Prevent span elements in table html

Prevent span elements from being created for each
newline character in the html string returned by the
markdown parser.

These span elements should not exist in the table
markup which is causing css rules to not work as expected.

Author: dhruvisompura

src/vs/base/common/htmlParser.ts

@@ -84,6 +84,24 @@ const voidElements: Record<string, boolean> = {
 	'wbr': true
 };
 
+/**
+ * Quick lookup table for table elements.
+ *
+ * Per HTML spec:
+ * - <table> can only contain: caption, colgroup, thead, tbody, tfoot, tr
+ * - <thead>, <tbody>, <tfoot> can only contain: tr
+ * - <tr> can only contain: td, th
+ * - <colgroup> can only contain: col
+ */
+const tableElements: Record<string, boolean> = {
+	'table': true,
+	'thead': true,
+	'tbody': true,
+	'tfoot': true,
+	'tr': true,
+	'colgroup': true
+};
+
 /**
  * Checks if a node is nested inside a <pre> element.
  * In <pre> elements, whitespace must be preserved per HTML spec.
@@ -107,6 +125,19 @@ function isInsidePreElement(node: HtmlNode | undefined): boolean {
 	return false;
 }
 
+/**
+ * Checks if a node is a table elemen
+ *
+ * @param node The node to check (typically parent of a text node)
+ * @returns true if node is a table element
+ */
+function isTableElement(node: HtmlNode | undefined): boolean {
+	if (!node || !node.name) {
+		return false;
+	}
+	return tableElements[node.name] === true;
+}
+
 /**
  * Parses a single HTML tag.
  *
@@ -321,14 +352,22 @@ export function parseHtml(html: string): Array<HtmlNode> {
 				nextChar &&
 				nextChar !== '<'
 			) {
-				// This is a text node; add it as a child node
-				if (current.children === undefined) {
-					current.children = [];
+				const textContent = html.slice(start, html.indexOf('<', start));
+				const isWhitespace = whitespaceRE.test(textContent);
+				// Check if we are in a table element context with whitespace-only text node
+				const whitespaceInTable = isWhitespace && isTableElement(current);
+
+				// Don't add whitespace-only text nodes if they are inside table elements
+				if (!whitespaceInTable) {
+					// This is a text node; add it as a child node
+					if (current.children === undefined) {
+						current.children = [];
+					}
+					current.children.push({
+						type: 'text',
+						content: decode(textContent),
+					});
 				}
-				current.children.push({
-					type: 'text',
-					content: decode(html.slice(start, html.indexOf('<', start))),
-				});
 			}
 
 			// if we're at root, push new base node
@@ -383,11 +422,14 @@ export function parseHtml(html: string): Array<HtmlNode> {
 					content = ' ';
 				}
 
-				// Don't add whitespace-only text nodes if they would be trailing text nodes
-				// or if they would be leading whitespace-only text nodes:
+				// Check if we are in a table element context with whitespace-only text node
+				const whitespaceInTable = whitespaceRE.test(content) && isTableElement(current);
+
+				// Don't add whitespace-only text nodes if they would be: trailing text nodes
+				// leading whitespace-only text nodes, or inside table elements:
 				//  * end > -1 indicates this is not a trailing text node
 				//  * leading node is when level is -1 and parent has length 0
-				if ((end > -1 && level + parent.length >= 0) || content !== ' ') {
+				if (!whitespaceInTable && ((end > -1 && level + parent.length >= 0) || content !== ' ')) {
 					parent.push({
 						type: 'text',
 						parent: current,