.. _setup:

Setup

=====

Setting up the application

--------------------------

First You'll ned to create RhodeCode config file. Run the following command 

to do this

::

 paster make-config RhodeCode production.ini

- This will create `production.ini` config inside the directory

  this config contains various settings for RhodeCode, e.g proxy port, 

  email settings, usage of static files, cache, celery settings and logging.

Next we need to create the database.

::

 paster setup-app production.ini

- This command will create all needed tables and an admin account. 

  When asked for a path You can either use a new location of one with already 

  existing ones. RhodeCode will simply add all new found repositories to 

  it's database. Also make sure You specify correct path to repositories.

- Remember that the given path for mercurial_ repositories must be write 

  accessible for the application. It's very important since RhodeCode web 

  interface will work even without such an access but, when trying to do a 

  push it'll eventually fail with permission denied errors. 

You are ready to use rhodecode, to run it simply execute

::

 paster serve production.ini

- This command runs the RhodeCode server the app should be available at the 

  127.0.0.1:5000. This ip and port is configurable via the production.ini 

  file created in previous step

- Use admin account you created to login.

- Default permissions on each repository is read, and owner is admin. So 

  remember to update these if needed. In the admin panel You can toggle ldap,

  anonymous, permissions settings. As well as edit more advanced options on 

  users and repositories

Setting up Whoosh full text search

----------------------------------

to enable full index rebuild. Without that indexing will run always in in

incremental mode.

 paster make-index production.ini --repo-location=<location for repos> 

 paster make-index production.ini -f --repo-location=<location for repos>

::

 /path/to/python/bin/paster /path/to/rhodecode/production.ini --repo-location=<location for repos> 

When using incremental(default) mode whoosh will check last modification date 

of each file and add it to reindex if newer file is available. Also indexing 

daemon checks for removed files and removes them from index. 

Sometime You might want to rebuild index from scratch. You can do that using 

the `-f` flag passed to paster command or, in admin panel You can check 

`build from scratch` flag.

Setting up LDAP support

-----------------------

RhodeCode starting from version 1.1 supports ldap authentication. In order

to use ldap, You have to install python-ldap package. This package is available

via pypi, so You can install it by running

::

 easy_install python-ldap

::

 pip install python-ldap

.. note::

   python-ldap requires some certain libs on Your system, so before installing 

   it check that You have at least `openldap`, and `sasl` libraries.

ldap settings are located in admin->ldap section,

Here's a typical ldap setup::

 Enable ldap  = checked                 #controls if ldap access is enabled

 Host         = host.domain.org         #actual ldap server to connect

 Port         = 389 or 689 for ldaps    #ldap server ports

 Enable LDAPS = unchecked               #enable disable ldaps

 Account      = <account>               #access for ldap server(if required)

 Password     = <password>              #password for ldap server(if required)

 Base DN      = uid=%(user)s,CN=users,DC=host,DC=domain,DC=org

`Account` and `Password` are optional, and used for two-phase ldap 

authentication so those are credentials to access Your ldap, if it doesn't 

support anonymous search/user lookups. 

Base DN must have %(user)s template inside, it's a placer where Your uid used

to login would go, it allows admins to specify not standard schema for uid 

variable

If all data are entered correctly, and `python-ldap` is properly installed

Users should be granted to access RhodeCode wit ldap accounts. When 

logging at the first time an special ldap account is created inside RhodeCode, 

so You can control over permissions even on ldap users. If such user exists 

already in RhodeCode database ldap user with the same username would be not 

able to access RhodeCode.

If You have problems with ldap access and believe You entered correct 

information check out the RhodeCode logs,any error messages sent from 

ldap will be saved there.

Setting Up Celery

-----------------

Since version 1.1 celery is configured by the rhodecode ini configuration files

simply set use_celery=true in the ini file then add / change the configuration 

variables inside the ini file.

Remember that the ini files uses format with '.' not with '_' like celery

so for example setting `BROKER_HOST` in celery means setting `broker.host` in

the config file.

In order to make start using celery run::

 paster celeryd <configfile.ini>

.. note::

   Make sure You run this command from same virtualenv, and with the same user

   that rhodecode runs.

