booque

Library to parse scopus-like boolean queries and translate them into elasticsearch ones.

Installation

git clone https://github.com/martijnvanbeers/booque

create/activate a conda/virtualenv environment

pip install .

or when you plan to make changes:

pip install -e .

Example

cat '"foo" W/3 ("bar" OR "baz")' | booque_parse |curl -s -XPOST 'http://localhost:9200/my_index/_search' -H 'Content-Type: application/json' -d '@-' |jq '.hits.hits[]' |less

or if you want to see the resulting json:

`cat '"foo" W/3 ("bar" OR "baz")' | booque_parse

Translating the aurora and elsevier queries with the provided scripts

Aurora

• clone the auroa repository: git clone https://github.com/Aurora-Network-Global/sdg-queries

• run booque_aurora_parse --indir /path/to/sdg-queries --outdir /path/to/output_directory

  This loops over the xml files from the sdg-queries repository and writes json files containing a elasticsearch query for each target per SDG and also a combined query for each SDG

Elsevier

• Download the Elsevier data

• run booque_elsevier_parse /path/to/SDG_queries_collated_20191010.xlsx --putdir /path/to/output_directory

Common option

• each script accepts a --help option that has summary information about how to call it

• the --outdir option defaults to ./es/, so if that's an acceptable location, there is no need to specify it

• they all read $HOME/.config/booque/config.json. Currently this can only contain a fields object/dict that maps the scopus field names to the elasticsearch field names. If this file doesn't exist, it will be filled with the default mapping