Question

today i’m trying to parse a ‘ps aux’ output (issued from a node.js server)

so i made an happy regex to parse the values, and i wanted to put all those fancy informations into a simple JSON object 🙂

problem is, the regex i’ve made has 3 match groups, doing multiple matches… and i don’t even know if it’s possible to have it all the matches returned.

var spawn = require("child_process").spawn;

function exec(http_req, http_resp){
    var re_ps = /(?:([\d\.\w\-\?\/\+]+)\s+){8}(?:((?:\d+\:\d+)|(?:\w+\d+))\s+){2}([^\n]+)/g,
        result_re_ps,
        array_re_ps = [];
    //[USER, PID, %CPU, %MEM, VSZ, RSS, TTY, STAT], dates[START, TIME], COMMAND
    var ps_process = spawn("ps", ["aux"]);
    ps_process.stdout.on("data", function(data){
        var split_data = data.toString().split('\n');
        for (var i=1; i < 2; i++){
            while ((result_re_ps = re_ps.exec(split_data[i])) != null){
                console.log("--- while iteration ---");
                console.dir(result_re_ps);
                //console.dir(result_re_ps);
            }
        }
        //http_resp.write(data);
    });
    ps_process.on("exit", function (code) {
        if (code !== 0) {
            console.log("ps_process exited with code " + code);
        }
        http_resp.end();
    });
}

exports.exec = exec;

the “— while iteration —” it’s just one per each line, instead, it should be multiples!

I will link a fancy screenshot to let you visually understand what kind of data i’m expecting

As you can see the above are 3 groups of data, each one with multiple occurrences

Any help? :p

EDITED TO CLARIFY:

This is the result of a line like:

root      5213  0.0  2.1   8688  2760 ?        Ss   11:33   0:01 sshd: root@pts/0 

As you can see, the resulting array is 1-dimensional, instead for the results to be contained entirely it HAS to be at least 2-dimensional

UPDATE:

For now i’ve solved the problem changin my code like this:

var spawn = require("child_process").spawn;

function exec(http_req, http_resp){
    var array_re_ps = [];
    var re_ps1 = /(?:([\d\.\w\-\?\/\+]+)\s+)/g,
        re_ps2 = /(?:((?:\d+\:\d+)|(?:\w+\d+))\s+)/g,
        re_ps3 = /([^\n]+)/g;
    //[USER, PID, %CPU, %MEM, VSZ, RSS, TTY, STAT], dates[START, TIME], COMMAND
    var ps_process = spawn("ps", ["aux"]);
    ps_process.stdout.on("data", function(data){
        var split_data = data.toString().split('\n');
        for (var i=20; i < 21; i++){
            array_re_ps[0] = [], array_re_ps[1] = [];
            for (var j=0; j<8; j++){
                array_re_ps[0].push(re_ps1.exec(split_data[i])[1]);
            }
            re_ps2.lastIndex = re_ps1.lastIndex;
            for (var j=0; j<2; j++){
                array_re_ps[1].push(re_ps2.exec(split_data[i])[1])
            }
            re_ps3.lastIndex = re_ps2.lastIndex;
            array_re_ps[2] = re_ps3.exec(split_data[i])[1];
            console.dir(array_re_ps);
        }
    });
    ps_process.on("exit", function (code) {
        if (code !== 0) {
            console.log("ps_process exited with code " + code);
        }
        http_resp.end();
    });
}

exports.exec = exec;

Basically i split my long regexp, in 3 smaller ones, and i do cycle through their results.
If you try the code i formatted the resulting array exactly as i wanted, but i still think that it’s a side-solution.

Since Regular Expressions lets you “multiple parse group multiple times”, it is possible to have that functionality in javascript?

Answer 1

Take a look at the return values of the exec function.

It returns the index where it first matched, the original string and the capture groups in the elements [0]...[n].

All this is available in your output, you just have to iterate through the single elements of the array.