+News for v1.4.0:
+================
+
+ * Make the images adapt to the screen width in feed readers which render the
+ HTML data in the description.
+ * Indicate if there is a GIF image in a tweet.
+ * Add option to enable or disable showing usernames in RSS items.
+ * Retry multiple times to retrieve a resource before giving up.
+ * Fix coding style.
+ * Add option to enable or disable showing multimedia content in RSS items.
+ * Fix generating enclosures for Dilbert.com
+ * Make enclosure elements validate with feedvalidator.org when the server
+ does not provide a Content-Length header.
+
News for v1.3.0:
================
-- re-evaluate the use of trigger_error() or use a custom error handler,
- because right now the code exists as soon as trigger_error() gets called and
- any following code is ignored.
-
- write better XSL stylesheets? I am not an XSL expert
- evaluate the use of the <ttl/> RSS element
- show cards directly in RSS items for twitter.com
<?php
+
/**
* @file
* Tweeper - some logic to allow tweeper to run with or without composer.
*/
class Tweeper {
- private static $userAgent = "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101 Firefox/52.0";
+ private static $userAgent = "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:60.0) Gecko/20100101 Firefox/60.0";
+ private static $maxConnectionTimeout = 5;
+ private static $maxConnectionRetries = 5;
/**
- * Constructor sets up {@link $generate_enclosure}.
+ * Create a new Tweeper object controlling optional settings.
+ *
+ * @param bool $generate_enclosure
+ * Enables the creation of <enclosure/> elements (disabled by default).
+ * @param bool $show_usernames
+ * Enables showing the username in front of the content for multi-user
+ * sites (enabled by default). Only some stylesheets supports this
+ * functionality (twitter, instagram, pump.io).
+ * @param bool $show_multimedia
+ * Enables showing multimedia content (images, videos) directly in the
+ * item description (enabled by default). Only some stylesheets supports
+ * this functionality (twitter, instagram, dilbert).
*/
- public function __construct($generate_enclosure = FALSE) {
+ public function __construct($generate_enclosure = FALSE, $show_usernames = TRUE, $show_multimedia = TRUE) {
$this->generate_enclosure = $generate_enclosure;
+ $this->show_usernames = $show_usernames;
+ $this->show_multimedia = $show_multimedia;
}
/**
}
/**
+ * Perform a cURL session multiple times when it fails with a timeout.
+ *
+ * @param resource $ch
+ * a cURL session handle.
+ */
+ private static function curlExec($ch) {
+ $ret = FALSE;
+ $attempt = 0;
+ do {
+ $ret = curl_exec($ch);
+ if (FALSE === $ret) {
+ trigger_error(curl_error($ch), E_USER_WARNING);
+ }
+ } while (curl_errno($ch) == CURLE_OPERATION_TIMEDOUT && ++$attempt < Tweeper::$maxConnectionRetries);
+
+ return $ret;
+ }
+
+ /**
* Get the contents from a URL.
*/
private static function getUrlContents($url) {
$ch = curl_init($url);
curl_setopt_array($ch, array(
CURLOPT_HEADER => FALSE,
+ CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
// Follow http redirects to get the real URL.
CURLOPT_FOLLOWLOCATION => TRUE,
CURLOPT_RETURNTRANSFER => TRUE,
CURLOPT_HTTPHEADER => array('Accept-language: en'),
CURLOPT_USERAGENT => Tweeper::$userAgent,
));
- $contents = curl_exec($ch);
- if (FALSE === $contents) {
- trigger_error(curl_error($ch));
- }
+ $contents = Tweeper::curlExec($ch);
curl_close($ch);
return $contents;
curl_setopt_array($ch, array(
CURLOPT_HEADER => TRUE,
CURLOPT_NOBODY => TRUE,
+ CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
// Follow http redirects to get the real URL.