Nginx virtual host example

--------------------------

Sample config for nginx using proxy::

 server {

    listen          80;

    server_name     hg.myserver.com;

import os

import sys

import traceback

from os.path import dirname as dn, join as jn

#to get the rhodecode import

sys.path.append(dn(dn(dn(os.path.realpath(__file__)))))

from rhodecode.model import init_model

from rhodecode.model.scm import ScmModel

from rhodecode.config.environment import load_environment

from rhodecode.lib.utils import BasePasterCommand, Command, add_cache

from shutil import rmtree

from webhelpers.html.builder import escape

from vcs.utils.lazy import LazyProperty

from sqlalchemy import engine_from_config

from whoosh.analysis import RegexTokenizer, LowercaseFilter, StopFilter

from whoosh.fields import TEXT, ID, STORED, Schema, FieldType

from whoosh.index import create_in, open_dir

from whoosh.formats import Characters

from whoosh.highlight import highlight, SimpleFragmenter, HtmlFormatter

#EXTENSIONS WE WANT TO INDEX CONTENT OFF

INDEX_EXTENSIONS = ['action', 'adp', 'ashx', 'asmx', 'aspx', 'asx', 'axd', 'c',

                    'cfg', 'cfm', 'cpp', 'cs', 'css', 'diff', 'do', 'el', 'erl',

                    'h', 'htm', 'html', 'ini', 'java', 'js', 'jsp', 'jspx', 'lisp',

                    'lua', 'm', 'mako', 'ml', 'pas', 'patch', 'php', 'php3',

                    'php4', 'phtml', 'pm', 'py', 'rb', 'rst', 's', 'sh', 'sql',

                    'tpl', 'txt', 'vim', 'wss', 'xhtml', 'xml', 'xsl', 'xslt',

                    'yaws']

#CUSTOM ANALYZER wordsplit + lowercase filter

ANALYZER = RegexTokenizer(expression=r"\w+") | LowercaseFilter()

#INDEX SCHEMA DEFINITION

SCHEMA = Schema(owner=TEXT(),

                repository=TEXT(stored=True),

                path=TEXT(stored=True),

                content=FieldType(format=Characters(ANALYZER),

                             scorable=True, stored=True),

                modtime=STORED(), extension=TEXT(stored=True))

IDX_NAME = 'HG_INDEX'

FORMATTER = HtmlFormatter('span', between='\n<span class="break">...</span>\n')

FRAGMENTER = SimpleFragmenter(200)

class MakeIndex(BasePasterCommand):

    max_args = 1

    min_args = 1

    usage = "CONFIG_FILE"

    summary = "Creates index for full text search given configuration file"

    group_name = "RhodeCode"

    takes_config_file = -1

    parser = Command.standard_parser(verbose=True)

    def command(self):

        from pylons import config

        add_cache(config)

        engine = engine_from_config(config, 'sqlalchemy.db1.')

        init_model(engine)

        index_location = config['index_dir']

        repo_location = self.options.repo_location

        #======================================================================

        # WHOOSH DAEMON

        #======================================================================

        from rhodecode.lib.pidlock import LockHeld, DaemonLock

        from rhodecode.lib.indexers.daemon import WhooshIndexingDaemon

        try:

            l = DaemonLock()

            WhooshIndexingDaemon(index_location=index_location,

                .run(full_index=self.options.full_index)

            l.release()

        except LockHeld:

            sys.exit(1)

    def update_parser(self):

        self.parser.add_option('--repo-location',

                          action='store',

                          dest='repo_location',

                          help="Specifies repositories location to index REQUIRED",

                          )

        self.parser.add_option('-f',

                          action='store_true',

                          dest='full_index',

                          help="Specifies that index should be made full i.e"

                                " destroy old and build from scratch",

                          default=False)

class ResultWrapper(object):

    def __init__(self, search_type, searcher, matcher, highlight_items):

        self.search_type = search_type

        self.searcher = searcher

        self.matcher = matcher

        self.highlight_items = highlight_items

        self.fragment_size = 200 / 2

    @LazyProperty

    def doc_ids(self):

        docs_id = []

        while self.matcher.is_active():

            docnum = self.matcher.id()

            chunks = [offsets for offsets in self.get_chunks()]

            docs_id.append([docnum, chunks])

            self.matcher.next()

        return docs_id

    def __str__(self):

        return '<%s at %s>' % (self.__class__.__name__, len(self.doc_ids))

    def __repr__(self):

        return self.__str__()

    def __len__(self):

        return len(self.doc_ids)

    def __iter__(self):

        """

        Allows Iteration over results,and lazy generate content

        *Requires* implementation of ``__getitem__`` method.

