Geting rid of multi recommended links

[CyJobes]

The boss gave me all kinds of terms to use for creating recommended links, but if someone put a couple of the same terms in, it would show the recommended link twice.

No can do says the boss, so I gotta find a way to use all the terms and yet display only one instance of the recommended link.

Here is my solution to the problem. It may not be helpful to everyone, and there might be some other things I am missing in regards to other recommended links appearing. But we need to use the terms, and we don’t need multiple instances of a recommended link.

E D I T

Please see my reply below for the revised code.
http://www.wrensoft.com/forum/showpo...77&postcount=3
The original solution was incorrect.

[Ray]

It is on our list of things to do for the next version (tendatively numbered V6.1 but is most likely to be V7 now given the big changes we've made).

We're looking into making it possible to specify multiple words/phrases for a single recommended link. This will prevent the need for multiple entries. It will also be better than the filtering method above because it means less entries in the index, and less search time.

[CyJobes]

I found an error with the solution I had created. This error was allowing only ONE recommended link to appear, even if there were words that could have been associated with other recommended links.

The following fix would allow multiple words to find other recommended links, without producing duplicates of the same recommended link.

Just before the RecLinkWordMatch function place the declaration of a new array like this:

$url_dup_array= array();
function RecLinkWordMatch($rec_word, $rec_idx)
{ 

Now we need to add this to the function as a global variable. Just after the function, and before the other global variables, place the following:

global $url_dup_array; 

In the end, the first part of the function should look like this:

$url_dup_array= array();
function RecLinkWordMatch($rec_word, $rec_idx)
{    
    global $url_dup_array; 

Next, scroll down within the same function and find the 'if' statement that looks like this:

        if ($result == 0)
        { 

Now we have to move a few lines. Look for

$pgdata = GetPageData($rec_idx);
$url = $pgdata[$PAGEDATA_URL]; 

and move them to just after the ‘if’ statement. It should end up looking like this:

        if ($result == 0)
        {
            $bRecLinkFound = true;
            $pgdata = GetPageData($rec_idx); 

Now we add a few lines of code that will keep the script from creating a duplicate of a recommended link. The new lines look like this:

            $url = $pgdata[$PAGEDATA_URL];
            
            if (in_array($pgdata[$PAGEDATA_URL], $url_dup_array))
            {                
                    continue;    
            }            
            $url_dup_array[] = $pgdata[$PAGEDATA_URL]; 

In the end, the repositioned code and new lines just under the ‘if’ statement should look like this:

        if ($result == 0)
        {
            $bRecLinkFound = true;
            $pgdata = GetPageData($rec_idx);
            $url = $pgdata[$PAGEDATA_URL];
            
            if (in_array($pgdata[$PAGEDATA_URL], $url_dup_array))
            {                
                    continue;    
            }            
            $url_dup_array[] = $pgdata[$PAGEDATA_URL]; 

Here is the entire revised function:

// matches the recommended link word against the user search query
$url_dup_array= array();
function RecLinkWordMatch($rec_word, $rec_idx)
{    
    global $url_dup_array;
    global $NumSearchWords, $queryForSearch, $queryForURL, $query, $num_rec_words;
    global $SearchWords, $UseWildCards, $SearchAsSubstring, $ToLowerSearchWords;
    global $OutputBuffers, $OUTPUT_RECOMMENDED;
    global $PAGEDATA_URL, $PAGEDATA_TITLE, $PAGEDATA_DESC, $PAGEDATA_IMG, $UseZoomImage;
    global $zoom_target, $GotoHighlight, $PdfHighlight;
    global $num_recs_found;
    global $STR_RECOMMENDED;

    $bRecLinkFound = false;

    for ($sw = 0; $sw <= $NumSearchWords; $sw++)
    {
        
        // if finished with last search word, check the full search query
        $result = 1;
        if ($sw == $NumSearchWords)
            $result = wordcasecmp($queryForSearch, $rec_word);
        else if (strlen($SearchWords[$sw]) > 0)
        {
            if ($UseWildCards[$sw] == 1)
            {
                $pattern = "/";

