pglaf.phh

<?php
require_once('db.php');
require_once('blocks.phh');

abstract class Mode
{
    const Development = 1;
    const Production  = 3;
}

if (str_endswith($_SERVER['SERVER_NAME'], 'pglaf.org'))
{
    define('MODE', Mode::Production);
}
else
{
    define('MODE', Mode::Development);
}


// $tpv_root_dir:
// Absolute path of directory containing (indirectly) all uploaded TP+V files.

// $tpv_root_url:
// Quasi-absolute URL that resolves to $tpv_root_dir.

// $spooldir:
// Absolute path of directory containing uploaded ebook files
// (and their accompanying info.txt files).

switch (MODE)
{
    case Mode::Development:
        $tpv_root_dir  = "/home/michael/pg/pglaf/web_dyn";
        $tpv_root_url  = "/pglaf.web_dyn";
        $spooldir = "/home/michael/pg/pglaf/priv_dyn/upload.spool";
        $timezone = 'America/New_York';
        break;

    case Mode::Production:
        $tpv_root_dir  = "/htdocs/copy";
        $tpv_root_url  = "//copy.pglaf.org";
        $spooldir = "/home/spool";
        $timezone = 'America/Los_Angeles';
        break;

    default:
        assert(FALSE);
}

$early_tpv_dir = "$tpv_root_dir/NEW";
// Absolute path of directory containing those TP+V files
// that users upload via 'Upload a Single File'
// (i.e., prior to submitting a clearance request).

$uplogfile = "/home/copy/upload.log"; // log upload events
$copylogfile = "/home/copy/copysubmit.log"; // log upload events
$tmpdir = "/tmp";
$clearedfile = "/home/cleared/cleared6.gbn"; // Feb 4 2014

// set the default timezone to use. gbn 20090816
date_default_timezone_set($timezone);

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

abstract class Status
{
    const Unadjudicated = 1;
    const Accepted      = 2;
    const Rejected      = 5;
    const Cancelled     = 8;
    const Text_Uploaded = 9;
}

// Store common stuff here so we don't need to look up static items:
$label_for_status_id_ = array(
    Status::Unadjudicated => "AwaitingClearance",
    Status::Accepted      => "Cleared OK",
    Status::Rejected      => "Not OK",
    Status::Cancelled     => "Cancelled",
    Status::Text_Uploaded => "Submitted",
);

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

class Config {
  function __construct() {
    $this->endtag                   = ""; // "" for html, " /" for xhtml
    $this->domain                   = "copy.pglaf.org";
    $this->css_url                  = "/css-server.php";
    $this->page_encoding            = "ISO-8859-1";
    $this->dtd_public               = "-//W3C//DTD HTML 4.01//EN";
    $this->dtd_system               = "http://www.w3.org/TR/html4/strict.dtd";

    setlocale (LC_ALL, 'en_US');
    set_include_path ('.:..');
  }
}

$config = new Config ();

class Page {
  function header ($caption) {
    global $page_header, $config;
    $search = array ("<script_title>", "<page_encoding>", "<dtd_public>", "<dtd_system>");
    $replace = array ($caption, $config->page_encoding,
      $config->dtd_public, $config->dtd_system);

    echo (str_replace ($search, $replace, $page_header));
  }

  function footer () {
    global $page_footer;
    echo ($page_footer);
  }
}

$page = new Page ();

function pageheader ($caption) {
  echo "
    <style>
      table.compact th {text-align: left;}
    </style>
  ";

  // Currently, $page->header doesn't produce any output.
  // Eventually, it will, but until then, I want some output during testing
  // so that I can detect any changes.
  if (MODE == Mode::Development) testing_comment("pageheadder:\n$caption\n");
  global $page;
  $page->header ($caption);
}

function pagefooter () {
  global $page;
  $page->footer ();
}

$script_abs_dir = dirname($_SERVER['SCRIPT_FILENAME']);
$template = file_get_contents("$script_abs_dir/template.html.inc");
$template = str_replace ("<script_date>", date ("Y-m-d H:i:s", getlastmod()), $template);

// gbn:
// list ($page_header, $page_footer) = explode ("<script_output>", $template);

