Merge pull request #326 from ArthurHoaro/bug-url

Fixes #325 - Shaarli does not recognize saved links
author: VirtualTam <virtualtam@flibidi.net> 2015-08-31 20:31:41 +0200
committer: VirtualTam <virtualtam@flibidi.net> 2015-08-31 20:31:41 +0200
commit: 6211c498f6e0bdc6d86152e9777bcc75955a5ec4 (patch)
tree: f7b48317044c151609b9d8e7d476cc5014916245
parent: ce8e248ab04a035c2824bee6af91aed49d623a6a (diff)
parent: 26c503460cd2b50c2c122dba73d62e09ee04b9c8 (diff)
download: Shaarli-6211c498f6e0bdc6d86152e9777bcc75955a5ec4.tar.gz
Shaarli-6211c498f6e0bdc6d86152e9777bcc75955a5ec4.tar.zst
Shaarli-6211c498f6e0bdc6d86152e9777bcc75955a5ec4.zip
3 files changed, 76 insertions, 38 deletions
diff --git a/application/Url.php b/application/Url.php
index 23356f39..02a4395d 100644..100755
--- a/application/Url.php
+++ b/application/Url.php
@@ -81,6 +81,10 @@ class Url
    public function __construct($url)
    {
        $this->parts = parse_url($url);
+        if (!empty($url) && empty($this->parts['scheme'])) {
+            $this->parts['scheme'] = 'http';
+        }
    }
    /**
@@ -147,4 +151,16 @@ class Url
        $this->cleanupFragment();
        return $this->__toString();
    }
+    /**
+     * Get URL scheme.
+     *
+     * @return string the URL scheme or false if none is provided.
+     */
+    public function getScheme() {
+        if (!isset($this->parts['scheme'])) {
+            return false;
+        }
+        return $this->parts['scheme'];
+    }
 }
diff --git a/index.php b/index.php
index a093a283..f4c7e781 100755
--- a/index.php
+++ b/index.php
@@ -1500,51 +1500,57 @@ function renderPage()
        $url->cleanup();
        $link_is_new = false;
-        $link = $LINKSDB->getLinkFromUrl($url); // Check if URL is not already in database (in this case, we will edit the existing link)
+        // Check if URL is not already in database (in this case, we will edit the existing link)
+        $link = $LINKSDB->getLinkFromUrl((string)$url);
        if (!$link)
        {
-            $link_is_new = true;  // This is a new link
+            $link_is_new = true;
            $linkdate = strval(date('Ymd_His'));
-            $title = (empty($_GET['title']) ? '' : $_GET['title'] ); // Get title if it was provided in URL (by the bookmarklet).
+            // Get title if it was provided in URL (by the bookmarklet).
-            $description = (empty($_GET['description']) ? '' : $_GET['description']); // Get description if it was provided in URL (by the bookmarklet). [Bronco added that]
+            $title = (empty($_GET['title']) ? '' : $_GET['title'] );
-            $tags = (empty($_GET['tags']) ? '' : $_GET['tags'] ); // Get tags if it was provided in URL
+            // Get description if it was provided in URL (by the bookmarklet). [Bronco added that]
-            $private = (!empty($_GET['private']) && $_GET['private'] === "1" ? 1 : 0); // Get private if it was provided in URL
+            $description = (empty($_GET['description']) ? '' : $_GET['description']);
-            if (($url!='') && parse_url($url,PHP_URL_SCHEME)=='') $url = 'http://'.$url;
+            $tags = (empty($_GET['tags']) ? '' : $_GET['tags'] );
-            // If this is an HTTP link, we try go get the page to extract the title (otherwise we will to straight to the edit form.)
+            $private = (!empty($_GET['private']) && $_GET['private'] === "1" ? 1 : 0);
-            if (empty($title) && parse_url($url,PHP_URL_SCHEME)=='http')
+            // If this is an HTTP(S) link, we try go get the page to extract the title (otherwise we will to straight to the edit form.)
-            {
+            if (empty($title) && strpos($url->getScheme(), 'http') !== false) {
                list($status,$headers,$data) = getHTTP($url,4); // Short timeout to keep the application responsive.
