X-Git-Url: https://git.immae.eu/?a=blobdiff_plain;f=application%2Fhttp%2FHttpUtils.php;h=ed1002b04770b1e5628581c8f9046b96c5cc950c;hb=b3bd8c3e8d367975980043e772f7cd78b7f96bc6;hp=9f4140735a695c4ab8e08b2c10a8e49eaa3527bb;hpb=458b6b9918ec27154dd45416947bb93bedb97109;p=github%2Fshaarli%2FShaarli.git diff --git a/application/http/HttpUtils.php b/application/http/HttpUtils.php index 9f414073..ed1002b0 100644 --- a/application/http/HttpUtils.php +++ b/application/http/HttpUtils.php @@ -6,12 +6,14 @@ use Shaarli\Http\Url; * GET an HTTP URL to retrieve its content * Uses the cURL library or a fallback method * - * @param string $url URL to get (http://...) - * @param int $timeout network timeout (in seconds) - * @param int $maxBytes maximum downloaded bytes (default: 4 MiB) - * @param callable|string $curlWriteFunction Optional callback called during the download (cURL CURLOPT_WRITEFUNCTION). - * Can be used to add download conditions on the - * headers (response code, content type, etc.). + * @param string $url URL to get (http://...) + * @param int $timeout network timeout (in seconds) + * @param int $maxBytes maximum downloaded bytes (default: 4 MiB) + * @param callable|string $curlHeaderFunction Optional callback called during the download of headers + * (CURLOPT_HEADERFUNCTION) + * @param callable|string $curlWriteFunction Optional callback called during the download (cURL CURLOPT_WRITEFUNCTION). + * Can be used to add download conditions on the + * headers (response code, content type, etc.). * * @return array HTTP response headers, downloaded content * @@ -35,8 +37,13 @@ use Shaarli\Http\Url; * @see http://stackoverflow.com/q/9183178 * @see http://stackoverflow.com/q/1462720 */ -function get_http_response($url, $timeout = 30, $maxBytes = 4194304, $curlWriteFunction = null) -{ +function get_http_response( + $url, + $timeout = 30, + $maxBytes = 4194304, + $curlHeaderFunction = null, + $curlWriteFunction = null +) { $urlObj = new Url($url); $cleanUrl = $urlObj->idnToAscii(); @@ -70,7 +77,8 @@ function get_http_response($url, $timeout = 30, $maxBytes = 4194304, $curlWriteF // General cURL settings curl_setopt($ch, CURLOPT_AUTOREFERER, true); curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true); - curl_setopt($ch, CURLOPT_HEADER, true); + // Default header download if the $curlHeaderFunction is not defined + curl_setopt($ch, CURLOPT_HEADER, !is_callable($curlHeaderFunction)); curl_setopt( $ch, CURLOPT_HTTPHEADER, @@ -81,25 +89,21 @@ function get_http_response($url, $timeout = 30, $maxBytes = 4194304, $curlWriteF curl_setopt($ch, CURLOPT_TIMEOUT, $timeout); curl_setopt($ch, CURLOPT_USERAGENT, $userAgent); - if (is_callable($curlWriteFunction)) { - curl_setopt($ch, CURLOPT_WRITEFUNCTION, $curlWriteFunction); - } - // Max download size management curl_setopt($ch, CURLOPT_BUFFERSIZE, 1024*16); curl_setopt($ch, CURLOPT_NOPROGRESS, false); + if (is_callable($curlHeaderFunction)) { + curl_setopt($ch, CURLOPT_HEADERFUNCTION, $curlHeaderFunction); + } + if (is_callable($curlWriteFunction)) { + curl_setopt($ch, CURLOPT_WRITEFUNCTION, $curlWriteFunction); + } curl_setopt( $ch, CURLOPT_PROGRESSFUNCTION, - function ($arg0, $arg1, $arg2, $arg3, $arg4 = 0) use ($maxBytes) { - if (version_compare(phpversion(), '5.5', '<')) { - // PHP version lower than 5.5 - // Callback has 4 arguments - $downloaded = $arg1; - } else { - // Callback has 5 arguments - $downloaded = $arg2; - } + function ($arg0, $arg1, $arg2, $arg3, $arg4) use ($maxBytes) { + $downloaded = $arg2; + // Non-zero return stops downloading return ($downloaded > $maxBytes) ? 