Getting Started

Installation

Create a Python virtual environment, and then issue the command:

pip install fizzysearch

Unfortunately, this is not the whole story...

This works when only using the Bloomtyper functionality, but not when you would like to use the fill SPARQL rewriting for the various search systems. Until we have a more documentation, you can look at the example Dockerfile

To build it, you can say:

docker build -t fizzysearch .

And then to run it, you could for example do something like:

docker run --rm -it -v $(PWD):/data -e FTS_SQLITE_PATH=/data/example.db -e INPUT_FILEPATH=/data/  fizzysearch

This will create a FTS index named example.db in your current directory, after looking for files to index from the directory that you ran the command in.

Running the tests

See the test file for some examples on how to build an index and run some test queries.

Examples

Let's say you have a n-triples files in the current directory name foobar.nt. To create a fulltext index of all the literals in this file, you can issue the command in the shell:

FTS_SQLITE_PATH=example.db python -m fizzysearch

If all goes well, this should read your file named foobar.nt, index it, and store the results in a SQLite file named example.db

Note

Running on the command line will 'walk' the entire current directory, and all sub-directories if you do not explicitly specify a path. So if you have many .nt files in the current directory, it will try to index them all!

Now that you have a fulltext index for your n-triple file, you could use it in a system like SHMARQL to query the file easily.