                // FIXME: Decode charset according to specified in either 1) HTTP response headers or 2) <head> in html
-                if (strpos($status,'200 OK')!==false)
+                if (strpos($status,'200 OK')!==false) {
-                                         {
+                    // Look for charset in html header.
-                        // Look for charset in html header.
+                    preg_match('#<meta .*charset=.*>#Usi', $data, $meta);
-                                                preg_match('#<meta .*charset=.*>#Usi', $data, $meta);
+                    // If found, extract encoding.
-                                                // If found, extract encoding.
+                    if (!empty($meta[0])) {
-                                                if (!empty($meta[0]))
+                        // Get encoding specified in header.
-                                                {
+                        preg_match('#charset="?(.*)"#si', $meta[0], $enc);
-                                                        // Get encoding specified in header.
+                        // If charset not found, use utf-8.
-                                                        preg_match('#charset="?(.*)"#si', $meta[0], $enc);
+                        $html_charset = (!empty($enc[1])) ? strtolower($enc[1]) : 'utf-8';
-                                                        // If charset not found, use utf-8.
+                    }
-                                                        $html_charset = (!empty($enc[1])) ? strtolower($enc[1]) : 'utf-8';
+                    else {
-                                                }
+                        $html_charset = 'utf-8';
-                                                else { $html_charset = 'utf-8'; }
+                    }
-                                                // Extract title
+                    // Extract title
-                                                $title = html_extract_title($data);
+                    $title = html_extract_title($data);
-                                                if (!empty($title))
+                    if (!empty($title)) {
-                                                {
+                        // Re-encode title in utf-8 if necessary.
-                                                        // Re-encode title in utf-8 if necessary.
+                        $title = ($html_charset == 'iso-8859-1') ? utf8_encode($title) : $title;
-                                                        $title = ($html_charset == 'iso-8859-1') ? utf8_encode($title) : $title;
+                    }
-                                                }
+                }
-                                        }
            }
-            if ($url=='') // In case of empty URL, this is just a text (with a link that points to itself)
+            if ($url == '') {
-            {
+                $url = '?' . smallHash($linkdate);
-                $url='?'.smallHash($linkdate);
+                $title = 'Note: ';
-                $title='Note: ';
            }
-            $link = array('linkdate'=>$linkdate,'title'=>$title,'url'=>$url,'description'=>$description,'tags'=>$tags,'private'=>$private);
+            $link = array(
+                'linkdate' => $linkdate,
+                'title' => $title,
+                'url' => (string)$url,
+                'description' => $description,
+                'tags' => $tags,
+                'private' => $private
+            );
        }
        $PAGE = new pageBuilder;
diff --git a/tests/UrlTest.php b/tests/UrlTest.php
index a39630f1..c848e88e 100644..100755
--- a/tests/UrlTest.php
+++ b/tests/UrlTest.php
@@ -151,4 +151,20 @@ class UrlTest extends PHPUnit_Framework_TestCase
            $url->cleanup()
        );
    }
+    /**
+     * Test default http scheme.
