Files
Moltbot/src/signal/format.chunking.test.ts
Hudson 1d6abddb9f fix(signal): outbound formatting and markdown IR rendering improvements (#9781)
* fix: Signal and markdown formatting improvements

Markdown IR fixes:
- Fix list-paragraph spacing (extra newline between list items and following paragraphs)
- Fix nested list indentation and newline handling
- Fix blockquote_close emitting redundant newline (inner content handles spacing)
- Render horizontal rules as visible ─── separator instead of silent drop
- Strip inner cell styles in code-mode tables to prevent overlapping with code_block span

Signal formatting fixes:
- Normalize URLs for dedup comparison (strip protocol, www., trailing slash)
- Render headings as bold text (headingStyle: 'bold')
- Add '> ' prefix to blockquotes for visual distinction
- Re-chunk after link expansion to respect chunk size limits

Tests:
- 51 new tests for markdown IR (spacing, lists, blockquotes, tables, HR)
- 18 new tests for Signal formatting (URL dedup, headings, blockquotes, HR, chunking)
- Update Slack nested list test expectation to match corrected IR output

* refactor: style-aware Signal text chunker

Replace indexOf-based chunk position tracking with deterministic
cursor tracking. The new splitSignalFormattedText:

- Splits at whitespace/newline boundaries within the limit
- Avoids breaking inside parentheses (preserves expanded link URLs)
- Slices style ranges at chunk boundaries with correct local offsets
- Tracks position via offset arithmetic instead of fragile indexOf

Removes dependency on chunkText from auto-reply/chunk.

Tests: 19 new tests covering style preservation across chunk boundaries,
edge cases (empty text, under limit, exact split points), and integration
with link expansion.

* fix: correct Signal style offsets with multiple link expansions

applyInsertionsToStyles() was using original coordinates for each
insertion without tracking cumulative shift from prior insertions.
This caused bold/italic/etc styles to drift to wrong text positions
when multiple markdown links expanded in a single message.

Added cumulative shift tracking and a regression test.

* test: clean up test noise and fix ineffective assertions

- Remove console.log from ir.list-spacing and ir.hr-spacing tests
- Fix ir.nested-lists.test.ts: remove ineffective regex assertion
- Fix ir.hr-spacing.test.ts: add actual assertions to edge case test

* refactor: split Signal formatting tests (#9781) (thanks @heyhudson)

---------

Co-authored-by: Hudson <258693705+hudson-rivera@users.noreply.github.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
2026-02-14 16:57:20 +01:00

