Migrate markdown scopes (#2043)

## Checklist

- [-] I have added
[tests](https://www.cursorless.org/docs/contributing/test-case-recorder/)
- [-] I have updated the
[docs](https://github.com/cursorless-dev/cursorless/tree/main/docs) and
[cheatsheet](https://github.com/cursorless-dev/cursorless/tree/main/cursorless-talon/src/cheatsheet)
- [-] I have not broken the cheatsheet

Author: AndreasArvidsson

packages/cursorless-engine/src/languages/getTextFragmentExtractor.ts

@@ -131,7 +131,6 @@ const textFragmentExtractors: Record<
     htmlStringTextFragmentExtractor,
   ),
   latex: fullDocumentTextFragmentExtractor,
-  markdown: fullDocumentTextFragmentExtractor,
   ruby: constructDefaultTextFragmentExtractor(
     "ruby",
     rubyStringTextFragmentExtractor,

packages/cursorless-engine/src/languages/markdown.ts

@@ -1,48 +1,19 @@
-import { Range, Selection, TextEditor } from "@cursorless/common";
+import { Range, SimpleScopeTypeType, TextEditor } from "@cursorless/common";
 import type { SyntaxNode } from "web-tree-sitter";
-import { getMatchesInRange } from "../util/getMatchesInRange";
-import { SimpleScopeTypeType } from "@cursorless/common";
 import {
   NodeFinder,
   NodeMatcherAlternative,
   SelectionWithContext,
 } from "../typings/Types";
+import { getMatchesInRange } from "../util/getMatchesInRange";
 import { leadingSiblingNodeFinder, patternFinder } from "../util/nodeFinders";
 import { createPatternMatchers, matcher } from "../util/nodeMatchers";
 import {
   extendUntilNextMatchingSiblingOrLast,
-  getNodeRange,
   selectWithLeadingDelimiter,
 } from "../util/nodeSelectors";
 import { shrinkRangeToFitContent } from "../util/selectionUtils";
 
-/**
- * Given a node representing the text of a section heading (without leading
- * marker), will return the content range as the text without the leading
- * whitespace, and the outside range includes the leading marker, so that
- * "chuck name" deletes the heading
- * @param editor The editor containing the node
- * @param node The node to extract from; will be the content of the heading without the leading marker
- * @returns The selection with context
- */
-function nameExtractor(
-  editor: TextEditor,
-  node: SyntaxNode,
-): SelectionWithContext {
-  const range = getNodeRange(node);
-  const contentRange = range.isEmpty
-    ? range
-    : range.with(range.start.translate(0, 1));
-  const removalRange = getNodeRange(node.parent!);
-
-  return {
-    selection: new Selection(contentRange.start, contentRange.end),
-    context: {
-      removalRange,
-    },
-  };
-}
-
 const HEADING_MARKER_TYPES = [
   "atx_h1_marker",
   "atx_h2_marker",
@@ -139,12 +110,6 @@ function itemExtractor(
 const nodeMatchers: Partial<
   Record<SimpleScopeTypeType, NodeMatcherAlternative>
 > = {
-  list: ["list"],
-  comment: "html_block",
-  name: matcher(
-    leadingSiblingNodeFinder(patternFinder("atx_heading[heading_content]")),
-    nameExtractor,
-  ),
   collectionItem: matcher(patternFinder("list_item.paragraph!"), itemExtractor),
   section: sectionMatcher("atx_heading"),
   sectionLevelOne: sectionMatcher("atx_heading.atx_h1_marker"),

queries/markdown.scm

@@ -0,0 +1,19 @@
+(document) @textFragment
+
+(html_block) @comment
+
+;;!! * hello * stuff
+;;!  ^^^^^^^^^^^^^^^
+(list) @list
+
+;;!! # Title
+;;!    ^^^^^
+;;!! xxxxxxx
+;;!  -------
+(section
+    (atx_heading
+        (_)
+        heading_content: (_) @name
+    ) @_.removal
+    (#shrink-to-match! @name "^\\s*(?<keep>.*)$")
+) @_.domain