html-parser - npm Package Compare versions

Comparing version 0.10.1 to 0.11.0

package.json

 {
 	"name": "html-parser",
-	"version": "0.10.1",
+	"version": "0.11.0",
 	"description": "HTML/XML parser with less explosions",
@@ -17,3 +17,4 @@ "keywords": [ "html", "xml", "parser", "explosion" ],
 		{ "name": "fiatjaf" },
-		{ "name": "Sergii Kliuchnyk" }
+		{ "name": "Sergii Kliuchnyk" },
+		{ "name": "Edwin Hoogerbeets" }
 	],
@@ -20,0 +21,0 @@

src/parser.js

@@ -6,9 +6,24 @@ var parseContext = require('./context');
 	var name = context.readRegex(context.regex.attribute);
-	var value = null;
+	var value = null, quote = '';
 	if (context.current === '=' || context.peekIgnoreWhitespace() === '=') {
 		context.readRegex(/\s*=\s*/);
-		var quote = /['"]/.test(context.current) ? context.current : '';
-		var attributeValueRegex = !quote
-			? /(.*?)(?=[\s>])/
-			: new RegExp(quote + '(.*?)' + quote);
+		var attributeValueRegex;
+		switch (context.current) {
+		case "'":
+			attributeValueRegex = /('(\\'|<%.*?%>|[^'])*?')/;
+			quote = "'";
+			break;
+		case '"':
+			attributeValueRegex = /("(\\"|<%.*?%>|[^"])*?")/;
+			quote = '"';
+			break;
+		case '<':
+			attributeValueRegex = (context.peek() === '%') ? 
+				/(<%.*?%>)/ :
+				/(.*?)(?=[\s><])/;
+			break;
+		default:
+			attributeValueRegex = /(.*?)(?=[\s><])/;
+			break;
+		} 
 
@@ -18,5 +33,13 @@ var match = attributeValueRegex.exec(context.substring) || [0, ''];
 		context.read(match[0].length);
+
+		if (value[0] === '"' || value[0] === "'") {
+			value = value.substring(1);
+		}
+
+		if (value[value.length-1] === '"' || value[value.length-1] === "'") {
+			value = value.substring(0, value.length-1);
+		}
 	}
 
-	context.callbacks.attribute(name, value);
+	context.callbacks.attribute(name, value, quote);
 }
@@ -33,10 +56,46 @@
 
-	var next = context.current;
+	var next = context.current, handled;
 	while (!context.isEof() && !isClosingToken()) {
-		if (context.regex.attribute.test(next)) {
-			readAttribute(context);
-			next = context.current;
+		handled = false;
+		if (context.current === '<') {
+			for (var callbackName in context.regex.dataElements) {
+				if (!context.regex.dataElements.hasOwnProperty(callbackName)) {
+					continue;
+				}
+
+				var dataElement = context.regex.dataElements[callbackName],
+					start = dataElement.start,
+					isValid = false;
+
+				switch (typeof start) {
+					case 'string':
+						isValid = context.substring.slice(0, start.length) === start;
+						break;
+					case 'object':
+						isValid = start.test(context.substring);
+						break;
+					case 'function':
+						isValid = start(context.substring) > -1;
+						break;
+				}
+
+				if (isValid) {
+					callbackText(context);
+					context.callbacks[callbackName](parseDataElement(context, dataElement));
+					next = context.current;
+					handled = true;
+					break;
+				} 
+				next = context.current;
+			}
 		}
-		else {
-			next = context.read();
+		
+		if (!handled) {
+			if (context.regex.attribute.test(next)) {
+				readAttribute(context);
+				next = context.current;
+			}
+			else {
+				next = context.read();
+			}
 		}
@@ -79,3 +138,3 @@ }
 
-	if (!/^(script|xmp)$/i.test(name)) {
+	if (!/^(script|xmp|style)$/i.test(name)) {
 		return;
@@ -439,3 +498,3 @@ }
 
-		attribute: function(name, value) {
+		attribute: function(name, value, quote) {
 			if (ignoreStack.length) {
@@ -452,3 +511,4 @@ return;
 			if (value) {
-				sanitized += '="' + value.replace(/"/g, '&quot;') + '"';
+				// reuse the existing quote style if possible
+				sanitized += '=' + quote + ((quote === '"') ? value.replace(/"/g, '&quot;') : value.replace(/'/g, '&apos;')) + quote;
 			}
@@ -477,3 +537,3 @@ },
 			for (var i = tagStack.length - 1; i >= 0; i--) {
-				if (tagStack[i] === 'script' || tagStack[i] === 'xmp') {
+				if (tagStack[i] === 'script' || tagStack[i] === 'xmp' || tagStack[i] === 'style') {
 					sanitized += value;
@@ -480,0 +540,0 @@ return;

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

25784

increased by6.43%

Lines of code

593

increased by10.02%

No dependency changes