freshcode: Artifact [2102c65069]

Artifact 2102c65069eb2ec38b06dc0948e151c7fa037dd5:

File aux.php — part of check-in [792720840e] at 2017-01-31 18:42:53 on branch trunk — Recent updates to basic spam blacklist. (user: mario size: 11326)
<?php
/**
 * api: freshmeat
 * title: template auxiliary code
 * description: A few utility functions and data for the templates
 * version: 0.5.2
 * license: AGPL
 *
 * This function asortment prepares some common output.
 * While a few are parsing helpers or DB query shortcuts.
 *
 */



#-- Additional input filters


// Project names may be alphanumeric, and contain dashes
function proj_name($s) {
    return preg_replace("/[^a-z0-9-_.]+|^[^a-z0-9]+|\W+$|\.(?!\w{1,7}$)|[-_.]+(?=[-_.])/", "", strtolower($s));
}

// Tags is a comma-separated list, yet sometimes delimited with something else; normalize..
function f_tags($s) {
    return
        preg_replace(     # exception for "c++" and "c#"
            ["~[-_.:/]+~", "/(([cflje]#|c\+\+)(?=[\s,-]))?[+#]*/", "/[,;|]+/", "/[^a-z0-9,+#\s-]+/", "/[,\s]+/", "/^\W+|\W+$/"],
            [  "-",            "$1",                               ",",             " ",             ", "    ,      ""      ],
            strtolower($s)
        );
}

// Version numbers, remove non-word characters, compact spaces, strip leftmost "v" prefix
function f_version($s) {
    return trim(preg_replace("/(?<=\s)\s+|^[vV](?![a-z])/", "", trim(input::words($s, "~+()#\/@:"))));
}


#-- Template helpers

// Wrap tag list into links
function wrap_tags($tags, $r="") {
    foreach (str_getcsv($tags) as $id) {
        $id = trim($id);
        $r .= "<a href=\"/search?tag=$id\">$id </a>";
    }
    return $r;    
}

// Return DAY MONTH and TIME or YEAR for older entries
function date_fmt($time) {
    $lastyear = time() - $time > 250*24*3600;
    return strftime($lastyear ? "%d %b %Y" : "%d %b %H:%M", $time);
}



/**
 * Substitute `$version` placeholders in URLs.
 * (Which is one of the most useful features on freshcode.club)
 *
 * Supported syntax variations:
 *    →  $version and $version$
 *    →  %version and %version%
 *
 * And for substituting $version-number dots:
 *    →  $-$version  for which 1.2.3 becomes 1-2-3
 *    →  $_$version  for which 2.3.4 becomes 2_3_4
 *
 * Or to interpolate a single version tuple:
 *    →  $version0  (becomes `7` for `7.1.2` input)
 *
 */
function versioned_url($url, $version) {
    $rx = "/
        ([ \$ % ])                  # var sigil
        ( (.?) \\1 )?+              # substitution prefix
        (version|Version|VERSION)   # 'version'
        (\d?)                       # suffix 0, 1, 2 to access version tuples
        (?= \\1 | \b | _ )          # optional ending sigil [%$], wordbreak, or underscore
    /x";
    // Check for and displace '$version'
    return preg_replace_callback(
        $rx,
        function ($m) use ($version) {
            // Optionally substitute dots in version string
            if (strlen($m[2])) {
                $version = strtr($version, ["." => $m[3]]);
            }
            // tuple access
            if (strlen($m[5])) {
                $version = preg_split("/[-._~]/", $version);
                return $version[$m[5]];
            }
            return $version;
        },
        $url
    );
}


/**
 * Convert "url1=, url2=, url3=" list into titled hyperlinks.
 *
 */
function proj_links($urls, $entry, $r="") {

    // unpack and filter
    $urls = p_key_value($urls, NULL);
    $urls = array_filter(array_map("input::url", $urls));

    // join into HTML list
    foreach ($urls as $title=>$url) {
    
        // normalize title and substitute $version placeholders
        $title = ucwords($title);
        $_title = strtolower($title);
        $url = input::html(versioned_url($url, $entry["version"]));
        
