A solution against shouting (for Firefox/Greasemonkey)

Brilliant !

I'm afraid to say it's a little buggy. descent(table) might be executed without table being set.

The bug doesn't cause any problems, but still:

// ==UserScript==
// @name        Anti Shout
// @namespace   kamikaze.bsdforen.de
// @include     http://www.keil.com/forum/*
// @description Scale down excessive shouting on the Keil forums
// ==/UserScript==

function descent(node) {
        if (node.data) {
                var volume = node.data.match(/[A-Z]/g);
                if (volume && volume.length / node.data.length > 0.5) {
                        node.parentNode.style.fontSize = "x-small";
                }
        }
        for (var i = 0; i < node.childNodes.length; i++) {
                descent(node.childNodes[i]);
        }
}

var tables = document.getElementsByTagName("table");
for (var i = 0; i < tables.length; i++) {
        if (tables[i].getAttribute("class") == "thd") {
                descent(tables[i]);
        }
}

This update gets rid of a false positive condition.

// ==UserScript==
// @name        Anti Shout
// @namespace   kamikaze.bsdforen.de
// @include     http://www.keil.com/forum/*
// @description Scale down excessive shouting on the Keil forums
// ==/UserScript==

function descent(node) {
        var length = 0;
        var shouted = 0;
        for (var i = 0; i < node.childNodes.length; i++) {
                var child = node.childNodes[i];
                if (child.data) {
                        var caps = child.data.match(/[A-Z]/g);
                        length += child.data.length;
                        if (caps) {
                                shouted += caps.length;
                        }
                } else {
                        descent(child);
                }
        }
        if (length && shouted / length > 0.5) {
                node.style.fontSize = "x-small";
        }
}

var tables = document.getElementsByTagName("td");
for (var i = 0; i < tables.length; i++) {
        if (tables[i].getAttribute("class") == "bod") {
                descent(tables[i]);
        }
}

you could get Keil to incorporate it.

Incorporating it locally will have no effect on the IDIOTs screen.

Erik

Sadly, the days when you could break into a company's web service without repercussions, as long as you didn't do any real harm, are long over.

I was afraid something like this would happen (false positives):
http://www.keil.com/forum/21786/

Though I expected it from preprocessor examples.

What should I do about it, any ideas?

what in the thread you link to do you consider a false positive?

Erik

If you use my script 2 blocks of ASM code are affected. I don't consider assembler code shouting.

.

Is that due to the total lack of comments?

You could say so.

Is that due to the total lack of comments?

in that case it would probably be OK to 'minimize'

Erik

Good point, I am appeased.