Trac uses a directory structure and a database for storing project data. The directory is referred to as the “environment”.
A new Trac environment is created using trac-admin's initenv:
$ trac-admin /path/to/myproject initenv
trac-admin will ask you for the name of the project and the
database connection string (explained below).
initenv, when using an svn repository, does not imply that trac-admin will perform svnadmin create for the specified repository path. You need to perform the svnadmin create prior to trac-admin initenv if you're creating a new svn repository altogether with a new trac environment, otherwise you will see a message "Warning: couldn't index the repository" when initializing the environment.
Since version 0.9, Trac supports both SQLite and PostgreSQL database backends. Preliminary support for MySQL was added in 0.10. The default is to use SQLite, which is probably sufficient for most projects. The database file is then stored in the environment directory, and can easily be backed up together with the rest of the environment.
The connection string for an SQLite database is:
sqlite:db/trac.db
where db/trac.db is the path to the database file within the Trac environment.
If you want to use PostgreSQL or MySQL instead, you'll have to use a
different connection string. For example, to connect to a PostgreSQL
database on the same machine called trac, that allows access to the
user johndoe with the password letmein, use:
postgres://johndoe:letmein@localhost/trac
Note that due to the way the above string is parsed, the "/" and "@" characters cannot be part of the password.
If PostgreSQL is running on a non-standard port (for example 9342), use:
postgres://johndoe:letmein@localhost:9342/trac
On UNIX, you might want to select a UNIX socket for the transport, either the default socket as defined by the PGHOST environment variable:
postgres://user:password@/database
or a specific one:
postgres://user:password@/database?host=/path/to/socket/dir
Note that with PostgreSQL you will have to create the database before running
trac-admin initenv.
See the PostgreSQL documentation for detailed instructions on how to administer PostgreSQL.
Generally, the following is sufficient to create a database user named tracuser, and a database named trac.
createuser -U postgres -E -P tracuser createdb -U postgres -O tracuser -E UTF8 trac
When running createuser you will be prompted for the password for the user 'tracuser'. This new user will not be a superuser, will not be allowed to create other databases and will not be allowed to create other roles. These privileges are not needed to run a trac instance. If no password is desired for the user, simply remove the -P and -E options from the createuser command. Also note that the database should be created as UTF8. LATIN1 encoding causes errors trac's use of unicode in trac. SQL_ASCII also seems to work.
Under some default configurations (debian) one will have run the createuser and createdb scripts as the postgres user. For example:
sudo su - postgres -c 'createuser -U postgres -S -D -R -E -P tracuser' sudo su - postgres -c 'createdb -U postgres -O tracuser -E UTF8 trac'
Trac uses the public schema by default but you can specify a different schema in the connection string:
postgres://user:pass@server/database?schema=yourschemaname
If you want to use MySQL instead, you'll have to use a
different connection string. For example, to connect to a MySQL
database on the same machine called trac, that allows access to the
user johndoe with the password letmein, the mysql connection string is:
mysql://johndoe:letmein@localhost:3306/trac
Since version 0.12, a single Trac environment can be connected to more than one repository. There are many different ways to connect repositories to an environment, see TracRepositoryAdmin. This page also details the various attributes that can be set for a repository (like type, url, description).
In Trac 0.12 trac-admin no longer asks questions related to repositories. Therefore, by default Trac is not connected to any source code repository, and the Browse Source toolbar item will not be displayed.
You can also explicitly disable the trac.versioncontrol.* components (which are otherwise still loaded)
[components] trac.versioncontrol.* = disabled
For some version control systems, it is possible to specify not only the path to the repository, but also a scope within the repository. Trac will then only show information related to the files and changesets below that scope. The Subversion backend for Trac supports this; for other types, check the corresponding plugin's documentation.
Example of a configuration for a Subversion repository used as the default repository:
[trac] repository_type = svn repository_dir = /path/to/your/repository
The configuration for a scoped Subversion repository would be:
[trac] repository_type = svn repository_dir = /path/to/your/repository/scope/within/repos
An environment directory will usually consist of the following files and directories:
README - Brief description of the environment.
VERSION - Contains the environment version identifier.
attachments - Attachments to wiki pages and tickets are stored here.
conf
trac.ini - Main configuration file. See TracIni.
db
trac.db - The SQLite database (if you're using SQLite).
htdocs - directory containing web resources, which can be referenced in Genshi templates using /htdocs/site/... URLs. (since 0.11)
log - default directory for log files, if logging is turned on and a relative path is given.
plugins - Environment-specific plugins (Python eggs or single file plugins, since 0.10)
templates - Custom Genshi environment-specific templates. (since 0.11)
site.html - method to customize header, footer, and style, described in TracInterfaceCustomization#SiteAppearance
This is a common beginners' mistake. It happens that the structure for a Trac environment is loosely modelled after the Subversion repository directory structure, but those are two disjoint entities and they are not and must not be located at the same place.
See also: TracAdmin, TracBackup, TracIni, TracGuide