+     */
+    public function testDefaultScheme() {
+        $url = new Url(self::$baseUrl);
+        $this->assertEquals('http', $url->getScheme());
+        $url = new Url('domain.tld');
+        $this->assertEquals('http', $url->getScheme());
+        $url = new Url('ssh://domain.tld');
+        $this->assertEquals('ssh', $url->getScheme());
+        $url = new Url('ftp://domain.tld');
+        $this->assertEquals('ftp', $url->getScheme());
+        $url = new Url('git://domain.tld/push?pull=clone#checkout');
+        $this->assertEquals('git', $url->getScheme());
+    }
 }
author	VirtualTam <virtualtam@flibidi.net>	2015-08-31 20:31:41 +0200
committer	VirtualTam <virtualtam@flibidi.net>	2015-08-31 20:31:41 +0200
commit	6211c498f6e0bdc6d86152e9777bcc75955a5ec4 (patch)
tree	f7b48317044c151609b9d8e7d476cc5014916245
parent	ce8e248ab04a035c2824bee6af91aed49d623a6a (diff)
parent	26c503460cd2b50c2c122dba73d62e09ee04b9c8 (diff)
download	Shaarli-6211c498f6e0bdc6d86152e9777bcc75955a5ec4.tar.gz Shaarli-6211c498f6e0bdc6d86152e9777bcc75955a5ec4.tar.zst Shaarli-6211c498f6e0bdc6d86152e9777bcc75955a5ec4.zip

diff --git a/application/Url.php b/application/Url.php index 23356f39..02a4395d 100644..100755 --- a/application/Url.php +++ b/application/Url.php
@@ -81,6 +81,10 @@ class Url
81	public function __construct($url)	81	public function __construct($url)
82	{	82	{
83	$this->parts = parse_url($url);	83	$this->parts = parse_url($url);
		84
		85	if (!empty($url) && empty($this->parts['scheme'])) {
		86	$this->parts['scheme'] = 'http';
		87	}
84	}	88	}
85		89
86	/**	90	/**
@@ -147,4 +151,16 @@ class Url
147	$this->cleanupFragment();	151	$this->cleanupFragment();
148	return $this->__toString();	152	return $this->__toString();
149	}	153	}
		154
		155	/**
		156	* Get URL scheme.
		157	*
		158	* @return string the URL scheme or false if none is provided.
		159	*/
		160	public function getScheme() {
		161	if (!isset($this->parts['scheme'])) {
		162	return false;
		163	}
		164	return $this->parts['scheme'];
		165	}
150	}	166	}


diff --git a/index.php b/index.php index a093a283..f4c7e781 100755 --- a/index.php +++ b/index.php
@@ -1500,51 +1500,57 @@ function renderPage()
1500	$url->cleanup();	1500	$url->cleanup();
1501		1501
1502	$link_is_new = false;	1502	$link_is_new = false;
1503	$link = $LINKSDB->getLinkFromUrl($url); // Check if URL is not already in database (in this case, we will edit the existing link)	1503	// Check if URL is not already in database (in this case, we will edit the existing link)
		1504	$link = $LINKSDB->getLinkFromUrl((string)$url);
1504	if (!$link)	1505	if (!$link)
1505	{	1506	{
1506	$link_is_new = true; // This is a new link	1507	$link_is_new = true;
1507	$linkdate = strval(date('Ymd_His'));	1508	$linkdate = strval(date('Ymd_His'));
1508	$title = (empty($_GET['title']) ? '' : $_GET['title'] ); // Get title if it was provided in URL (by the bookmarklet).	1509	// Get title if it was provided in URL (by the bookmarklet).
1509	$description = (empty($_GET['description']) ? '' : $_GET['description']); // Get description if it was provided in URL (by the bookmarklet). [Bronco added that]	1510	$title = (empty($_GET['title']) ? '' : $_GET['title'] );
1510	$tags = (empty($_GET['tags']) ? '' : $_GET['tags'] ); // Get tags if it was provided in URL	1511	// Get description if it was provided in URL (by the bookmarklet). [Bronco added that]
1511	$private = (!empty($_GET['private']) && $_GET['private'] === "1" ? 1 : 0); // Get private if it was provided in URL	1512	$description = (empty($_GET['description']) ? '' : $_GET['description']);
1512	if (($url!='') && parse_url($url,PHP_URL_SCHEME)=='') $url = 'http://'.$url;	1513	$tags = (empty($_GET['tags']) ? '' : $_GET['tags'] );
1513	// If this is an HTTP link, we try go get the page to extract the title (otherwise we will to straight to the edit form.)	1514	$private = (!empty($_GET['private']) && $_GET['private'] === "1" ? 1 : 0);
1514	if (empty($title) && parse_url($url,PHP_URL_SCHEME)=='http')	1515	// If this is an HTTP(S) link, we try go get the page to extract the title (otherwise we will to straight to the edit form.)
1515	{	1516	if (empty($title) && strpos($url->getScheme(), 'http') !== false) {
1516	list($status,$headers,$data) = getHTTP($url,4); // Short timeout to keep the application responsive.	1517	list($status,$headers,$data) = getHTTP($url,4); // Short timeout to keep the application responsive.
