fix: refine paragraph boundaries for inline contents

[3w36zj6]

cf. #31 (comment)

[Copilot]

Pull request overview

This PR refines how Typst inline/script content is paragraphized in the generated textlint AST, addressing the paragraph-boundary bugs surfaced by the integration tests added in #31 (comments, nested list items, term lists, and inline content blocks).

Changes:

• Reworked paragraphization to better split/merge inline vs non-inline nodes (including hash statements, term list bodies, and content blocks) and updated node conversions for labels/refs and linebreak escapes.
• Updated integration tests that were previously marked as expected failures, and added/updated fixture-based unit coverage for the new paragraph boundary behavior.
• Regenerated/adjusted expected AST fixture outputs to match the refined paragraphization.

Reviewed changes

Copilot reviewed 32 out of 32 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
src/typstToTextlintAst.ts	Core changes: refine paragraph boundary detection/splitting, content-block handling, list/term normalization, and inline label/ref conversions.
test/integration/linting.test.ts	Enables previously failing integration tests (comments immediately followed by elements, nested lists, term list violations).
test/unit/paragraphizedTextlintAstObject.json	Updates expected paragraphized AST snapshot to reflect trimming/refined paragraph boundaries.
test/unit/fixtures/unsupported-node-paragraph-split/input.typ	Adds fixture input for splitting paragraphs around unsupported nodes in content blocks.
test/unit/fixtures/unsupported-node-paragraph-split/output.json	Expected AST output for unsupported-node paragraph splitting behavior.
test/unit/fixtures/paragraph-between-term-list/output.json	Updates expected term list handling so term bodies are paragraphized consistently.
test/unit/fixtures/list-item-split-hash-statement/input.typ	Adds fixture input for list item splitting around hash statements.
test/unit/fixtures/list-item-split-hash-statement/output.json	Expected AST output for list-item hash-statement splitting.
test/unit/fixtures/inline-text-space-linebreak/input.typ	Adds fixture input for forced line break (\\ + newline) staying inline within a paragraph.
test/unit/fixtures/inline-text-space-linebreak/output.json	Expected AST output for inline forced line break handling.
test/unit/fixtures/inline-raw-and-equation/input.typ	Adds fixture input for inline raw/code and inline equations staying within one paragraph.
test/unit/fixtures/inline-raw-and-equation/output.json	Expected AST output for inline raw/equation paragraphization.
test/unit/fixtures/inline-label-ref/input.typ	Adds fixture input for inline label + reference use in same paragraph.
test/unit/fixtures/inline-label-ref/output.json	Expected AST output for label/ref nodes represented inline (as code nodes).
test/unit/fixtures/inline-escape-shorthand-smartquote/input.typ	Adds fixture input covering escapes, shorthand, and smart quotes in a single paragraph.
test/unit/fixtures/inline-escape-shorthand-smartquote/output.json	Expected AST output for escape/shorthand/smartquote inline behavior.
test/unit/fixtures/inline-empty-content-block-inline/input.typ	Adds fixture input for empty inline content block #[] behavior.
test/unit/fixtures/inline-empty-content-block-inline/output.json	Expected AST output for empty inline content block behavior.
test/unit/fixtures/inline-content-block-with-non-inline/input.typ	Adds fixture input for inline content block that contains non-inline nodes (e.g., table).
test/unit/fixtures/inline-content-block-with-non-inline/output.json	Expected AST output for inline content block splitting around non-inline content.
test/unit/fixtures/inline-content-block-inline/input.typ	Adds fixture input for a purely inline content block #[inline content].
test/unit/fixtures/inline-content-block-inline/output.json	Expected AST output for purely inline content block paragraphization.
test/unit/fixtures/figure-table-block/input.typ	Adds fixture input for figure/table block behavior and boundaries.
test/unit/fixtures/figure-table-block/output.json	Expected AST output for figure/table block structure under new paragraph rules.
test/unit/fixtures/content-blocks/output.json	Updates expected AST for #{ ... } content blocks (avoids wrapping code blocks in Paragraph).
test/unit/fixtures/comment-followed-hash-statement/input.typ	Adds fixture input for comment immediately followed by a hash statement.
test/unit/fixtures/comment-followed-hash-statement/output.json	Expected AST output for comment + following hash-statement boundary handling.
test/unit/fixtures/Strong/output.json	Updates expected output for strong/emphasis + forced linebreak behavior within paragraphs.
test/unit/fixtures/Emphasis/output.json	Updates expected output for emphasis + forced linebreak behavior within paragraphs.
test/unit/fixtures/Break/output.json	Updates expected output to keep forced line breaks inline rather than separate Break nodes.
test/unit/fixtures/Equation/output.json	Updates expected output for equation-adjacent paragraph boundaries.
test/unit/fixtures/List/output.json	Updates expected list-related outputs impacted by paragraphization changes.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 32 out of 32 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 33 out of 33 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 33 out of 33 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (1)

src/typstToTextlintAst.ts:1078

• includesLineBreak treats any Str containing \n as a statement boundary. After this PR, forced inline line breaks are represented as Str nodes (e.g. "\" + "\n"), so a #... expression immediately after a forced line break may be mis-detected as a new hash statement and split out of the paragraph. Consider teaching isStatementBoundaryBefore to ignore the "\\" + "\n" inline-linebreak pattern (or use a dedicated node/type/flag for inline line breaks) so only real statement boundaries trigger hash-statement collection/splitting.

	const includesLineBreak = (n: Content): boolean => {
		if (n.type === ASTNodeTypes.Str) {
			return n.raw.includes("\n");
		}
		if (n.type === ASTNodeTypes.Break) {
			return n.raw.includes("\n");
		}
		return false;
	};

	const isStatementBoundaryBefore = (
		arr: Content[],
		index: number,
	): boolean => {
		if (index === 0) {
			return true;
		}
		return includesLineBreak(arr[index - 1]);
	};

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 33 out of 33 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 35 out of 35 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[3w36zj6]

Could you please review this? @r4ai

[r4ai]

Could you please review this? @r4ai

LGTM!