        //gimmick: append HTML link and <audio> for theme song
        $r .= "\t   &rarr; <a href=\"$url\">$title</a>"
            . ($_title == "theme-song" ? "<audio autoplay onclick='this.paused ? this.play() : this.pause()'>♫<source type=\"audio/ogg\" src=\"$url\"></audio>" : "")
            . "<br>\n";
    }
    return $r;
}




// Project listing output preparation;
// HTML context escapaing, versioned urls, formatted date string
function prepare_output(&$entry) {

    // versioned URLs
    $entry["download"] = versioned_url($entry["download"], $entry["version"]);
    
    // project screenshots
    if (TRUE or empty($entry["image"])) {
        if (file_exists($fn = "img/screenshot/$entry[name].jpeg")) {
            $entry["image"] = "/$fn?" . filemtime($fn);
        }
        else {
            $entry["image"] = "/img/nopreview.png";
        }
    }
    
    //
    $entry["formatted_date"] = date_fmt($entry["t_published"]);
    
    // HTML context
    $entry = array_map("input::_html", $entry);

    // user image
    $entry["submitter_img"] = submitter_gravatar($entry["submitter_image"]);
}


/**
 * Convert email@xyz to gravatar or identicon,
 * keep raw URLs, or use default image for empty fields.
 *
 */
function submitter_gravatar($img, $size=24) {
    
    // capture+strip email
    if (is_int(strpos($img, "@"))) {
        $img = "//www.gravatar.com/avatar/" . md5($img) . "?s=$size&d=identicon&r=pg";
    }
    elseif (empty($img)) {
        $img = "/img/user.png";
    }
    
    // return html <img> snippet
    return "<img src=\"$img\" width=$size height=$size class=gravatar>";
}


/**
 * Convert special user@service monikers into URLs,
 * keep everything else as supposed gravatar link.
 *
 */
function url_user_icon($mail) {

    // special project hosters
    if (preg_match("/^([\w.-]+)@(github|sourceforge|launchpad)/i", $mail, $user)) {

        // user image lookup rules    
        $map = [
            "github" => [
                #"page" => ["URL" => "http://api.github.com/users/$user[1]", "USERPWD" => GITHUB_API_PW],
                #"rx" => "~\"avatar_url\":\s*\"(.*?)\"~",
                "page" => "http://github.com/$user[1]",
                "rx" => "~(//avatars\d*.githubusercontent.com/u/\d+[?=\w&;]+)~",
                "default" => "//identicons.github.com/$user[1].png",
            ],
            "sourceforge" => [
                "page" => "http://sourceforge.net/u/$user[1]/profile",
                "rx" => "~(http[:/\w.]+/avatar/[[:xdigit:]]+)~",
                "default" => "$user[1]@users.sourceforge.net",
            ],
            "launchpad" => [
                "page" => "http://launchpad.net/~$user[1]",
                "rx" => "~(//launchpadlibrarian.net/\d+/[^\"\'\>\)]+)~",
                "default" => "$user[1]@launchpad.net",
            ],
        ];
        extract($map[strtolower($user[2])]);
        
        // fetch user page on service, check for gravatar-id or url
        if (preg_match($rx, $src = curl($page)->timeout(2.5)->exec(), $url))
        {
            $mail = $url[1];
        }
        // fallback
        else {
            $mail = $default;
        }
    }

    // keep as-is, and have it resolved as standard gravatar
    return($mail);
}



// Social media share links
function social_share_links($name, $url) {
    $c = array("google"=>0, "facebook"=>0, "twitter"=>0, "reddit"=>0, "linkedin"=>0, "stumbleupon"=>0, "delicious"=>0);
    return <<<HTML
      <span class=social-share-links>
         <a href="https://plus.google.com/share?url=$url" title=google+> g+ </a>
         <a href="https://www.facebook.com/sharer/sharer.php?u=$url" title=facebook> fb </a>
         <a href="https://twitter.com/intent/tweet?url=$url" title=twitter> tw </a>
         <a href="http://reddit.com/submit?url=$url" title=reddit> rd </a>
         <a href="https://www.linkedin.com/shareArticle?mini=true&amp;url=$url" title=linkedin> in </a>
         <a href="https://www.stumbleupon.com/submit?url=$url" title=stumbleupon> su </a>
         <a href="https://del.icio.us/post?url=$url" title=delicious> dl </a>
      </span>
HTML;
}
function social_share_count($num) {
    return empty($num) ? "" : "<var class=social-share-count>$num</var>";
}