391 lines
16 KiB
TypeScript

import { describe, expect, it } from "vitest";
import { markdownToSignalTextChunks } from "./format.js";
describe("splitSignalFormattedText", () => {
// We test the internal chunking behavior via markdownToSignalTextChunks with
// pre-rendered SignalFormattedText. The helper is not exported, so we test
// it indirectly through integration tests and by constructing scenarios that
// exercise the splitting logic.
describe("style-aware splitting - basic text", () => {
it("text with no styles splits correctly at whitespace", () => {
// Create text that exceeds limit and must be split
const limit = 20;
const markdown = "hello world this is a test";
const chunks = markdownToSignalTextChunks(markdown, limit);
expect(chunks.length).toBeGreaterThan(1);
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
// Verify all text is preserved (joined chunks should contain all words)
const joinedText = chunks.map((c) => c.text).join(" ");
expect(joinedText).toContain("hello");
expect(joinedText).toContain("world");
expect(joinedText).toContain("test");
});
it("empty text returns empty array", () => {
// Empty input produces no chunks (not an empty chunk)
const chunks = markdownToSignalTextChunks("", 100);
expect(chunks).toEqual([]);
});
it("text under limit returns single chunk unchanged", () => {
const markdown = "short text";
const chunks = markdownToSignalTextChunks(markdown, 100);
expect(chunks).toHaveLength(1);
expect(chunks[0].text).toBe("short text");
});
});
describe("style-aware splitting - style preservation", () => {
it("style fully within first chunk stays in first chunk", () => {
// Create a message where bold text is in the first chunk
const limit = 30;
const markdown = "**bold** word more words here that exceed limit";
const chunks = markdownToSignalTextChunks(markdown, limit);
expect(chunks.length).toBeGreaterThan(1);
// First chunk should contain the bold style
const firstChunk = chunks[0];
expect(firstChunk.text).toContain("bold");
expect(firstChunk.styles.some((s) => s.style === "BOLD")).toBe(true);
// The bold style should start at position 0 in the first chunk
const boldStyle = firstChunk.styles.find((s) => s.style === "BOLD");
expect(boldStyle).toBeDefined();
expect(boldStyle!.start).toBe(0);
expect(boldStyle!.length).toBe(4); // "bold"
});
it("style fully within second chunk has offset adjusted to chunk-local position", () => {
// Create a message where the styled text is in the second chunk
const limit = 30;
const markdown = "some filler text here **bold** at the end";
const chunks = markdownToSignalTextChunks(markdown, limit);
expect(chunks.length).toBeGreaterThan(1);
// Find the chunk containing "bold"
const chunkWithBold = chunks.find((c) => c.text.includes("bold"));
expect(chunkWithBold).toBeDefined();
expect(chunkWithBold!.styles.some((s) => s.style === "BOLD")).toBe(true);
// The bold style should have chunk-local offset (not original text offset)
const boldStyle = chunkWithBold!.styles.find((s) => s.style === "BOLD");
expect(boldStyle).toBeDefined();
// The offset should be the position within this chunk, not the original text
const boldPos = chunkWithBold!.text.indexOf("bold");
expect(boldStyle!.start).toBe(boldPos);
expect(boldStyle!.length).toBe(4);
});
it("style spanning chunk boundary is split into two ranges", () => {
// Create text where a styled span crosses the chunk boundary
const limit = 15;
// "hello **bold text here** end" - the bold spans across chunk boundary
const markdown = "hello **boldtexthere** end";
const chunks = markdownToSignalTextChunks(markdown, limit);
expect(chunks.length).toBeGreaterThan(1);
// Both chunks should have BOLD styles if the span was split
const chunksWithBold = chunks.filter((c) => c.styles.some((s) => s.style === "BOLD"));
// At least one chunk should have the bold style
expect(chunksWithBold.length).toBeGreaterThanOrEqual(1);
// For each chunk with bold, verify the style range is valid for that chunk
for (const chunk of chunksWithBold) {
for (const style of chunk.styles.filter((s) => s.style === "BOLD")) {
expect(style.start).toBeGreaterThanOrEqual(0);
expect(style.start + style.length).toBeLessThanOrEqual(chunk.text.length);
}
}
});
it("style starting exactly at split point goes entirely to second chunk", () => {
// Create text where style starts right at where we'd split
const limit = 10;
const markdown = "abcdefghi **bold**";
const chunks = markdownToSignalTextChunks(markdown, limit);
expect(chunks.length).toBeGreaterThan(1);
// Find chunk with bold
const chunkWithBold = chunks.find((c) => c.styles.some((s) => s.style === "BOLD"));
expect(chunkWithBold).toBeDefined();
// Verify the bold style is valid within its chunk
const boldStyle = chunkWithBold!.styles.find((s) => s.style === "BOLD");
expect(boldStyle).toBeDefined();
expect(boldStyle!.start).toBeGreaterThanOrEqual(0);
expect(boldStyle!.start + boldStyle!.length).toBeLessThanOrEqual(chunkWithBold!.text.length);
});
it("style ending exactly at split point stays entirely in first chunk", () => {
const limit = 10;
const markdown = "**bold** rest of text";
const chunks = markdownToSignalTextChunks(markdown, limit);
// First chunk should have the complete bold style
const firstChunk = chunks[0];
if (firstChunk.text.includes("bold")) {
const boldStyle = firstChunk.styles.find((s) => s.style === "BOLD");
expect(boldStyle).toBeDefined();
expect(boldStyle!.start + boldStyle!.length).toBeLessThanOrEqual(firstChunk.text.length);
}
});
it("multiple styles, some spanning boundary, some not", () => {
const limit = 25;
// Mix of styles: italic at start, bold spanning boundary, monospace at end
const markdown = "_italic_ some text **bold text** and `code`";
const chunks = markdownToSignalTextChunks(markdown, limit);
expect(chunks.length).toBeGreaterThan(1);
// Verify all style ranges are valid within their respective chunks
for (const chunk of chunks) {
for (const style of chunk.styles) {
expect(style.start).toBeGreaterThanOrEqual(0);
expect(style.start + style.length).toBeLessThanOrEqual(chunk.text.length);
expect(style.length).toBeGreaterThan(0);
}
}
// Collect all styles across chunks
const allStyles = chunks.flatMap((c) => c.styles.map((s) => s.style));
