Say you have a large PHP project and suddenly, when attempting to run it, you just end up with a blank page. The script terminates and you want to find exactly where that is with as little effort as possible.
Is there a tool/program/command/IDE that can, on PHP script termination, tell you the location of a script exit?
Note: I can't mark my own post as "accepted answer" so look at the bottom to see my solution. If you come up with a better solution I will mark your post as the answer.
I use the following code and need no special debugging environment. Note that this might take really long; you can set the ticks count higher - that makes it faster, but blurry.
function shutdown_find_exit()
{
var_dump($GLOBALS['dbg_stack']);
}
register_shutdown_function('shutdown_find_exit');
function write_dbg_stack()
{
$GLOBALS['dbg_stack'] = debug_backtrace();
}
register_tick_function('write_dbg_stack');
declare(ticks=1);
With some inspiration from the nonworking but still right-direction answer from RoBorg, I used the following code in the beginning:
function shutdown() {
global $dbg_stack_a;
print_r($dbg_stack_a);
}
register_shutdown_function('shutdown');
And then I made a global conditional breakpoint (global = breakpoint is evaluated on each row), exploiting the fact that it can run code trough eval(), with the following "condition":
eval('
global $dbg_stack_a, $dbg_stack_b, $dbg_stack_c;
$dbg_stack_a = $dbg_stack_b;
$dbg_stack_b = $dbg_stack_c;
$dbg_stack_c = debug_backtrace();
return false;
')
Probably not fast but does the trick! Using this I was able to determine the exact file and line location that raised die(). (This example works in NuSphere.)
grep -n die filename
Don't forget to grep for "exit" too.
Add this to the top of the file:
function shutdown()
{
print_r(debug_backtrace());
}
register_shutdown_function('shutdown');
See register_ shutdown_function()
You can use an interactive debugger to step through the code until you reach the exit point. Other than that, I think you're down to grep'ing the code for exit|die.
xdebug has a nice trace feature that'll allow you to see all the entire trace of your php app execution and it should give you give clue as to where your exit is.
but for the quick and dirty solution a grep/find as mentioned above will do rightly.
Also check the error___logs for "memory_limit" errors in the Apache error_log.
Memory Limit >= 10M Warning: (Set this to 10M or larger in your php.ini file)
In my experience, scripts suddenly end without warning or notice when this happens.
Make sure that errors are displayed in your development environment (not production).
Related
I have made a function that finds all the URLs within an html file and repeats the same process for each html content linked to the discovered URLs. The function is recursive and can go on endlessly. However, I have put a limit on the recursion by setting a global variable which causes the recursion to stop after 100 recursions.
However, php returns this error:
Fatal error: Maximum function nesting level of '100' reached,
aborting! in
D:\wamp\www\crawler1\simplehtmldom_1_5\simple_html_dom.php on line
1355
I found a solution here: Increasing nesting function calls limit but this is not working in my case.
I am quoting one of the answers from the link mentioned above. Please do consider it.
"Do you have Zend, IonCube, or xDebug installed? If so, that is probably where you are getting this error from.
I ran into this a few years ago, and it ended up being Zend putting that limit there, not PHP. Of course removing it will let >you go past the 100 iterations, but you will eventually hit the memory limits."
Is there a way to increase the maximum function nesting level in PHP
Increase the value of xdebug.max_nesting_level in your php.ini
A simple solution solved my problem. I just commented this line:
zend_extension = "d:/wamp/bin/php/php5.3.8/zend_ext/php_xdebug-2.1.2-5.3-vc9.dll
in my php.ini file. This extension was limiting the stack to 100 so I disabled it. The recursive function is now working as anticipated.
Another solution is to add xdebug.max_nesting_level = 200 in your php.ini
Rather than going for a recursive function calls, work with a queue model to flatten the structure.
$queue = array('http://example.com/first/url');
while (count($queue)) {
$url = array_shift($queue);
$queue = array_merge($queue, find_urls($url));
}
function find_urls($url)
{
$urls = array();
// Some logic filling the variable
return $urls;
}
There are different ways to handle it. You can keep track of more information if you need some insight about the origin or paths traversed. There are also distributed queues that can work off a similar model.
