feat(language-server): add offset-preserving virtual HTML document

2026-08-18 20:34:24 +05:30
parent 7d481df652
commit 7301849a7f
2 changed files with 225 additions and 0 deletions
@@ -0,0 +1,116 @@
+import type { TextDocument } from "./index.ts";
+
+export interface HtmlRegion {
+  start: number;
+  end: number;
+}
+
+/**
+ * Byte ranges of the markup inside each `view { }` block.
+ *
+ * This is a tolerant scanner rather than the parser: completion fires while
+ * the document is being typed, which is exactly when it does not parse.
+ */
+export function viewRegions(text: string): HtmlRegion[] {
+  const regions: HtmlRegion[] = [];
+  const pattern = /\bview\s*\{/g;
+  let match: RegExpExecArray | null;
+
+  while ((match = pattern.exec(text))) {
+    const bodyStart = match.index + match[0].length;
+    const end = matchingBrace(text, bodyStart);
+    regions.push({ start: bodyStart, end });
+    pattern.lastIndex = end;
+  }
+  return regions;
+}
+
+/**
+ * Offset of the brace closing the block that starts at `from`, or the end of
+ * the text when it is never closed (an unterminated block is normal mid-edit).
+ *
+ * Quotes are only tracked inside a tag, never in text content: `<p>it's</p>`
+ * would otherwise open a string that never closes and swallow the rest of the
+ * file.
+ */
+function matchingBrace(text: string, from: number): number {
+  let depth = 1;
+  let inTag = false;
+  let quote: string | null = null;
+
+  for (let index = from; index < text.length; index += 1) {
+    const char = text[index]!;
+
+    if (quote) {
+      if (char === quote) quote = null;
+      continue;
+    }
+    if (inTag && (char === '"' || char === "'")) {
+      quote = char;
+      continue;
+    }
+    if (char === "<") inTag = true;
+    else if (char === ">") inTag = false;
+    else if (char === "{") depth += 1;
+    else if (char === "}") {
+      depth -= 1;
+      if (depth === 0) return index;
+    }
+  }
+  return text.length;
+}
+
+/**
+ * A parallel document containing only the markup.
+ *
+ * Everything outside a view block becomes whitespace of the same length, and
+ * newlines are preserved, so an offset in the source is the same offset here.
+ * That removes the need for a mapping table entirely.
+ */
+export function virtualHtmlDocument(document: TextDocument): {
+  uri: string;
+  languageId: "html";
+  text: string;
+} {
+  const source = document.text;
+  const keep = new Array<boolean>(source.length).fill(false);
+  for (const region of viewRegions(source)) {
+    for (let index = region.start; index < region.end; index += 1) keep[index] = true;
+  }
+
+  let text = "";
+  for (let index = 0; index < source.length; index += 1) {
+    const char = source[index]!;
+    text += keep[index] || char === "\n" ? char : char === "\r" ? "\r" : " ";
+  }
+
+  return { uri: `${document.uri}.html`, languageId: "html", text };
+}
+
+export function isInsideHtml(document: TextDocument, offset: number): boolean {
+  return regionsFor(document).some((region) => offset >= region.start && offset <= region.end);
+}
+
+/**
+ * Regions for a document, cached by uri and version.
+ *
+ * A single keystroke produces a burst of completion, hover, and tag-close
+ * requests; without this each one rescans the file.
+ */
+const regionCache = new Map<string, { version: number; regions: HtmlRegion[] }>();
+
+function regionsFor(document: TextDocument): HtmlRegion[] {
+  const version = document.version ?? -1;
+  const cached = regionCache.get(document.uri);
+  if (cached && cached.version === version) return cached.regions;
+
+  const regions = viewRegions(document.text);
+  regionCache.set(document.uri, { version, regions });
+  return regions;
+}
+
+/** Drops a document's cached regions. Call when a document closes. */
+export function clearHtmlRegionCache(uri?: string): void {
+  if (uri) regionCache.delete(uri);
+  else regionCache.clear();
+}