~jan0sch/smederee
Clone this repository
- read-only
- read-write
-
Currently only the owner of a repository can use it via ssh.
- Downloads
- Source code (.tar.gz)
Currently only the owner of a repository can use it via ssh.
A frugal platform which is dedicated to help people build great software together.
This code is licensed under the EUPL version 1.2 or later, see the LICENSE file for details.
Etc/UTC
!For development and testing you need to install sbt. Please see CONTRIBUTING.md for details how to to contribute to the project.
During development you can start (and restart) the application via the reStart
sbt task provided by the sbt-revolver plugin.
Cloning via ssh has the following constraints currently:
scp
To make darcs use the custom port either set the SSH_PORT
environment
variable (SSH_PORT=30983 darcs clone ...
) or configure your ssh client
accordingly in ~/.ssh/config
. Something like this should be sufficient:
# Only use configured identities even if there are more available.
IdentitiesOnly yes
IdentityFile ~/.ssh/id_ed25519
Host darcs.smeder.ee
IdentityFile ~/.ssh/id_rsa
Port 30983
User darcs
With such a configuration you should be able to use darcs clone
etc.
without having to set the environment variable.
If an endpoint is intended to be used only by guests (users that are not logged in) then it should be implemented two times:
HttpRoutes
) with the intended implementationAuthedRoutes
) with:
SeeOther
) or error pageEndpoints that provide functionality for both guests and users (logged in) should have their common logic extracted into a separate function and there should be two endpoints:
AuthedRoutes
) which calls the logic functionHttpRoutes
) which calls the logic functionTests are included in the project. You can run them via the sbt task test
.
Be aware that the some tests might need a working database. These should be
tagged accordingly and can be excluded or included via the following
parameters for the test task:
--exclude-tags=NAME
(do not run any tests labelled with NAME
)--include-tags=NAME
(do only run tests labelled with NAME
)For example to not run any tests needing a database connection in the hub module:
sbt:smederee> hub/testOnly -- --exclude-tags=NeedsDatabase
Please ensure that the system requirements are met and that you have the credentials needed to access the database.
On the sbt console run the task hub/Universal/packageXzTarball
which will
create a package under modules/hub/target/universal
. This archive can be
extracted to a place of your liking. Afterwards you have to create the
file conf/application.conf
in which you configure the service.
The easiest way is to use the provided sample configuration file:
% cat application.conf.sample > application.conf
Another option is to copy the default configuration:
% cat modules/hub/src/main/resources/reference.conf > application.conf
The first and foremost settings are the ones related to the database. Please ensure that both database settings (hub and tickets) are set to proper values.
Secondly take a look at the external
configuration settings for both
hub and tickets and configure them accordingly to be able to reach them
depending on your setup. We assume that you will be running them behind a
reverse proxy which also does take care of encryption (https). Depending on
your setup you can either decide to run both endpoints under one domain or on
different subdomains e.g. hub on "example.com" and tickets on
"tickets.example.com".
The default case is that the tickets simply use the configuration from the hub
section which means you only have to care for the setup in the hub section.
The hub section has a ticket-integration
section into which you have to
enter the base uri under which the ticket endpoints are reachable. The
ticket section has a hub-integration
section with the analogue settings.
The hub section has a cookie-secret
setting that MUST be set to a randomly
generated sequence of at least 64 alphanumeric characters.
Next you might want to check the pre-configured paths and make adjustments were needed.
The application uses schemas which may reside in a single database or not but it is important that the configured database users have all necessary permissions on that schemas. This can be achieved in a multitude of ways:
CREATE SCHEMA "hub" AUTHORIZATION smederee_hub;
Please note that different users (modules) sharing the same database (which
is possible due to the schema separation) need CONNECT
and CREATE
permissions on the database.
GRANT CONNECT, CREATE ON DATABASE TO ;
It is advised to run the service behind a reverse proxy like nginx^6. The configuration is pretty straight forward and should not require any tweaking:
server {
# Your server configuration (servername, ssl, etc.)
# ...
# The actual reverse proxy configuration:
location / {
proxy_pass http://localhost:8080;
proxy_set_header X-Forwarded-Host $http_host;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
proxy_http_version 1.1;
}
}
Please note that you will have to adjust the proxy_pass
part (hostname and
port) if you're running the service on another interface or port.
colordiff
$SMEDEREE_USER/.darcs/defaults
file to contain the following line:
diff diff-command colordiff -rN -u %1 %2