LinkDB: move to a proper file, add test coverage

Relates to #71

LinkDB
 - move to application/LinkDB.php
 - code cleanup
   - indentation
   - whitespaces
   - formatting
 - comment cleanup
   - add missing documentation
   - unify formatting

Test coverage for LinkDB
 - constructor
 - public / private access
 - link-related methods

Shaarli utilities (LinkDB dependencies)
 - move startsWith() and endsWith() functions to application/Utils.php
 - add test coverage

Dev utilities
 - Composer: add PHPUnit to dev dependencies
 - Makefile:
    - update lint targets
    - add test targets
    - generate coverage reports

Signed-off-by: VirtualTam <virtualtam@flibidi.net>

1 files changed, 10 insertions, 249 deletions

diff --git a/index.php b/index.php
index 9561f63b..ed18c7f9 100644
--- a/index.php
+++ b/index.php
@@ -68,6 +68,10 @@ checkphpversion();
 error_reporting(E_ALL^E_WARNING);  // See all error except warnings.
 //error_reporting(-1); // See all errors (for debugging only)
 
+// Shaarli library
+require_once 'application/LinkDB.php';
+require_once 'application/Utils.php';
+
 include "inc/rain.tpl.class.php"; //include Rain TPL
 raintpl::$tpl_dir = $GLOBALS['config']['RAINTPL_TPL']; // template directory
 raintpl::$cache_dir = $GLOBALS['config']['RAINTPL_TMP']; // cache directory
@@ -268,21 +272,6 @@ function nl2br_escaped($html)
     return str_replace('>','&gt;',str_replace('<','&lt;',nl2br($html)));
 }
 
-/* Returns the small hash of a string, using RFC 4648 base64url format
-   e.g. smallHash('20111006_131924') --> yZH23w
-   Small hashes:
-     - are unique (well, as unique as crc32, at last)
-     - are always 6 characters long.
-     - only use the following characters: a-z A-Z 0-9 - _ @
-     - are NOT cryptographically secure (they CAN be forged)
-   In Shaarli, they are used as a tinyurl-like link to individual entries.
-*/
-function smallHash($text)
-{
-    $t = rtrim(base64_encode(hash('crc32',$text,true)),'=');
-    return strtr($t, '+/', '-_');
-}
-
 // In a string, converts URLs to clickable links.
 // Function inspired from http://www.php.net/manual/en/function.preg-replace.php#85722
 function text2clickable($url)
@@ -536,20 +525,6 @@ function getMaxFileSize()
     return $maxsize;
 }
 
-// Tells if a string start with a substring or not.
-function startsWith($haystack,$needle,$case=true)
-{
-    if($case){return (strcmp(substr($haystack, 0, strlen($needle)),$needle)===0);}
-    return (strcasecmp(substr($haystack, 0, strlen($needle)),$needle)===0);
-}
-
-// Tells if a string ends with a substring or not.
-function endsWith($haystack,$needle,$case=true)
-{
-    if($case){return (strcmp(substr($haystack, strlen($haystack) - strlen($needle)),$needle)===0);}
-    return (strcasecmp(substr($haystack, strlen($haystack) - strlen($needle)),$needle)===0);
-}
-
 /*  Converts a linkdate time (YYYYMMDD_HHMMSS) of an article to a timestamp (Unix epoch)
     (used to build the ADD_DATE attribute in Netscape-bookmarks file)
     PS: I could have used strptime(), but it does not exist on Windows. I'm too kind. */
@@ -711,220 +686,6 @@ class pageBuilder
 }
 