1 : 0; } @@ -489,6 +493,46 @@ function is_https($server) return ! empty($server['HTTPS']); } +/** + * Get cURL callback function for CURLOPT_WRITEFUNCTION + * + * @param string $charset to extract from the downloaded page (reference) + * @param string $curlGetInfo Optionally overrides curl_getinfo function + * + * @return Closure + */ +function get_curl_header_callback( + &$charset, + $curlGetInfo = 'curl_getinfo' +) { + $isRedirected = false; + + return function ($ch, $data) use ($curlGetInfo, &$charset, &$isRedirected) { + $responseCode = $curlGetInfo($ch, CURLINFO_RESPONSE_CODE); + $chunkLength = strlen($data); + if (!empty($responseCode) && in_array($responseCode, [301, 302])) { + $isRedirected = true; + return $chunkLength; + } + if (!empty($responseCode) && $responseCode !== 200) { + return false; + } + // After a redirection, the content type will keep the previous request value + // until it finds the next content-type header. + if (! $isRedirected || strpos(strtolower($data), 'content-type') !== false) { + $contentType = $curlGetInfo($ch, CURLINFO_CONTENT_TYPE); + } + if (!empty($contentType) && strpos($contentType, 'text/html') === false) { + return false; + } + if (!empty($contentType) && empty($charset)) { + $charset = header_extract_charset($contentType); + } + + return $chunkLength; + }; +} + /** * Get cURL callback function for CURLOPT_WRITEFUNCTION * @@ -507,9 +551,8 @@ function get_curl_download_callback( &$description, &$keywords, $retrieveDescription, - $curlGetInfo = 'curl_getinfo' + $tagsSeparator ) { - $isRedirected = false; $currentChunk = 0; $foundChunk = null; @@ -524,37 +567,19 @@ function get_curl_download_callback( * * @return int|bool length of $data or false if we need to stop the download */ - return function (&$ch, $data) use ( + return function ($ch, $data) use ( $retrieveDescription, - $curlGetInfo, + $tagsSeparator, &$charset, &$title, &$description, &$keywords, - &$isRedirected, &$currentChunk, &$foundChunk ) { + $chunkLength = strlen($data); $currentChunk++; - $responseCode = $curlGetInfo($ch, CURLINFO_RESPONSE_CODE); - if (!empty($responseCode) && in_array($responseCode, [301, 302])) { - $isRedirected = true; - return strlen($data); - } - if (!empty($responseCode) && $responseCode !== 200) { - return false; - } - // After a redirection, the content type will keep the previous request value - // until it finds the next content-type header. - if (! $isRedirected || strpos(strtolower($data), 'content-type') !== false) { - $contentType = $curlGetInfo($ch, CURLINFO_CONTENT_TYPE); - } - if (!empty($contentType) && strpos($contentType, 'text/html') === false) { - return false; - } - if (!empty($contentType) && empty($charset)) { - $charset = header_extract_charset($contentType); - } + if (empty($charset)) { $charset = html_extract_charset($data); } @@ -562,6 +587,10 @@ function get_curl_download_callback( $title = html_extract_title($data); $foundChunk = ! empty($title) ? $currentChunk : $foundChunk; } + if (empty($title)) { + $title = html_extract_tag('title', $data); + $foundChunk = ! empty($title) ? $currentChunk : $foundChunk; + } if ($retrieveDescription && empty($description)) { $description = html_extract_tag('description', $data); $foundChunk = ! empty($description) ? $currentChunk : $foundChunk; @@ -571,10 +600,10 @@ function get_curl_download_callback( if (! empty($keywords)) { $foundChunk = $currentChunk; // Keywords use the format tag1, tag2 multiple words, tag - // So we format them to match Shaarli's separator and glue multiple words with '-' - $keywords = implode(' ', array_map(function($keyword) { - return implode('-', preg_split('/\s+/', trim($keyword))); - }, explode(',', $keywords))); + // So we split the result with `,`, then if a tag contains the separator we replace it by `-`. + $keywords = tags_array2str(array_map(function(string $keyword) use ($tagsSeparator): string { + return tags_array2str(tags_str2array($keyword, $tagsSeparator), '-'); + }, tags_str2array($keywords, ',')), $tagsSeparator); } } @@ -591,6 +620,6 @@ function get_curl_download_callback( return false; } - return strlen($data); + return $chunkLength; }; }