handling patch - Thread: https://externals.io/message/22714 - Original archive: https://news-web.php.net/php.internals/22714 - Started by Andrei Zmievski on 2006-04-11T18:25:31+00:00 - Messages: 12 - Reply by email to internals@lists.php.net - More for agents: search with https://externals.io/search.md?q={query} · JSON API at https://externals.io/api · documentation at https://externals.io/llms.txt --- ## #22714 — Andrei Zmievski — 2006-04-11T18:25:31+00:00 URL: https://externals.io/message/22714#22714 I made a patch (against HEAD currently only) that makes it possible to set runtime limits on backtracking and recursion. It also implements `preg_last_error()` that lets users check whether there was an execution error (such as limit exceeded or bad UTF-8 or whatever). Ilia, do you think it might make it into 5.1.3? http://www.php.net/~andrei/pcre.diff -Andrei --- ## #22716 — Jani Taskinen — 2006-04-11T18:35:07+00:00 URL: https://externals.io/message/22714#22716 · In reply to: #22714 Sounds to me like a bug fix..all those recursion bugs in the db.. :) --Jani > I made a patch (against HEAD currently only) that makes it possible to set > runtime limits on backtracking and recursion. It also implements > `preg_last_error()` that lets users check whether there was an execution error > (such as limit exceeded or bad UTF-8 or whatever). > > Ilia, do you think it might make it into 5.1.3? > > http://www.php.net/~andrei/pcre.diff > > -Andrei --- ## #22717 — Nuno Lopes — 2006-04-11T18:51:43+00:00 URL: https://externals.io/message/22714#22717 · In reply to: #22714 I would say that 100.000 recursions is too much. When I made my patch it was segfaulting with such value. 50.000 is a much safer choice (at least on my gentoo box, 8 MB of stack). Nuno ----- Original Message ----- >I made a patch (against HEAD currently only) that makes it possible to set >runtime limits on backtracking and recursion. It also implements >preg_last_error() that lets users check whether there was an execution >error (such as limit exceeded or bad UTF-8 or whatever). > > Ilia, do you think it might make it into 5.1.3? > > http://www.php.net/~andrei/pcre.diff > > -Andrei --- ## #22718 — Andrei Zmievski — 2006-04-11T19:25:29+00:00 URL: https://externals.io/message/22714#22718 · In reply to: #22717 It's hard to pick a default that will work for everyone. I can probably see setting recursion limit somewhat lower than backtracking one, though. Maybe 100,000/50,000 or something like that. -Andrei > I would say that 100.000 recursions is too much. When I made my patch > it was segfaulting with such value. 50.000 is a much safer choice (at > least on my gentoo box, 8 MB of stack). > > Nuno > > > ----- Original Message ----- > > […] --- ## #22733 — Ilia Alshanetsky — 2006-04-12T13:05:38+00:00 URL: https://externals.io/message/22714#22733 · In reply to: #22714 No, lets wait with it till the next PHP 5 release. Ilia > I made a patch (against HEAD currently only) that makes it possible > to set runtime limits on backtracking and recursion. It also > implements `preg_last_error()` that lets users check whether there > was an execution error (such as limit exceeded or bad UTF-8 or > whatever). > > Ilia, do you think it might make it into 5.1.3? > > http://www.php.net/~andrei/pcre.diff > > -Andrei --- ## #22720 — Andrei Zmievski — 2006-04-12T16:07:22+00:00 URL: https://externals.io/message/22714#22720 · In reply to: #22733 5.1.4? -Andrei > No, lets wait with it till the next PHP 5 release. > > Ilia > > > > > > […] --- ## #22721 — Lukas Smith — 2006-04-12T16:17:13+00:00 URL: https://externals.io/message/22714#22721 · In reply to: #22720 Andrei Zmievski wrote: > 5.1.4? I have the following items as future todos: 1. new functions: ext/date: date_sun_info (derick) 2. Switch for disabling/enabling materialized cursors in mysqli (georg) 3. fix __toString() (marcus) 4. add support for files >2GB (wez) 5. Add input filter extension via a symlink from pecl into core as experimental (derick) 6. Add