How to search string in a directory? 2

[spicymango]

Hi,

I need to write a program which go over all the files in my directory/sub directories and search for a string and echo the line where it exits,... has anyone writen a programm which does somthing like this? Is it possible to that in php ............ i know u can do quite easily in unix

 

[Zhris]

Hi,

I once had to perform a similar task. Heres my example:

<?php
function recur($path,$search_pattern) {
	$handle = opendir($path);
	while($file = readdir($handle)) {
		if ($file == '.' || $file == '..') { continue; }
		if(is_dir("$path/$file")) {
			recur("$path/$file",$search_pattern);
		}
		else {
			#####			
			$var=file_get_contents("$path/$file");
			$var=explode("\n", $var);
			foreach ($var as $line=>$data) {
				if (preg_match($search_pattern,$data)) {
					echo "Match Found > Line $line > $path/$file<br>";
				}
			} 
			#####			
		}
	}
	closedir($handle);
}

$path = getcwd();
$search_pattern = '/^searchpattern/';
recur($path,$search_pattern);
?>

Chris

 

[Zhris]

Just a note. I didn't need to use a reccurssive function when I performed this task, therefore the reccurssive nature of the example above is mostly untested and may not work as expected, although can be quickly fixed.

 

[Zhris]

Just 1 final note, as I have tested. Seems to work as expected. I changed:

echo "Match Found > Line $line > $path/$file<br>";

to:

echo "Match Found > Line " . ($line+1) . " > $path/$file<br>";
in order to print the actual line number (as $line will read line 1 as line 0).

Also, its most likely you want to remove the caret ^ symbol from $search_pattern ( $search_pattern = '/searchpattern/'; )

Chris

 

[jpadie]

could you not just use grep with a system call?

 

[spicymango]

thanks a lot Zhris, I will try

 

[spicymango]

I tried it works good. But only in the current folder or directory, if there are subdirectories it does not go into subdirectories

 

[Zhris]

It should work fine. Have you changed $path = getcwd();, to e.g. $path = 'path/to/first/directory'; ? And you've removed the ^ in $search_pattern ? If you've made any notable changed, post your code here.

Chris

 

[spicymango]

i tried again it work great.

But if directory is big it is very slow process. It times out after 30 sec.

unix grep is much faster. but i don't have unix box so i was hoping to achieve it using php

 

[jpadie]

you might find this code a bit quicker. it won't time out either but it might well get upset if you run out of available memory.

<?php
//this example will search recursively downwards from the current directory for the case insensitive string 'pwd' 

new folderGrep('.', '/pwd/i');

class folderGrep{
	
	private $debug = false; //change to true for verbose output
	private $ignoreHidden = true; //change if you want to include files prefixed with '.'
	
	public function __construct($dir, $pattern){
		$this->pattern = $pattern;
		$this->ignore = array('.' ,'..');
		$this->memLimit = $this->returnBytes(ini_get('memory_limit'));
		$this->dir = $this->file = $this->lineNum = $this->line = array();
		$this->start = microtime(true);
		$this->scanDir(realpath($dir));
		$this->displayResults();
	}
	
	private function scanDir($dir){
		set_time_limit(60);
		$dir = trim($dir);
		if (substr($dir, -1) !== '/'){
			$dir .= '/';
		}
		
		if (is_dir($dir)){
			if (is_readable ($dir) ){
				$this->log ($dir, 'opening directory');
				$dh = opendir($dir);
				$this->log($dir, 'directory open');
				$this->log($dir, 'about to scan directory');
				while (FALSE !== ($file = readdir($dh))){
					if (!in_array($file, $this->ignore)){
						if (substr($file,0,1) == '.' && $this->ignoreHidden){
							continue;
						}
						if (is_dir($dir . $file)){
							$this->log($dir.$file, 'recurse as directory');
							$this->scanDir($dir.$file);
						} else {
							if (is_readable($dir.$file)){
								$this->log($dir.$file, 'File is readable');
								if ($this->tooBig($dir.$file)){
									$this->log($dir.$file, 'File is too big to be read in one gulp');
									$fh = fopen($dir.$file, 'rbt');
									$this->log($dir.$file, 'opening file for line by line reading');
									$cnt = 0;
									while (!feof($fh)){
										$line = fgets($fh);
										if (preg_match($this->pattern, $line)){
											$this->log($dir.$file, 'Pattern match found');
											$this->output($dir,$file, $cnt, $line);
										} else {
											$this->log($dir.$file, 'Patern match NOT found');
										}
										$cnt++;
									}
									unset($line);
								} else {
									$lines = file($dir.$file);
									$results = preg_grep($this->pattern, $lines);
									if (is_array($results)){
										foreach ($results as $key=>$result){
											$this->output($dir, $file, $key, $result);
										}
									}
									unset ($lines);
									unset($results);
								}
							}
						}
					} 
				}
				
