mini-jtbl

JTable is a dead-simple and intuitive format for tables, based on JSON.

mini-jtbl is a set of utilities for the JTable ecosystem, developed as a demonstration of its power.

For example:

$ jcsv  "https://datahub.io/core/country-list/r/data.csv" | jwhere "Code.slice(0,1)=='C'" | jlimit 5 | cat
{
    "source": "https://datahub.io/core/country-list/r/data.csv",
    "header": ["Name", "Code"],
    "rows": [
        ["Cameroon", "CM"],
        ["Canada", "CA"],
        ["Cape Verde", "CV"],
        ["Central African Republic", "CF"],
        ["Chile", "CL"]
    ]
}

Install

Download the repository in a local directory, and type:

npm install mini-jtbl

To install the following utilities: jcsv, jwhere, jcut, jlimit and jout on your system

npm link

It also provides a utility jschema to test compliance of a JTable file with the JTable schema.

Example

$ jcsv "https://datahub.io/world-bank/en.pop.dnst/r/data.csv" | jwhere "Year < 1964" | jlimit 5
Apply expression on each row: Year < 1964
Limit rows: 5
| Country Name           | Country Code | Year | Value            |
| ---------------------- | ------------ | ---- | ---------------- |
| Arab World             | ARB          | 1961 | 6.97623904408321 |
| Arab World             | ARB          | 1962 | 7.16985259355326 |
| Arab World             | ARB          | 1963 | 7.37014441614233 |
| Caribbean small states | CSS          | 1961 | 10.567692687747  |
| Caribbean small states | CSS          | 1962 | 10.7651828063241 |

Documentation

mini-jtbl has its own documentation page, which is part of the documentation on the JTable format.

All command-line utilities provide a help, accessible with the options -h or --help.

Limitations

The purpose of mini-jtbl is not to become a production tool. Its purpose is to provide a minimal model for the development of similar tools.

It is workable and sometimes suprisingly resilient, but do not expect that all corner cases will be covered or that it will be expanded to cover all cases.

Also, it was never developed with efficiency or large datasets in mind.

Should you find obvious errors, or better ways to do things, please create an issue.

You are invited to contibute to the JTable ecosystem!

The success of a table format depends on the creation of an ecosystem. The JTable ecosystem starts with a protocol for command lines interfaces, based on pipes.

Anyone can develop their own filters for the JTable ecosystem, in any language: Javascript, Python, compiled languages, etc..

For a command-line filter to be part of the CLI ecosystem, the following things are usually sufficient:

1. Be able to read compliant JTable files from stdin.
2. Perform whatever transformations.
3. Be able to output compliant JTable files to stdout.
4. When the output is a TTY, then output it by default in a human-readable format.