Support spaces in angle bracket link fragments

CHANGELOG.md

@@ -1,5 +1,9 @@
 # Changelog
 
+## 0.5.0-alpha.8 — Unreleased
+- Fix incorrect detection of a multiline link definition #192
+- Support spaces in angle brackets inside links: `[text](<#heading with spaces>)`
+
 ## 0.5.0-alpha.7 — July 25, 2024
 - Strip markup from header document symbols. This makes them more readable.
 

src/languageFeatures/definitions.ts

@@ -74,7 +74,7 @@ export class MdDefinitionProvider {
 		}
 
 		const toc = await this.#tocProvider.get(resolvedResource);
-		return toc?.lookup(sourceLink.href.fragment)?.headerLocation;
+		return toc?.lookupByFragment(sourceLink.href.fragment)?.headerLocation;
 	}
 
 	#getDefinitionOfRef(ref: string, allLinksInFile: readonly MdLink[]) {

src/languageFeatures/diagnostics.ts

@@ -9,7 +9,7 @@ import * as lsp from 'vscode-languageserver-protocol';
 import { URI } from 'vscode-uri';
 import { LsConfiguration } from '../config';
 import { ILogger, LogLevel } from '../logging';
-import { MdTableOfContentsProvider } from '../tableOfContents';
+import { MdTableOfContentsProvider, TableOfContents } from '../tableOfContents';
 import { HrefKind, InternalHref, LinkDefinitionSet, MdLink, MdLinkDefinition, MdLinkKind, MdLinkSource, ReferenceLinkMap } from '../types/documentLink';
 import { translatePosition } from '../types/position';
 import { modifyRange } from '../types/range';
@@ -224,12 +224,11 @@ export class DiagnosticComputer {
 
 		const diagnostics: lsp.Diagnostic[] = [];
 		for (const link of links) {
-
 			if (link.href.kind === HrefKind.Internal
 				&& link.source.hrefText.startsWith('#')
 				&& isSameResource(link.href.path, getDocUri(doc))
 				&& link.href.fragment
-				&& !toc.lookup(link.href.fragment)
+				&& !tocLookupByLink(toc, { source: link.source, fragment: link.href.fragment })
 			) {
 				// Don't validate line number links
 				if (parseLocationInfoFromFragment(link.href.fragment)) {
@@ -408,7 +407,7 @@ export class DiagnosticComputer {
 									continue;
 								}
 
-								if (!toc?.lookup(link.fragment) && !this.#isIgnoredLink(options, link.source.pathText) && !this.#isIgnoredLink(options, link.source.hrefText)) {
+								if (!(toc && tocLookupByLink(toc, link)) && !this.#isIgnoredLink(options, link.source.pathText) && !this.#isIgnoredLink(options, link.source.hrefText)) {
 									const range = (link.source.fragmentRange && modifyRange(link.source.fragmentRange, translatePosition(link.source.fragmentRange.start, { characterDelta: -1 }), undefined)) ?? link.source.hrefRange;
 									diagnostics.push({
 										code: DiagnosticCode.link_noSuchHeaderInFile,
@@ -667,3 +666,7 @@ export class DiagnosticsManager extends Disposable implements IPullDiagnosticsMa
 		this.#linkWatcher.deleteDocument(uri);
 	}
 }
+
+function tocLookupByLink(toc: TableOfContents, link: { readonly source: MdLinkSource; readonly fragment: string; }) {
+	return link.source.isAngleBracketLink ? toc.lookupByHeading(link.fragment) : toc.lookupByFragment(link.fragment);
+}

src/languageFeatures/documentHighlights.ts

@@ -57,7 +57,7 @@ export class MdDocumentHighlightProvider {
 		const docUri = getDocUri(document);
 		for (const link of links) {
 			if (link.href.kind === HrefKind.Internal
-				&& toc.lookup(link.href.fragment) === header
+				&& toc.lookupByFragment(link.href.fragment) === header
 				&& link.source.fragmentRange
 				&& isSameResource(link.href.path, docUri)
 			) {
@@ -106,7 +106,7 @@ export class MdDocumentHighlightProvider {
 		const fragment = href.fragment.toLowerCase();
 
