Merge branch 'master' of github.com:hellolwq/nproxy

README

@@ -1,7 +1,17 @@
 nproxy基于node.js开发，用于实现http协议级别的Web代理。
 目前支持简单的的HTML代理，完整的HTML、JS等重写正在完成中，问题会比较多，如果有发现，请反馈到hellolwq@gmail.com
 
-一、安装步骤
+一、特性
+1、支持HTML重写，可重写静态URL链接。
+2、...
+
+二、待支持特性
+1、支持服务端javascript重写。
+2、支持客户端javascript、html等脚本重写。
+3、支持css脚本重写。
+4、支持cookie保存。
+
+三、安装步骤
 1、安装node.js环境 http://nodejs.org/#download
 2、下载nproxy文件 https://github.com/hellolwq/nproxy/downloads
 3、修改nproxy.js配置监听端口。（之后会统一放到配置文件）

lib/htmlparser.js

@@ -12,12 +12,64 @@
  * });
  *
  * // or to get an XML string:
- * HTMLtoXML(htmlString);
+ * RewriteHtml(htmlString);
  *
  */
- exports.parseHTML = HTMLtoXML
+ exports.RewriteHtml = function( html ,request,tran_url_fun) {
+ //RewriteHtml = function( html ,request,tran_url_fun) {
+	var results = "";
+
+	HTMLParser(html, {
+		start: function( tag, attrs, unary ) {
+			results += "<" + tag;
+
+			for ( var i = 0; i < attrs.length; i++ )
+			{
+                if(g_html_rules[tag] && g_html_rules[tag][attrs[i].name])
+                {
+                    if(tag.toLowerCase() == "meta" && attrs[i].name.toLowerCase() == 'content')
+                    {
+                        var isRefresh = false;
+                        for(var j=0;j<attrs.length;j++)
+                        {
+                            if(attrs[j].name.toLowerCase() == 'http-equiv' && attrs[j].escaped.toLowerCase() == 'refresh')
+                            {
+                                isRefresh = true;
+                                break;
+                            }
+
+                        }
+
+                        if(isRefresh)
+                        {
+                            attrs[i].escaped = attrs[i].escaped.replace(/(url=)(.+)$/i, function($0,$1,$2){return ($1+tran_url_fun($2,request));});
+                            results += " " + attrs[i].name + '="' + attrs[i].escaped + '"';
+                            continue;
+                        }
+                    }
+                    attrs[i].escaped = tran_url_fun(attrs[i].escaped,request);
+                }
+
+                results += " " + attrs[i].name + '="' + attrs[i].escaped + '"';
+            }
+
+			results += (unary ? "/" : "") + ">";
+		},
+		end: function( tag ) {
+			results += "</" + tag + ">";
+		},
+		chars: function( text ) {
+			results += text;
+		},
+		comment: function( text ) {
+			results += "<!--" + text + "-->";
+		}
+	});
+
+	return results;
+};
 // Regular Expressions for parsing tags and attributes
-var startTag = /^<([-A-Za-z0-9_]+)((?:\s+\w+(?:\s*=\s*(?:(?:"[^"]*")|(?:'[^']*')|[^>\s]+))?)*)\s*(\/?)>/,
+var startTag = /^<([-A-Za-z0-9_]+)((?:\s+[-\w]+(?:\s*=\s*(?:(?:"[^"]*")|(?:'[^']*')|[^>\s]+))?)*)\s*(\/?)>/,
 	endTag = /^<\/([-A-Za-z0-9_]+)[^>]*>/,
 	attr = /([-A-Za-z0-9_]+)(?:\s*=\s*(?:(?:"((?:\\.|[^"])*)")|(?:'((?:\\.|[^'])*)')|([^>\s]+)))?/g;
 
@@ -40,6 +92,12 @@ var fillAttrs = makeMap("checked,compact,declare,defer,disabled,ismap,multiple,n
 // Special Elements (can contain anything)
 var special = makeMap("script,style");
 
