Captain Go - Simple Fuzzy Text Searching API

For creating a very simple keyword search for mapping input strings against a predefined set. Uses Levenshtein distance (https://en.wikipedia.org/wiki/Levenshtein_distance).

./captaingo --source=my_data.json --port=8080

Usage ( Single ):

GET /match?q=London&keywords=England&keywords=United+Kingdom
Content-Type: application/x-www-form-urlencoded

Out:

{
    "total": 2,
    "matches": [
        {
            "name": "City of London",
            "id": 12345,
            "keywords": "City of London, Greater London, England, United Kingdom",
            "score": 0
        },
        ...
    ]
}

Usage ( Multi ):

POST /match
Content-Type: application/json

{
    "max_score": 1,
    "queries": [
        {
            "q": "London",
            "keywords": [ "Greater London", "England" ],
            "id": 554433
        },
        ...
    ]
}

Out:

{
    "queries": [
        {
            "input": {
                "q": "London",
                "keywords": [ "Greater London", "England" ],
                "id": 554433
            },
            "total": 2,
            "matches": [
                {
                    "name": "City of London",
                    "id": 12345,
                    "keywords": "City of London, Greater London, England, United Kingdom",
                    "score": 0
                },
                ...
            ]
        }
    ]
}

Input source should be in format

[
    {
        "name": "City of London",
        "id": 12345,
        "keywords": ["City of London", "Greater London", "England", "United Kingdom"]
    },
    ...
]

Algorithm

Matcher uses a simple levenshtein algorithm ( look in matcher/matcher.go ) but offers other option(s)... You can use the textheater levenshtein implementation (https://github.com/texttheater/golang-levenshtein) by adding the algorithm plugin into your main import

E.g.

import (
    "github.com/patmooney/captaingo/matcher"
    _ "github.com/patmooney/captaingo/matcher/algorithm/levenshtein"
);