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function get_application()
// This code services two 'applications',
// one at copy.pglaf.org and one at upload.pglaf.org.
// Sometimes the code needs to know which one it's being used for.
// (But we should probably try to reduce those cases.)
{
    // On *.pglaf.org, the discriminator is in the server_name.
    // Testing at home, the server is always 'localhost',
    // so the discriminator is in the script_name
    // (specifically, in the directory path, not the name of the script file necessarily).
    // So we concatenate the two and look for the discriminator in that.
    $quasi_request_url = $_SERVER['SERVER_NAME'] . $_SERVER['SCRIPT_NAME'];

    if (str_contains($quasi_request_url, 'upload'))
    {
        return 'upload';
    }
    elseif (str_contains($quasi_request_url, 'copy'))
    {
        return 'copy';
    }
    else
    {
        // Can't happen?
        return 'copy';
    }
}

function show_footer()
{
    $app = get_application();

    echo "
        <hr>

        <p>This system uses cookies to track who you are.
        Be sure to <a href=\"logout.php\">logout</a> when done, if you are using
        a public terminal.
        </p>

        <p>Trouble or questions?
    ";

    if ($app == 'copy')
    {
        echo "Please email <i>copyright2024 AT pglaf.org,</i>
            including details such as your clearance username and the item ID";
    }
    elseif ($app == 'upload')
    {
        echo "Please email <i>copyright2024 AT pglaf.org</i> for clearance questions,
            or <i>pgww AT lists.pglaf.org</i> for production or status questions.";
    }
    else
    {
        assert(FALSE);
    }
    echo "
        </p>
    ";

    $discussion_url = "https://www.pgdp.net/phpBB3/viewtopic.php?f=27&t=63371";
    echo "
        <p>
            <span style='color:red'>New!</span> (May 2017)
            The web apps at copy.pglaf.org and upload.pglaf.org are undergoing some changes.
            If you encounter a bug,
            or have a comment about the software,
            you're welcome to join the discussion at
            <a href='$discussion_url'>this forum topic</a>.
            (If you're not registered at pgdp.net,
            you can send your feedback to <i>copyright2024 AT pglaf.org</i>.)
        </p>
    ";

    echo "
        <ul>
            <li>2018-01-01: You can now supply a scan-set URL when submitting a clearance request.</li>
            <li>2017-05-29: New feature: If you use HTTP to access the login and new-user screens, they will now offer you a link to login/signup via HTTPS instead.</li>
            <li>2017-05-12: New feature: <a href='//copy.pglaf.org/status.php'>status.php</a> now lets you fetch a clearance record if you know its key.</li>
            <li>2017-05-04: Maintenance install: future-proofing and bugfixing. Users should see little to no difference.</li>
        </ul>
    ";

    echo "
        </body>
        </html>
    ";
}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

// People use some weird-ass filenames.  Make more Unix-friendly:
// Note that this function is also referenced from /htdocs/epubmaker/*
function fix_filename($inname) {

  // Get just the filename part, not the directory part (if any):
  $inname = str_replace("\\", "/", $inname); // Any dos directories?
  $inname = basename($inname); // Just the filename part

#  $inname = ereg_replace("[^-_a-zA-Z0-9.]", "", $inname);
  $inname = preg_replace("/[^-_a-zA-Z0-9.]/", "", $inname);
  // Safety: If that removed everything, just give it a random
  if (0 == strlen($inname)) {
    $inname = time() . rand(1000,9999); //gmp_strval(gmp_random(1));
  }

  return($inname);
}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function require_logged_in_user()
{
    if (isset($_SESSION['pgusername']))
    {
        // The sender is a logged-in user.
        return;
    }

    // The sender is not logged-in.
    // Don't let them proceed.

    $script_name = $_SERVER['SCRIPT_NAME'];
    if (str_endswith($script_name, 'copy.php'))
    {
        print "<p>You are not logged in.</p>\n\n";
        echo_login_form();
        print "<a href='usernew.php'>New user?</a>\n";
        show_footer();
    }
    elseif (str_endswith($script_name,'upload.php'))
    {
        print "<p>You are not logged in.</p>\n\n";
        echo_login_form(TRUE);
        print "<a href='//copy.pglaf.org/usernew.php'>New user?</a>\n";
        show_footer();
    }
    else
    {
        print "<p><font color='red'>You need to <a href='copy.php'>login</a> first.</p>\n";
    }
    exit;
}