			} else {
				$this->log($dir , "cannot read directory", 'error');
			}
		} else {
			$this->log($dir, "Not a Directory", 'error');
		}
		
	}
	private function log ($item, $message, $type=null){
		if ($this->debug){
			echo "<p><pre>$item \t $message</pre></p>";
		} else {
			if ($type == 'error'){
				echo "<p><pre>$item \t $message</pre></p>";
			}
		}
	}
	private function output($dir, $file, $lineNum, $line){
		//split into multiple arrays for easy sorting
		$this->dir[] = $dir;
		$this->file[] = $file;
		$this->lineNum[] = $lineNum;
		$this->line[] = $line;
	}
	
	private function tooBig($file){
		if (function_exists('memory_get_usage')){
			$curMemory = memory_get_usage(true);
			if ($curMemory + filesize($file) > $this->memLimit * 0.8){
				return true;
			}	else {
				return false;
			}
		} else {
			if (filesize($file) > ($this->memLimit * 0.5)){
				return true;
			} else {
				return false;
			}
		}
	}
	private function returnBytes($val){
	    $val = trim($val);
	    $last = strtolower(substr($val, -1));
	    switch($last) {
	        // The 'G' modifier is available since PHP 5.1.0
	        case 'g':
	            $val *= 1024;
	        case 'm':
	            $val *= 1024;
	        case 'k':
	            $val *= 1024;
	    }
	
	    return $val;
	}
	private function displayResults(){
		//first order the arrays
		if (count($this->dir) ===0 ){
			echo "No results found";
		} else {
			array_multisort($this->dir, SORT_ASC, SORT_STRING, $this->file, SORT_ASC, SORT_STRING, $this->lineNum, SORT_ASC, SORT_NUMERIC, $this->line);
			$this->stop = microtime(true);
			echo "This operation took " . ($this->stop - $this->start) . " seconds to complete <hr/>";
			//now iterate
			echo '<table width="800px" border="1">';
			echo <<<HTML
		<tr>
			<th scope="col" width="200px">Directory</th>
			<th scope="col" widt="100px">File</th>
			<th scope="col" width="50px">Line Num</th>
			<th scope="col" >Line</th>
		</tr>
HTML;
			$cDir = '';
			$cFile = '';
			foreach ($this->dir as $k=>$d){
				if ($cDir == $d){
					$_d = '&nbsp';
				} else {
					$_d = $cDir = $d;
				} 
				if ($cFile == $this->file[$k]){
					$_f = '&nbsp;';
				} else {
					$_f = $cFile = $this->file[$k];
				}
				$line = htmlentities(nl2br(wordwrap($this->line[$k], 250)));
				echo <<<HTML
		<tr>
			<td>$_d</td>
			<td>$_f</td>
			<td>{$this->lineNum[$k]}</td>
			<td>$line</td>
		</tr>
HTML;
			}
			echo "</table>";
			if (!empty($this->log)) var_dump($this->log);
		}
	}
}
?>

 

[spicymango]

Thanks a lot jpadie, i will try

 

[spicymango]

jpadie your program is a beauty ..

I will try to understand the logic and may ask you few questions later !!

 

[spicymango]

I want to loop through a text file and search all the strings that I have in the file in the directory. So I added some code to your code. Unless I delete previous object before creating new one.

$f=fopen("welcome.txt","r") or exit("Unable to open file!");

while (!feof($f)) 
{ 

$x=fgets($f); 
new folderGrep('.', '/'.$x.'/i');
}
fclose($f);

But this way I will keep creating objects, suppose their are 1000 strings meaning I will create 1000 objects. I am not a OOP person, I was wondering if so many objects will create memory issue.

FULL CODE

<?php
//this example will search recursively downwards from the current directory for the case insensitive string 'pwd' 

$f=fopen("welcome.txt","r") or exit("Unable to open file!");

while (!feof($f)) 
{ 

$x=fgets($f); 
new folderGrep('.', '/'.$x.'/i');
}
fclose($f);

class folderGrep{
    
   private $debug = false; //change to true for verbose output
    private $ignoreHidden = true; //change if you want to include files prefixed with '.'
    
    function __destruct() {
        echo "Person5 Object Released\n";
    }

    
    
    public function __construct($dir, $pattern){
        $this->pattern = $pattern;
        $this->ignore = array('.' ,'..');
        $this->memLimit = $this->returnBytes(ini_get('memory_limit'));
        $this->dir = $this->file = $this->lineNum = $this->line = array();
        $this->start = microtime(true);
        $this->scanDir(realpath($dir));
        $this->displayResults();
    }
    
    private function scanDir($dir){
        set_time_limit(60);
        $dir = trim($dir);
        if (substr($dir, -1) !== '/'){
            $dir .= '/';
        }
        