Rather than disabling the xdebug, you can set the higher limit like
xdebug.max_nesting_level=500
It's also possible to fix this directly in php, for example in the config file of your project.
ini_set('xdebug.max_nesting_level', 200);
Go into your php.ini configuration file and change the following line:
xdebug.max_nesting_level=100
to something like:
xdebug.max_nesting_level=200
on Ubuntu using PHP 5.59 :
got to `:
/etc/php5/cli/conf.d
and find your xdebug.ini in that dir, in my case is 20-xdebug.ini
and add this line `
xdebug.max_nesting_level = 200
or this
xdebug.max_nesting_level = -1
set it to -1 and you dont have to worry change the value of the nesting level.
`
probably happened because of xdebug.
Try commenting the following line in your "php.ini" and restart your server to reload PHP.
";xdebug.max_nesting_level"
Try looking in /etc/php5/conf.d/ to see if there is a file called xdebug.ini
max_nesting_level is 100 by default
If it is not set in that file add:
xdebug.max_nesting_level=300
to the end of the list so it looks like this
xdebug.remote_enable=on
xdebug.remote_handler=dbgp
xdebug.remote_host=localhost
xdebug.remote_port=9000
xdebug.profiler_enable=0
xdebug.profiler_enable_trigger=1
xdebug.profiler_output_dir=/home/drupalpro/websites/logs/profiler
xdebug.max_nesting_level=300
you can then use #Andrey's test before and after making this change to see if worked.
php -r 'function foo() { static $x = 1; echo "foo ", $x++, "\n"; foo(); } foo();'
php.ini:
xdebug.max_nesting_level = -1
I'm not entirely sure if the value will ever overflow and reach -1, but it'll either never reach -1, or it'll set the max_nesting_level pretty high.
You could convert your recursive code into an iterative code, which simulates the recursion. This means that you have to push the current status (url, document, position in document etc.) into an array, when you reach a link, and pop it out of the array, when this link has finished.
Check recursion from command line:
php -r 'function foo() { static $x = 1; echo "foo ", $x++, "\n"; foo(); } foo();'
if result > 100 THEN check memory limit;
You could try to wiggle down the nesting by implementing parallel workers (like in cluster computing) instead of increasing the number of nesting function calls.
For example: you define a limited number of slots (eg. 100) and monitor the number of "workers" assigned to each/some of them. If any slots become free, you put the waiting workers "in them".
<?php
ini_set('xdebug.max_nesting_level', 9999);
... your code ...
P.S. Change 9999 to any number you want.
Stumbled upon this bug as well during development.
However, in my case it was caused by an underlying loop of functions calling eachother - as a result of continuous iterations during development.
For future reference by search engines - the exact error my logs provided me with was:
Exception: Maximum function nesting level of '256' reached, aborting!
If, like in my case, the given answers do not solve your problem, make sure you're not accidentally doing something along the lines of the following simplified situation:
function foo(){
// Do something
bar();
}
function bar(){
// Do something else
foo();
}
In this case, even if you set ini_set('xdebug.max_nesting_level', 9999); it will still print out the same error message in your logs.
If you're using Laravel, do
composer update
This should be work.
In your case it's definitely the crawler instance is having more Xdebug limit to trace error and debug info.
But, in other cases also errors like on PHP or core files like CodeIgniter libraries will create such a case and if you even increase the x-debug level setting it would not vanish.
So, look into your code carefully :) .
Here was the issue in my case.
I had a service class which is library in CodeIgniter. Having a function inside like this.
class PaymentService {
private $CI;
public function __construct() {
$this->CI =& get_instance();
}
public function process(){
//lots of Ci referencing here...
}
My controller as follow:
$this->load->library('PaymentService');
$this->process_(); // see I got this wrong instead it shoud be like
Function call on last line was wrong because of the typo, instead it should have been like below:
$this->Payment_service->process(); //the library class name
Then I was keeping getting the exceed error message. But I disabled XDebug but non helped. Any way please check you class name or your code for proper function calling.
