Jump to content

whats wrong with this code. No result

Help!php

By Help!php
in PHP Coding Help

 Share

Recommended Posts

Help!php Member

Help!php

Posted
Posted

This code should read a sitemap and print results on a text file but for some reason it doesnt. any idea why or whats wrong with it

 

require_once 'simplehtmldom/simple_html_dom.php';

$_ECHO = FALSE;
$html = new simple_html_dom();
$printerListFileName = "pl_list.txt";
$outputFileName = "pl_new.txt";

//////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//
//		cleanBadChars()
//
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////
function cleanBadChars( $plFieldValue )
{
	$badChars = array('â', '€', '™', 'â', 'Â', '“');

	$cleanFieldValue = str_replace( $badChars, "", $plFieldValue );

	return( trim( $cleanFieldValue ) );
}

//////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//
//		cleanDBField()
//
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////
function cleanDBField( $CLPField, $plFieldValueDirty )
{
	global $_ECHO;

	if($_ECHO) echo "$CLPField=>$plFieldValueDirty<br />";

	$plFieldValueClean = cleanBadChars( $plFieldValueDirty );

	switch( $CLPField )
	{
		case "productid":
			$clpFieldValueClean = preg_replace( "/Item Number:  /", "", $plFieldValueClean );
			$clpFieldValueClean = str_replace( "/", "_", $clpFieldValueClean );
		break;

		case "name":
			$clpFieldValueClean = trim( strstr( $plFieldValueClean, " " ) );
			$clpFieldValueClean = strstr( $clpFieldValueClean, "Minolta" ) ? strstr( $clpFieldValueClean, " " ) : $clpFieldValueClean;
		break;

		case "manufacturer":
			$clpFieldValueClean = ucfirst( strtolower( substr( $plFieldValueClean, 0, strpos( $plFieldValueClean, " " ) ) ) );
			$clpFieldValueClean = strstr( $clpFieldValueClean, "Konica" ) ? $clpFieldValueClean . " Minolta" : $clpFieldValueClean;
			$clpFieldValueClean = strstr( $clpFieldValueClean, "Hp" ) ? "HP" : $clpFieldValueClean;
		break;

		case "format":
			$clpFieldValueClean = $plFieldValueClean;
		break;

		case "platform":
			$arr = array();
			$arr[] = strstr( $plFieldValueClean, "Windows" ) ? "Windows" : "";
			$arr[] = strstr( $plFieldValueClean, "Mac" ) ? "Mac" : "";
			$clpFieldValueClean = count( $arr ) == 2 ? implode( " / ", $arr ) : implode( "", $arr );
		break;

		case "bwppm":
			$clpFieldValueClean = strstr( $plFieldValueClean, "i" ) ? substr( $plFieldValueClean, 0, strpos( $plFieldValueClean, "i" ) ) : substr( $plFieldValueClean, 0, strpos( $plFieldValueClean, "p" ) );
			$clpFieldValueClean = ( $num = intval( $clpFieldValueClean ) ) > 0 ? $num : "NULL";
		break;

		case "cppm":
			$clpFieldValueClean = strstr( $plFieldValueClean, "i" ) ? substr( $plFieldValueClean, 0, strpos( $plFieldValueClean, "i" ) ) : substr( $plFieldValueClean, 0, strpos( $plFieldValueClean, "p" ) );
			$clpFieldValueClean = ( $num = intval( $clpFieldValueClean ) ) > 0 ? $num : "NULL";
		break;

		case "resolution":
			$clpFieldValueClean = substr( $plFieldValueClean, 0, strpos( $plFieldValueClean, "dpi"  ) );
		break;

		case "ram":
			$ramValues = explode( " ", $plFieldValueClean );
			$clpFieldValueClean = str_replace( array( "MB", "GB", "KB" ), "", $ramValues[ 0 ] );
		break;

		case "maxram":
			$clpFieldValueClean = $plFieldValueClean;
		break;

		case "ethernet":
			$clpFieldValueClean = !strstr( $plFieldValueClean, "Yes") ? "NULL" : "Yes";
		break;

