| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Comments:
<0> he needs ['extension']
<0> how do you set up casting behaviour for your objects? I'm not PHP 5 savvy
<1> Ciaran: You don't.
<1> Zend handles that
<2> TML: at least i might be able to come up with something that will help someone else solve the real issue
<0> actually he should just use PATHINFO_EXTENSION
<3> so path = pathinfo($_FILES['pic']['name'])
<0> oh..so where does the int come from?
<3> and then path['extension']
<1> Ciaran: Z_OBJ_HT_P(op)->cast_object
<0> if you do pathinfo(filename, PATHINFO_EXTENSION)
<0> it will return the extension
<0> http://bugs.php.net/bug.php?id=33999
<1> Ciaran: Yes
<1> That's part of the issue we're discussing
<0> sorry, thought you'd just discovered the bug
<2> TML: all i have to do is diff the cast_object from 5.0.4 and 5.1.3RC and i should be able to see if anything's changed/come unstuck
<1> Ciaran: No, we're discussing it
<4> hi
<3> thanks :)
<4> 8-)
<5> so, here's an example of broken nested output buffering with virtual : http://www.owid.org/php/test_ob.php
<5> there could easily be something I'm not understanding.
<1> earth: Wait...what does virtual() have to do with output buffering?
<1> earth: Note that it's never been considered a terribly good idea to request PHP scripts using virtual()
<5> right, it is effectively broken all the way around
<5> but, it totally borks ob
<1> It doesn't just bork OB. It borks the entire engine.
<5> so, anyway, that's an example of one of the problems with ob :)
<5> if the call shouldn't be used, it should be removed from the source. I've never understood why it doesn't actually generate a new httpd call.
<1> earth: It's not a problem with OB, it's a problem with vortual.
<5> heh
<1> s/vor/vir/
<1> The call can be used, but it shouldn't be used to call another instance of the PHP engine, because bad things can happen
<5> you note that there is no indication of that in the documentation.
<5> but, either way, I don't need to argue about it
<5> there are a couple of main problems i have with virtual and one of them is ob
<6> is it something that can't be done with include() or require()?
<5> it really ought to just throw it out as a real httpd request, don't try to do smart twisty internal things. make a new hit to httpd. but whatever, I had this argument about 4 years ago with php devs :)
<5> I don't need to go back over it.
<1> earth: If you want a real httpd request, use include
<1> include 'http://...'; will do precisely that
<5> my more recent minor issue was line returns with XML bits parsed and then using appendChild: http://www.owid.org/php/test_dom.php
<1> virtual is there specifically for the "smarty twisty internal things" that include doesn't cover
<5> right, except it is borked :)
<1> earth: Yes, it has bugs. Mainly because no one uses it, so no one can fix the bugs.
<5> it is not always possible to easily avoid calling php scripts. shtml include virtual was a sort of duct tape solution that works wonders
<1> I don't understand what you mean by that.
<5> one of the sites I run has tomcat, zope, php, perl, text files, shtml files, remote SOAP crap.. all barely held together. I often have no idea what the next layer down is doing.
<5> well, php's virtual is an attempted copy of shtml include virtual
<5> and that allows one to drop subrequests in that save the environment and can grab output from another cgi or server applet
<5> s/save/same/
<5> anyway, understood. it is little used and kinda broken. I drop it.
<5> I don't mean to pick a fight, was just showing the example code I wrote to show the problem I have.
<1> earth: I just want to make sure it's clear: there's not a problem in OB, there's a problem in virtual. You haven't demonstrated a problem in OB that I can see.
<5> heh
<5> now you're picking the fight ;]
<5> ob breaks when virtual is called and there's ob_ usage in the script called by virtual. I don't exactly care, as a user, where the problem is :)
<5> so I accept what you say
<7> i have a PHP script used to cache remote images to display on my site, any comments on how to make it more efficient would be appreciated : http://hashphp.org/pastebin?pid=6510
<5> in my case, the problem is solved not by removing virtual, but by removing ob_ usage in all but php scripts at the layer called directly by the user
<1> earth: No, I'm not picking a fight, I'm saying it's incorrect to ***ert that there are any problems at all in ob_*(), because there aren't. There are problems in virtual(). And, as a user, you *should* care where the problem is, because ***erting it's in ob_*() eliminates a whole slate of PHP functions that aren't actually broken.
<5> but, as you say, that is a very niche use related to legacy code
<5> well, great. again, I accept that the real problem is in virtual.
<5> you have convinced me.
<5> and I agree that I should make the differentiation.
<8> Hi! Could anyone point me in the right direction here please .. :)
<1> earth: As for your DOM thing, it's not really a bug.