I had a error when i was installing many plugins So the error 100 showed including the location of the last plugin that i installed C:\wamp\www\mysite\wp-content\plugins\"..." so i deleted this plugin folder on the C: drive then everything was back to normal.I think i have to limit the amount of plug-in i install or have activated .good luck i hope it helps
I had this issue with WordPress on cloud9. It turns out it was the W3 Caching plugin. I disabled the plugin and it worked fine.
Another solution if you are running php script in CLI(cmd)
The php.ini file that needs edit is different in this case. In my WAMP installation the php.ini file that is loaded in command line is:
\wamp\bin\php\php5.5.12\php.ini
instead of \wamp\bin\apache\apache2.4.9\bin\php.ini which loads when php is run from browser
You can also modify the {debug} function in modifier.debug_print_var.php, in order to limit its recursion into objects.
Around line 45, before :
$results .= '<br>' . str_repeat(' ', $depth * 2)
. '<b> ->' . strtr($curr_key, $_replace) . '</b> = '
. smarty_modifier_debug_print_var($curr_val, ++$depth, $length);
After :
$max_depth = 10;
$results .= '<br>' . str_repeat(' ', $depth * 2)
. '<b> ->' . strtr($curr_key, $_replace) . '</b> = '
. ($depth > $max_depth ? 'Max recursion depth:'.(++$depth) : smarty_modifier_debug_print_var($curr_val, ++$depth, $length));
This way, Xdebug will still behave normally: limit recursion depth in var_dump and so on.
As this is a smarty problem, not a Xdebug one!
I had the same problem and I resolved it like this:
Open MySQL my.ini file
In [mysqld] section, add the following line: innodb_force_recovery =
1
Save the file and try starting MySQL
Remove that line which you just added and Save
In a php script I have some test and after the script the html page.
When a test fail i call die("Test 1 failed");
If no test fail the php script reach the end ?> and then load the html code after the php script.
Is this a good procedure? Or I need to write die() or exit() before the end of php script?
No you don't have to write that and this is not best practice. If the script reaches the end without fatal errros it will exit.
If this means "testing" for you, you're wrong. Testing should be done using unit tests. For php there is phpunit. Give it a try, that's the proper way of testing your code.
Edit: As CompuChip says in a comment, the only useful use case for exit is when you're writing a php based shell script that should return an error code. See the parameter section of the documentation for the exit() function.
You should never be using die() or exit in your production PHP scripts except in very specific cases. Instead, re-work your code paths to simply show an error message to the user rather than exiting the script early.
No you don't need that, but when writing console PHP scripts, you might want to check with for example Bash if the script completed everything in the right way. That's when you use exit() or die()
Is the die() or exit() function needed in the end of a php script?
No, PHP will end the script itself. If the script is an included file (called from another file) then it will end script in the included file and then continue with any code in the original file after where you included (if there is any code).
So you put die() or exit() where ever you want or need it.
For testing, put it after each block of code you test. I use them in some parts of testing if I just want PHP to show me something then stop, such as print out an array to make sure it's being constructed correctly etc.
eg:
print_r($array);
exit();
For other code tests, I sometimes just echo "Section A worked", etc, such as within if/else. If I want to know if a particular part of code is working or if some criteria is being met or not (basically, it lets you trace where PHP itself is going within your code).
All that said, don't use die() or exit() in production code. You should use a more friendly and controlled messaging setup. For security reasons and visual, as you could potentially give them some info like "ERROR Failed to load SomethingSecret". Also it doesn't look pretty when you page only half loads and then puts out an on screen error message which likely means nothing to the end user.
Have a read through this:
PHP Error handling: die() Vs trigger_error() Vs throw Exception
No !
This is not recommanded to use it
Use trigger_error or error_log to log the tests in your error.log. Then check it.
No you don't have to use these functions at the end of the script, because it exists anyway at the end of the script.
No need to put a die or an exit at the end of the scipt.
But you may use exit to terminate your script with a specific exit code (by default it's 0).