CURLOPT_FOLLOWLOCATION => TRUE,
CURLOPT_RETURNTRANSFER => TRUE,
CURLOPT_SSL_VERIFYPEER => FALSE,
CURLOPT_USERAGENT => Tweeper::$userAgent,
));
- curl_exec($ch);
+
+ $ret = Tweeper::curlExec($ch);
+ if (FALSE === $ret) {
+ curl_close($ch);
+ return FALSE;
+ }
+
$url_info = curl_getinfo($ch);
if (FALSE === $url_info) {
- trigger_error(curl_error($ch));
+ trigger_error(curl_error($ch), E_USER_WARNING);
}
curl_close($ch);
);
$url_info = Tweeper::getUrlInfo($url);
+ if (FALSE === $url_info) {
+ trigger_error("Failed to retrieve info for URL: " . $url, E_USER_WARNING);
+ return '';
+ }
$supported = in_array($url_info['content_type'], $supported_content_types);
if (!$supported) {
- error_log("Unsupported enclosure content type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url']);
+ trigger_error("Unsupported enclosure content type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url'], E_USER_WARNING);
return '';
}
// See http://sourceforge.net/p/feedvalidator/bugs/72/
$http_url = preg_replace("/^https/", "http", $url_info['url']);
+ // When the server does not provide a Content-Length header,
+ // curl_getinfo() would return a negative value for
+ // "download_content_length", however RSS recommends to use 0 when the
+ // enclosure's size cannot be determined.
+ // See: https://www.feedvalidator.org/docs/error/UseZeroForUnknown.html
+ $length = max($url_info['download_content_length'], 0);
+
$dom = new DOMDocument();
$enc = $dom->createElement('enclosure');
$enc->setAttribute('url', $http_url);
- $enc->setAttribute('length', $url_info['download_content_length']);
+ $enc->setAttribute('length', $length);
$enc->setAttribute('type', $url_info['content_type']);
return $enc;
$output .= " line $error->line";
- error_log($output);
+ trigger_error($output, E_USER_WARNING);
}
/**
$xml_data = $serializer->serialize($data, 'xml', $serializer_options);
if (!$xml_data) {
- trigger_error("Cannot serialize data", E_USER_ERROR);
+ trigger_error("Cannot serialize data", E_USER_WARNING);
return NULL;
}
$json_match_expr = '/window._sharedData = (.*);/';
$ret = preg_match($json_match_expr, $html, $matches);
if ($ret !== 1) {
- trigger_error("Cannot match expression: $json_match_expr\n", E_USER_ERROR);
+ trigger_error("Cannot match expression: $json_match_expr\n", E_USER_WARNING);
return NULL;
}
+ $data = json_decode($matches[1], $assoc = TRUE);
+
// The "qe" object contains elements which will result in invalid XML
// element names, so remove it.
- $data = json_decode($matches[1], $assoc = TRUE);
unset($data["qe"]);
+
+ // The "knobs" object contains elements with undefined namespaces, so
+ // remove it to silence an error message.
+ unset($data["knobs"]);
+
$json = json_encode($data);
return Tweeper::jsonToXml($json, 'instagram');
private function loadStylesheet($host) {
$stylesheet = "file://" . __DIR__ . "/rss_converter_" . $host . ".xsl";
if (FALSE === file_exists($stylesheet)) {
- trigger_error("Conversion to RSS not supported for $host ($stylesheet not found)", E_USER_ERROR);
+ trigger_error("Conversion to RSS not supported for $host ($stylesheet not found)", E_USER_WARNING);
return NULL;
}
$stylesheet_contents = Tweeper::getUrlContents($stylesheet);
+ if (FALSE === $stylesheet_contents) {
+ trigger_error("Cannot open $stylesheet", E_USER_WARNING);
+ return NULL;
+ }
$xslDoc = new DOMDocument();
$xslDoc->loadXML($stylesheet_contents);
$xsltProcessor = new XSLTProcessor();
$xsltProcessor->registerPHPFunctions();
$xsltProcessor->setParameter('', 'generate-enclosure', $this->generate_enclosure);
+ $xsltProcessor->setParameter('', 'show-usernames', $this->show_usernames);
+ $xsltProcessor->setParameter('', 'show-multimedia', $this->show_multimedia);
$xsltProcessor->importStylesheet($xslDoc);
return $xsltProcessor;
/**
* Convert the site content to RSS.
*/
- public function tweep($src_url, $host=NULL, $validate_scheme=TRUE) {
+ public function tweep($src_url, $host = NULL, $validate_scheme = TRUE) {
$url = parse_url($src_url);
if (FALSE === $url) {
- trigger_error("Invalid URL: $src_url", E_USER_ERROR);
+ trigger_error("Invalid URL: $src_url", E_USER_WARNING);
return NULL;
}
if (TRUE === $validate_scheme) {
$scheme = $url["scheme"];
if (!in_array($scheme, array("http", "https"))) {
- trigger_error("unsupported scheme: $scheme", E_USER_ERROR);
+ trigger_error("unsupported scheme: $scheme", E_USER_WARNING);
return NULL;
}
}
- // if the host is not given derive it from the URL
+ // If the host is not given derive it from the URL.
if (NULL === $host) {
if (empty($url["host"])) {
- trigger_error("Invalid host in URL: $src_url", E_USER_ERROR);
+ trigger_error("Invalid host in URL: $src_url", E_USER_WARNING);
return NULL;
}
// Strip the leading www. to be more forgiving on input URLs.