 		if (isSameResource(targetDoc, getDocUri(document))) {
-			const header = toc.lookup(fragment);
+			const header = toc.lookupByFragment(fragment);
 			if (header) {
 				yield { range: header.headerLocation.range, kind: lsp.DocumentHighlightKind.Write };
 			}

src/languageFeatures/documentLinks.ts

@@ -732,7 +732,7 @@ export class MdLinkProvider extends Disposable {
 
 		if (doc) {
 			const toc = await this.#tocProvider.getForContainingDoc(doc, token);
-			const entry = toc.lookup(linkFragment);
+			const entry = toc.lookupByFragment(linkFragment);
 			if (entry) {
 				return { kind: 'file', uri: URI.parse(entry.headerLocation.uri), position: entry.headerLocation.range.start, fragment: linkFragment };
 			}

src/languageFeatures/references.ts

@@ -228,7 +228,7 @@ export class MdReferencesProvider extends Disposable {
 
 		if (resolvedResource && this.#isMarkdownPath(resolvedResource) && sourceLink.href.fragment && sourceLink.source.fragmentRange && rangeContains(sourceLink.source.fragmentRange, triggerPosition)) {
 			const toc = await this.#tocProvider.get(resolvedResource);
-			const entry = toc?.lookup(sourceLink.href.fragment);
+			const entry = toc?.lookupByFragment(sourceLink.href.fragment);
 			if (entry) {
 				references.push({
 					kind: MdReferenceKind.Header,

src/tableOfContents.ts

@@ -219,8 +219,13 @@ export class TableOfContents {
 		this.#slugifier = slugifier;
 	}
 
-	public lookup(fragment: string): TocEntry | undefined {
-		const slug = this.#slugifier.fromFragment(fragment);
+	public lookupByFragment(fragmentText: string): TocEntry | undefined {
+		const slug = this.#slugifier.fromFragment(fragmentText);
+		return this.entries.find(entry => entry.slug.equals(slug));
+	}
+
+	public lookupByHeading(text: string): TocEntry | undefined {
+		const slug = this.#slugifier.fromHeading(text);
 		return this.entries.find(entry => entry.slug.equals(slug));
 	}
 }

src/test/diagnostic.test.ts

@@ -524,6 +524,24 @@ suite('Diagnostic Computer', () => {
 		const diagnostics = await getComputedDiagnostics(store, doc, workspace, {});
 		assertDiagnosticsEqual(diagnostics, []);
 	}));
+
+	test('Should support angle bracket links with spaces in fragment', withStore(async (store) => {
+		const doc1 = new InMemoryDocument(workspacePath('doc space1.md'), joinLines(
+			`[i](<doc space2.md>)`,
+			`[i](<doc space2.md#fr agment>)`,
+			`[i](<#fr agment>)`,
+			``,
+			'# FR agment',
+
+		));
+		const doc2 = new InMemoryDocument(workspacePath('doc space2.md'), joinLines(
+			'# FR agment',
+		));
+		const workspace = new InMemoryWorkspace([doc1, doc2]);
+
+		const diagnostics = await getComputedDiagnostics(store, doc1, workspace, {});
+		assertDiagnosticsEqual(diagnostics, []);
+	}));
 });
 
 

src/test/documentLinks.test.ts

@@ -495,43 +495,43 @@ suite('Link computer', () => {
 		]);
 	});
 
-	test('Should find link only within angle brackets.', async () => {
+	test('Should find link only within angle brackets', async () => {
 		const links = await getLinksForText(joinLines(
 			`[link](<path>)`
 		));
 		assertLinksEqual(links, [lsp.Range.create(0, 8, 0, 12)]);
 	});
 