 // ------------------------------------------------------------------------------------------
-/* Data storage for links.
-   This object behaves like an associative array.
-   Example:
-      $mylinks = new linkdb();
-      echo $mylinks['20110826_161819']['title'];
-      foreach($mylinks as $link)
-         echo $link['title'].' at url '.$link['url'].' ; description:'.$link['description'];
-
-   Available keys:
-       title : Title of the link
-       url : URL of the link. Can be absolute or relative. Relative URLs are permalinks (e.g.'?m-ukcw')
-       description : description of the entry
-       private : Is this link private? 0=no, other value=yes
-       linkdate : date of the creation of this entry, in the form YYYYMMDD_HHMMSS (e.g.'20110914_192317')
-       tags : tags attached to this entry (separated by spaces)
-
-   We implement 3 interfaces:
-     - ArrayAccess so that this object behaves like an associative array.
-     - Iterator so that this object can be used in foreach() loops.
-     - Countable interface so that we can do a count() on this object.
-*/
-class linkdb implements Iterator, Countable, ArrayAccess
-{
-    private $links; // List of links (associative array. Key=linkdate (e.g. "20110823_124546"), value= associative array (keys:title,description...)
-    private $urls;  // List of all recorded URLs (key=url, value=linkdate) for fast reserve search (url-->linkdate)
-    private $keys;  // List of linkdate keys (for the Iterator interface implementation)
-    private $position; // Position in the $this->keys array. (for the Iterator interface implementation.)
-    private $loggedin; // Is the user logged in? (used to filter private links)
-
-    // Constructor:
-    function __construct($isLoggedIn)
-    // Input : $isLoggedIn : is the user logged in?
-    {
-        $this->loggedin = $isLoggedIn;
-        $this->checkdb(); // Make sure data file exists.
-        $this->readdb();  // Then read it.
-    }
-
-    // ---- Countable interface implementation
-    public function count() { return count($this->links); }
-
-    // ---- ArrayAccess interface implementation
-    public function offsetSet($offset, $value)
-    {
-        if (!$this->loggedin) die('You are not authorized to add a link.');
-        if (empty($value['linkdate']) || empty($value['url'])) die('Internal Error: A link should always have a linkdate and URL.');
-        if (empty($offset)) die('You must specify a key.');
-        $this->links[$offset] = $value;
-        $this->urls[$value['url']]=$offset;
-    }
-    public function offsetExists($offset) { return array_key_exists($offset,$this->links); }
-    public function offsetUnset($offset)
-    {
-        if (!$this->loggedin) die('You are not authorized to delete a link.');
-        $url = $this->links[$offset]['url']; unset($this->urls[$url]);
-        unset($this->links[$offset]);
-    }
-    public function offsetGet($offset) { return isset($this->links[$offset]) ? $this->links[$offset] : null; }
-
-    // ---- Iterator interface implementation
-    function rewind() { $this->keys=array_keys($this->links); rsort($this->keys); $this->position=0; } // Start over for iteration, ordered by date (latest first).
-    function key() { return $this->keys[$this->position]; } // current key
-    function current() { return $this->links[$this->keys[$this->position]]; } // current value
-    function next() { ++$this->position; } // go to next item
-    function valid() { return isset($this->keys[$this->position]); }    // Check if current position is valid.
-
-    // ---- Misc methods
-    private function checkdb() // Check if db directory and file exists.
-    {
-        if (!file_exists($GLOBALS['config']['DATASTORE'])) // Create a dummy database for example.
-        {
-             $this->links = array();
-             $link = array('title'=>'Shaarli - sebsauvage.net','url'=>'http://sebsauvage.net/wiki/doku.php?id=php:shaarli','description'=>'Welcome to Shaarli ! This is a bookmark. To edit or delete me, you must first login.','private'=>0,'linkdate'=>'20110914_190000','tags'=>'opensource software');
-             $this->links[$link['linkdate']] = $link;
-             $link = array('title'=>'My secret stuff... - Pastebin.com','url'=>'http://sebsauvage.net/paste/?8434b27936c09649#bR7XsXhoTiLcqCpQbmOpBi3rq2zzQUC5hBI7ZT1O3x8=','description'=>'SShhhh!!  I\'m a private link only YOU can see. You can delete me too.','private'=>1,'linkdate'=>'20110914_074522','tags'=>'secretstuff');
-             $this->links[$link['linkdate']] = $link;
-             file_put_contents($GLOBALS['config']['DATASTORE'], PHPPREFIX.base64_encode(gzdeflate(serialize($this->links))).PHPSUFFIX); // Write database to disk
-        }
-    }
-
-    // Read database from disk to memory
-    private function readdb()
-    {
-        // Read data
-        $this->links=(file_exists($GLOBALS['config']['DATASTORE']) ? unserialize(gzinflate(base64_decode(substr(file_get_contents($GLOBALS['config']['DATASTORE']),strlen(PHPPREFIX),-strlen(PHPSUFFIX))))) : array() );
-        // Note that gzinflate is faster than gzuncompress. See: http://www.php.net/manual/en/function.gzdeflate.php#96439
-
-        // If user is not logged in, filter private links.
