jasontasker / search_locality

Search Locality Problem

Search Locality Problem

Details

As a staff researcher, you must repeatedly search through a set of plain text documents in order to do your work.

A common need is to find documents where two different topics or search terms are discussed in the same context, for example: oncology and morbidity, or virtualization and agility.

"Context" within a document can be ambiguous or ill-defined, but in this case, "context" is defined as the two search terms occurring within N words of each other irrespective of sentence, paragraph or page boundaries.

Using a programming language of your choice, develop a piece of software that can iterate over all the plain text documents in a directory and return the set of documents where two search terms occur within N words of each other.

Extra credit:

In addition to allowing two search terms, allow the user to specify phrases instead of terms, e.g. "Dr. Fowler" and "Continuous Delivery"

Requirements

• Python >= 2.7 re, os, glob, and argparse modules

Default Settings

• Min: "0" - Minimum words between contexts needed
• Max: "10" - Maximum words between contexts needed
• Context1: "context1" - Required Define first context to search for
• Context2: "context2" - Required Define second context to search for
• Directory: "." - Define directory to search for files
• Extension: "txt" - Define file extension to search for

Installation

Download the project:

git clone https://github.com/jasontasker/search_locality.git

Execution

cd search_locality
python search.py --context1 "vim tutor" --context2 "follow the instructions" --directory /usr/share/vim --min 0 --max 5 --extension txt

Usage

usage: search.py [-h] [--min MIN_WORDS] [--max MAX_WORDS] --context1 CONTEXT1
                 --context2 CONTEXT2 [--directory DIRECTORY]
                 [--extension EXTENSION]

Sample Output

This file matches the 2 contexts:
/usr/share/vim/vim73/tutor/README.txt