		case "usb":
			$clpFieldValueClean = strlen ( $plFieldValueClean ) > 1 ? "Yes" : "NULL";
		break;

		case "firstprint":
			$clpFieldValueClean = explode( " ",  $plFieldValueClean );
			$clpFieldValueClean = $clpFieldValueClean[ 0 ];
		break;

		case "parallel":
			$clpFieldValueClean = strstr( $plFieldValueClean, "Parallel" ) ? "Yes" : "NULL";
		break;

		case "duplex":
			$clpFieldValueClean = strlen( $plFieldValueClean ) < 2 ? "Manual" : $plFieldValueClean;
		break;

		case "printmethod":
			$clpFieldValueClean = preg_replace( "/ Printer/", "", $plFieldValueClean );
			$clpFieldValueClean = preg_replace( "/ Fax/", "", $clpFieldValueClean );
		break;

		case "category":
			if ( strstr( $plFieldValueClean, "Multifunction" ) )
				$clpFieldValueClean = "Multifunction";
			elseif ( strstr( $plFieldValueClean, "Laser" ) )
				$clpFieldValueClean = strstr( $plFieldValueClean, "Colour" ) ? "Colour Laser" : "Mono Laser";
			elseif ( strstr( $plFieldValueClean, "Fax" ) )
				$clpFieldValueClean = "Fax";
			elseif ( strstr( $plFieldValueClean, "Dot Matrix" ) )
				$clpFieldValueClean = "Dot Matrix";
			elseif ( strstr( $plFieldValueClean, "Inkjet" ) )
				$clpFieldValueClean = "Inkjet";
			elseif ( strstr( $plFieldValueClean, "Label" ) )
				$clpFieldValueClean = "Label";
			elseif ( strstr( $plFieldValueClean, "Scanner" ) )
				$clpFieldValueClean = "Scanner";
			elseif ( strstr( $plFieldValueClean, "Thermal" ) )
				$clpFieldValueClean = "Thermal";
		break;

		case "description":
			$clpFieldValueClean = $plFieldValueClean;
		break;

		case "rrp":
			$clpFieldValueClean = strstr( $plFieldValueClean, "£" );
			$clpFieldValueClean = preg_replace( "/£/", "", $clpFieldValueClean );
		break;

		case "paper":
			$clpFieldValueClean = $plFieldValueClean;
		break;

		case "additional":
			$clpFieldValueClean = $plFieldValueClean;
		break;

		case "offertext":
			if ( $plFieldValueClean == "" )
				$clpFieldValueClean = "New Low price On This Printer";
			else
			{					
				$clpFieldValueClean = str_replace( "£", "pound;", $plFieldValueClean );
				$clpFieldValueClean = str_replace( "&", "amp;", $clpFieldValueClean );			
				$clpFieldValueClean = str_replace( "pound;", "£", $clpFieldValueClean );
				$clpFieldValueClean = str_replace( "amp;", "&", $clpFieldValueClean );			

				$clpFieldValueClean .= " <br /><br />Offer ends";
			}
		break;

		case "specialhead":
			if ( $plFieldValueClean == "" )
				$clpFieldValueClean = "New Low price On This Printer";
			else
			{					
				$clpFieldValueClean = str_replace( "£", "pound;", $plFieldValueClean);
				$clpFieldValueClean = str_replace( "&", "amp;", $clpFieldValueClean );			
				$clpFieldValueClean = str_replace( "pound;", "£", $clpFieldValueClean );
				$clpFieldValueClean = str_replace( "amp;", "&", $clpFieldValueClean );
			}
		break;

		default:
		break;
	}

	$clpFieldValueClean = preg_replace( "/\"/", "", $clpFieldValueClean );

	if($_ECHO) echo "$CLPField=>$clpFieldValueClean<br />";

	return( trim( $clpFieldValueClean ) );	
}

//////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//
//		fetchPrinterDetails()
//
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////
function fetchPrinterDetails( $printerURL )
{
	global $html;
	global $outputFileName;
	global $_ECHO;

