$bookCount = 0; # Parse all the tables on the web page into an array $divs = parse_array($web_page['FILE'], "<div", "</div>"); foreach ($divs as $div) { $divClass = 'views-field-field-image-cache-fid'; if (stristr($div, $divClass)) { $imgTag = parse_array($div, '<img', '/>'); if ($cloudflare == 1) { $books[$bookCount]['imageUrl'] = resolve_address(str_replace("www", "ftp", get_attribute($imgTag[0], $attribute = "src")), $page_base); } else { $books[$bookCount]['imageUrl'] = resolve_address(get_attribute($imgTag[0], $attribute = "src"), $page_base); } if (!is_dir($save_image_directory)) { mkpath($save_image_directory); } $this_image_file = download_binary_file($books[$bookCount]['imageUrl'], $ref = ""); if (stristr($books[$bookCount]['imageUrl'], ".jpg") || stristr($books[$bookCount]['imageUrl'], ".gif") || stristr($books[$bookCount]['imageUrl'], ".png")) { file_put_contents($save_image_directory . basename($books[$bookCount]['imageUrl']), $this_image_file); } } $divClass = 'views-field-title'; if (stristr($div, $divClass)) { $books[$bookCount]['title'] = trim(strip_tags($div)); $aTag = parse_array($div, '<a', '</a>'); if ($cloudflare == 1) { $books[$bookCount]['bookUrl'] = resolve_address(str_replace("www", "ftp", get_attribute($aTag[0], $attribute = "href")), $page_base); } else { $books[$bookCount]['bookUrl'] = resolve_address(get_attribute($aTag[0], $attribute = "href"), $page_base); } $bookPage[$bookCount] = http_get($books[$bookCount]['bookUrl'], $target); $bookDivs[$bookCount] = parse_array($bookPage[$bookCount]['FILE'], "<div class=\"product-body\"", "</div>");
function download_images_for_page($target) { echo "target = {$target}\n"; # Download the web page $web_page = http_get($target, $referer = ""); # Update the target in case there was a redirection $target = $web_page['STATUS']['url']; # Strip file name off target for use as page base $page_base = get_base_page_address($target); # Identify the directory where iamges are to be saved $save_image_directory = "saved_images_" . str_replace("http://", "", $page_base); # Parse the image tags $img_tag_array = parse_array($web_page['FILE'], "<img", ">"); if (count($img_tag_array) == 0) { echo "No images found at {$target}\n"; exit; } # Echo the image source attribute from each image tag for ($xx = 0; $xx < count($img_tag_array); $xx++) { $image_path = get_attribute($img_tag_array[$xx], $attribute = "src"); echo " image: " . $image_path; $image_url = resolve_address($image_path, $page_base); if (get_base_domain_address($page_base) == get_base_domain_address($image_url)) { # Make image storage directory for image, if one doesn't exist $directory = substr($image_path, 0, strrpos($image_path, "/")); $directory = str_replace(":", "-", $directory); $image_path = str_replace(":", "-", $image_path); clearstatcache(); // clear cache to get accurate directory status if (!is_dir($save_image_directory . "/" . $directory)) { mkpath($save_image_directory . "/" . $directory); } # Download the image, report image size $this_image_file = download_binary_file($image_url, $ref = ""); echo " size: " . strlen($this_image_file); # Save the image if (stristr($image_url, ".jpg") || stristr($image_url, ".gif") || stristr($image_url, ".png")) { $fp = fopen($save_image_directory . "/" . $image_path, "w"); fputs($fp, $this_image_file); fclose($fp); echo "\n"; } } else { echo "\nSkipping off-domain image.\n"; } } }
function save_binary_file($url) { $binary_file = download_binary_file($url, $refer = ""); file_put_contents("temp.jpg", $binary_file); }