gnu-social/classes/File_redirection.php
Alexei Sorokin b1b1d2af93 [DATABASE] Update "modified" in Managed_DataObject instead of a DBMS trigger
Instead of relying on the MariaDB's ON UPDATE CURRENT_TIMESTAMP trigger update
"modified" attributes in Managed_DataObject. Every raw query that needs
adjusting is adjusted, as they won't update "modified" automatically anymore.

The main goal behind this change is to fix "modified" updates on PostgreSQL.
2021-07-16 19:44:37 +01:00

488 lines
18 KiB
PHP

<?php
// This file is part of GNU social - https://www.gnu.org/software/social
//
// GNU social is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// GNU social is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with GNU social. If not, see <http://www.gnu.org/licenses/>.
defined('GNUSOCIAL') || die();
/**
* Table Definition for file_redirection
*/
class File_redirection extends Managed_DataObject
{
###START_AUTOCODE
/* the code below is auto generated do not remove the above tag */
public $__table = 'file_redirection'; // table name
public $urlhash; // varchar(64) primary_key not_null
public $url; // text
public $file_id; // int(4)
public $redirections; // int(4)
public $httpcode; // int(4)
public $modified; // timestamp() not_null default_CURRENT_TIMESTAMP
/* the code above is auto generated do not remove the tag below */
###END_AUTOCODE
protected $file; /* Cache the associated file sometimes */
public static function schemaDef()
{
return array(
'fields' => array(
'urlhash' => array('type' => 'varchar', 'length' => 64, 'not null' => true, 'description' => 'sha256 hash of the URL'),
'url' => array('type' => 'text', 'description' => 'short URL (or any other kind of redirect) for file (id)'),
'file_id' => array('type' => 'int', 'description' => 'short URL for what URL/file'),
'redirections' => array('type' => 'int', 'description' => 'redirect count'),
'httpcode' => array('type' => 'int', 'description' => 'HTTP status code (20x, 30x, etc.)'),
'modified' => array('type' => 'timestamp', 'not null' => true, 'description' => 'date this record was modified'),
),
'primary key' => array('urlhash'),
'foreign keys' => array(
'file_redirection_file_id_fkey' => array('file', array('file_id' => 'id')),
),
);
}
public static function getByUrl($url)
{
return self::getByPK(array('urlhash' => File::hashurl($url)));
}
public static function _commonHttp($url, $redirs)
{
$request = new HTTPClient($url);
$request->setConfig(array(
'connect_timeout' => 10, // # seconds to wait
'max_redirs' => $redirs, // # max number of http redirections to follow
'follow_redirects' => false, // We follow redirects ourselves in lib/httpclient.php
'store_body' => false, // We won't need body content here.
));
return $request;
}
/**
* Check if this URL is a redirect and return redir info.
*
* Most code should call File_redirection::where instead, to check if we
* already know that redirection and avoid extra hits to the web.
*
* The URL is hit and any redirects are followed, up to 10 levels or until
* a protected URL is reached.
*
* @param string $in_url
* @return mixed one of:
* string - target URL, if this is a direct link or can't be followed
* array - redirect info if this is an *unknown* redirect:
* associative array with the following elements:
* code: HTTP status code
* redirects: count of redirects followed
* url: URL string of final target
* type (optional): MIME type from Content-Type header
* size (optional): byte size from Content-Length header
* time (optional): timestamp from Last-Modified header
*/
public static function lookupWhere($short_url, $redirs = 10, $protected = false)
{
if ($redirs < 0) {
return false;
}
if (strpos($short_url, '://') === false) {
return $short_url;
}
try {
$request = self::_commonHttp($short_url, $redirs);
// Don't include body in output
$request->setMethod(HTTP_Request2::METHOD_HEAD);
$response = $request->send();
if (405 == $response->getStatus() || 204 == $response->getStatus()) {
// HTTP 405 Unsupported Method
// Server doesn't support HEAD method? Can this really happen?
// We'll try again as a GET and ignore the response data.
//
// HTTP 204 No Content
// YFrog sends 204 responses back for our HEAD checks, which
// seems like it may be a logic error in their servers. If
// we get a 204 back, re-run it as a GET... if there's really
// no content it'll be cheap. :)
$request = self::_commonHttp($short_url, $redirs);
$response = $request->send();
} elseif (400 == $response->getStatus()) {
throw new Exception('Got error 400 on HEAD request, will not go further.');
}
} catch (Exception $e) {
// Invalid URL or failure to reach server
common_log(LOG_ERR, "Error while following redirects for $short_url: " . $e->getMessage());
return $short_url;
}
// if last url after all redirections is protected,
// use the url before it in the redirection chain
if ($response->getRedirectCount() && File::isProtected($response->getEffectiveUrl())) {
$return_url = $response->redirUrls[$response->getRedirectCount() - 1];
} else {
$return_url = $response->getEffectiveUrl();
}
$ret = array('code' => $response->getStatus()
, 'redirects' => $response->getRedirectCount()
, 'url' => $return_url);
$type = $response->getHeader('Content-Type');
if ($type) {
$ret['type'] = $type;
}
if ($protected) {
$ret['protected'] = true;
}
$size = $response->getHeader('Content-Length'); // @fixme bytes?
if ($size) {
$ret['size'] = $size;
}
$time = $response->getHeader('Last-Modified');
if ($time) {
$ret['time'] = strtotime($time);
}
return $ret;
}
/**
* Check if this URL is a redirect and return redir info.
