Newer jslint editions has a 0 index for lines and columns

[nomego]

If you switch to edition "latest", "2015-05-08" or "es6" in the current node-jslint, the indexes for line and column in error reporting is now 0.

The default edition is still an old jslint from 2013 though, so I guess SublimeLinter-contrib-jslint needs a configuration option whether to adjust the lines and columns with one or not.

[devdoc]

Thanks for reporting. Indeed the new JSLint seems to use zero-based indexing for columns and lines. It does include a new option ("fudge", see http://www.jslint.com/help.html) that seems to provide a way to revert to the usual 1-based indexing used by most text editors. However, this option does not seem to be supported by node-jslint yet.

I tried using fudge as an inline setting (by including /*jslint fudge: true in a .js-file to be linted), but it didn't seem to affect what was reported by node-jslint. Not sure why that is, but on jslint.com the inline fudge directive seems to work as advertised.

I would argue that this feature should really be implemented in node-jslint rather than SublimeLinter-jslint (node-jslint is the immediate wrapper for jslint, which already seems to support 1-based line/column numbering through the fudge option), so I don't think I will officially release a new version of SublimeLinter-jslint at this stage (I did create an issue at reid/node-jslint#135). That said, I did try the following modifications locally, which seemed to work with node-jslint's "latest" edition, so if you want to manually install SublimeLinter-jslint you could do these modifications locally:

In linter.py you should add at the bottom:

    def split_match(self, match):
        """
        We override this method because new versions of jslint report line/col
        numbers starting with 0 instead of 1, necessiating use of a setting to
        allow transforming the numbers into matching line/col numbers in
        Sublime Text
        """

        startFromZero = self.get_view_settings().get('startFromZero', False)

        match, line, col, error, warning, message, near = super().split_match(match)

        if startFromZero:
            line = line + 1
            col = col + 1

        return match, line, col, error, warning, message, near

Then, in SublimeLinter.sublime-settings (find them e.g. via Preferences > Package Settings > SublimeLinter > Settings - User) you should modify the jslint section as follows:

            "jslint": {
                "@disable": false,
                "args": [
                    "--edition=latest"
                ],
                "startFromZero": true,
                "excludes": []
            },

The startFromZero option will indicate to the modified linter.py that row/column numbers need to be transformed to 1-based numbers before being handed over to SublimeLinter.

[pmw57]

You'll find that jslint.com is fudging the reported line/col numbers from the report.js script.

node-jslint would need to adjust their own reporter.js script to make use of the data.option.fudge value.

[devdoc]

node-jslint now (as of version 0.10.0) supports fudging thanks to a recently merged PR (reid/node-jslint#150) by @pmw57, which means you can edit the jslint section in you SublimeLinter.sublime-settings (find them e.g. via Preferences > Package Settings > SublimeLinter > Settings - User) as follows to make it use the latest (es6 enabled) version of JSLint and still get 1-based line numbers which allow linting errors to be highlighted at the correct spots in your code.

            "jslint": {
                "@disable": false,
                "args": [
                    "--edition=latest",
                    "--fudge"
                ],
                "excludes": []
            },

Kudos to @pmw57 for expediting fudge support!