	$line = "";
	$fileContents = file_get_contents( $printerURL );

	if($_ECHO) echo "Scraping details started for " . $printerURL . "<br />";

	$html->load( $fileContents );

	$stop = FALSE;
	$metas = $html->find( "meta[name=Keywords]" );

	if ( isset( $metas[ 0 ] ) )
	{
		$stop = 	strstr( strtoupper( $metas[ 0 ]->content ), "EXDEMO" ) || 
					strstr( strtoupper( $metas[ 0 ]->content ), "BOXOPEN" )|| 
					strstr( strtoupper( $metas[ 0 ]->content ), "BOX OPEN" )|| 
					strstr( strtoupper( $metas[ 0 ]->content ), "DISCONTINUED" );

		$stop = $stop ? $stop : !( strstr( strtoupper( $metas[ 0 ]->content ), "PRINTER" ) || strstr( strtoupper( $metas[ 0 ]->content ), "FAX" ) );
		$stop = $stop ? $stop :    strstr( strtoupper( $metas[ 0 ]->content ), "ACCESSORIES" );
	}
	else
		if ( $_ECHO ) echo "<meta> tag NOT FOUND<br />";

	// Dont bother with non-current  printers 
	if ( $stop )
	{
		if ( $_ECHO ) echo "Ignoring $printerURL<br />";
		return;
	}

	$DBFields = array 
	( 
		"productid" => "#ctl00_placeholderMain_lblItem", 
		"name" => "#ctl00_placeholderMain_lblProductHead", 
		"manufacturer" => "h1", 
		"format" => "@Product Group Output",
		"platform" => "@Operating Systems Supported", 
		"height" => "NULL", 
		"width" => "NULL", 
		"depth" => "NULL", 
		"weight" => "NULL", 
		"bwppm" => "@Speed Monochrome", 
		"cppm" => "@Speed Colour",
		"resolution" => "Printer Resolution@Printer Enhanced Resolution",
		"ram" => "@Memory (Maximum)",
		"maxram" => "NULL",
		"ethernet" => "@Network Ready",
		"parallel" => "@Interface Type(s)", 
		"usb" => "USB Port@USB Ports", 
		"firstprint" => "First Page@Print First Page", 
		"warmupprint" => "NULL", 
		"duplex" => "@Double Sided Printing", 
		"printmethod" => "@Technology", 
		"relability" => "NULL", 
		"standby" => "NULL", 
		"running" => "NULL", 
		"category" => "h1", 
		"description" => ".productdescriptioncontainer", 
		"rrp" => "#ctl00_placeholderMain_lbltxtProductPrice", 
		"printspeed" => "NULL", 
		"large" => "NULL", 
		"discont" => "NULL", 
		"pdf" => "DEFAULT=1", 
		"paper" => "@Paper Handling Input 1", 
		"multi" => "NULL", 
		"additional" => "@Paper Handling Input 2", 
		"CPppma3" => "NULL", 
		"CPppm" => "NULL", 
		"CPram" => "NULL", 
		"CPmaxram" => "NULL", 
		"CPresolution" => "NULL", 
		"Fmodem" => "NULL", 
		"Fresolution" => "NULL", 
		"Fcompatability" => "NULL", 
		"Fram" => "NULL", 
		"Fmaxram" => "NULL", 
		"SCspeed" => "NULL", 
		"SCresolution" => "NULL", 
		"SCmodes" => "NULL", 
		"specialid" => "DEFAULT=1", 
		"offertext" => "#ctl00_placeholderMain_lblMareketingText", 
		"image" => "NULL", 
		"promo" => "NULL", 
		"metatag" => "NULL", 
		"metadescrip" => "NULL", 
		"pricerunner" => "NULL", 
		"google" => "NULL", 
		"offerdate" => "NULL", 
		"specialtext" => "NULL", 
		"specialhead" => "#ctl00_placeholderMain_lblMareketingText" 
	);

	foreach( $DBFields as $CLPField => $PLField )
	{ echo $PLField;
		if ( $PLField == "NULL" )
			$line .= '"' . trim( $PLField ) . '",';
		elseif ( strstr( $PLField, "DEFAULT=" ) )
			$line .= '"' . str_replace( "DEFAULT=", "", $PLField ) . '",';
		else
		{
			// This is a Spec field so we will need to work out which one
			if ( strstr( $PLField, "@" ) != FALSE )
			{
				// Get all the spec titles
				$specTitles = $html->find( ".specleftitem" );

