Logic problem in a loop

[ftrudeau]

Greetings community !

I've build a simple script that parses HTML content, which is written in a plain text file.

In each HTML file, there is a single item that I want to grab, always delimited by specifc HTML code. Problem is, those delimiters may change, and if its the case, I want to be warned, and that my script does not end looping forever, which is the case right now.

What im doing is quite simple. Im opening the file, reading the content line by line, and trying to match a Regex patern on each line to try and find delimiter_start of the item.

If it is found, I loop through remaining lines, fetch and manipulate HTML content, until I find the delimiter_end.

So here are bits and pieces.

$f = fopen ($file_path, "r");

while (!feof($f))
{

   /* Get file content, one line at a time */
   $line = fgets($f, 1024);

   /* First HTML line we will seek */
   if (preg_match("/$item_delimiter_start/", $line))
   {

   /* Read until this line */
  while (!preg_match("/$item_delimiter_end/", $line))
  {
     // HTML manipulation, striping tags, fetching content and etc.
  }

   // content manipulation goes here...

   }

}

I think there might be a problem of logic in my script, but need help figuring out where, and why.

In a much better case, I would need to stop if delimiter_start cannot be found, and warn if delimiter_end did not match any remaining lines.

Any pointers would be appreciated !

Thanks in advance.

[Weedpacket]

Your problem is that you only ever look at one line at a time. If what you're looking for spreads over more than one line you're out of luck. So what you're wanting to do is tack each piece of the file on the end of what you've already got.

$line .= fgets($f, 1024); // Note the .

Couple of things: if you're not actually using regular expressions to identify the tags, then following the tip given on the [man]preg_match[/man] page, you'd be much better off using [man]strpos[/man]. Also, wouldn't [man]file_get_contents[/man] be a much more convenient way of getting the page? Sure you'll end up loading the entire page at once instead of doing so piece by piece, but it would greatly cut down your code.

[ftrudeau]

Thanks M. Packet ; )

However, I like the idea of getting the file content line by line, and the current structure of the code, for the simple reason that i get only exactly what I need out of the file, that is the content of the lines between $item_delimiter_start and $item_delimiter_end.

Maybe im wrong to proceed that way, you, or anyone else, tell me if its the case.

My goal here is not to clean my code, but make sure it is failure free, and that it is as fast as possible.

Thanks for your support.

[dewen]

fgets by default read the line or 1024 bytes, what if you have a line really long?

so try this:

$line = fgets($f, 5000);

[douceur]

Hm, for line by line reading of a file, isn't the file() function easiest?

$file = file($file_path);
foreach ($file as $line) {
   //...
}

Or, if you need the indices,

$file = file($file_path);
for ($i = 0; $i < count($file); $i++) {
   $line = $file[$i];
   //...
}

The first one uses a line less, and the 2nd one includes the index of the line in the array. I don't know, just a thought.

[dewen]

the file size can be big if it's something like log, read them in to array might slow down the system

[ftrudeau]

Thanks for your input guys.

I think I may have found a better solution. I found out that in all cases, it was better to read the entire file using the file() function.

I can then traverse each element of the array, trying to find the first delimiter, and then getting all remaining lines until the end of the file, or until the line matches the ending delimiter.

function ripHTML($file_id)
{
	$delimiter_start_matched = false;
        $delimiter_end_matched = false;

    /* Get all lines from file */
    $lines = file($this->_source_dir . '/' . $file_id);
    $lines_count = count($lines);

    $this->_log->log("[ripHTML] $lines_count lines [" . __LINE__ . ']', PEAR_LOG_DEBUG);

    foreach ($lines as $num => $line)
    {

	/* Trim line */
        	$line = trim($line);

        /* Skip empty line */
	if (empty($line)) continue;

	/* First HTML line we will seek */
	if (preg_match("/$this->_delimiter_start/", $line))
	{
		$this->_log->log("[ripHTML] Start delimiter matches on line $num [" . __LINE__ . ']', PEAR_LOG_DEBUG);
		$delimiter_start_matched = true;

		for ($i = $num; $i <= $lines_count; $i++)
		{
			if (preg_match("/$this->_delimiter_end/", $lines[$i]))
			{   
				$this->_log->log("[ripHTML] End delimiter matches on line $i [" . __LINE__ . ']', PEAR_LOG_DEBUG);
				$delimiter_end_matched = true;
				break;
			}

			/* Trim line */
			$line = trim($lines[$i]);

			/* Strip all HTML tags but IMG */
			$line = strip_tags($line, '<img>');

			/* Skip empty line */    
			if (empty($line)) continue;

			$f_cont[] = $line;                                                                                                                        }                                                                                                                         


		}                                                                                                                                                                                                                                                                           
	}                                                                                                                                 
}                                                                                                                                     

    if (!$delimiter_start_matched)                                                                                                        
    {                                                                                                                                     
	$this->_log->log("[ripHTML] Start delimiter did not match in file [" . __LINE__ . ']', PEAR_LOG_CRIT);                            
	return false;                                                                                                                     
    }                                                                                                                                     

    if (!$delimiter_end_matched)                                                                                                          
    {                                                                                                                                     
	$this->_log->log("[ripHTML] End delimiter did not match in file [" . __LINE__ . ']', PEAR_LOG_WARNING);                           
    }                                                                                                                                     

    return $f_cont;

}