// -----------------------------------------------------------------------------

function echo_login_form($upload_kludge=FALSE)
{
    encourage_https_for('login');

    print "<form method='POST' accept-charset='UTF-8'>\n";
    print "Username: <input type='text' size=50 name='username'><br>\n";
    print "Password: <input type='password' size=50 name='userpass'><br>\n";
    print "<input type='submit' name='login' value='login'>\n";
    print "<input type='submit' name='logout' value='logout'>\n";
    if ($upload_kludge)
    {
        $esc_in_ok_key = eha(@$_REQUEST['in_ok_key']);
        print "<input type='hidden' name='in_ok_key' value='$esc_in_ok_key'>\n";
    }
    print "</form>\n\n";
}

function encourage_https_for($verb)
{
    if (@$_SERVER['REQUEST_SCHEME'] != 'https')
    {
        $url = "https://" . $_SERVER['SERVER_NAME'] . $_SERVER['REQUEST_URI'];
        $esc_url = eha($url);
        print "<p style='width: 32em; padding: 3px; border: solid red 1px'>\n";
        print "You can use the form on this page to $verb using HTTP,\n";
        print "but we encourage you to securely <a href='$esc_url'>$verb using HTTPS</a> instead.\n";
        print "</p>\n";
    }
}

// -----------------------------------------------------------------------------

function handle_login_attempt()
// If the current request is an attempt to login
// (i.e., a submission from the login form),
// handle it.
{
    if (isset ($_SESSION['pgusername']))
    {
        // The sender is a logged-in user,
        // so the current request can't be a login attempt.
        return;
    }

    $username = get_username_param($_REQUEST, 'username');
    $userpass = get_password_param($_REQUEST, 'userpass');

    if ($username == '')
    {
        // The sender didn't provide login credentials,
        // so they're just viewing the page as a not-logged-in visitor.
        return;
    }

    $row = confirm_credentials($username, $userpass);

    $username_id=$row[0];
    $firstname=$row[3];

    unset($_REQUEST['login']);

    $expire = time() + 24 * 3600 * 365; // 1 year
    setcookie ("pgusername", $username, $expire, "/", "pglaf.org", 0);
    setcookie ("pgusernameid", $username_id, $expire, "/", "pglaf.org", 0);
    setcookie ("pgfirstname", $firstname, $expire, "/", "pglaf.org", 0);
    $_SESSION['pgusername'] = $username;
    $_SESSION['pgusernameid'] = $username_id;
    $_SESSION['pgfirstname'] = $firstname;

    // Reload the page with new values:
    if (@$_REQUEST['in_ok_key'] != '') {
        header ("Location: " . $_SERVER['PHP_SELF'] . "?in_ok_key=" . $_REQUEST['in_ok_key']);
    } else {
        header ("Location: " . $_SERVER['PHP_SELF']);
    }

    exit;
    // Because after you send a "Location" header, there's no point sending anything else.
}

// -----------------------------------------------------------------------------

function confirm_credentials($username, $password=NULL)
{
    if (is_null($password))
    {
        // Don't need to check password.
        // I.e., just checking that $username is (still) in the 'users' table.
        $condition = "(username='$username')";
    }
    else
    {
        $esc_password = db_escape(htmlspecialchars($password));
        // It doesn't make sense to send it through 'htmlspecialchars',
        // but we have to, because their password was sent through 'htmlspecialchars'
        // when they signed up, so that's what's in the database.
        // And we can't 'fix' the database, because what's there is a *hash* of the password,
        // and there's no way to convert that to a hash of a 'fixed' password
        // without knowing the password, which we don't.
        // And we can't ask the users to reset their passwords,
        // because there's no mechanism for them to do that.
        $condition = "(username='$username') and (password=password('$esc_password'))";
    }

    $query = "SELECT * from users where $condition";

    $result = db_query_od($query);

    $row = db_fetch_row($result);
    if (!$row)
    {
        // The result-set has no rows.
        // I.e., the 'users' table has no row with the given username+password.
        print "<p><font color=\"red\">Username/password not found[1].  Please <a href='{$_SERVER['SCRIPT_NAME']}'>try again</a>, or create a <a href=\"//copy.pglaf.org/usernew.php\">new username</a></font></p>";
        exit;
    }

    if ($row[1] != $username)
    {
        // The 'users' table *does* have a matching row,
        // but its 'username' field is not PHP-equal to $username.
        // I think this can only happen if the username given just now
        // is a 'case-variant' of the username given at signup.
        // (I.e., they're equal only if you ignore case,
        // which is what MySQL does by default.)
        print "<p><font color=\"red\">Username/password not found[0].  Please try again, or create a <a href=\"//copy.pglaf.org/usernew.php\">new username</a></font></p>";
        exit;
    }

    return $row;
}

// -----------------------------------------------------------------------------

function require_admin_user()
{
    $login_username=$_SESSION['pgusername'];
    if (!is_admin_user($login_username)) {
        print "<p><font color=\"red\">Administrative access denied for $login_username</font></p>";
        exit;
    }
}