-        if (!$this->loggedin)
-        {
-            $toremove=array();
-            foreach($this->links as $link) { if ($link['private']!=0) $toremove[]=$link['linkdate']; }
-            foreach($toremove as $linkdate) { unset($this->links[$linkdate]); }
-        }
-
-        // Keep the list of the mapping URLs-->linkdate up-to-date.
-        $this->urls=array();
-        foreach($this->links as $link) { $this->urls[$link['url']]=$link['linkdate']; }
-    }
-
-    // Save database from memory to disk.
-    public function savedb()
-    {
-        if (!$this->loggedin) die('You are not authorized to change the database.');
-        file_put_contents($GLOBALS['config']['DATASTORE'], PHPPREFIX.base64_encode(gzdeflate(serialize($this->links))).PHPSUFFIX);
-        invalidateCaches();
-    }
-
-    // Returns the link for a given URL (if it exists). False if it does not exist.
-    public function getLinkFromUrl($url)
-    {
-        if (isset($this->urls[$url])) return $this->links[$this->urls[$url]];
-        return false;
-    }
-
-    // Case insensitive search among links (in the URLs, title and description). Returns filtered list of links.
-    // e.g. print_r($mydb->filterFulltext('hollandais'));
-    public function filterFulltext($searchterms)
-    {
-        // FIXME: explode(' ',$searchterms) and perform a AND search.
-        // FIXME: accept double-quotes to search for a string "as is"?
-        // Using mb_convert_case($val, MB_CASE_LOWER, 'UTF-8') allows us to perform searches on
-        // Unicode text. See https://github.com/shaarli/Shaarli/issues/75 for examples.
-        $filtered=array();
-        $s = mb_convert_case($searchterms, MB_CASE_LOWER, 'UTF-8');
-        foreach($this->links as $l)
-        {
-            $found=   (strpos(mb_convert_case($l['title'],       MB_CASE_LOWER, 'UTF-8'),$s) !== false)
-                   || (strpos(mb_convert_case($l['description'], MB_CASE_LOWER, 'UTF-8'),$s) !== false)
-                   || (strpos(mb_convert_case($l['url'],         MB_CASE_LOWER, 'UTF-8'),$s) !== false)
-                   || (strpos(mb_convert_case($l['tags'],        MB_CASE_LOWER, 'UTF-8'),$s) !== false);
-            if ($found) $filtered[$l['linkdate']] = $l;
-        }
-        krsort($filtered);
-        return $filtered;
-    }
-
-    // Filter by tag.
-    // You can specify one or more tags (tags can be separated by space or comma).
-    // e.g. print_r($mydb->filterTags('linux programming'));
-    public function filterTags($tags,$casesensitive=false)
-    {
-        // Same as above, we use UTF-8 conversion to handle various graphemes (i.e. cyrillic, or greek)
-        // TODO: is $casesensitive ever true ?
-        $t = str_replace(',',' ',($casesensitive?$tags:mb_convert_case($tags, MB_CASE_LOWER, 'UTF-8')));
-        $searchtags=explode(' ',$t);
-        $filtered=array();
-        foreach($this->links as $l)
-        {
-            $linktags = explode(' ',($casesensitive?$l['tags']:mb_convert_case($l['tags'], MB_CASE_LOWER, 'UTF-8')));
-            if (count(array_intersect($linktags,$searchtags)) == count($searchtags))
-                $filtered[$l['linkdate']] = $l;
-        }
-        krsort($filtered);
-        return $filtered;
-    }
-
-    // Filter by day. Day must be in the form 'YYYYMMDD' (e.g. '20120125')
-    // Sort order is: older articles first.
-    // e.g. print_r($mydb->filterDay('20120125'));
-    public function filterDay($day)
-    {
-        $filtered=array();
-        foreach($this->links as $l)
-        {
-            if (startsWith($l['linkdate'],$day)) $filtered[$l['linkdate']] = $l;
-        }
-        ksort($filtered);
-        return $filtered;
-    }
-    // Filter by smallHash.
-    // Only 1 article is returned.
-    public function filterSmallHash($smallHash)
-    {
-        $filtered=array();
-        foreach($this->links as $l)
-        {
-            if ($smallHash==smallHash($l['linkdate'])) // Yes, this is ugly and slow
-            {
-                $filtered[$l['linkdate']] = $l;
-                return $filtered;
-            }
-        }
-        return $filtered;
-    }
-
-    // Returns the list of all tags
-    // Output: associative array key=tags, value=0
-    public function allTags()
-    {
-        $tags=array();
-        foreach($this->links as $link)
-            foreach(explode(' ',$link['tags']) as $tag)
-                if (!empty($tag)) $tags[$tag]=(empty($tags[$tag]) ? 1 : $tags[$tag]+1);
-        arsort($tags); // Sort tags by usage (most used tag first)
-        return $tags;
-    }
-
-    // Returns the list of days containing articles (oldest first)
-    // Output: An array containing days (in format YYYYMMDD).
-    public function days()
-    {
-        $linkdays=array();
-        foreach(array_keys($this->links) as $day)
-        {
-            $linkdays[substr($day,0,8)]=0;
-        }
-        $linkdays=array_keys($linkdays);
-        sort($linkdays);
-        return $linkdays;
-    }
-}
-
-// ------------------------------------------------------------------------------------------
 // Output the last N links in RSS 2.0 format.
 function showRSS()
 {
@@ -941,7 +702,7 @@ function showRSS()
     $cached = $cache->cachedVersion(); if (!empty($cached)) { echo $cached; exit; }
 