+var g_html_rules = {
+    img:{src:1},
+    href:{href:1},
+    meta:{content:1}
+}
+
 var HTMLParser = this.HTMLParser = function( html, handler ) {
 	var index, chars, match, stack = [], last = html;
 	stack.last = function(){
@@ -75,9 +133,22 @@ var HTMLParser = this.HTMLParser = function( html, handler ) {
 
 			// start tag
 			} else if ( html.indexOf("<") == 0 ) {
-				match = html.match( startTag );
-
-				if ( match ) {
+				var docTypeTag = "<!DOCTYPE "
+                if(html.substr(0,docTypeTag.length).toLowerCase() == docTypeTag.toLowerCase()){
+                    index = html.indexOf('>')
+                    if(index < 0)
+                        throw "Parse Error:doctype invalid!";
+
+                    ++index;
+                    var text = html.substring( 0, index );
+
+                    if ( handler.chars )
+                        handler.chars( text );
+
+                    html = html.substr( index );
+                    chars = false;
+                }
+                else if ( match = html.match( startTag )) {
 					html = html.substring( match[0].length );
 					match[0].replace( startTag, parseStartTag );
 					chars = false;
@@ -178,31 +249,6 @@ var HTMLParser = this.HTMLParser = function( html, handler ) {
 	}
 };
 
-this.HTMLtoXML = function( html ) {
-	var results = "";
-
-	HTMLParser(html, {
-		start: function( tag, attrs, unary ) {
-			results += "<" + tag;
-
-			for ( var i = 0; i < attrs.length; i++ )
-				results += " " + attrs[i].name + '="' + attrs[i].escaped + '"';
-
-			results += (unary ? "/" : "") + ">";
-		},
-		end: function( tag ) {
-			results += "</" + tag + ">";
-		},
-		chars: function( text ) {
-			results += text;
-		},
-		comment: function( text ) {
-			results += "<!--" + text + "-->";
-		}
-	});
-
-	return results;
-};
 
 function makeMap(str){
 	var obj = {}, items = str.split(",");

nproxy.js

@@ -1,6 +1,7 @@
 var HTTP = require('http');
 var FS = require('fs');
 var URL = require('url');
+var HtmlParser = require('./lib/htmlparser.js');
 
 var SVR_PORT = 8080;
 var LOG_DIR = "./logs/";
@@ -155,12 +156,18 @@ function proxy_request(urlObj,callback)
 function proxy_rewrite(data,request)
 {
     //log(LOG_DBG,"typeof(data):",typeof(data));
-    //print_r(data);
-	//log(LOG_DBG,"proxy_rewrite:" + data);
-	//log(LOG_DBG,"proxy_rewrite request.orig_url" + request.orig_url);
-    return data.replace(/(\s(src|href)=['"])([^'"]*)(['"])/g, function() { 
+    log(LOG_DBG,"proxy_rewrite before:" + data);
+    var ret = data;
+   // try{
+        ret = HtmlParser.RewriteHtml(data,request,helper_trans_url);
+    //}catch(e){
+    //    ret = data;
+    //}
+    log(LOG_DBG,"proxy_rewrite after:" + ret);
+    return ret;
+    /*return data.replace(/(\s(src|href)=['"])([^'"]*)(['"])/g, function() { 
         return (arguments[1]+helper_trans_url(arguments[3],request)+arguments[4]);
-    });
+    });*/
 }
 
 function helper_trans_url(url,request)

tset.html

@@ -0,0 +1,25 @@
+<html>
+<head>
+<script src="lib/htmlparser.js"> </script>
+</head>
+<body>
+    <textarea id="input" rows="20" style="width:100%">
+    </textarea>
+    <button onClick="onRewrite()" style="width:200px;">ok</button>
+    <textarea id="output" rows="20" style="width:100%">
+    </textarea>
+    <script>
+        function onRewrite()
+        {
+            var input = document.getElementById("input").innerText;
+            var output = RewriteHtml(input,null,tran_url);
+            document.getElementById("output").innerText = output;
+        }
+
+        function tran_url(url)
+        {
+            return url+"##";
+        }
+    </script>
+</body>
+</html>