// -----------------------------------------------------------------------------

function is_admin_user($username)
{
    // Hard-code who is allowed to admin:
    $admin_usernames = array(
        'gbnewby',
        'JulietS',
        'jmdyck',
        'srjfoo',
        'widger',
        'ccx074',
    );
    return in_array($username, $admin_usernames, TRUE);
}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function echo_nav_bar($menu_items=array())
{
    assert(isset($_SESSION['pgusername']));

    echo "<p>You are: <font color='green'>Logged in</font>"
        . " as " . ehe($_SESSION['pgfirstname'])
        . " [" . $_SESSION['pgusername'] . "]";

    $menu_items = array('logout.php' => 'logout') + $menu_items;
    foreach ($menu_items as $url => $label)
    {
        echo " (<a href='$url'>$label</a>)";
    }
    echo "</p>\n\n";
}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function echo_clearance_action_form()
{
    echo <<< _EOS_

<form method="POST" action="copy.php" accept-charset="UTF-8">

<input type="radio" name="userchoice" value="upload">
<a href="/cgi-bin/upfiles">Upload groups of page images for clearance</a><br>

<input type="radio" name="userchoice" value="newentry" checked>
<a href="submit.php">Submit a new clearance request</a><br>

<input type="radio" name="userchoice" value="status">
<a href="status.php">Get status of my prior clearance requests</a><br>

<input type="submit" name="selected" value="Go!">

</form>

_EOS_;
}

function handle_clearance_action_form()
// If the current request is a submission from the clearance action form,
// handle it.
{
    if (isset ($_REQUEST['selected'])) {
        if ($_REQUEST['userchoice'] == "upload") {
            header ("Location: /cgi-bin/upfiles");
            exit;
        }
        if ($_REQUEST['userchoice'] == "newentry") {
            header ("Location: submit.php");
            exit;
        }
        if ($_REQUEST['userchoice'] == "status") {
            header ("Location: status.php");
            exit;
        }
    }
}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function echo_filtering_form($default_username)
{
    print "<form action=\"admin.php\" method=\"get\" accept-charset=\"UTF-8\">\n";
    print "<p>Menu:\n";
    print "\n<br>Get username: <input type=\"text\" size=20 name=\"alternate_username\"";
    if (strlen($default_username) > 0) {
        print " value=$default_username";
    }
    print ">\n";
    global $label_for_status_id_;
    foreach ($label_for_status_id_ as $status_id => $label)
    {
        if ($label == 'Submitted')
            continue;
            // Why do we skip this option?
            // Is it because the list might be long?

        print "<input type=\"radio\" name=\"status_id\" value=\"$status_id\">$label\n";
    }
    if (!str_endswith($_SERVER['SCRIPT_NAME'], 'viewall.php'))
    {
        print "... <a href=\"viewall.php\">Everything</a>";
    }
    print "<input type=\"submit\" value=\"Submit\">\n";
    print "</form>\n";
}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function echo_clearance_key_form()
{
    print "<form method=\"POST\" action=\"upload.php\" accept-charset=\"UTF-8\">\n";
    print "<p>Provide your <i>Clearance OK key</i> to upload a completed eBook:";
    print "<br><input type=\"text\" name=\"in_ok_key\" size=\"32\">\n";
    print "<input type=\"submit\" name=\"submit\" value=\"Go!\">\n";
    print "</form>\n";
}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function get_username_param($array, $key)
{
    return get_param_by_regex('/^[-_0-9a-zA-Z]+$/', $array, $key, FALSE, '', TRUE, '');
}

function get_password_param($array, $key)
{
    return get_param_by_regex('/^.+$/', $array, $key, FALSE, '', TRUE, '');
}

function get_request_id_param($array, $key)
{
    return get_param_by_regex('/^\d+$/', $array, $key, FALSE, '', TRUE, 'DIE');
}