1517	// FIXME: Decode charset according to specified in either 1) HTTP response headers or 2) <head> in html	1518	// FIXME: Decode charset according to specified in either 1) HTTP response headers or 2) <head> in html
1518	if (strpos($status,'200 OK')!==false)	1519	if (strpos($status,'200 OK')!==false) {
1519	{	1520	// Look for charset in html header.
1520	// Look for charset in html header.	1521	preg_match('#<meta .charset=.>#Usi', $data, $meta);
1521	preg_match('#<meta .charset=.>#Usi', $data, $meta);	1522
1522		1523	// If found, extract encoding.
1523	// If found, extract encoding.	1524	if (!empty($meta[0])) {
1524	if (!empty($meta[0]))	1525	// Get encoding specified in header.
1525	{	1526	preg_match('#charset="?(.*)"#si', $meta[0], $enc);
1526	// Get encoding specified in header.	1527	// If charset not found, use utf-8.
1527	preg_match('#charset="?(.*)"#si', $meta[0], $enc);	1528	$html_charset = (!empty($enc[1])) ? strtolower($enc[1]) : 'utf-8';
1528	// If charset not found, use utf-8.	1529	}
1529	$html_charset = (!empty($enc[1])) ? strtolower($enc[1]) : 'utf-8';	1530	else {
1530	}	1531	$html_charset = 'utf-8';
1531	else { $html_charset = 'utf-8'; }	1532	}
1532		1533
1533	// Extract title	1534	// Extract title
1534	$title = html_extract_title($data);	1535	$title = html_extract_title($data);
1535	if (!empty($title))	1536	if (!empty($title)) {
1536	{	1537	// Re-encode title in utf-8 if necessary.
1537	// Re-encode title in utf-8 if necessary.	1538	$title = ($html_charset == 'iso-8859-1') ? utf8_encode($title) : $title;
1538	$title = ($html_charset == 'iso-8859-1') ? utf8_encode($title) : $title;	1539	}
1539	}	1540	}
1540	}
1541	}	1541	}
1542	if ($url=='') // In case of empty URL, this is just a text (with a link that points to itself)	1542	if ($url == '') {
1543	{	1543	$url = '?' . smallHash($linkdate);
1544	$url='?'.smallHash($linkdate);	1544	$title = 'Note: ';
1545	$title='Note: ';
1546	}	1545	}
1547	$link = array('linkdate'=>$linkdate,'title'=>$title,'url'=>$url,'description'=>$description,'tags'=>$tags,'private'=>$private);	1546	$link = array(
		1547	'linkdate' => $linkdate,
		1548	'title' => $title,
		1549	'url' => (string)$url,
		1550	'description' => $description,
		1551	'tags' => $tags,
		1552	'private' => $private
		1553	);
1548	}	1554	}
1549		1555
1550	$PAGE = new pageBuilder;	1556	$PAGE = new pageBuilder;


diff --git a/tests/UrlTest.php b/tests/UrlTest.php index a39630f1..c848e88e 100644..100755 --- a/tests/UrlTest.php +++ b/tests/UrlTest.php
@@ -151,4 +151,20 @@ class UrlTest extends PHPUnit_Framework_TestCase
151	$url->cleanup()	151	$url->cleanup()
152	);	152	);
153	}	153	}
		154
		155	/**
		156	* Test default http scheme.
		157	*/
		158	public function testDefaultScheme() {
		159	$url = new Url(self::$baseUrl);
		160	$this->assertEquals('http', $url->getScheme());
		161	$url = new Url('domain.tld');
		162	$this->assertEquals('http', $url->getScheme());
		163	$url = new Url('ssh://domain.tld');
		164	$this->assertEquals('ssh', $url->getScheme());
		165	$url = new Url('ftp://domain.tld');
		166	$this->assertEquals('ftp', $url->getScheme());
		167	$url = new Url('git://domain.tld/push?pull=clone#checkout');
		168	$this->assertEquals('git', $url->getScheme());
		169	}
154	}	170	}