Quit Store

Build status of master branch:

The Quit Store (stands for <em>Qu</em>ads in G<em>it</em>) provides a workspace for distributed collaborative Linked Data knowledge engineering. You are able to read and write RDF Datasets (aka. multiple Named Graphs) through a standard SPARQL 1.1 Query and Update interface. To collaborate you can create multiple branches of the Dataset and share your repository with your collaborators as you know it from Git.

If you want to read more about the Quit Store we can recommend our paper:

Decentralized Collaborative Knowledge Management using Git by Natanael Arndt, Patrick Naumann, Norman Radtke, Michael Martin, and Edgard Marx in Journal of Web Semantics, 2018 [@sciencedirect] [@arXiv]

Getting Started

To get the Quit Store you have three options:

Install via pipx
Clone it with Git from our repository: https://github.com/AKSW/QuitStore
Use Docker and see the section Docker in the README
The binary self-contained releasses created with pyinstaller are currently broken #291 and #302. I'm happy about help.

Installation via pipx

$ pipx install git+https://github.com/AKSW/QuitStore.git
  installed package quit 0.25.4, installed using Python 3.12.1
  These apps are now globally available
    - quitstore
done! ✨ 🌟 ✨
$ quitstore --help

Installation from Source

Install poetry.

Get the Quit Store source code:

$ git clone https://github.com/AKSW/QuitStore.git
$ cd QuitStore

If you are using virtualenvwrapper:

$ poetry install
$ poetry run quitstore --help

Git configuration

Configure your name and email for Git. This information will be stored in each commit you are creating with Git and the Quit Store on your system. It is relevant so people know which contribution is coming from whom. Execute the following command if you haven't done that before.

$ git config --global user.name "Your Name"
$ git config --global user.email "you@e-mail-provider.org"

Start with Existing Data (Optional)

If you already have data which you want to use in the quit store follow these steps:

Create a repository which will contain your RDF data.

$ git init /path/to/repo

Put your RDF data formatted as N-Triples and sorted (e.g. using cat data-in.nt | LC_ALL=C sort -u > data-out.nt) into files like <graph>.nt into this directory.
For each <graph>.nt file create a corresponding <graph>.nt.graph file which must contain the IRI for the respective graph. (These .graph files are also used by the Virtuoso bulk loading process).
Add the data to the repository and create a commit.

$ git add …
$ git commit -m "init repository"

To ingest further versions of your data into the Quit Store you can add further commits by going through steps 2.-4.. Alternatively you are also able to execute SPARQL 1.1. Update operations to create new versions on the Quit Store.

Start the Quit Store

If you are using the binary:

$ chmod +x quit #
$ ./quit -t /path/to/repo

If you have it installed from the sources:

$ poetry run quitstore -t /path/to/repo

Open your browser and go to http://localhost:5000/.

Have a lot of fun!

For more command line options check out the section Command Line Options in the README.

Command Line Options

-b, --basepath

Specify a base path/application root. This will work with WSGI and docker only.

-t, --targetdir

Specify a target directory where the repository can be found or will be cloned (if remote is given) to.

-r, -repourl

Specify a link/URL to a remote repository.

-c, --configfile

Specify a path to a configuration file. (Defaults to ./config.ttl)

-nv, --disableversioning

Run Quit-Store without versioning activated

-f, --features

This option enables additional features of the store:

provenance - Enable browsing interfaces for provenance information.
persistance - Store all internal data as RDF graph.
garbagecollection - Enable garbage collection. With this feature enabled, git will check for garbage collection after each commit. This may slow down response time but will keep the repository size small.

-v, --verbose and -vv, --verboseverbose

Set the log level for the standard output to verbose (INFO) respective extra verbose (DEBUG).

-l, --logfile

Write the log output to the given path. The path is interpreted relative to the current working directory. The log level for the logfile is always extra verbose (DEBUG).

Configuration File

deprecated (we plan to remove the configuration file feature)

If you want to work with configuration files you can create a config.ttl file. This configuration file consists of two parts, the store configuration and the graph configuration. The store configuration manages everything related to initializing the software, the graph configuration maps graph files to their graph IRIs. The graph configuration in the config.ttl is an alternative to using <graph>.nt.graph files next to the graphs. Make sure you put the correct path to your git repository ("../store") and the IRI of your graph (<http://example.org/>) and name of the file holding this graph ("example.nt").

conf:store a <YourQuitStore> ;
    <pathOfGitRepo> "../store" ; # Set the path to the repository that contains the files .
    <origin> "git:github.com/your/repository.git" . # Optional a git repo that will be cloned into dir given in line above on startup.


conf:example a <Graph> ; # Define a Graph resource for a named graph
    <graphUri> <http://example.org/> ; # Set the IRI of named graph
    <isVersioned> 1 ; # Defaults to True, future work
    <graphFile> "example.nt" . # Set the filename

API

The Quit-Store comes with three kinds of interfaces, a SPARQL update and query interface, a provenance interface, and a Git management interface.

SPARQL Update and Query Interface

The SPARQL interface support update and select queries and is meant to adhere to the SPARQL 1.1 Protocol. You can find the interface to query the current HEAD of your repository under http://your-quit-host/sparql. To access any branch or commit on the repository you can query the endpoints under http://your-quit-host/sparql/<branchname> resp. http://your-quit-host/sparql/<commitid>. Since the software is still under development there might be some missing features or strange behavior. If you are sure that the store does not follow the W3C recommendation please file an issue.

Examples

Execute a select query with curl

curl -d "select ?s ?p ?o ?g where { graph ?g { ?s ?p ?o} }" -H "Content-Type: application/sparql-query" http://your-quit-host/sparql
curl -d "select ?s ?p ?o ?g where { graph ?g { ?s ?p ?o} }" -H "Content-Type: application/sparql-query" http://your-quit-host/sparql/develop

If you are interested in a specific result mime type you can use the content negotiation feature of the interface:

curl -d "select ?s ?p ?o ?g where { graph ?g { ?s ?p ?o} }" -H "Content-Type: application/sparql-query" -H "Accept: application/sparql-results+json" http://your-quit-host/sparql

Execute an update query with curl

curl -d "insert data { graph <http://example.org/> { <urn:a> <urn:b> <urn:c> } }" -H "Content-Type: application/sparql-update"  http://your-quit-host/sparql

Provenance Interface

To use the provenance browsing feature you have to enable it with the argument --feature=provenance. The provenance browsing feature extracts provenance meta data for the revisions and makes it available through a SPARQL endpoint and the blame interface. The provenance interface is available under the following two URLs:

http://your-quit-host/provenance which is a SPARQL query interface (see above) to query the provenance graph
http://your-quit-host/blame to get a git blame like output per statement in the store

Git Management Interface

The git management interface allows access to some operations of quit in conjunction with the underlying git repository. You can access them with your browser at the following paths.

/commits: See commits, messages, committer, and date of commits.
/branch, /merge: allows to manage branches and merge branches with different strategies.
/pull, /fetch, /push work similar to the respective git commands. (These operations will only works if you have configured remotes on the repository.)

Docker

We provide a Docker image for the Quit Store on the public docker hub as well as on the github docker registry. The image exp

QuitStore

Install / Use

README