Indent tracking

dm-lezer/src/context.js

@@ -0,0 +1,25 @@
+import { ContextTracker } from "@lezer/lr";
+import { dedent, indent } from "./parser.terms.js";
+
+class IndentLevel {
+  constructor(parent, depth) {
+    this.parent = parent;
+    this.depth = depth;
+    this.hash =
+      (parent ? (parent.hash + parent.hash) << 8 : 0) + depth + (depth << 4);
+  }
+}
+
+export const ctx = new ContextTracker({
+  start: { indent: new IndentLevel(null, 0) },
+  shift(context, term, stack, input) {
+    if (term === indent)
+      return {
+        ...context,
+        indent: new IndentLevel(context, stack.pos - input.pos),
+      };
+    if (term === dedent) return { ...context, indent: context.parent };
+    return context;
+  },
+  hash: (context) => context.hash,
+});

dm-lezer/src/dm.grammar

@@ -1,3 +1,10 @@
+@context ctx from "./context"
+@external tokens indentation from "./tokens.js" {
+  indent
+  dedent
+  blankLineStart
+}
+
 @top File {
   (
     preprocessor |

dm-lezer/src/tokens.js

@@ -0,0 +1,32 @@
+import { ExternalTokenizer } from "@lezer/lr";
+import { blankLineStart, dedent, indent } from "./parser.terms";
+
+const newline = "\n".charCodeAt(0),
+  space = " ".charCodeAt(0),
+  tab = "\t".charCodeAt(0),
+  hash = "#".charCodeAt(0),
+  slash = "/".charCodeAt(0),
+  star = "*".charCodeAt(0);
+
+export const indentation = new ExternalTokenizer((input, stack) => {
+  let prev = input.peek(-1);
+  if (prev !== -1 && prev !== newline) return;
+  let spaces = 0;
+  while (input.next === space || input.next === tab) {
+    input.advance();
+    spaces++;
+  }
+  if (
+    input.next === newline ||
+    input.next === hash ||
+    (input.next === slash &&
+      (input.peek(1) === slash || input.peek(1) === star))
+  ) {
+    if (stack.canShift(blankLineStart))
+      input.acceptToken(blankLineStart, -spaces);
+  } else if (spaces > stack.context.indent.depth) {
+    input.acceptToken(indent);
+  } else if (spaces < stack.context.indent.depth) {
+    input.acceptToken(dedent, -spaces);
+  }
+});