function get_request_key_param($array, $key)
{
    return get_param_by_regex('/^[0-9a-z-]+$/', $array, $key, FALSE, '', TRUE, 'DIE');
}

function get_status_id_param($array, $key)
{
    return get_param_by_regex('/^[12589]$/', $array, $key, FALSE, strval(Status::Unadjudicated), TRUE, 'DIE');
}

function get_param_by_regex(
    $regex, $array, $key,
    $complain_if_unset, $result_if_unset,
    $complain_if_bad,   $result_if_bad)
{
    assert_options(ASSERT_BAIL, TRUE);
    assert(is_string($regex));
    assert(is_array($array));
    assert(is_string($key));
    assert(is_bool($complain_if_unset));
    assert(is_string($result_if_unset));
    assert(is_bool($complain_if_bad));
    assert(is_string($result_if_bad));

    if (isset($array[$key]))
    {
        $arg = $array[$key];

        // Trim whitespace (because copy-and-paste sometimes includes extra whitespace)
        $arg = trim($arg);

        if (is_string($arg) && preg_match($regex, $arg))
        {
            return $arg;
        }
        else
        {
            if ($complain_if_bad)
            {
                $esc_arg = ehe($arg);
                echo "<p style='color:red'>Value passed to parameter '$key' is invalid: '$esc_arg'</p>\n";
            }
            if ($result_if_bad == 'DIE')
            {
                exit;
            }
            else
            {
                return $result_if_bad;
            }
        }
    }
    else
    {
        if ($complain_if_unset)
        {
            echo "<p style='color:red'>Parameter '$key' was not set.</p>\n";
        }
        if ($result_if_unset == 'DIE')
        {
            exit;
        }
        else
        {
            return $result_if_unset;
        }
    }

}

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function send_email($to_addr, $subject, $body, $more_headers)
{
    if (MODE == Mode::Development)
    {
        testing_comment("The following email would have been sent:
================================================================================
To: $to_addr
Subject: $subject
$more_headers

$body
================================================================================
");
    }
    else
    {
        mail($to_addr, $subject, $body, $more_headers);
    }
}

// -----------------------------------------------------------------------------

function append_to_log($filename, $log_entry)
{
    if (MODE == Mode::Development)
    {
        testing_comment("would have appended to $filename:\n$log_entry");
    }
    else
    {
        $handle = fopen($filename, "a");  if (!$handle) return FALSE;
        $w = fwrite($handle, $log_entry); if (!$w) return FALSE;
        fclose($handle);
    }
    return TRUE;
}

// -----------------------------------------------------------------------------

function create_file($filename, $permissions, $content)
{
    if (MODE == Mode::Development)
    {
        testing_comment("would have created file $filename with permissions $permissions and content:\n$content");
    }
    else
    {
        $handle = fopen($filename, "w");
        fwrite($handle, $content);
        fclose($handle);
        chmod($filename, $permissions);
    }
}

// -----------------------------------------------------------------------------

function testing_comment($content)
{
    assert(MODE == Mode::Development);
    $escaped_content = preg_replace('/--/', '~~', $content);
    echo "<!--\nTESTING: $escaped_content-->\n";
}

// -----------------------------------------------------------------------------

function str_startswith($subject, $prefix)
// Return TRUE iff $subject starts with $prefix.
{
    return ( strncmp( $subject, $prefix, strlen($prefix) ) == 0 );
}

function str_endswith($subject, $suffix)
// Return TRUE iff $subject ends with $suffix.
{
    return ( substr($subject, -strlen($suffix)) == $suffix );
}

// gbn 2022-10-26: PHP 8.1 complained about this. It's a library
// function, so I don't know why it's also defined here.
// //function str_contains($haystack, $needle)
// // Return TRUE iff $haystack contains $needle.
// // {
// //     return ( strpos($haystack, $needle) !== FALSE );
// // }

// XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

function eha($s)
// Escape for HTML Attribute
// Entity-encode $s so that it's safe to use as the value of an attribute in HTML
// (assuming that the result will elsewhere be embedded in either single- or double-quotes).
{
    return htmlspecialchars($s, ENT_QUOTES, 'ISO-8859-1');
}

function ehe($s)
// Escape for HTML Element
// Entity-encode $s so that it's safe to use in textual content of an HTML element.
{
    return htmlspecialchars($s, ENT_NOQUOTES, 'ISO-8859-1');
}

// vim: sw=2 ts=2 expandtab