Line wrap doesn't wrap comments

I’m inheriting a node.js app, and it’s ridden with code and comments that exceed 79 characters, so I executed the following on all files:

js-beautify -r -s 2 -a -j -B -w 79 -n --good-stuff <file>

When I run this on all files, I’m finding that there’s 379 lines in comments that aren’t wrapped in the application:

sed -n '/\(.\)\{80\}/p' <file>

If I put this into a file test.js (and set the screen width to 79):

/*
 * This is a test line that is more than seventy-nine characters in length, inc
luding the comment start and period.
 * This is another line that is less than seventy-nine characters in length.
 *
 */

var test = 'This is a test string that is more than seventy characters in lengt
h, even without the var declaration.';

And then I run js-beautify -r -s 2 -a -j -B -w 79 -n --good-stuff test.js then the following is the modified file:

/*
 * This is a test line that is more than seventy-nine characters in length, inc
luding the comment start and period.
* This is another line that is less than seventy-nine characters in length.
*
*/

var test =
  'This is a test string that is more than seventy characters in length, even w
ithout the var declaration.';

It appears that long strings don’t get broken up either.

Is this intentional or a bug?

[bitwiseman]

Hey, thanks for asking.

I agree #421 covers your feature request for comment formatting, but (as you pointed out) not the behavior you described for string formatting. Leaving this open makes.

(I’ve moved your examples into the first comment and cleaned up the thread a bit. )

Unfortunately, the beautifier uses a rather simplistic engine and focuses on rearranging whitespace around tokens, rather than on rearranging tokens themselves. The string formatting you suggest is more involved requiring rewriting of tokens themselves, not just rearranging whitespace. So, it is out of scope at this time. If the beautifier reaches sufficient sophistication to be able to do token modification and we get the line wrapping from “best effort” to fully baked, this would certainly be one of the first token modification features I’d look at adding.

Thanks!

[karma0]

Perhaps a desired output could be:

/*
 * This is a test line that is more than seventy-nine characters in length,
 * including the comment start and period.
 * This is another line that is less than seventy-nine characters in length.
 *
 */

var test =
  'This is a test string that is more than seventy characters in length, '+
  'even without the var declaration.';

Or similar.