				// Look for the field title
				if ( isset( $specTitles[ 0 ] ) )
				{
					$clpFieldValue = "NULL";
					$possFields = explode( "@", $PLField );

					// Loop thru all spec items
					foreach( $specTitles as $specTitle )
					{
						// Check all poss fields for a match
						foreach( $possFields as $possField )
						{
							if ( trim( $specTitle->plaintext ) == $possField )
							{
								$clpFieldValue = $specTitle->next_sibling()->plaintext;
								$line .= '"' . cleanDBField( $CLPField, $clpFieldValue ) . '",';
								break;
							}
						}

						if ( $clpFieldValue != "NULL" )
							break;
					}

					if ( $clpFieldValue == "NULL" )
						$line .= '"' . $clpFieldValue . '",';
				}
			}
			else
			{
				$plFieldValue = $html->find( $PLField );

				// Found the field in the PL page ?
				if ( isset( $plFieldValue[ 0 ] ) )
				{
					$clpFieldValue = $plFieldValue[ 0 ]->plaintext;
					$line .= '"' . cleanDBField( $CLPField, $clpFieldValue ) . '",';
				}
				else
				{
					$line .= '"NULL",';
				}
			}
		}
	}

	$line = preg_replace( "/,$/", "\n", trim( $line ) );

	$fp = fopen( $outputFileName, "a" );
	fputs( $fp, $line );
	fclose( $fp );
	//echo "stop:". $stop;
	if($_ECHO) echo "Scraping details completed for " . $printerURL . "<br />";
}

//////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//
//		scrapePrinters()
//
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////
function scrapePrinters()
{
	global $printerListFileName;
	global $outputFileName;
	global $_ECHO;

	set_time_limit( 0 );

	if ( file_exists( $printerListFileName ) == FALSE )
	{
		if($_ECHO) echo "Cannnot find $printerListFileName so quitting...<br />";
		exit(0);
	}

	if($_ECHO) echo "Deleting existing $outputFileName file...<br />";

	if ( file_exists( $outputFileName ) == TRUE )
		unlink( $outputFileName );

	// List of PL printers taken from sitemap page of PL website
	$fp = fopen( $printerListFileName, "r" );

	if($_ECHO) echo "Fetching printer details started...<br />";

	while ( $printerURL = fgets( $fp ) )	
		fetchPrinterDetails( trim( $printerURL ) );

	if($_ECHO) echo "Fetching printer details completed...<br />";

	fclose( $fp );
}

scrapePrinters();

function test()
{
	global $html;

	$files = array( "OKI-C810n-Box-Opened--P110692.aspx", "HP-1320-P4453.aspx", "Waste-Toner-Cleaner-Pack-12-000-Pages--P48796.aspx", "Lexmark-C543dn-P6117.aspx", "Brother-FAX-T104-P11767.aspx", "Black-Toner-3500-pages--P110364.aspx", "Lexmark-X544dn-P9732.aspx", "EB-05-IEEE-1394-Expansion-Board-P30721.aspx", "Kodak-Photo-Paper-Gloss-A4-210-x-297mm-20-Sheets-165gsm--P13998.aspx", "Xerox-7600-P13571.aspx" );

	echo "STARTING...<br />";

	foreach( $files as $file )
	{
		fetchPrinterDetails( trim( $file ) );
	}

	echo "DONE...<br />";
}

Link to comment
Share on other sites
This thread is more than a year old. Please don't revive it unless you have something important to add.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
 Share
Go to topic listing
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.