Handling Hashbang/Shebang

In a Unix environment, a shell script often has its first line marked by a hashbang or a shebang, #!. A common example is a utility intended to be executed by Node.js, which may look like the following:

#!/usr/bin/env node
console.log('Hello from Node.js!');

If Esprima parser is being used to process the content of the above file, the parser will throw an exception. This is because that hashbang is not valid in JavaScript. A quick Node.js REPL session to illustrate the point:

$ node
> var esprima = require('esprima')
> var src = ['#!/usr/bin/env node', 'answer = 42'].join('\n')
> esprima.parseScript(src)
Error: Line 1: Unexpected token ILLEGAL

The workaround for this problem is to remove the first line completely before passing it to the parser. One way to do that is to use a regular expression, as shown below:

$ node
> var esprima = require('esprima')
> var src = ['#!/usr/bin/env node', 'answer = 42'].join('\n')
> src = src.replace(/^#!(.*\n)/, '')
'answer = 42'
> esprima.parseScript(src)
Script {
  type: 'Program',
  body: [ ExpressionStatement { type: 'ExpressionStatement', expression: [Object] } ],
  sourceType: 'script' }

Note that the above approach will shorten the source string. If the string length needs to be preserved, e.g. to facilitate an exact location mapping to the original version, then a series of whitespaces need to be padded to the beginning. A modified approach looks like the following:

$ node
> var esprima = require('esprima')
> var src = ['#!/usr/bin/env node', 'answer = 42'].join('\n')
> src = src.replace(/(^#!.*)/, function(m) { return Array(m.length + 1).join(' ') });
> esprima.parseScript(src, { range: true })
Script {
  type: 'Program',
  body:
   [ ExpressionStatement {
       type: 'ExpressionStatement',
       expression: [Object],
       range: [Object] } ],
  sourceType: 'script',
  range: [ 15, 26 ] }