Jump to content

User:Interiot/Tool2/code.js

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by Interiot (talk | contribs) at 22:23, 27 December 2005. The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.
Note: After saving, you have to bypass your browser's cache to see the changes. Google Chrome, Firefox, Microsoft Edge and Safari: Hold down the ⇧ Shift key and click the Reload toolbar button. For details and instructions about other browsers, see Wikipedia:Bypass your cache.
// see http://paperlined.org/apps/wikipedia/Tool2/ for instructions on adding this to your monobook.js

// To run this tool on other servers:
//	1. copy this script to the target server (this is required because of javascript cross-site security restrictions)

//	2. update the following URL
//		for example: "User:Interiot/Tool2/code.js"
var tool2_url = "User:Interiot/Tool2/code.js";

//	3. update this namespace list, extracted from something like http://en.wikiquote.org/wiki/Special:Export//
var namespaces = [
"Talk",
"User",
"User talk",
"Wikiquote",
"Wikiquote talk",
"Image",
"Image talk",
"MediaWiki",
"MediaWiki talk",
"Template",
"Template talk",
"Help",
"Help talk",
"Category",
"Category talk"
];


// TODO:
//	- the current document.location method doesn't work when the page is accessed sans-mod_rewrite
//	- test with non-ASCII characters
//		- non-ascii usernames
//		- ??



var prefix = "";
var params = parse_params();

addOnloadFunction(function() {
  var path_len = document.location.pathname.length;
  // trigger once we view the right page
  if (document.location.pathname.substring(path_len - tool2_url.length, path_len) == tool2_url) {
    // get the prefix (needs to be fixed to work sans-mod_rewrite
    prefix = document.location.protocol + "//" + document.location.host + "/"
            + document.location.pathname.substring(1, path_len - tool2_url.length);

    // blank the inner contents of the page
    var bodyContent = document.getElementById("bodyContent");
    while (bodyContent.childNodes.length > 0) bodyContent.removeChild(bodyContent.lastChild);

    if (document.location.search.length == 0) {
      generate_input_form(bodyContent);
    } else {
      generate_main_report(bodyContent);
    }
  }
});


function generate_input_form(bodyContent) {
  bodyContent.innerHTML =
            "<form><table><tr><td>Username <td><input maxlength=128 name=username value='' title='username'>" +
            "             <tr><td>         <td><input type=submit value='Submit'>" +
            "</table></form>";

  var form = bodyContent.getElementsByTagName("form")[0];
  form.method = "get";
  form.action = document.location;
}

function generate_main_report() {
  fetch_data(params["username"], "", output_main_report, 0, []);
}

function output_main_report(data) {
}



var offset_regexp = /href="[^"]+:Contributions[^"]+offset=(\d+)/gi;
function fetch_data(username, end_date, handler, offset, page_list) {
	var url = prefix + "Special:Contributions/" + username + "?offset=" + offset + "&limit=100";
	loadXMLDoc(url, 
		function (request) {
			var next_offset = 0;
			if (request.readyState != 4)   return;
			if (request.status == 200) {
				page_list.push(request.responseText);
				//dump_text(request.responseText);

				// see if there's another pageful to get
				var matches = map( function(p){
						return p.match( /(\d+)$/ )[0];
					}, request.responseText.match( offset_regexp ) );
				for (var i=0; i<matches.length; i++) {
					var v = matches[i] * 1;
					if (v != 0 && (offset == 0 || v < offset)) {
						next_offset = v;
						break;
					}
				}
			}

			next_offset = 0;			// for testing, only retrieve one page of results

			if (next_offset == 0) {
				parse_data(page_list, handler);
			} else {
				// tail recurse
				fetch_data(username, end_date, handler, next_offset, page_list);
			}
		});
}


// input: a list strings, each string containing the HTML from a single page
// output: a list, where each individual entry is a specific edit from history
function parse_data(page_list, handler) {
	//var total_len = 0;
	//for (var i=0; i<page_list.length; i++) total_len += page_list[i].length;
	//alert("parsing " + page_list.length + " pages comprising " + total_len + " total bytes");

	var edit_history = new Array();
	for (var pagecnt=0; pagecnt<page_list.length; pagecnt++) {
		var matches = page_list[pagecnt].match( /^<li>[^(]+\(<a href="[^"]+action=history.*/gim );
		//dump_lines(matches);
		for (var matchcnt=0; matchcnt<matches.length; matchcnt++) {
			var history_text = matches[matchcnt];

			var history_entry = new Array();
			//history_entry["date_text"] = history_text.match( /^<li>([^(<]+)/is )[1];
			history_entry["title"] = history_text.match( /title="([^"]+)"/i )[1]
					.replace( /&quot;/g, "\"")
					.replace( /&amp;/g, "&");
			var find_comment = history_text.replace(/<span class="autocomment">.*?<\/span> ?/, "");
			history_entry["comment"] = ifmatch(find_comment.match( /<span class='comment'>(.*?)<\/span>/ ))
					.replace(/^\((.*)\)$/, "$1");
			history_entry["minor"] = /<span class="minor"/.test(history_text);

			//dump_lines(history_entry);
			dump_text(history_entry["title"]);
		}
	}
}

function ifmatch(ary) {
	if (ary && ary.length >= 2) {
		return ary[1];
	} else {
		return "";
	}
}



// ===================================== test/debug functions =========================================

function dump_text(text) {
  //alert("dump_text, with text of size " + text.length);

  var pre = document.createElement("pre");

  var div = document.createElement("div");
  div.style.width = "60em";
  div.style.maxHeight = "40em";
  div.style.overflow = "auto";

  pre.appendChild(document.createTextNode(text));
  div.appendChild(pre);
  document.getElementById("bodyContent").appendChild(div);
}

function dump_lines(ary) {
  dump_text("--> " + ary.join("\n--> "));
}


// ===================================== utility functions =========================================

function addOnloadFunction(f) {
  if (window.addEventListener) window.addEventListener("load",f,false);
  else if (window.attachEvent) window.attachEvent("onload",f);
  else {
    var oldOnload='_old_onload_'+addOnloadFunction.uid;
    addOnloadFunction[oldOnload] = window.onload ? window.onload : function () {};
    window.onload = function() { addOnloadFunction[oldOnload]();  f(); }
    ++addOnloadFunction.uid;
  }
}

function parse_params() {
  var pairs = document.location.search.substring(1).split("&");
  var ret = [];
  for (var i=0; i < pairs.length; i++) {
    var values = pairs[i].split("=");
    ret[values[0]] = unescape(values[1]);
  }
  return ret; 
}

function loadXMLDoc(url, handler)
{
    // branch for native XMLHttpRequest object
    if (window.XMLHttpRequest) {
        req = new XMLHttpRequest();
	req.onreadystatechange = function () {handler(req)};
        req.open("GET", url, true);
        req.send(null);
    // branch for IE/Windows ActiveX version
    } else if (window.ActiveXObject) {
        req = new ActiveXObject("Microsoft.XMLHTTP");
        if (req) {
            req.onreadystatechange = function () {handler(req)};
            req.open("GET", url, true);
            req.send();
        }
    }
}

// see http://search.cpan.org/dist/perl/pod/perlfunc.pod#map
function map (handler, list) {
  var ret = new Array();
  for (var i=0; i<list.length; i++) {
    ret[i] = handler( list[i] );
    // ret.push( handler( list[i] ) );
  }
  return ret;
}