* If a File record is present for this URL, it is not considered a redirect.
* If a File_redirection record is present for this URL, the recorded target is returned.
*
* If no File or File_redirect record is present, the URL is hit and any
* redirects are followed, up to 10 levels or until a protected URL is
* reached.
*
* @param string $in_url
* @param boolean $discover true to attempt dereferencing the redirect if we don't know it already
* @return File_redirection
*/
public static function where($in_url, $discover = true)
{
$redir = new File_redirection();
$redir->url = $in_url;
$redir->urlhash = File::hashurl($redir->url);
$redir->redirections = 0;
try {
$r = File_redirection::getByUrl($in_url);
try {
$f = File::getByID($r->file_id);
$r->file = $f;
$r->redir_url = $f->url;
} catch (NoResultException $e) {
// Invalid entry, delete and run again
common_log(
LOG_ERR,
'Could not find File with id=' . $r->file_id . ' referenced in File_redirection, deleting File redirection entry and and trying again...'
);
$r->delete();
return self::where($in_url);
}
// File_redirecion and File record found, return both
return $r;
} catch (NoResultException $e) {
// File_redirecion record not found, but this might be a direct link to a file
try {
$f = File::getByUrl($in_url);
$redir->file_id = $f->id;
$redir->file = $f;
return $redir;
} catch (NoResultException $e) {
// nope, this was not a direct link to a file either, let's keep going
}
}
if ($discover) {
// try to follow redirects and get the final url
$redir_info = File_redirection::lookupWhere($in_url);
if (is_string($redir_info)) {
$redir_info = array('url' => $redir_info);
}
// the last url in the redirection chain can actually be a redirect!
// this is the case with local /attachment/{file_id} links
// in that case we have the file id already
try {
$r = File_redirection::getByUrl($redir_info['url']);
$f = File::getKV('id', $r->file_id);
if ($f instanceof File) {
$redir->file = $f;
$redir->redir_url = $f->url;
} else {
// Invalid entry in File_redirection, delete and run again
common_log(
LOG_ERR,
'Could not find File with id=' . $r->file_id . ' referenced in File_redirection, deleting File_redirection entry and trying again...'
);
$r->delete();
return self::where($in_url);
}
} catch (NoResultException $e) {
// save the file now when we know that we don't have it in File_redirection
try {
$redir->file = File::saveNew($redir_info, $redir_info['url']);
} catch (ServerException $e) {
common_log(LOG_ERR, $e);
}
}
// If this is a redirection and we have a file to redirect to, save it
// (if it doesn't exist in File_redirection already)
if ($redir->file instanceof File && $redir_info['url'] != $in_url) {
try {
$file_redir = File_redirection::getByUrl($in_url);
} catch (NoResultException $e) {
$file_redir = new File_redirection();
$file_redir->urlhash = File::hashurl($in_url);
$file_redir->url = $in_url;
$file_redir->file_id = $redir->file->getID();
$file_redir->insert();
$file_redir->redir_url = $redir->file->url;
}
$file_redir->file = $redir->file;
return $file_redir;
}
}
return $redir;
}
/**
* Shorten a URL with the current user's configured shortening
* options, if applicable.
*
* If it cannot be shortened or the "short" URL is longer than the
* original, the original is returned.
*
* If the referenced item has not been seen before, embedding data
* may be saved.
*
* @param string $long_url
* @param User $user whose shortening options to use; defaults to the current web session user
* @return string
*/
public static function makeShort($long_url, $user = null)
{
$canon = File_redirection::_canonUrl($long_url);
$short_url = File_redirection::_userMakeShort($canon, $user);
// Did we get one? Is it shorter?
return !empty($short_url) ? $short_url : $long_url;
}
/**
* Shorten a URL with the current user's configured shortening
* options, if applicable.
*
* If it cannot be shortened or the "short" URL is longer than the
* original, the original is returned.
*
* If the referenced item has not been seen before, embedding data
* may be saved.