-	test('Should find link within angle brackets even with link title.', async () => {
+	test('Should find link within angle brackets even with link title', async () => {
 		const links = await getLinksForText(joinLines(
 			`[link](<path> "test title")`
 		));
 		assertLinksEqual(links, [lsp.Range.create(0, 8, 0, 12)]);
 	});
 
-	test('Should find link within angle brackets even with surrounding spaces.', async () => {
+	test('Should find link within angle brackets even with surrounding spaces', async () => {
 		const links = await getLinksForText(joinLines(
 			`[link]( <path> )`
 		));
 		assertLinksEqual(links, [lsp.Range.create(0, 9, 0, 13)]);
 	});
 
-	test('Should find link within angle brackets for image hyperlinks.', async () => {
+	test('Should find link within angle brackets for image hyperlinks', async () => {
 		const links = await getLinksForText(joinLines(
 			`![link](<path>)`
 		));
 		assertLinksEqual(links, [lsp.Range.create(0, 9, 0, 13)]);
 	});
 
-	test('Should find link with spaces in angle brackets for image hyperlinks with titles.', async () => {
+	test('Should find link with spaces in angle brackets for image hyperlinks with titles', async () => {
 		const links = await getLinksForText(joinLines(
 			`![link](< path > "test")`
 		));
 		assertLinksEqual(links, [lsp.Range.create(0, 9, 0, 15)]);
 	});
 
 
-	test('Should not find link due to incorrect angle bracket notation or usage.', async () => {
+	test('Should not find link due to incorrect angle bracket notation or usage', async () => {
 		const links = await getLinksForText(joinLines(
 			`[link](<path )`,
 			`[link](<> path>)`,
@@ -540,14 +540,22 @@ suite('Link computer', () => {
 		assertLinksEqual(links, []);
 	});
 
-	test('Should find link within angle brackets even with space inside link.', async () => {
+	test('Should find link within angle brackets even with space inside link', async () => {
 		const links = await getLinksForText(joinLines(
 			`[link](<pa th>)`
 		));
 
 		assertLinksEqual(links, [lsp.Range.create(0, 8, 0, 13)]);
 	});
 
+	test('Should find link within angle brackets with spaces in fragment', async () => {
+		const links = await getLinksForText(joinLines(
+			`[link](<pa th#fr agment>)`
+		));
+
+		assertLinksEqual(links, [lsp.Range.create(0, 8, 0, 23)]);
+	});
+
 	test('Should find links with titles', async () => {
 		const links = await getLinksForText(joinLines(
 			`[link](<no such.md> "text")`,

src/test/tableOfContents.test.ts

@@ -25,8 +25,8 @@ suite('Table of contents', () => {
 		const doc = new InMemoryDocument(testFileName, '');
 		const provider = await createToc(doc);
 
-		assert.strictEqual(provider.lookup(''), undefined);
-		assert.strictEqual(provider.lookup('foo'), undefined);
+		assert.strictEqual(provider.lookupByFragment(''), undefined);
+		assert.strictEqual(provider.lookupByFragment('foo'), undefined);
 	});
 
 	test('Lookup should not return anything for document with no headers', async () => {
@@ -36,10 +36,10 @@ suite('Table of contents', () => {
 		));
 		const provider = await createToc(doc);
 
-		assert.strictEqual(provider.lookup(''), undefined);
-		assert.strictEqual(provider.lookup('foo'), undefined);
-		assert.strictEqual(provider.lookup('a'), undefined);
-		assert.strictEqual(provider.lookup('b'), undefined);
+		assert.strictEqual(provider.lookupByFragment(''), undefined);
+		assert.strictEqual(provider.lookupByFragment('foo'), undefined);
+		assert.strictEqual(provider.lookupByFragment('a'), undefined);
+		assert.strictEqual(provider.lookupByFragment('b'), undefined);
 	});
 
 	test('Lookup should return basic #header', async () => {
@@ -51,15 +51,15 @@ suite('Table of contents', () => {
 		const provider = await createToc(doc);
 