E.g
$ php -r "/* does nothing */;"
$ echo $?
0
$ php -r "exit(123);"
$ echo $?
123
http://php.net/exit
From the documentation:
The link to the server will be closed as soon as the execution of the
script ends, unless it's closed earlier by explicitly calling
mysql_close().
https://secure.php.net/function.mysql-connect
Nope, you don't need to call die() or exit(0 if you have another code to run, like you HTML code
Well this isn't true, I'm sure there's a reason, but I can't find it!!
I have a script that can take around 10 minutes to execute. It does a lot of communicating with an api on a service that we have that use. It pulls a bit of a fingerprint of everything every 24 hours. So what it's doing is pretty aside from the point. the probm I'm finding is the script stops executing somewhat randomly!!
I can't find any errors that would cause my script to stop executing, even with
//for debugging
error_reporting(E_ALL);
ini_set('display_errors', '1');
on for debugging, it's all clean. I've also used
set_time_limit(0);
so that it shouldn't ever time out.
With that said, I'm not sure how to get any more debug info to figure out what it's stopping. I can say that the script should NOT be hitting any memory limits or anything. I mean that should throw an error, and I've gone through and cleaned this script up as much as I can see to clean it up.
So my Question is: What are common causes for a cron ending when it shouldn't? How can I debug this more effectively?
You could try using a register_shutdown_function() to define a codeblock that will execute when the script shuts down. Then create a variable across the main code execution points in the cron with details of what is going on. In the shutdown function write this into a log and check your log to see what state the program was in when it stopped. Of course, this is based on the assumption that your code is not totally erroring out.
You could also redirect the standard echo statements and logs into a log file by using
/path/to/cron.php > /path/to/log.txt 2>&1
2>&1 indicates that the standard error (2>) is redirected to the same file descriptor that is pointed by standard output (&1).So, both standard output and error will be redirected to /path/to/log.txt
UPDATE:
Below is a function/flow that I usually use in my crons:
function addLog($msg)
{
if(empty($msg)) return;
$handle = fopen('log.txt', 'a');
$msg = $msg."\r\n";
fwrite($handle,$msg);
fclose($handle);
}
Then I use it like so:
addLog("Initializing...");
init();
addLog("Finished initializing...");
addLog("Calling blah-blah API...");
$result = callBlahBlah();
addLog("blah-blah API returned value". $result);
It is more tedious to have all these logs, but when cron messes up, it really helps!
For eg. when you look at your log.txt and if you see something like:
Initializing...
Finished initializing...
Calling blah-blah API...
And there is no entry which says blah-blah API returned value, then you know that the function call to blah-blah messed up.
What are common causes for a cron ending when it shouldn't?
The most common in my experience is that the cron user has different permissions or different environment variables than the way that you're executing it from the command line.
Make your cronned program dump its environment to a temporary file and see if it's what you expect.
In a php script I have some test and after the script the html page.
When a test fail i call die("Test 1 failed");
If no test fail the php script reach the end ?> and then load the html code after the php script.
Is this a good procedure? Or I need to write die() or exit() before the end of php script?
No you don't have to write that and this is not best practice. If the script reaches the end without fatal errros it will exit.
If this means "testing" for you, you're wrong. Testing should be done using unit tests. For php there is phpunit. Give it a try, that's the proper way of testing your code.
Edit: As CompuChip says in a comment, the only useful use case for exit is when you're writing a php based shell script that should return an error code. See the parameter section of the documentation for the exit() function.
You should never be using die() or exit in your production PHP scripts except in very specific cases. Instead, re-work your code paths to simply show an error message to the user rather than exiting the script early.
No you don't need that, but when writing console PHP scripts, you might want to check with for example Bash if the script completed everything in the right way. That's when you use exit() or die()
Is the die() or exit() function needed in the end of a php script?
No, PHP will end the script itself. If the script is an included file (called from another file) then it will end script in the included file and then continue with any code in the original file after where you included (if there is any code).
So you put die() or exit() where ever you want or need it.
