Counting Lines of code

[ober]

Does anyone know of a tool that could count the number of lines of code in a project or just in a bunch of files on a server?

I've been working on a number of applications for work that are all web-based and I'm just curious. For example, one of my backend ajax files is almost 900 lines of code by itself. If I had to guess, I'd say I have close to 20,000 - 30,000 lines of code in all the various applications. And I've written 99% of it myself (I'd say .2% is javascript I've found elsewhere and another .8% is PHP that I picked up on here or other places.

[AndyB]

Not exactly a tool, but with a short script you could read all .php files in a folder, read each file into an array, count array size, add to total, do next file. A few lines of code should be able to handle that.

[ober]

That's kinda what I thought about doing... I'll have to play around with it.

[AndyB]

Quick and dirty in the folder where you want to count:
[code]<?php
$lines = 0;
foreach (glob("*.php") as $filename) {
    $thefile = file($filename);
    $lines+= count($thefile);
}
echo "Lines of code: ". $lines;
?> [/code]

[Guest askjames01]

How about [b]TextPad[/b] ober?
It will make your life easier i think? :)

[a href=\"http://www.textpad.com/\" target=\"_blank\"]http://www.textpad.com/[/a]

it has Document Select for multiple files... and it can count lines of codes... and it's [b]FREE[/b].


try it!
and GOOD LUCK!


-/james

[ober]

Apparently I underestimated myself. Using Andy's code, I'm showing 44,167 lines of code. And that's only the top level of each application, but it also includes a little over 2000 lines that I didn't write. So if it all balances out, I'd say I've written at least 43000 lines of PHP/JS/HTML in the past year and a half.

Crazy.

[redbullmarky]

[!--quoteo(post=354505:date=Mar 13 2006, 02:37 PM:name=ober)--][div class=\'quotetop\']QUOTE(ober @ Mar 13 2006, 02:37 PM) [snapback]354505[/snapback][/div][div class=\'quotemain\'][!--quotec--]
Apparently I underestimated myself. Using Andy's code, I'm showing 44,167 lines of code. And that's only the top level of each application, but it also includes a little over 2000 lines that I didn't write. So if it all balances out, I'd say I've written at least 43000 lines of PHP/JS/HTML in the past year and a half.

Crazy.
[/quote]

who at the zend offices decides to make a function called 'glob'? lol lovely, but andy - thank you, i've been looking for something like that for another little project i'm testing out.

ober, my guess is that if you want a more accurate guess, then unless you neve have blank lines between your lines of code, it maybe a good idea to do a check for lines that are literally just empty.

[ober]

I don't have many blank lines, and I think it'd be hard to differentiate between blank lines and comment lines.

And I'd guess that if you took both of those out, I'd still be above 40,000.

Alright... I had to know:

[code]$lines = 0;
foreach (glob("*.php") as $filename) {
    $thefile = file($filename);
    foreach($thefile as $key => $val){
        if(trim($val) != '')
            $lines++;
    }
}
echo "Lines of code: ". $lines;[/code]

That takes one of my biggest applications from 9474 lines of code to 8630, about an 8.5% drop, which as I guessed, still leaves me about 40000 lines if the ratio is similar in the other apps. Keep in mind that this doesn't remove comment lines.

[redbullmarky]

[!--quoteo(post=354639:date=Mar 13 2006, 08:20 PM:name=ober)--][div class=\'quotetop\']QUOTE(ober @ Mar 13 2006, 08:20 PM) [snapback]354639[/snapback][/div][div class=\'quotemain\'][!--quotec--]
Alright... I had to know:
[/quote]
lol curiosity always wins in the end...

[zq29]

Aahhh, why not check for commented lines too while you're at it...
[code]$lines = 0;
foreach (glob("*.php") as $filename) {
    $thefile = file($filename);
    foreach($thefile as $key => $val){
        if(trim($val) != '' || substr(trim($val),0,2) != "//")
            $lines++;
    }
}
echo "Pretty accurate estimate lines of code: ". $lines;[/code]

[ober]

Your || should be an &&.

Well let's just go full out and list both ;-)

[code]
$comment_lines = 0;
$code_lines = 0;
foreach (glob("*.php") as $filename) {
    $thefile = file($filename);
    foreach($thefile as $key => $val){
        if(trim($val) != '' && substr(trim($val),0,2) != "//" && substr(trim($val),0,2) != "/*")
            $code_lines++;
        if(trim($val) != '' && (substr(trim($val),0,2) == "//" || substr(trim($val),0,2) == "/*"))
            $comment_lines++;
    }
}
echo "Lines of code: $code_lines<br/>";
echo "Lines of comments: $comment_lines<br/>";
[/code]

And we could get more tricky with it... look for lines that have series of "========", "---------", etc... and you could single out full comment blocks as well, looking for the beginning and ending /* */

Crazy.

