Reading HtmlFeed user agent from config file

2 files changed, 15 insertions, 3 deletions

diff --git a/bin/get-user-agent.sh b/bin/get-user-agent.sh
new file mode 100755
index 0000000..4d348d9
--- /dev/null
+++ b/bin/get-user-agent.sh
@@ -0,0 +1,10 @@
+#!/bin/bash
+ACCESSLOG="$1"
+PATTERN="$2"
+OUTPUT="$3"
+LINE=`grep "$PATTERN" "$ACCESSLOG" | grep Chrome | tail -n 1`
+if [ -n "$LINE" ]
+then
+    COLS=`echo "$LINE" | csvtool -t ' ' width -`
+    echo "$LINE" | csvtool -t ' ' col $COLS - | sed 's/"//g' > $OUTPUT
+fi
diff --git a/providers/HtmlFeed.php b/providers/HtmlFeed.php
index 411cddf..64d6259 100644
--- a/providers/HtmlFeed.php
+++ b/providers/HtmlFeed.php
@@ -7,10 +7,12 @@ require_once('../lib/querypath/src/qp.php');
 
 abstract class HtmlFeed extends \Providers\XmlFeed {
 
-    const HTML_FEED_USER_AGENT = 'Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.0.11) Gecko/2009060215 Firefox/3.0.11 (.NET CLR 3.5.30729)';
-
     private $_encoding;
 
+    private function __getUserAgent() {
+        return trim(file_get_contents('../config/user-agent'));
+    }
+
     private function __getHttpContent($feedUrl) {
         $header = array();
         $header[] = 'Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5';
@@ -20,7 +22,7 @@ abstract class HtmlFeed extends \Providers\XmlFeed {
         $header[] = 'Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7';
         $header[] = 'Pragma: ';
         $ch = curl_init($feedUrl);
-        curl_setopt($ch, CURLOPT_USERAGENT, self::HTML_FEED_USER_AGENT);
+        curl_setopt($ch, CURLOPT_USERAGENT, $this->__getUserAgent());
         curl_setopt($ch, CURLOPT_HTTPHEADER, $header);
         curl_setopt($ch, CURLOPT_AUTOREFERER, TRUE);
         curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);