For testing, put it after each block of code you test. I use them in some parts of testing if I just want PHP to show me something then stop, such as print out an array to make sure it's being constructed correctly etc.
eg:
print_r($array);
exit();
For other code tests, I sometimes just echo "Section A worked", etc, such as within if/else. If I want to know if a particular part of code is working or if some criteria is being met or not (basically, it lets you trace where PHP itself is going within your code).
All that said, don't use die() or exit() in production code. You should use a more friendly and controlled messaging setup. For security reasons and visual, as you could potentially give them some info like "ERROR Failed to load SomethingSecret". Also it doesn't look pretty when you page only half loads and then puts out an on screen error message which likely means nothing to the end user.
Have a read through this:
PHP Error handling: die() Vs trigger_error() Vs throw Exception
No !
This is not recommanded to use it
Use trigger_error or error_log to log the tests in your error.log. Then check it.
No you don't have to use these functions at the end of the script, because it exists anyway at the end of the script.
No need to put a die or an exit at the end of the scipt.
But you may use exit to terminate your script with a specific exit code (by default it's 0).
E.g
$ php -r "/* does nothing */;"
$ echo $?
0
$ php -r "exit(123);"
$ echo $?
123
http://php.net/exit
From the documentation:
The link to the server will be closed as soon as the execution of the
script ends, unless it's closed earlier by explicitly calling
mysql_close().
https://secure.php.net/function.mysql-connect
Nope, you don't need to call die() or exit(0 if you have another code to run, like you HTML code
Is there any way to make PHP wait until a function returns before continuing?
This is my code:
<?php
set_time_limit(0);
function waitforchange($nof) {
$lfilemod=filemtime($nof);
while(filemtime($nof) == $lfilemod) {
clearstatcache();
usleep(10000);
}
}
waitforchange('./blahblah.txt')
sleep(5);
echo 'done';
?>
It is supposed to wait until blahblah.txt changes, wait another five seconds after that, then print out "done", however, it prints out "done" after five seconds, regardless whether the file actually changed.
PHP is single-threaded, which means that it executes one instruction at a time before moving on. In other words, PHP naturally waits for a function to finish executing before it continues with the next statement.
I tried executing your code on my machine, and it properly waited and waited until I modified the file before completing the function and displaying the message.
I can't even see anything in your code that might be failing. Since the creation of $lfilemod is performed in the same way as the check in the while loop, the condition of that loop would return TRUE and execute even if there was a problem with the file (filemtime would return FALSE on both errors, so the condition would read (FALSE == FALSE, which is obviously TRUE).
Does your PHP script modify the file at all before running that loop? If it does, then the initial value returned by filemtime might be the modification time from when the script originally started. When you run clearstatcache in the loop, you'll pick up the new modification time caused by your changes earlier in the script.
My advice:
Try running clearstatcache before setting the value of $lfilemod so that you know the value is clean, and that you're comparing apples-to-apples with what is being checked in the loop.
Make sure the file really isn't being modified. Try placing a couple debugging lines at the start and end of your code which prints out the last modification time for the file, then comparing the two yourself to see if PHP is reporting seeing a change in modification time.
This should go without saying, but you should make sure that PHP is configured to display all errors during development, so you are immediately shown when and how things go wrong. Make sure that the display_errors setting is turned On in your php.ini file (or use ini_set() if you can't modify the file itself), and that your error_reporting() is set to E_ALL | E_STRICT for PHP <= 5.3, or E_ALL for PHP 5.4 (E_STRICT is part of E_ALL as of that version). A better way is to just set your error reporting to -1 which effectively turns on all error reporting regardless of PHP version.
Try running your code again with these modifications. If you see that the file really is being modified, then you know that your code works. If the file isn't being modified, you should at least have an error that you can look up, or ask us here.
Try assigning the function call to a variable. I think it doesn't wait for the return otherwise.
The code looks good; save for a missing semicolon (;) after the waitforchange line. I tested putting the (;) in there and the script behaved as intended. Perhaps that is the culprit? I am at loss though to explain how you got your code to execute at all with that error in there.