                // match entire word
                if ($SearchAsSubstring == 0)
                    $pattern = $pattern . "\A";

                $pattern = $pattern . $RegExpSearchWords[$sw];

                if ($SearchAsSubstring == 0)
                    $pattern = $pattern . "\Z";

                if ($ToLowerSearchWords != 0)
                    $pattern = $pattern . "/i";
                else
                    $pattern = $pattern . "/";

                $result = !(preg_match($pattern, $rec_word));
            }
            else if ($SearchAsSubstring == 0)
            {
                $result = wordcasecmp($SearchWords[$sw], $rec_word);
            }
            else
            {
                if (mystristr($rec_word, $SearchWords[$sw]) == FALSE)
                    $result = 1;    // not matched
                else
                    $result = 0;    // matched
            }

            if ($result != 0)
            {
                // if not matched, we check if the word is a wildcard
                if (strpos($rec_word, "*") !== false || strpos($rec_word, "?") !== false)
                {
                    $RecWordRegExp = "/\A" . pattern2regexp($rec_word) . "\Z/i";
                    $result = !(preg_match($RecWordRegExp, $SearchWords[$sw]));
                }
            }
        }


            
            
        if ($result == 0)
        {
            $bRecLinkFound = true;
            $pgdata = GetPageData($rec_idx);
            $url = $pgdata[$PAGEDATA_URL];
            
            if (in_array($pgdata[$PAGEDATA_URL], $url_dup_array))
            {                
                    continue;    
            }            
            $url_dup_array[] = $pgdata[$PAGEDATA_URL];
            
            if ($num_recs_found == 0)
            {
                $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<div class=\"recommended\">\n";
                $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<div class=\"recommended_heading\">$STR_RECOMMENDED</div>\n";
            }
            
            
            $title = $pgdata[$PAGEDATA_TITLE];
            $description = $pgdata[$PAGEDATA_DESC];
            if ($UseZoomImage)
                $image = $pgdata[$PAGEDATA_IMG];

            $urlLink = $url;
            //$urlLink = rtrim($urls[$rec_idx]);

            if ($GotoHighlight == 1)
            {
                if ($SearchAsSubstring == 1)
                    $urlLink = RecLinkAddParamToURL($urlLink, "zoom_highlightsub=".$queryForURL);
                else
                    $urlLink = RecLinkAddParamToURL($urlLink, "zoom_highlight=".$queryForURL);
            }
            if ($PdfHighlight == 1)
            {
                if (stristr($urlLink, ".pdf") != FALSE)
                    $urlLink = $urlLink."#search=&quot;".str_replace("\"", "", $query)."&quot;";
            }
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<div class=\"recommend_block\">\n";
            if ($UseZoomImage)
            {
                if (strlen($image) > 0)
                {
                    $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<div class=\"recommend_image\">";
                    $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<a href=\"".$urlLink."\"" . $zoom_target . "><img src=\"$image\" alt=\"\" class=\"recommend_image\"></a>";
                    $OutputBuffers[$OUTPUT_RECOMMENDED] .= "</div>";
                }
            }
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<div class=\"recommend_title\">";
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<a href=\"".$urlLink."\"" . $zoom_target . ">";
            if (strlen($title) > 1)
                $OutputBuffers[$OUTPUT_RECOMMENDED] .= PrintHighlightDescription($title);
            else
                $OutputBuffers[$OUTPUT_RECOMMENDED] .= PrintHighlightDescription($pgdata[$PAGEDATA_URL]);
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "</a></div>\n";
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<div class=\"recommend_description\">";
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= PrintHighlightDescription($description);
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "</div>\n";
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "<div class=\"recommend_infoline\">$url</div>\n";
            $OutputBuffers[$OUTPUT_RECOMMENDED] .= "</div>";
            $num_recs_found++;
            break;
        }
    }
    return $bRecLinkFound;
}