<0> anyone here vbulletin savvy? just want to know if something is possible
<8> say .. if I have a php script running .. it takes approx 15mins to complete .. what if I start the same script from crontab every one minutes.. how can I prevent the script from being run more than one time at the time.. ? :)
<5> perhaps it is something I'm doing wrong.
<1> earth: It is. Allow me to explain.
<8> The obious way would be to write to a file .. like .. run.lock .. if that file exists, just exit..
<2> nobb: i use lock files
<8> but i wanna know if there are any .. standard ways of synchronizing things :)
<5> use mysql :]
<8> dools: Cool. How do you use them?
<5> look at flock documentation
<8> dodo, simpyl write to a file or use an API?
<2> nobb: exactly as you describe
<1> earth: Here's the thing: Anything after the close of the document element is considered noise. You're creating a new document, rooted at <event>. When you importNode() that document fragment, it's free to discard the trailing newline, because it's content that lives outside the document root.
<5> but flock fails across many network file systems :\
<8> dools, great.. thanks for your input. I'll try that then and see how it works out.. :)
<2> geez, a framework for a lock file haha
<5> nobb: check flock documentation
<1> Better would be to acquire a semaphore
<5> I use mysql for all of my synchronizing, since I can use table and row locks.
<1> That's what they're for
<5> :)
<1> nobb: Check out php.net/sem
<8> earth, ah.. didn't realize that you were talking to me :) I will check that out, thanks
<1> It's a better solution than flock
<5> ok, TML, do you have a suggestion for how to retain the line returns after the nodes when importing XML
<8> TML, and thanks to you too .. :)
<1> earth: $mParentNode->appendChild($mainDocument->createTextNode("\n"));
<1> earth: As the last thing in the loop.
<5> ok, thanks.
<1> You're appending a textnode that contains \n
<5> oh, yeah, if you're able to stay all in PHP, shared memory can work.
<1> Actually, the nice thing about a semaphore is that anything on the machine can see it
<1> earth: But you're doing it in the mainDocument, so that it doesn't get "trimmed"
<1> earth: Make sense?
<5> yes, actually.
<1> earth: the loadXML is one of those cases where it's actually *correct* to disregard the whitespace...because it's "outside" the document itself
<5> that makes sense to me.
<9> hi. question: why does var_dump('11111111111111111' == '11111111111111112'); return true? makes only sense if php does an implicit cast to int, but i dont see the reason...
<5> ok, good. that solves the problem. I'm unclear whether that is preferred or the docfragment solution is preferred.
<9> btw, var_dump('1111111111111111' == '1111111111111112'); returns false ?!
<1> earth: I don't think it really makes a difference.
<1> earth: Whichever looks better to you
<5> thanks
<5> Perhaps a comment in the documentation for loadXML that suggests that solution?
<1> earth: Go ahead. I don't see it as actually being neccesary, because DOM shouldn't ever be used for pretty printing XML anyways, but I'm not opposed to seeing a note.
<10> how do i translate string into UTF-8?
<1> mrix: php.net/utf8_encode
<1> Anyone else having issues with google?
<5> not I.
<11> hi, if i define an array like: $TmpError=array(array()); and add elements with: $TmpError[]=$content; I got an element 0 with no content, how can I avoid this?
<1> Badcel: By not defining the array as array(array())
<12> idd
<13> what is this sort of ***ignment called? : $array[] = $info;
<12> try to print_r($TmpError) and see if u have the structure you wanted
<1> Jmax: What do you mean?
<1> !tell z1n about u
<12> i am i so bad at that
<12> sorry
<1> z1n: Most clients can be configured to fix that for you
<13> TML: the [] part
<13> TML: what's that called and do you know of a reference for it?
<12> cool
<13> the empty bracket ***igning heh
<1> Jmax: It's appending to an array, I'm not sure what you mean by "a reference for it"
<12> [] is a place in the array
<13> oh
<13> ok
<11> works TML :)
<1> Actually, it's not a place in the array...not until you put something in there, anyways
<12> [0] is place 0 ofcourse .. by doing [] you add to the array and make a new place
<12> TML, :P
<13> ahh ok
<12> if you only have 1 place ([0]) [] would be place [1]
<5> I think it'
<5> doh. I think it's equivalent to http://us3.php.net/manual/en/function.array-push.php
<1> earth: It is
<5> or, in perl $array[$#array + 1] :)
<13> :)
<13> thanks
<12> yay you now works :P
<12> was reading the guidelines ... what is xyz ?
Return to
#php
or
Go to some related
logs:
installing mupen64 on ubuntu
sil3112 debian mount
wg111v2 ubuntu
TuxNotes
lvm2 preinstallation error
rpm exlusive lock linux
USE +fortran +gentoo
nma_dbus_init acquire service
#web
tcp_keepalive_time torrent
|
|