js-cli

Command line script to transform stdout from shell scripts using a JS lambda or expression. Inspired by fx but meant to be more general for parsing either raw text or JSON. CSV, TOML and YAML support is provided in the scope of the lambda or expression by the corresponding global variables CSV, TOML and YAML, each offering a similar parse/stringify interface as that of the native JSON (with some caveats, refer to example code below to see where parse or stringify might accept a second options argument for a given data type).

Installation

npm i -g @ryantate/js-cli

Usage

js (-s | -h | -v) 'handler_code'
    -s | --stream - Streaming mode, handle input line by line
    -h | --help - Show this help message and quit
    -v | --version Show the version number and quit

Pipe data in to js and supply a lambda or expression to parse it with. The invoked expression or lambda will have this bound to either all of stdin when processing line by line, or the individual line when in streaming mode. The value returned from a lambda or the result of an inline expression will be logged to the console. Undefined or null values will not be logged. Promises returned will be resolved and then logged.

Packages installed either globally or in the current working directory are accessible via require in this scope (see examples below).

Examples

Working with CSV, TOML or YAML data

Powered by the Papa Parse, @ltd/j-toml and yaml libraries respectively. Refer to external documentation for specifics as to their functionality. Each library is available to handler functions as the global variables CSV, TOML and YAML. The CSV global is actually a wrapper around Papa Parse to enable header parsing by default for csv data and to offer a stringify function (vs. the upstream unparse), bringing it in line with the interface available in JSON and the other supported data formats.

# csv with header row
cat file.csv | js 'CSV.parse(this)'
# csv without headers
cat file.csv | js 'CSV.parse(this, {headers: false})'
# output csv data
cat array_of_objects.json | js 'CSV.stringify(JSON.parse(this))' > out.csv # keys of first object become header row
cat array_of_arrays.json | js 'CSV.stringify(JSON.parse(this))' > out.csv # no headers in output csv

# reading TOML data
cat Cargo.toml | js 'TOML.parse(this)'
# outputting TOML
curl my.api | js 'TOML.stringify(JSON.parse(this), {newline: "\n"})' # note that it is necessary to specify newline character

# single yaml document
cat file.yaml | js 'YAML.parse(this)'
# multi-document yaml
cat multi.yaml | js 'YAML.parseAllDocuments(this).map(doc => doc.toJSON())'
# outputting YAML
cat data.csv | js 'YAML.stringify(CSV.parse(this))'

Sum all numbers in a file

js 'this.trim().split(/\s/+).map(Number).reduce((a,b) => a+b, 0)' < file_of_numbers.txt

Curl-like

npm i -g node-fetch
cat urls.txt | js --stream 'require("node-fetch")(this).then(res => res.json())'

HTML Parsing

npm i -g cheerio
curl https://some_website.org | js 'html => require("cheerio").load(html).find("p").length'

Largest number per line in a CSV file

cat data.csv | js --stream 'line => Math.max(...line.split(",").map(val => Number(val) || -Infinity))'

Create a directory index

find . -type f | sort | js 'JSON.stringify(this.split("\n").filter(Boolean), null, 2)' > index.json

Update all dependencies in package.json

cat package.json | js 'Object.keys(JSON.parse(this).dependecies).forEach(d => console.log(d))' | xargs npm install

Concatenate MP3 files with ffmpeg

ffmpeg \
	-i \
	concat:$(find . -type f -name '*.mp3' | js 'this.split("\n").sort().join("|")') \
	-codec:a libmp3lame \
	concatenated.mp3

Convert CSV to JSON

cat data.csv | js 'JSON.stringify(CSV.parse(this))'