        """

        for docid in self.doc_ids:

            yield self.get_full_content(docid)

    def __getslice__(self, i, j):

        """

        Slicing of resultWrapper

        """

        slice = []

        for docid in self.doc_ids[i:j]:

            slice.append(self.get_full_content(docid))

        return slice

    def get_full_content(self, docid):

        res = self.searcher.stored_fields(docid[0])

        f_path = res['path'][res['path'].find(res['repository']) \

                             + len(res['repository']):].lstrip('/')

        content_short = self.get_short_content(res, docid[1])

        res.update({'content_short':content_short,

                    'content_short_hl':self.highlight(content_short),

                    'f_path':f_path})

        return res

    def get_short_content(self, res, chunks):

        return ''.join([res['content'][chunk[0]:chunk[1]] for chunk in chunks])

    def get_chunks(self):

        """

        Smart function that implements chunking the content

        but not overlap chunks so it doesn't highlight the same

        close occurrences twice.

        @param matcher:

        @param size:

        """

        memory = [(0, 0)]

        for span in self.matcher.spans():

            start = span.startchar or 0

            end = span.endchar or 0

            start_offseted = max(0, start - self.fragment_size)

            end_offseted = end + self.fragment_size

            if start_offseted < memory[-1][1]:

                start_offseted = memory[-1][1]

            memory.append((start_offseted, end_offseted,))

            yield (start_offseted, end_offseted,)

    def highlight(self, content, top=5):

        if self.search_type != 'content':

            return ''

        hl = highlight(escape(content),

                 self.highlight_items,

                 analyzer=ANALYZER,

                 fragmenter=FRAGMENTER,

# -*- coding: utf-8 -*-

"""

    rhodecode.lib.indexers.daemon

    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    A deamon will read from task table and run tasks

    :created_on: Jan 26, 2010

    :author: marcink

    :copyright: (C) 2009-2010 Marcin Kuzminski <marcin@python-works.com>    

    :license: GPLv3, see COPYING for more details.

"""

# This program 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; version 2

# of the License or (at your opinion) any later version of the license.

# 

# This program 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 Street, Fifth Floor, Boston,

# MA  02110-1301, USA.

import sys

import os

import traceback

from os.path import dirname as dn

from os.path import join as jn

#to get the rhodecode import

project_path = dn(dn(dn(dn(os.path.realpath(__file__)))))

sys.path.append(project_path)

from rhodecode.model.scm import ScmModel

from rhodecode.lib.helpers import safe_unicode

from whoosh.index import create_in, open_dir

from shutil import rmtree

from rhodecode.lib.indexers import INDEX_EXTENSIONS, SCHEMA, IDX_NAME

from time import mktime

from vcs.exceptions import ChangesetError, RepositoryError

import logging

log = logging.getLogger('whooshIndexer')

# create logger

log.setLevel(logging.DEBUG)

log.propagate = False

# create console handler and set level to debug

ch = logging.StreamHandler()

ch.setLevel(logging.DEBUG)

# create formatter

formatter = logging.Formatter("%(asctime)s - %(name)s - %(levelname)s - %(message)s")

# add formatter to ch

ch.setFormatter(formatter)

# add ch to logger

log.addHandler(ch)

class WhooshIndexingDaemon(object):

    """

    Deamon for atomic jobs

    """

    def __init__(self, indexname='HG_INDEX', index_location=None,

        self.indexname = indexname

        self.index_location = index_location

        if not index_location:

            raise Exception('You have to provide index location')

        self.repo_location = repo_location

        if not repo_location:

            raise Exception('You have to provide repositories location')

        self.repo_paths = ScmModel(sa).repo_scan(self.repo_location, None)

        self.initial = False

        if not os.path.isdir(self.index_location):

            os.makedirs(self.index_location)

            log.info('Cannot run incremental index since it does not'