$html = Tweeper::getUrlContents($src_url);
if (FALSE === $html) {
+ trigger_error("Failed to retrieve $src_url", E_USER_WARNING);
return NULL;
}
}
$output = $xsltProcessor->transformToXML($xmlDoc);
-
if (FALSE === $output) {
- trigger_error('XSL transformation failed.', E_USER_ERROR);
+ trigger_error('XSL transformation failed.', E_USER_WARNING);
return NULL;
}
+
return $output;
}
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:php="http://php.net/xsl"
- xsl:extension-element-prefixes="php"
exclude-result-prefixes="php">
+ <xsl:param name="generate-enclosure"/>
+ <xsl:param name="show-multimedia"/>
+
<xsl:output method="xml" indent="yes"/>
<xsl:variable name="BaseURL" select="//meta[@property='og:url']/@content"/>
</pubDate>
<description>
<xsl:text disable-output-escaping="yes"><![CDATA[</xsl:text>
- <img src="{$picture-url}" alt="{$picture-title}"/>
+ <xsl:if test="$show-multimedia = 1">
+ <img src="{$picture-url}" alt="{$picture-title}"/>
+ </xsl:if>
<xsl:text disable-output-escaping="yes">]]></xsl:text>
</description>
<xsl:if test="$generate-enclosure = 1">
- <xsl:copy-of select="php:functionString('Tweeper\Tweeper::generateEnclosure', $picture-url)"/>
+ <!--
+ Dilbert.com uses protocol-relative urls for pictures but
+ generateEnclosure() relies on curl which cannot work
+ without a schema.
+
+ Use http as protocol because curl gives some error when
+ attempting TLS negotiation with the server where
+ Dilbert.com assets are.
+ -->
+ <xsl:copy-of select="php:functionString('Tweeper\Tweeper::generateEnclosure', concat('http:', $picture-url))"/>
</xsl:if>
</item>
</xsl:template>
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:php="http://php.net/xsl"
- xsl:extension-element-prefixes="php"
exclude-result-prefixes="php">
<xsl:output method="xml" indent="yes"/>
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:php="http://php.net/xsl"
- xsl:extension-element-prefixes="php"
exclude-result-prefixes="php">
<xsl:param name="generate-enclosure"/>
+ <xsl:param name="show-usernames"/>
+ <xsl:param name="show-multimedia"/>
<xsl:output method="xml" indent="yes"/>
<item>
<title>
<xsl:variable name="title-length" select="140"/>
- <xsl:variable name="item-content-title" select="normalize-space(concat($screen-name, ': ', $item-content-caption))"/>
+ <xsl:variable name="item-content-title">
+ <xsl:if test="$show-usernames = 1">
+ <xsl:value-of select="concat($screen-name, ': ')"/>
+ </xsl:if>
+ <xsl:value-of select="normalize-space($item-content-caption)"/>
+ </xsl:variable>
<!-- ellipsize, inspired from http://stackoverflow.com/questions/13622338 -->
<xsl:choose>
<xsl:when test="string-length($item-content-title) > $title-length">
</xsl:if>
<xsl:value-of select="$item-content-caption"/>
</p><br />
- <a href="{$item-permalink}"><img src="{$item-content-image}" style="max-width: 100%"/></a>
+ <xsl:if test="$show-multimedia = 1">
+ <a href="{$item-permalink}"><img src="{$item-content-image}" style="max-width: 100%"/></a>
+ </xsl:if>
<xsl:text disable-output-escaping="yes">]]></xsl:text>
</description>
<xsl:if test="$generate-enclosure = 1">
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:php="http://php.net/xsl"
- xsl:extension-element-prefixes="php"
exclude-result-prefixes="php">
+ <xsl:param name="generate-enclosure"/>
+ <xsl:param name="show-usernames"/>
+
<xsl:output method="xml" indent="yes"/>
<xsl:variable name="domain-name" select="substring-after(//div[@id='profile-block']/@data-profile-id, '@')"/>
<xsl:variable name="item-permalink" select=".//p[@class='muted']/small/a/@href"/>
<item>
<title>
- <xsl:value-of select="concat($user-name, ': ', normalize-space($item-content))"/>
+ <xsl:if test="$show-usernames = 1">
+ <xsl:value-of select="concat($user-name, ': ')"/>
+ </xsl:if>
+ <xsl:value-of select="normalize-space($item-content)"/>
</title>
<link>
<xsl:value-of select="$item-permalink"/>
<xsl:value-of select="php:functionString('Tweeper\Tweeper::strToRssDate', .//abbr[@class='easydate']/@title)"/>
</pubDate>
<description>
