fix(parser): Correctly handle literal less-than signs in text

The previous parsing logic identified any literal less-than character ('<') as the start of an tag.
This caused text containing the symbol (e.g., "p < 0.05") to be prematurely truncated when edited.

src/frontend/HighlightedTextbox.svelte

@@ -119,14 +119,14 @@
 		});
 		for (let i = 0; i < clean_marked_text.length; i++) {
 			let char = clean_marked_text[i];
-			if (char === "<") {
+			if (char === "<" && (i+5) <= clean_marked_text.length && clean_marked_text.slice(i+1,i+5) === "mark") {
 				in_tag = true;
 				if (text) {
 					new_value.push([text, category]);
 				}
 				text = "";
 				category = null;
-			} else if (char === ">") {
+			} else if (char === ">" && in_tag) {
 				in_tag = false;
 				if (tag.slice(0, 4) === "mark") {
 					let match = /class="hl ([^"]+)"/.exec(tag);