                     ' yet exist running full build')

            self.initial = True

    def get_paths(self, repo):

        """recursive walk in root dir and return a set of all path in that dir

        based on repository walk function

        """

        index_paths_ = set()

        try:

            for topnode, dirs, files in repo.walk('/', 'tip'):

                for f in files:

                    index_paths_.add(jn(repo.path, f.path))

                for dir in dirs:

                    for f in files:

                        index_paths_.add(jn(repo.path, f.path))

        except RepositoryError, e:

            log.debug(traceback.format_exc())

            pass

        return index_paths_

    def get_node(self, repo, path):

        n_path = path[len(repo.path) + 1:]

        node = repo.get_changeset().get_node(n_path)

        return node

    def get_node_mtime(self, node):

        return mktime(node.last_changeset.date.timetuple())

    def add_doc(self, writer, path, repo):

        """Adding doc to writer this function itself fetches data from

        the instance of vcs backend"""

        node = self.get_node(repo, path)

        #we just index the content of chosen files, and skip binary files

        if node.extension in INDEX_EXTENSIONS and not node.is_binary:

            u_content = node.content

            if not isinstance(u_content, unicode):

                log.warning('  >> %s Could not get this content as unicode '

                          'replacing with empty content', path)

                u_content = u''

            else:

                log.debug('    >> %s [WITH CONTENT]' % path)

        else:

            log.debug('    >> %s' % path)

            #just index file name without it's content

            u_content = u''

        writer.add_document(owner=unicode(repo.contact),

                        repository=safe_unicode(repo.name),

                        path=safe_unicode(path),

                        content=u_content,

                        modtime=self.get_node_mtime(node),

                        extension=node.extension)

    def build_index(self):

        if os.path.exists(self.index_location):

            log.debug('removing previous index')

            rmtree(self.index_location)

        if not os.path.exists(self.index_location):

            os.mkdir(self.index_location)

        idx = create_in(self.index_location, SCHEMA, indexname=IDX_NAME)

        writer = idx.writer()

            log.debug('building index @ %s' % repo.path)

            for idx_path in self.get_paths(repo):

                self.add_doc(writer, idx_path, repo)

        log.debug('>> COMMITING CHANGES <<')

        writer.commit(merge=True)

        log.debug('>>> FINISHED BUILDING INDEX <<<')

    def update_index(self):

        log.debug('STARTING INCREMENTAL INDEXING UPDATE')

        idx = open_dir(self.index_location, indexname=self.indexname)

        # The set of all paths in the index

        indexed_paths = set()

        # The set of all paths we need to re-index

        to_index = set()

        reader = idx.reader()

        writer = idx.writer()

        # Loop over the stored fields in the index

        for fields in reader.all_stored_fields():

            indexed_path = fields['path']

            indexed_paths.add(indexed_path)

            repo = self.repo_paths[fields['repository']]

            try:

                node = self.get_node(repo, indexed_path)

            except ChangesetError:

                # This file was deleted since it was indexed

                log.debug('removing from index %s' % indexed_path)

                writer.delete_by_term('path', indexed_path)

            else:

                # Check if this file was changed since it was indexed

                indexed_time = fields['modtime']

                mtime = self.get_node_mtime(node)

                if mtime > indexed_time:

                    # The file has changed, delete it and add it to the list of

                    # files to reindex

                    log.debug('adding to reindex list %s' % indexed_path)

                    writer.delete_by_term('path', indexed_path)

                    to_index.add(indexed_path)

        # Loop over the files in the filesystem

        # Assume we have a function that gathers the filenames of the

        # documents to be indexed

        for repo in self.repo_paths.values():

            for path in self.get_paths(repo):

                if path in to_index or path not in indexed_paths:

                    # This is either a file that's changed, or a new file

                    # that wasn't indexed before. So index it!

                    self.add_doc(writer, path, repo)

                    log.debug('re indexing %s' % path)

        log.debug('>> COMMITING CHANGES <<')

        writer.commit(merge=True)

        log.debug('>>> FINISHED REBUILDING INDEX <<<')

    def run(self, full_index=False):

        """Run daemon"""

        if full_index or self.initial:

            self.build_index()

        else:

            self.update_index()