- <xsl:value-of select="concat($user-name, ': ')"/>
<xsl:text disable-output-escaping="yes"><![CDATA[</xsl:text>
+ <xsl:if test="$show-usernames = 1">
+ <xsl:value-of select="concat($user-name, ': ')"/>
+ </xsl:if>
<xsl:copy-of select="$item-content/node()"/>
<xsl:text disable-output-escaping="yes">]]></xsl:text>
</description>
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:php="http://php.net/xsl"
- xsl:extension-element-prefixes="php"
exclude-result-prefixes="php">
<xsl:param name="generate-enclosure"/>
+ <xsl:param name="show-usernames"/>
+ <xsl:param name="show-multimedia"/>
<xsl:output method="xml" indent="yes"/>
<!-- Identity transform -->
<xsl:template match="@*|node()">
<xsl:copy>
+ <!--
+ Strip the style attribute while copying elements because it may be
+ dangerous, see:
+ https://validator.w3.org/feed/docs/warning/DangerousStyleAttr.html
+ -->
<xsl:apply-templates select="@*[not(name() = 'style')]|node()"/>
</xsl:copy>
</xsl:template>
too instead of the t.co redirections.
-->
<xsl:template match="a[@data-pre-embedded='true']">
- <!-- Prepend and append a white space for aestethic reasons -->
- <xsl:text> </xsl:text>
- <a>
- <xsl:attribute name="href">
+ <xsl:if test="$show-multimedia = 1">
+ <!-- Prepend and append a white space for aestethic reasons -->
+ <xsl:text> </xsl:text>
+ <a>
+ <xsl:attribute name="href">
+ <xsl:value-of select="concat('https://', .)"/>
+ </xsl:attribute>
<xsl:value-of select="concat('https://', .)"/>
- </xsl:attribute>
- <xsl:value-of select="concat('https://', .)"/>
- </a>
- <xsl:text> </xsl:text>
+ </a>
+ <xsl:text> </xsl:text>
+ </xsl:if>
</xsl:template>
<!-- Present images in a more convenient way -->
<xsl:attribute name="href">
<xsl:value-of select="concat(@data-image-url, ':orig')"/>
</xsl:attribute>
- <img>
+ <img style="max-width: 100%">
<xsl:attribute name="src">
<xsl:value-of select="@data-image-url"/>
</xsl:attribute>
<xsl:copy>
<xsl:apply-templates select="@*"/>
<xsl:attribute name="style">
- <xsl:value-of select="concat(@style, '; background-repeat: no-repeat')"/>
+ <xsl:value-of select="concat(@style, '; background-repeat: no-repeat; background-size: 100% auto')"/>
</xsl:attribute>
<xsl:apply-templates select="node()"/>
</xsl:copy>
<xsl:variable name="item-permalink" select="concat($BaseURL, .//div[@data-permalink-path]/@data-permalink-path)"/>
<xsl:variable name="item-has-video" select="$item-media//*[contains(@class, 'PlayableMedia--video')]"/>
+ <xsl:variable name="item-has-gif" select="$item-media//*[contains(@class, 'PlayableMedia--gif')]"/>
<item>
<title>
- <xsl:value-of select="concat($user-name, ': ')"/>
+ <xsl:if test="($show-usernames = 1) or ($screen-name != $user-name)">
+ <xsl:value-of select="concat($user-name, ': ')"/>
+ </xsl:if>
<xsl:if test="$item-has-video">
<xsl:text>(Video) </xsl:text>
</xsl:if>
</pubDate>
<description>
<xsl:text disable-output-escaping="yes"><![CDATA[</xsl:text>
- <xsl:value-of select="concat($user-name, ':')"/>
- <xsl:element name="br"/>
+ <xsl:if test="($show-usernames = 1) or ($screen-name != $user-name)">
+ <xsl:value-of select="concat($user-name, ':')"/>
+ <xsl:element name="br"/>
+ </xsl:if>
<xsl:if test="$item-has-video">
<xsl:text> (Video)</xsl:text>
<xsl:element name="br"/>
</xsl:if>
+ <xsl:if test="$item-has-gif">
+ <xsl:text> (GIF)</xsl:text>
+ <xsl:element name="br"/>
+ </xsl:if>
<xsl:element name="span">
<xsl:attribute name="style">white-space: pre-wrap;</xsl:attribute>
<xsl:apply-templates select="$item-content/node()"/>
</xsl:element>
- <xsl:apply-templates select="$item-media/node()"/>
+ <xsl:if test="$show-multimedia = 1">
+ <xsl:apply-templates select="$item-media/node()"/>
+ </xsl:if>
<xsl:text disable-output-escaping="yes">]]></xsl:text>
</description>
<xsl:if test="$generate-enclosure = 1">
*-e*::
show links to supported media files in the RSS <enclosure/> element
+*-m <0|1>*::
+ enable or disable showing multimedia content (e.g. Twitter or Instagram
+ pictures) directly inside the item description. Default is 1 (enable).