        if (is_dir($dir)){
            if (is_readable ($dir) ){
                $this->log ($dir, 'opening directory');
                $dh = opendir($dir);
                $this->log($dir, 'directory open');
                $this->log($dir, 'about to scan directory');
                while (FALSE !== ($file = readdir($dh))){
                    if (!in_array($file, $this->ignore)){
                        if (substr($file,0,1) == '.' && $this->ignoreHidden){
                            continue;
                        }
                        if (is_dir($dir . $file)){
                            $this->log($dir.$file, 'recurse as directory');
                            $this->scanDir($dir.$file);
                        } else {
                            if (is_readable($dir.$file)){
                                $this->log($dir.$file, 'File is readable');
                                if ($this->tooBig($dir.$file)){
                                    $this->log($dir.$file, 'File is too big to be read in one gulp');
                                    $fh = fopen($dir.$file, 'rbt');
                                    $this->log($dir.$file, 'opening file for line by line reading');
                                    $cnt = 0;
                                    while (!feof($fh)){
                                        $line = fgets($fh);
                                        if (preg_match($this->pattern, $line)){
                                            $this->log($dir.$file, 'Pattern match found');
                                            $this->output($dir,$file, $cnt, $line);
                                        } else {
                                            $this->log($dir.$file, 'Patern match NOT found');
                                        }
                                        $cnt++;
                                    }
                                    unset($line);
                                } else {
                                    $lines = file($dir.$file);
                                    $results = preg_grep($this->pattern, $lines);
                                    if (is_array($results)){
                                        foreach ($results as $key=>$result){
                                            $this->output($dir, $file, $key, $result);
                                        }
                                    }
                                    unset ($lines);
                                    unset($results);
                                }
                            }
                        }
                    } 
                }
                
            } else {
                $this->log($dir , "cannot read directory", 'error');
            }
        } else {
            $this->log($dir, "Not a Directory", 'error');
        }
        
    }
    private function log ($item, $message, $type=null){
        if ($this->debug){
            echo "<p><pre>$item \t $message</pre></p>";
        } else {
            if ($type == 'error'){
                echo "<p><pre>$item \t $message</pre></p>";
            }
        }
    }
    private function output($dir, $file, $lineNum, $line){
        //split into multiple arrays for easy sorting
        $this->dir[] = $dir;
        $this->file[] = $file;
        $this->lineNum[] = $lineNum;
        $this->line[] = $line;
    }
    
    private function tooBig($file){
        if (function_exists('memory_get_usage')){
            $curMemory = memory_get_usage(true);
            if ($curMemory + filesize($file) > $this->memLimit * 0.8){
                return true;
            }    else {
                return false;
            }
        } else {
            if (filesize($file) > ($this->memLimit * 0.5)){
                return true;
            } else {
                return false;
            }
        }
    }
    private function returnBytes($val){
        $val = trim($val);
        $last = strtolower(substr($val, -1));
        switch($last) {
            // The 'G' modifier is available since PHP 5.1.0
            case 'g':
                $val *= 1024;
            case 'm':
                $val *= 1024;
            case 'k':
                $val *= 1024;
        }
    
        return $val;
    }
    private function displayResults(){
        //first order the arrays
        if (count($this->dir) ===0 ){
            echo "No results found";
        } else {
            array_multisort($this->dir, SORT_ASC, SORT_STRING, $this->file, SORT_ASC, SORT_STRING, $this->lineNum, SORT_ASC, SORT_NUMERIC, $this->line);
            $this->stop = microtime(true);
            echo "This operation took " . ($this->stop - $this->start) . " seconds to complete <hr/>";
            //now iterate
            echo '<table width="800px" border="1">';
            echo <<<HTML
        <tr>
            <th scope="col" width="200px">Directory</th>
            <th scope="col" widt="100px">File</th>
            <th scope="col" width="50px">Line Num</th>
            <th scope="col" >Line</th>
        </tr>
HTML;
            $cDir = '';
            $cFile = '';
            foreach ($this->dir as $k=>$d){
                if ($cDir == $d){
                    $_d = '&nbsp';
                } else {
                    $_d = $cDir = $d;
                } 
                if ($cFile == $this->file[$k]){
                    $_f = '&nbsp;';
                } else {
                    $_f = $cFile = $this->file[$k];
                }
                $line = htmlentities(nl2br(wordwrap($this->line[$k], 250)));
                echo <<<HTML
        <tr>
            <td>$_d</td>
            <td>$_f</td>
            <td>{$this->lineNum[$k]}</td>
            <td>$line</td>
        </tr>
HTML;
            }
            echo "</table>";
            if (!empty($this->log)) var_dump($this->log);
        }
    }
}
?>

 

[jpadie]

i don't think it will get intrinsically unwieldy. but why not instead pass an array of patterns into the script and adjust the loops to iterate on each pattern?

 

[spicymango]

Yes array might be a better choice. Just a OOP question, when we create objects do they get deleted on their own when programe done executing . Or we have to put code to delete them.

 

[jpadie]

they are like any variable. they disappear once their scope evaporates.