[TuxToaster]

Here's my updated version, that checks for comment blocks... I'm sure there's probably a better way to do this, but none that I could think of at the moment. It's a quick hack anyways ;-)

[code]
<?php
$comment_lines = 0;
$code_lines = 0;
$in_comment = false;
foreach (glob("*.php") as $filename) {
    $thefile = file($filename);
    foreach($thefile as $key => $val){
        if (trim($val) != '') {
            if ($in_comment) {
                if (trim($val) != "*/") {
                    $comment_lines++;
                }else{
                    $comment_lines++;
                    $in_comment = false;
                }
            }else{
                if (substr(trim($val),0,2) == "//") {
                    $comment_lines++;
                }else if (substr(trim($val),0,2) == "/*") {
                    $comment_lines++;
                    if (substr(trim($val), -2) != "*/") {
                        $in_comment = true;
                    }
                }else{
                    $code_lines++;
                }
            }
        }
    }
}
echo "Lines of code: $code_lines\n";
echo "Lines of comments: $comment_lines\n";
?>
[/code]

[TuxToaster]

Ok, one more fix, this one includes checks to see how much code is between <?php and ?> tags... again there's probably a better or more accurate way, but this at least gives a rough idea. Used it because the project I'm doing includes alot of HTML outside the tags and I don't have time yet to get it separated like I'd like to. It also totals everything up at the end.

Here it is:

[code]
<?php
// Code Counter
$comment_lines = 0;
$code_lines = 0;
$other_lines = 0;
$total_lines = 0;
$in_comment = false;
foreach (glob("*.php") as $filename) {
    $thefile = file($filename);
    foreach($thefile as $key => $val){
        if (trim($val) != '') {
            if (preg_match('/\<\?php/', trim($val))) {
                $in_php = true;
            }
            if ($in_php) {
                if ($in_comment) {
                    if (trim($val) != "*/") {
                        $comment_lines++;
                    }else{
                        $comment_lines++;
                        $in_comment = false;
                    }
                }else{
                    if (substr(trim($val),0,2) == "//") {
                        $comment_lines++;
                    }else if (substr(trim($val),0,2) == "/*") {
                        $comment_lines++;
                        if (substr(trim($val), -2) != "*/") {
                            $in_comment = true;
                        }
                    }else{
                        $code_lines++;
                    }
                }
            }else{
                $other_lines++;
            }
            if (preg_match('/\?\>/', trim($val))) {
                $in_php = false;
            }
        }
    }
}
echo "Lines of PHP code: $code_lines\n";
echo "Lines of other code: $other_lines\n";
echo "Lines of comments: $comment_lines\n";
$total_lines = $code_lines + $other_lines;
echo "---------------------------------------------\n";
echo "Total Lines of code: $total_lines\n";
$total_lines += $comment_lines;
echo "Total Lines(including comments): $total_lines\n";
?>
[/code]

[ober]

Awesome work!

[obsidian]

[!--quoteo(post=364124:date=Apr 12 2006, 01:35 PM:name=ober)--][div class=\'quotetop\']QUOTE(ober @ Apr 12 2006, 01:35 PM) [snapback]364124[/snapback][/div][div class=\'quotemain\'][!--quotec--]
Awesome work!
[/quote]


i just realized that with NONE of the above are you checking for comments starting with "#"... just figured i'd throw that into the mix ;-)

sorry, couldn't resist

now, for the trick i'm working on... make this recursive to read all the sub-folders and keep a log of each folder as well as a running total... that would be awesome

one other problem i just realized... i use quick echoes quite often ("<?= ?>"), and those as well as straight "<?" tags won't be recognized by the above code... something else to keep in mind.

[AndyB]

I happy to see everyone working at turning one little glob into the All_Powerful_Code_Counting_Tool :)

[ypirc]

Also, to just count lines of numerous files of any type in a directory, you might also want to check out the *nix command "wc".

[!--quoteo--][div class=\'quotetop\']QUOTE[/div][div class=\'quotemain\'][!--quotec--]
%wc -l /etc/passwd /etc/services
37 /etc/passwd
2114 /etc/services
2151 total
[/quote]