CoverMonkey is a code-coverage tool for the SpiderMonkey JavaScript
interpreter from Mozilla.  Using it requires a recent debug build of
SpiderMonkey. To obtain a debug build, follow the "Advanced Build"
directions here:
https://developer.mozilla.org/En/SpiderMonkey/Build_Documentation .
For correct dead-code analysis, CoverMonkey requires a patch that just
went into the SpiderMonkey source tree (2011-08-22), so make sure you
have a recent snapshot of the source.

To build CoverMonkey, type:
   
   make

This just assembles a CoverMonkey node script from the files in src/

To use CoverMonkey, run your code in SpiderMonkey with the -D option
set, and pipe the output (-D generates a *lot* of output) to
CoverMonkey.

For example:

  js -D -f my_library.js -f my_tests.js | CoverMonkey -b 

CoverMonkey uses node, so you'll need to get that from
http://nodejs.org if you don't already have it. I've used the
util.format() function that is new in Node 0.5, so you'll need a
recent version of node.

Usage: CoverMonkey [options]

    CoverMonkey reads the output generated by SpiderMonkey's -D option
    (supported by debug builds of SpiderMonkey only) and analyzes it to
    detect dead code and report code coverage.

    CoverMonkey reads -D data from stdin by default. Use the -d option
    to make it read from a file instead.

    CoverMonkey writes basic code coverage statistics to stdout.
    Use -q, -p, -c and -l to modify the information displayed.

    CoverMonkey can generate an HTML file that annotates and colors your 
    source code to highlight uncovered and dead code and indicate how
    many times each line executed.  Use the -h, -a, and -b options to 
    control HTML generation.

    By default, CoverMonkey will output coverage information for all 
    source files that were run. Use one or more -t options to specify
    which files CoverMonkey should analyze.

 Options:

    -d <file> Read -D data from the specified file instead of stdin

    -t <file> Analyze coverage for the specified target file.
              Multiple -t options are allowed.

    -q        Quiet: don't display any output to stdout

    -p        Percent: only display coverage percentage

    -c        Display the coverage statistics in compact tabular form

    -l        List the line numbers of all uncovered lines

    -h <file> Output annotated source code, in HTML format, to the 
              specified file.

    -a        Include assembly code for each line in the HTML file.
              This may be useful for understanding partially-covered
              lines.

    -b        Automatically launch a browser window to display the
              HTML. If no -h option, writes HTML to a temporary file.
              This option may only work on MacOS.

    --atlines Honor //@line comments in the source
              This option is probably not generally useful

    -v        Display the CoverMonkey version number and exit

    --help    Display this message and exit