Commits
-------
159fc0e [Validator] Added symbols to IDNs validation
c827faf [Validator] Add support for IDNs and custom TLDs
Discussion
----------
[Validator] Add support for IDNs and custom TLDs
Minor changes to allow for IDNs and [custom TLDs](http://news.softpedia.com/news/ICANN-Approves-New-Custom-Generic-Top-Level-Domains-Like-google-bank-206977.shtml). This is the only sane way to support everything in a timeless manner.
---------------------------------------------------------------------------
by stealth35 at 2011/06/20 04:32:09 -0700
maybe it should be check the host with idn_to_ascii (if function exists, maybe it's should recreate un punycode en/decoder in the stub)
---------------------------------------------------------------------------
by mvrhov at 2011/06/20 04:40:10 -0700
/me :faceslap.
Haven't seen the link in PR
---------------------------------------------------------------------------
by Seldaek at 2011/06/20 04:40:40 -0700
@mvrhov: Yup, that's what pushed me to reconsider adding this.
@stealth35: I'm not sure if this is needed. I don't want this to be too strict, with another validator or with an extra option I think we can make a check that the domain actually exists, or do a GET / on it or something, but this just checks validity of the syntax.
---------------------------------------------------------------------------
by stealth35 at 2011/06/20 04:48:05 -0700
I understand :)
what about funny IDN like : [☎.com] (http://xn--y3h.com/) ?
---------------------------------------------------------------------------
by Seldaek at 2011/06/20 04:53:19 -0700
@stealth35: Fixed
---------------------------------------------------------------------------
by stealth35 at 2011/06/20 04:56:18 -0700
it's seem great,for acceptable chars [RFC] (http://www.faqs.org/rfcs/rfc3490.html) said (with UseSTD3ASCIIRules option) :
(a) Verify the absence of non-LDH ASCII code points; that is, the
absence of 0..2C, 2E..2F, 3A..40, 5B..60, and 7B..7F.
Commits
-------
58b6403 Adjusted UPDATE
7350109 Renamed core.* events to kernel.* and CoreEvents to KernelEvents
edbdf7b Rename kernel.listener to kernel.event_listener
Discussion
----------
Rename kernel.listener to kernel.event_listener
Better consistency with doctrine.event_listener
---------------------------------------------------------------------------
by schmittjoh at 2011/06/20 07:39:18 -0700
I think "symfony.event_listener" would be even better then, or just short "event_listener".
There are some more inconsistencies:
- "core.response" -> "kernel.response" / "http_kernel.response"
- "core.request" -> "kernel.request" / "http_kernel.request"
- "core.controller" -> "kernel.controller" / "http_kernel.controller"
- "core.view" -> "kernel.view" / "http_kernel.view"
---------------------------------------------------------------------------
by Seldaek at 2011/06/20 07:45:53 -0700
I proposed `event_listener` but @fabpot didn't like it, `symfony.event_listener` could do, but usual we refer to global stuff as framework more than symfony, so `framework.event_listener`?
The point is that the kernel "owns" the event dispatcher, even if anyone can dispatch his stuff through it. You still listen on the kernel dispatcher. So it's not that wrong imo.
As for the others. I agree that the "core" name is a bit strange. kernel would be more suitable imo. @fabpot?
---------------------------------------------------------------------------
by stof at 2011/06/20 07:50:35 -0700
@Seldaek the issue with having ``kernel`` in the tag name is that some user think that they need to define their own tag when they want to dispatch their own events (I faced it a few times on #symfony and it was also the case in #1298)
---------------------------------------------------------------------------
by Seldaek at 2011/06/20 07:51:55 -0700
Well yeah that's why I liked the idea of having only `event_listener`. IMO even with `framework.event_listener` the confusion can still happen.
---------------------------------------------------------------------------
by fabpot at 2011/06/20 08:00:23 -0700
Renaming `core.response` to `kernel.response` is fine with me.
---------------------------------------------------------------------------
by Seldaek at 2011/06/21 07:53:13 -0700
Ok, added another commit that does that renaming. Updated the two related PRs in -docs and -standard as well.
Commits
-------
e272d56 [WebProfilerBundle] Fixes toolbar content check
Discussion
----------
[WebProfilerBundle] Fixes toolbar content check
It appears that some html optimizers trim the comments, therefore the old check was not working. This is more robust.
Fixes the issue reported in 8541a5bcbc
Commits
-------
abd60ac [WebProfilerBundle] Do not display toolbar loading result if it's not a valid toolbar
406c8d8 [WebProfilerBundle] Make toolbar loading non-blocking
Discussion
----------
Non-blocking WDT & prevents garbage to slip in the page
I made the loading non-blocking so that it's not preventing normal operation of the page when the WDT takes a bit long to come up (happens sometimes when the machine is busy).
The second commit also checks that the response looks correct, to prevent stack traces and such to appear in the page if there was a problem. The main issue is not really stack traces though it's mostly with security and intercept_redirect enabled, if you look at a fully secured site you get twice the redirect intercept message to the login page.
Tested in IE7/9/FF4/Opera11
Commits
-------
f315ad9 [WebProfilerBundle] Make sure the toolbar closes properly
Discussion
----------
[WebProfilerBundle] Make sure the toolbar closes properly
Due to the whitespace element between the div which clears and the toolbar div, in some browsers it was left over after you close the toolbar, this doesn't happen anymore.
Tested in IE7/9/FF4/Opera11
Commits
-------
6c46a3b [FrameworkBundle] Prevent breakage when an array callback is not callable
Discussion
----------
[FrameworkBundle] Prevent breakage when an array callback is not callable
W/o this you get warnings that objects can't be converted to strings.
Commits
-------
f400124 [HttpFoundation] Small optimization of Session flash handling
Discussion
----------
[HttpFoundation] Small optimization of Session flash handling
Due to copy-on-write, this is faster and takes less memory unless the flashes are changed during the request, but that's not very likely as typically you set flashes, redirect, then show them, and at that point you do not modify them again.
Commits
-------
72d0ebe9 [WebProfilerBundle] Added the support of the the logging context in the template
410b3e0 [HttpKernel] Added the context in the LoggerInterface
Discussion
----------
context in the LoggerInterface
This adds the context in the LoggerInterface. The change is totally BC for people using the logger. However this affects people implementing the interface.
Note that this require Seldaek/monolog#33 for the implementation
---------------------------------------------------------------------------
by Seldaek at 2011/06/17 04:24:18 -0700
@fabpot: just ping me when you are merging this one, so I can merge in monolog and we avoid out-of-sync issues.
---------------------------------------------------------------------------
by stof at 2011/06/17 04:49:05 -0700
@Seldaek you can merge in Monolog when you want. Monolog is BC so merging it before the PR in Symfony2 does not break things.
---------------------------------------------------------------------------
by Seldaek at 2011/06/17 05:08:34 -0700
Ah right, I thought the interfaces wouldn't match, but PHP allows extra args it seems so I'll merge right now.
---------------------------------------------------------------------------
by stof at 2011/06/17 05:32:58 -0700
PHP allows extra *optionnal* args and it is the case here :)
---------------------------------------------------------------------------
by Seldaek at 2011/06/17 05:35:00 -0700
Well yes otherwise you break the interface. Anyway it's merged so @fabpot, anytime :)
Commits
-------
3cfff4b [Console] fixed output formatter if content has line breaks
Discussion
----------
[Console] fixed formatter if content has line breaks
Fixes the Symfony2 ASCII art in shell.