+
+*-u <0|1>*::
+ enable or disable showing usernames in front of the item for hosts which
+ supports it (Twitter.com/Instagram.com). Default is 1 (enable).
+
*-h, --help*::
show the help message
<?php
+
/**
* @file
* Tweeper - a Twitter to RSS web scraper.
*/
function usage($argv) {
if (is_cli()) {
- $usage = "{$argv[0]} [-e|-h|--help] <src_url>\n";
+ $usage = "{$argv[0]} [-e|-m <0|1>|-u <0|1>|-h|--help] <src_url>\n";
}
else {
- $usage = htmlentities("{$_SERVER['SCRIPT_NAME']}?src_url=<src_url>&generate_enclosure=<0|1>");
+ $usage = htmlentities("{$_SERVER['SCRIPT_NAME']}?src_url=<src_url>&generate_enclosure=<0|1>&show_usernames=<0|1>&show_multimedia=<0|1>");
}
return "usage: $usage";
function parse_options_cli($argv, $argc) {
$options = array(
'generate_enclosure' => FALSE,
+ 'show_usernames' => TRUE,
+ 'show_multimedia' => TRUE,
);
if ($argc < 2) {
return $options;
}
- $cli_options = getopt("eh", array("help"));
+ $cli_options = getopt("em:u:h", array("help"));
foreach ($cli_options as $opt => $val) {
switch ($opt) {
case 'e':
$options['generate_enclosure'] = TRUE;
break;
+ case 'm':
+ $ret = filter_var($val, FILTER_VALIDATE_BOOLEAN, FILTER_NULL_ON_FAILURE);
+ if (NULL === $ret) {
+ fwrite(STDERR, "Invalid argument for the -m option.\n");
+ fwrite(STDERR, usage($argv));
+ exit(1);
+ }
+ $options['show_multimedia'] = $val;
+ break;
+
+ case 'u':
+ $ret = filter_var($val, FILTER_VALIDATE_BOOLEAN, FILTER_NULL_ON_FAILURE);
+ if (NULL === $ret) {
+ fwrite(STDERR, "Invalid argument for the -u option.\n");
+ fwrite(STDERR, usage($argv));
+ exit(1);
+ }
+ $options['show_usernames'] = $val;
+ break;
+
case 'h':
case 'help':
echo usage($argv);
}
}
- $options['src_url'] = $argv[count($cli_options) + 1];
+ // For now assume that the URL is the lest argument, in the future we could
+ // switch to PHP >= 7.1 and use the $optind argument of getopt().
+ $options['src_url'] = array_pop($argv);
return $options;
}
function parse_options_query_string() {
$options = array(
'generate_enclosure' => FALSE,
+ 'show_usernames' => TRUE,
+ 'show_multimedia' => TRUE,
);
if (isset($_GET['src_url'])) {
$options['generate_enclosure'] = $_GET['generate_enclosure'] == 1;
}
+ if (isset($_GET['show_multimedia'])) {
+ $options['show_multimedia'] = $_GET['show_multimedia'] != 0;
+ }
+
+ if (isset($_GET['show_usernames'])) {
+ $options['show_usernames'] = $_GET['show_usernames'] != 0;
+ }
+
return $options;
}
exit(1);
}
-$tweeper = new Tweeper($options['generate_enclosure']);
+$tweeper = new Tweeper($options['generate_enclosure'], $options['show_usernames'], $options['show_multimedia']);
$output = $tweeper->tweep($options['src_url']);
if (is_null($output)) {
exit(1);
projects / tweeper.git / commitdiff