     // If cached was not found (or not usable), then read the database and build the response:
-    $LINKSDB=new linkdb(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);  // Read links from database (and filter private links if user it not logged in).
+    $LINKSDB = new LinkDB(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);
 
     // Optionally filter the results:
     $linksToDisplay=array();
@@ -1019,7 +780,7 @@ function showATOM()
     $cached = $cache->cachedVersion(); if (!empty($cached)) { echo $cached; exit; }
     // If cached was not found (or not usable), then read the database and build the response:
 
-    $LINKSDB=new linkdb(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);  // Read links from database (and filter private links if used it not logged in).
+    $LINKSDB = new LinkDB(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);
 
 
     // Optionally filter the results:
@@ -1104,7 +865,7 @@ function showDailyRSS()
     $cache = new pageCache(pageUrl(),startsWith($query,'do=dailyrss') && !isLoggedIn());
     $cached = $cache->cachedVersion(); if (!empty($cached)) { echo $cached; exit; }
     // If cached was not found (or not usable), then read the database and build the response:
-    $LINKSDB=new linkdb(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);  // Read links from database (and filter private links if used it not logged in).
+    $LINKSDB = new LinkDB(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);
 
     /* Some Shaarlies may have very few links, so we need to look
        back in time (rsort()) until we have enough days ($nb_of_days).
@@ -1172,7 +933,7 @@ function showDailyRSS()
 // "Daily" page.
 function showDaily()
 {
-    $LINKSDB=new linkdb(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);  // Read links from database (and filter private links if used it not logged in).
+    $LINKSDB = new LinkDB(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);
 
 
     $day=Date('Ymd',strtotime('-1 day')); // Yesterday, in format YYYYMMDD.
@@ -1240,7 +1001,7 @@ function showDaily()
 // Render HTML page (according to URL parameters and user rights)
 function renderPage()
 {
-    $LINKSDB=new linkdb(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);  // Read links from database (and filter private links if used it not logged in).
+    $LINKSDB = new LinkDB(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);
 
     // -------- Display login form.
     if (isset($_SERVER["QUERY_STRING"]) && startswith($_SERVER["QUERY_STRING"],'do=login'))
@@ -1822,7 +1583,7 @@ HTML;
 function importFile()
 {
     if (!(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI'])) { die('Not allowed.'); }
-    $LINKSDB=new linkdb(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);  // Read links from database (and filter private links if used it not logged in).
+    $LINKSDB = new LinkDB(isLoggedIn() || $GLOBALS['config']['OPEN_SHAARLI']);
     $filename=$_FILES['filetoupload']['name'];
     $filesize=$_FILES['filetoupload']['size'];
     $data=file_get_contents($_FILES['filetoupload']['tmp_name']);