zeffii / precursor Goto Github PK

A simple tool to speed up writing for loops, and similar syntax heavy code constructs, that can be described in a more concise way. It only takes a small amount of information to describe what you want, but for the JavaScript engine to understand you extra syntax is required.

I've bound this to ctrl+shift+[

Precursor is to code what stenography is to reporting. Precursor and Snippets aim to achieve the same end goal, each using a different interaction paradigm.

Mostly through a cascade of Regex matching, if the content of a line doesn't match a first pattern it moves on to the next candidate, till a match is found. In this case the line can be rewritten from shorthand to longform. The current level of indentation is detected and used to make sure the rewrite happens where expected.

An example .json is provided with the extension, it includes the two main regexs that I use to avoid writing for-loops. It unfortunately requires you to escape backslashed regex commands. This means if you need a \S you need to write \\S, the same applies to any other command that starts with a \, they are listed here.

Regexs are tested in the order of appearance in prototyper.json, think of this as their order of precedence.

{
    "patterns": [
        {
            "pattern_name": "varname:iterations",
            "pattern_regex": "^(\\S+):(\\w+)$",
            "lines": [
                "for (var {0} = 0; {0} < {1}; {0} += 1) {",
                "{indent}{0};",
                "}"
            ]
        },
        {
            "pattern_name": "varname:iterable",
            "pattern_regex": "^(\\S+):(\\w+)\\[\\]?$",
            "lines": [
                "for (var {0} = 0; {0} < {1}.length; {0} += 1) {",
                "{indent}{1}[{0}];",
                "}"
            ]
        }
    ]
}

You should test if the addition you make to the prototyper.json results in a valid json file before attempting to run the extension. Use something like jsonlint.com if you aren't sure.

The regex will contain at least one match group ( ), but shouldn't use nested matches. Each set of ( ) will correspond to their numbered counterpart in the line property of the pattern. For example ^(\\S+):(\\w+)$ will match 2 parts of the input string, they can be accessed for the rewrite by using {0} and {1}.

In the case of no matches, there are a few potential reasons:

• the regex should be moved up in the prototyper.json because likely some other regex is catching the content of the line before it reaches the intended regex.
• the regex is not valid for the content of the line.
• The regex is passed a trimmed version of the line you initiate the command on. Using ^ and $ at either end of your regex will help you declare an unambiguous pattern.

Please use the issue tracker, I'll respond when time permits.