indefero/src/IDF/Scm.php
Thomas Keller 39c29dbe10 Started on issue 544, extended commit details
* Scm.php: new SCM method "getChanges" which returns all available
  change information grouped by type
* Monotone.php: implement getChanges via get_revision
* <other scms>: rename "changes" member for getCommit to "diff" which
  matches better
* Source.php: query the commit's changes and set them in the template
* commit.html: render the changes, type-by-type. Link to the tree or
  the individual diff if applicable
* styles.css: some initial style sheet work
2010-12-08 01:48:26 +01:00

463 lines
13 KiB
PHP

<?php
/* -*- tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
/*
# ***** BEGIN LICENSE BLOCK *****
# This file is part of InDefero, an open source project management application.
# Copyright (C) 2008 Céondo Ltd and contributors.
#
# InDefero is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# InDefero 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 General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
#
# ***** END LICENSE BLOCK ***** */
/**
* Manage differents SCM systems.
*
* This is the base class with the different required methods to be
* implemented by the SCMs. Each SCM backend need to extend this
* class. We are not using an interface because this is not really
* needed.
*
* The philosophy behind the interface is not to provide a wrapper
* around the different SCMs but to provide methods to retrieve in the
* most efficient way the informations to be displayed/needed in the
* web interface. This means that each SCM can use the best options,
* including caching to retrieve the informations.
*
* Note on caching: You must not cache ephemeral information like the
* changelog, but you can cache the commit info (except with
* subversion where you can change commit info...). It is ok to do
* some caching for the lifetime of the IDF_Scm object, for example
* not to retrieve several times the list of branches, etc.
*
* All the output of the methods must be serializable. This means that
* if you are parsing XML you need to correctly cast the results as
* string when needed.
*/
class IDF_Scm
{
/**
* String template for consistent error messages.
*/
public $error_tpl = 'Error command "%s" returns code %d and output: %s';
/**
* Path to the repository.
*/
public $repo = '';
/**
* Corresponding project object.
*/
public $project = null;
/**
* Cache storage.
*
* It must only be used to store data for the lifetime of the
* object. For example if you need to get the list of branches in
* several functions, better to try to get from the cache first.
*/
protected $cache = array();
/**
* Returns an instance of the correct scm backend object.
*
* @param IDF_Project
* @return Object
*/
public static function get($project)
{
// Get scm type from project conf ; defaults to git
// We will need to cache the factory
$scm = $project->getConf()->getVal('scm', 'git');
$scms = Pluf::f('allowed_scm');
return call_user_func(array($scms[$scm], 'factory'), $project);
}
/**
* Run exec and log some information.
*
* @param $caller Calling method
* @param $cmd Command to run
* @param &$out Array of output
* @param &$return Return value
* @return string Last line of the command
*/
public static function exec($caller, $cmd, &$out=null, &$return=null)
{
Pluf_Log::stime('timer');
$ret = exec($cmd, $out, $return);
Pluf_Log::perf(array($caller, $cmd, Pluf_Log::etime('timer', 'total_exec')));
Pluf_Log::debug(array($caller, $cmd, $out));
Pluf_Log::inc('exec_calls');
return $ret;
}
/**
* Run shell_exec and log some information.
*
* @param $caller Calling method
* @param $cmd Command to run
* @return string The output
*/
public static function shell_exec($caller, $cmd)
{
Pluf_Log::stime('timer');
$ret = shell_exec($cmd);
Pluf_Log::perf(array($caller, $cmd, Pluf_Log::etime('timer', 'total_exec')));
Pluf_Log::debug(array($caller, $cmd, $ret));
Pluf_Log::inc('exec_calls');
return $ret;
}
/**
* Return the size of the repository in bytes.
*
* @return int Size in byte, -1 if the size cannot be evaluated.
*/
public function getRepositorySize()
{
return -1;
}
/**
* Returns the URL of the git daemon.
*
* @param IDF_Project
* @return string URL
*/
public static function getAnonymousAccessUrl($project)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns the URL for SSH access
*
* @param IDF_Project
* @param Pluf_User
* @return string URL
*/
public static function getAuthAccessUrl($project, $user)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Check if the backend is available for display.
*
* @return bool Available
*/
public function isAvailable()
{
throw new Pluf_Exception_NotImplemented();
}
const REVISION_VALID = 0;
const REVISION_INVALID = 1;
const REVISION_AMBIGUOUS = 2;
/**
* Check if a revision or commit is valid, invalid or ambiguous.
*
* @param string Revision or commit
* @return int One of REVISION_VALID, REVISION_INVALID or REVISION_AMBIGIOUS
*/
public function validateRevision($rev)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns an array of single commit objects for ambiguous commit identifiers
*
* @param string Ambiguous commit identifier
* @return array of objects
*/
public function disambiguateRevision($commit)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns in which branches a commit/path is.
*
* A commit can be in several branches and some of the SCMs are
* managing branches using subfolders (like Subversion).
*
* This means that to know in which branch we are at the moment,
* one needs to have both the path and the commit.