*
* @param string $long_url
* @return string
*/
public static function forceShort($long_url, $user)
{
$canon = File_redirection::_canonUrl($long_url);
$short_url = File_redirection::_userMakeShort($canon, $user, true);
// Did we get one? Is it shorter?
return !empty($short_url) ? $short_url : $long_url;
}
public static function _userMakeShort($long_url, User $user = null, $force = false)
{
$short_url = common_shorten_url($long_url, $user, $force);
if (!empty($short_url) && $short_url != $long_url) {
$short_url = (string)$short_url;
// store it
try {
$file = File::getByUrl($long_url);
} catch (NoResultException $e) {
// Check if the target URL is itself a redirect...
// This should already have happened in processNew in common_shorten_url()
$redir = File_redirection::where($long_url);
$file = $redir->file;
}
// Now we definitely have a File object in $file
try {
$file_redir = File_redirection::getByUrl($short_url);
} catch (NoResultException $e) {
$file_redir = new File_redirection();
$file_redir->urlhash = File::hashurl($short_url);
$file_redir->url = $short_url;
$file_redir->file_id = $file->getID();
$file_redir->insert();
}
return $short_url;
}
return null;
}
/**
* Basic attempt to canonicalize a URL, cleaning up some standard variants
* such as funny syntax or a missing path. Used internally when cleaning
* up URLs for storage and following redirect chains.
*
* Note that despite being on File_redirect, this function DOES NOT perform
* any dereferencing of redirects.
*
* @param string $in_url input URL
* @param string $default_scheme if given a bare link; defaults to 'http://'
* @return string
*/
public static function _canonUrl($in_url, $default_scheme = 'http://')
{
if (empty($in_url)) {
return false;
}
$out_url = $in_url;
$p = parse_url($out_url);
if (empty($p['host']) || empty($p['scheme'])) {
list($scheme) = explode(':', $in_url, 2);
switch (strtolower($scheme)) {
case 'fax':
case 'tel':
$out_url = str_replace('.-()', '', $out_url);
break;
// non-HTTP schemes, so no redirects
case 'bitcoin':
case 'mailto':
case 'aim':
case 'jabber':
case 'xmpp':
// don't touch anything
break;
// URLs without domain name, so no redirects
case 'magnet':
// don't touch anything
break;
// URLs with coordinates, not browsable domain names
case 'geo':
// don't touch anything
break;
default:
$out_url = $default_scheme . ltrim($out_url, '/');
$p = parse_url($out_url);
if (empty($p['scheme'])) {
return false;
}
break;
}
}
if (('ftp' == $p['scheme']) || ('ftps' == $p['scheme']) || ('http' == $p['scheme']) || ('https' == $p['scheme'])) {
if (empty($p['host'])) {
return false;
}
if (empty($p['path'])) {
$out_url .= '/';
}
}
return $out_url;
}
public static function saveNew($data, $file_id, $url)
{
$file_redir = new File_redirection;
$file_redir->urlhash = File::hashurl($url);
$file_redir->url = $url;
$file_redir->file_id = $file_id;
$file_redir->redirections = intval($data['redirects']);
$file_redir->httpcode = intval($data['code']);
$file_redir->insert();
}
public static function beforeSchemaUpdate()
{
$table = strtolower(get_called_class());
$schema = Schema::get();
$schemadef = $schema->getTableDef($table);
// 2015-02-19 We have to upgrade our table definitions to have the urlhash field populated
if (isset($schemadef['fields']['urlhash']) && in_array('urlhash', $schemadef['primary key'])) {
// We already have the urlhash field, so no need to migrate it.
return;
}
echo "\nFound old $table table, upgrading it to contain 'urlhash' field...";
// We have to create a urlhash that is _not_ the primary key,
// transfer data and THEN run checkSchema
$schemadef['fields']['urlhash'] = [
'type' => 'varchar',
'length' => 64,
'not null' => true,
'description' => 'sha256 hash of the URL',
];
$schemadef['fields']['url'] = [
'type' => 'text',
'description' => 'short URL (or any other kind of redirect) for file (id)',
];
unset($schemadef['primary key']);
$schema->ensureTable($table, $schemadef);
echo "DONE.\n";
$classname = ucfirst($table);
$tablefix = new $classname;
// urlhash is hash('sha256', $url) in the File table
echo "Updating urlhash fields in $table table...";
switch (common_config('db', 'type')) {
case 'pgsql':
$url_sha256 = 'encode(sha256(CAST("url" AS bytea)), \'hex\')';
break;
case 'mysql':
$url_sha256 = 'sha2(`url`, 256)';
break;
default:
throw new ServerException('Unknown DB type selected.');
}
$tablefix->query(sprintf(
'UPDATE %1$s SET urlhash = %2$s, modified = CURRENT_TIMESTAMP;',
$tablefix->escapedTableName(),
$url_sha256
));
echo "DONE.\n";
echo "Resuming core schema upgrade...";
}
public function getFile()
{
if (!$this->file instanceof File) {
$this->file = File::getByID($this->file_id);
}
return $this->file;
}
}