 		{
-			const entry = provider.lookup('a');
+			const entry = provider.lookupByFragment('a');
 			assert.ok(entry);
 			assert.strictEqual(entry!.line, 0);
 		}
 		{
-			assert.strictEqual(provider.lookup('x'), undefined);
+			assert.strictEqual(provider.lookupByFragment('x'), undefined);
 		}
 		{
-			const entry = provider.lookup('c');
+			const entry = provider.lookupByFragment('c');
 			assert.ok(entry);
 			assert.strictEqual(entry!.line, 2);
 		}
@@ -72,9 +72,9 @@ suite('Table of contents', () => {
 		));
 		const provider = await createToc(doc);
 
-		assert.strictEqual((provider.lookup('fOo'))!.line, 0);
-		assert.strictEqual((provider.lookup('foo'))!.line, 0);
-		assert.strictEqual((provider.lookup('FOO'))!.line, 0);
+		assert.strictEqual((provider.lookupByFragment('fOo'))!.line, 0);
+		assert.strictEqual((provider.lookupByFragment('foo'))!.line, 0);
+		assert.strictEqual((provider.lookupByFragment('FOO'))!.line, 0);
 	});
 
 	test('should handle special characters #44779', async () => {
@@ -84,7 +84,7 @@ suite('Table of contents', () => {
 		));
 		const provider = await createToc(doc);
 
-		assert.strictEqual((provider.lookup('indentação'))!.line, 0);
+		assert.strictEqual((provider.lookupByFragment('indentação'))!.line, 0);
 	});
 
 	test('should handle special characters 2, #48482', async () => {
@@ -94,7 +94,7 @@ suite('Table of contents', () => {
 		));
 		const provider = await createToc(doc);
 
-		assert.strictEqual((provider.lookup('инструкция---делай-раз-делай-два'))!.line, 0);
+		assert.strictEqual((provider.lookupByFragment('инструкция---делай-раз-делай-два'))!.line, 0);
 	});
 
 	test('should handle special characters 3, #37079', async () => {
@@ -109,12 +109,12 @@ suite('Table of contents', () => {
 
 		const provider = await createToc(doc);
 
-		assert.strictEqual((provider.lookup('header-2'))!.line, 0);
-		assert.strictEqual((provider.lookup('header-3'))!.line, 1);
-		assert.strictEqual((provider.lookup('Заголовок-2'))!.line, 2);
-		assert.strictEqual((provider.lookup('Заголовок-3'))!.line, 3);
-		assert.strictEqual((provider.lookup('Заголовок-header-3'))!.line, 4);
-		assert.strictEqual((provider.lookup('Заголовок'))!.line, 5);
+		assert.strictEqual((provider.lookupByFragment('header-2'))!.line, 0);
+		assert.strictEqual((provider.lookupByFragment('header-3'))!.line, 1);
+		assert.strictEqual((provider.lookupByFragment('Заголовок-2'))!.line, 2);
+		assert.strictEqual((provider.lookupByFragment('Заголовок-3'))!.line, 3);
+		assert.strictEqual((provider.lookupByFragment('Заголовок-header-3'))!.line, 4);
+		assert.strictEqual((provider.lookupByFragment('Заголовок'))!.line, 5);
 	});
 
 	test('Lookup should support suffixes for repeated headers', async () => {
@@ -126,17 +126,17 @@ suite('Table of contents', () => {
 		const provider = await createToc(doc);
 
 		{
-			const entry = provider.lookup('a');
+			const entry = provider.lookupByFragment('a');
 			assert.ok(entry);
 			assert.strictEqual(entry!.line, 0);
 		}
 		{
-			const entry = provider.lookup('a-1');
+			const entry = provider.lookupByFragment('a-1');
 			assert.ok(entry);
 			assert.strictEqual(entry!.line, 1);
 		}
 		{
-			const entry = provider.lookup('a-2');
+			const entry = provider.lookupByFragment('a-2');
 			assert.ok(entry);
 			assert.strictEqual(entry!.line, 2);
 		}