*
* @param string Commit
* @param string Path
* @return array Branches
*/
public function inBranches($commit, $path)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns the list of branches.
*
* The return value must be a branch indexed array with the
* optional path to access the branch as value. For example with
* git you would get (note that some people are using / in the
* name of their git branches):
*
* <pre>
* array('master' => '',
* 'foo-branch' => '',
* 'design/feature1' => '')
* </pre>
*
* But with Subversion, as the branches are managed as subfolder
* with a special folder for trunk, you would get something like:
*
* <pre>
* array('trunk' => 'trunk',
* 'foo-branch' => 'branches/foo-branch',)
* </pre>
*
* @return array Branches
*/
public function getBranches()
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns the list of tags.
*
* The format is the same as for the branches.
*
* @see self::getBranches()
*
* @return array Tags
*/
public function getTags()
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns in which tags a commit/path is.
*
* A commit can be in several tags and some of the SCMs are
* managing tags using subfolders (like Subversion).
*
* This means that to know in which tag we are at the moment,
* one needs to have both the path and the commit.
*
* @param string Commit
* @param string Path
* @return array Tags
*/
public function inTags($commit, $path)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns the main branch.
*
* The main branch is the one displayed by default. For example
* master, trunk or tip.
*
* @return string
*/
public function getMainBranch()
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns the list of files in a given folder.
*
* The list is an array of standard class objects with attributes
* for each file/directory/external element.
*
* This is the most important method of the SCM backend as this is
* the one conveying the speed feeling of the application. All the
* dirty optimization tricks are allowed there.
*
* @param string Revision or commit
* @param string Folder ('/')
* @param string Branch (null)
* @return array
*/
public function getTree($rev, $folder='/', $branch=null)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Get commit details.
*
* @param string Commit or revision number
* @param bool Get commit diff (false)
* @return stdClass
*/
public function getCommit($commit, $getdiff=false)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Returns all recorded changes which lead to the particular commit
* or revision.
*
* Example output:
*
* stdClass object {
* 'additions' => array('path/to/file', 'path/to/directory', ...),
* 'deletions' => array('path/to/file', 'path/to/directory', ...),
* 'renames' => array('old/path/to/file' => 'new/path/to/file', ...)
* 'patches' => array('path/to/file', ...),
* 'properties' => array('path/to/file' => array(
* 'propname' => 'propvalue', 'deletedprop' => null, ...)
* ),
* ...)
* }
*
* Each member of the returned object is mandatory, but may contain
* an empty array if no changes were recorded.
*
* @param string A commit identifier
* @return object with arrays of individual changes
*/
public function getChanges($commit)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Get latest changes.
*
* It default to the main branch. If possible you should code in a
* way to avoid repetitive calls to getCommit. Try to be
* efficient.
*
* @param string Branch (null)
* @param int Number of changes (25)
* @return array List of commits
*/
public function getChangeLog($branch=null, $n=10)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Given the string describing the author from the log find the
* author in the database.
*
* If the input is an array, it will return an array of results.
*
* @param mixed string/array Author
* @return mixed Pluf_User or null or array
*/
public function findAuthor($author)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Given a revision and a file path, retrieve the file content.
*
* The $cmd_only parameter is to only request the command that is
* used to get the file content. This is used when downloading a
* file at a given revision as it can be passed to a
* Pluf_HTTP_Response_CommandPassThru reponse. This allows to
* stream a large response without buffering it in memory.
*
* The file definition is coming from getPathInfo().
*
* @see self::getPathInfo()
*
* @param stdClass File definition
* @param bool Returns command only (false)
* @return string File content
*/
public function getFile($def, $cmd_only=false)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Get information about a file or a path.
*
* @param string File or path
* @param string Revision (null)
* @return mixed False or stdClass with info
*/
public function getPathInfo($file, $rev=null)
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Given a revision and possible path returns additional properties.
*
* @param string Revision
* @param string Path ('')
* @return mixed null or array of properties
*/
public function getProperties($rev, $path='')
{
return null;
}
/**
* Generate a zip archive at a given commit, wrapped in a HTTP response, suitable for pushing to client.
*
* @param string Commit
* @param string Prefix ('repository/')
* @return Pluf_HTTP_Response The HTTP Response containing the zip archive
*/
public function getArchiveStream($commit, $prefix='repository/')
{
throw new Pluf_Exception_NotImplemented();
}
/**
* Sync the changes in the repository with the timeline.
*
*/
public static function syncTimeline($project, $force=false)
{
$cache = Pluf_Cache::factory();
$key = 'IDF_Scm:'.$project->shortname.':lastsync';
if ($force or null === ($res=$cache->get($key))) {
$scm = IDF_Scm::get($project);
if ($scm->isAvailable()) {
foreach ($scm->getChangeLog($scm->getMainBranch(), 25) as $change) {
IDF_Commit::getOrAdd($change, $project);
}
$cache->set($key, true, (int)(Pluf::f('cache_timeout', 300)/2));
}
}
}
/**
* Given a path, encode everything but the /
*/
public static function smartEncode($path)
{
return str_replace('%2F', '/', rawurlencode($path));
}
}