/**
 * Write out pseudo pagination links.
 * This is just appended no matter the actually available entries.
 * The db() queries themselves handle the LIMIT/OFFSET, depending on a page param.
 *
 */
function pagination($page_no, $GET_param="n") {
    print "<p class=pagination-links> »";
    foreach (range($page_no-2, $page_no+9) as $n) if ($n > 0) {
        print " <a " . ($n==$page_no ? "class=current " : ""). "href=\"?n=$n\">$n</a> ";
    }
    print "« </p>";
}


/**
 * Output a list of select <option>s
 *
 * - Either accepts a option,value,field list.
 * - Or an associative array.
 *
 */
function form_select_options($names, $value=NULL, $r="") {

    // Transform comma-separated string into array
    $map = is_string($names) ? array_combine($names = str_getcsv($names), $names) : $names;
    
    // Add currently active value if missing
    if ($value and !isset($map[$value]) and $value !== NULL) {
        $map[$value] = $value;
    }
    
    // Output <option> fields
    foreach ($map as $id=>$title) {
        // optgroup
        if (is_array($title)) {
            $r .= "<optgroup label=\"$id\">" . form_select_options($title, $value) . "</optgroup>";
        }
        // plain value field
        else {
            $r .= "<option" . ($id == $value ? " selected" : "")
                . " value=\"$id\" title=\"$title\">$id</option>";
        }
    }
    return $r;
}


/**
 * CSRF token generation/verification.
 *
 * Is only used for logged-in users though. Here they're mainly to prevent
 * remotely initiated requests against other users, not general form nonces.
 */
function csrf($probe=false) {

    // Tokens are stored in session, reusable, but only for an hour
    $store = & $_SESSION["csrf"];
    foreach ($store as $id=>$time) {
        if ($time < time()) { unset($store[$id]); }
    }
    
    // Test presence
    if ($probe) {
        if (empty($_SESSION["openid"])) {
            return TRUE;
        }
        if ($id = $_REQUEST->name["_ct"]) {
            return isset($store[$id]);
        }
    }
    
    // Create new entry, output form field for token
    else {
        // server ENV already contained Apache unique request id etc.
        $id = sha1(serialize($_SERVER->__vars));
        $store[$id] = time() + 3600;  // timeout
        return "<input type=hidden name=.ct value=$id>";
    }
}


/**
 * Trivial check against well-known project spam.
 *
 */
function data_blacklisted($release) {
    $rules = array(
        "submitter" => "/wilmix j/i",
        "homepage" => "~wilmix|zeesmovie|softwarereviews|diclofenac|\/\/buy-~i",
        "name" => "/DOLLAR|JDollar|Jehovah|Millionaire *s/",
        "description" => "/viagra|levitra|tetracycline*s|quickbooks/",
    );
    foreach ($rules as $field => $rx) {
        if (preg_match($rx, $release[$field])) {
            return TRUE;
        }
    }
    return false;
}




#-- Some string parsing


/**
 *  Plain comma-separated list
 *
 */
function p_csv($str) {
    return preg_split("/\s*,\s*/", trim($str));
}


/**
 *  Extracts key = value list.
 *  Keys may be wrapped in $, % or []
 *  Values may not contain spaces
 *
 */
function p_key_value($str, $case=CASE_LOWER, $match="\S+") {
    preg_match_all(
        "@
           [[%$]*  ([-\w]+)  []%$]*
              \h*  [:=>]+  \h*
                   ($match)
           (?<![,.;])
        @imsx",
        $str, $m
    );
    $r = array_combine($m[1], $m[2]);
    return is_int($case) ? array_change_key_case($r, $case) : $r;
}




?>