json extension via a symlink from pecl into core as experimental 7. add PDO::defaultFetchMode attribute (pierre) 8. PCRE: make it possible to set runtime limits on backtracking and recursion and add `preg_last_error()` Especially for Nr. 3 I guess we would have to move to 5.2. That is why I would favor 5.2 over 5.1.4 and reserve 5.1.4 to a quick security fix release if one becomes necessary. regards, Lukas --- ## #22722 — Jani Taskinen — 2006-04-12T16:48:27+00:00 URL: https://externals.io/message/22714#22722 · In reply to: #22721 > Andrei Zmievski wrote: > > […] > > I have the following items as future todos: > > 1. new functions: ext/date: date_sun_info (derick) > 2. Switch for disabling/enabling materialized cursors in mysqli (georg) > 3. fix __toString() (marcus) Wasn't that fixed already? Or was that just in HEAD? > 4. add support for files >2GB (wez) > 5. Add input filter extension via a symlink from pecl into core as > experimental (derick) > 6. Add json extension via a symlink from pecl into core as experimental > 7. add PDO::defaultFetchMode attribute (pierre) > 8. PCRE: make it possible to set runtime limits on backtracking and > recursion and add `preg_last_error()` > > Especially for Nr. 3 I guess we would have to move to 5.2. > That is why I would favor 5.2 over 5.1.4 and reserve 5.1.4 to a quick security > fix release if one becomes necessary. I could not agree more. +1 for 5.2. We can't wait for 6.0 forever to get some new features in. It's still at least year from now til it's release I bet. --Jani --- ## #22723 — Andrei Zmievski — 2006-04-12T18:01:41+00:00 URL: https://externals.io/message/22714#22723 · In reply to: #22722 The problem is that with 5.2 you'll get all kinds of old threads resurrected, because people think that it's time to add some new features.. Like namespaces. -Andrei > I could not agree more. +1 for 5.2. We can't wait for 6.0 forever > to get > some new features in. It's still at least year from now til it's > release > I bet. --- ## #22724 — Jani Taskinen — 2006-04-12T18:56:06+00:00 URL: https://externals.io/message/22714#22724 · In reply to: #22723 I have this thing called "delete" in my MUA which I can use to get rid of unwanted noise. :) --Jani > The problem is that with 5.2 you'll get all kinds of old threads resurrected, > because people think that it's time to add some new features.. Like > namespaces. > > -Andrei > > > > > […] > > -- Give me your money at @ Donating money may make me happier and friendlier for a limited period! Death to all 4 letter abbreviations starting with P! --- ## #22725 — Marcus Boerger — 2006-04-12T19:40:11+00:00 URL: https://externals.io/message/22714#22725 · In reply to: #22723 Hello Andrei, no namespaces in 5.2 (.) that is for head. Wednesday, April 12, 2006, 8:01:41 PM, you wrote: > The problem is that with 5.2 you'll get all kinds of old threads > resurrected, because people think that it's time to add some new > features.. Like namespaces. > -Andrei > > > […] Best regards, Marcus --- ## #22726 — Marcus Boerger — 2006-04-12T19:41:59+00:00 URL: https://externals.io/message/22714#22726 · In reply to: #22721 Hello Lukas, Wednesday, April 12, 2006, 6:17:13 PM, you wrote: > Andrei Zmievski wrote: > > […] > I have the following items as future todos: > 1. new functions: ext/date: date_sun_info (derick) > 2. Switch for disabling/enabling materialized cursors in mysqli (georg) > 3. fix __toString() (marcus) fixed in head, will be in 5.2 > 4. add support for files >2GB (wez) > 5. Add input filter extension via a symlink from pecl into core as > experimental (derick) > 6. Add json extension via a symlink from pecl into core as experimental > 7. add PDO::defaultFetchMode attribute (pierre) s/pierre/marcus/; fixed in head will be in 5.2 > 8. PCRE: make it possible to set runtime limits on backtracking and > recursion and add `preg_last_error()` Most of the other stuff is either bigger feature additions or requires api changes so i'd say all of them in 5.2. regards marcus