// We should have at least italic, bold, and monospace somewhere
expect(allStyles).toContain("ITALIC");
expect(allStyles).toContain("BOLD");
expect(allStyles).toContain("MONOSPACE");
});
});
describe("style-aware splitting - edge cases", () => {
it("handles zero-length text with styles gracefully", () => {
// Edge case: empty markdown produces no chunks
const chunks = markdownToSignalTextChunks("", 100);
expect(chunks).toHaveLength(0);
});
it("handles text that splits exactly at limit", () => {
const limit = 10;
const markdown = "1234567890"; // exactly 10 chars
const chunks = markdownToSignalTextChunks(markdown, limit);
expect(chunks).toHaveLength(1);
expect(chunks[0].text).toBe("1234567890");
});
it("preserves style through whitespace trimming", () => {
const limit = 30;
const markdown = "**bold** some text that is longer than limit";
const chunks = markdownToSignalTextChunks(markdown, limit);
// Bold should be preserved in first chunk
const firstChunk = chunks[0];
if (firstChunk.text.includes("bold")) {
expect(firstChunk.styles.some((s) => s.style === "BOLD")).toBe(true);
}
});
it("handles repeated substrings correctly (no indexOf fragility)", () => {
// This test exposes the fragility of using indexOf to find chunk positions.
// If the same substring appears multiple times, indexOf finds the first
// occurrence, not necessarily the correct one.
const limit = 20;
// "word" appears multiple times - indexOf("word") would always find first
const markdown = "word **bold word** word more text here to chunk";
const chunks = markdownToSignalTextChunks(markdown, limit);
// Verify chunks are under limit
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
// Find chunk(s) with bold style
const chunksWithBold = chunks.filter((c) => c.styles.some((s) => s.style === "BOLD"));
expect(chunksWithBold.length).toBeGreaterThanOrEqual(1);
// The bold style should correctly cover "bold word" (or part of it if split)
// and NOT incorrectly point to the first "word" in the text
for (const chunk of chunksWithBold) {
for (const style of chunk.styles.filter((s) => s.style === "BOLD")) {
const styledText = chunk.text.slice(style.start, style.start + style.length);
// The styled text should be part of "bold word", not the initial "word"
expect(styledText).toMatch(/^(bold( word)?|word)$/);
expect(style.start).toBeGreaterThanOrEqual(0);
expect(style.start + style.length).toBeLessThanOrEqual(chunk.text.length);
}
}
});
it("handles chunk that starts with whitespace after split", () => {
// When text is split at whitespace, the next chunk might have leading
// whitespace trimmed. Styles must account for this.
const limit = 15;
const markdown = "some text **bold** at end";
const chunks = markdownToSignalTextChunks(markdown, limit);
// All style ranges must be valid
for (const chunk of chunks) {
for (const style of chunk.styles) {
expect(style.start).toBeGreaterThanOrEqual(0);
expect(style.start + style.length).toBeLessThanOrEqual(chunk.text.length);
}
}
});
it("deterministically tracks position without indexOf fragility", () => {
// This test ensures the chunker doesn't rely on finding chunks via indexOf
// which can fail when chunkText trims whitespace or when duplicates exist.
// Create text with lots of whitespace and repeated patterns.
const limit = 25;
const markdown = "aaa **bold** aaa **bold** aaa extra text to force split";
const chunks = markdownToSignalTextChunks(markdown, limit);
// Multiple chunks expected
expect(chunks.length).toBeGreaterThan(1);
// All chunks should respect limit
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
// All style ranges must be valid within their chunks
for (const chunk of chunks) {
for (const style of chunk.styles) {
expect(style.start).toBeGreaterThanOrEqual(0);
expect(style.start + style.length).toBeLessThanOrEqual(chunk.text.length);
// The styled text at that position should actually be "bold"
if (style.style === "BOLD") {
const styledText = chunk.text.slice(style.start, style.start + style.length);
expect(styledText).toBe("bold");
}
}
}
});
});
});
describe("markdownToSignalTextChunks", () => {
describe("link expansion chunk limit", () => {
it("does not exceed chunk limit after link expansion", () => {
// Create text that is close to limit, with a link that will expand
const limit = 100;
// Create text that's 90 chars, leaving only 10 chars of headroom
const filler = "x".repeat(80);
// This link will expand from "[link](url)" to "link (https://example.com/very/long/path)"
const markdown = `${filler} [link](https://example.com/very/long/path/that/will/exceed/limit)`;
const chunks = markdownToSignalTextChunks(markdown, limit);
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
});
it("handles multiple links near chunk boundary", () => {
const limit = 100;
const filler = "x".repeat(60);
const markdown = `${filler} [a](https://a.com) [b](https://b.com) [c](https://c.com)`;
const chunks = markdownToSignalTextChunks(markdown, limit);
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
});
});
describe("link expansion with style preservation", () => {
it("long message with links that expand beyond limit preserves all text", () => {
const limit = 80;
const filler = "a".repeat(50);
const markdown = `${filler} [click here](https://example.com/very/long/path/to/page) more text`;
const chunks = markdownToSignalTextChunks(markdown, limit);
// All chunks should be under limit
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
// Combined text should contain all original content
const combined = chunks.map((c) => c.text).join("");
expect(combined).toContain(filler);
expect(combined).toContain("click here");
expect(combined).toContain("example.com");
});
it("styles (bold, italic) survive chunking correctly after link expansion", () => {
const limit = 60;
const markdown =
"**bold start** text [link](https://example.com/path) _italic_ more content here to force chunking";
const chunks = markdownToSignalTextChunks(markdown, limit);
// Should have multiple chunks
expect(chunks.length).toBeGreaterThan(1);
// All style ranges should be valid within their chunks
for (const chunk of chunks) {
for (const style of chunk.styles) {
expect(style.start).toBeGreaterThanOrEqual(0);
expect(style.start + style.length).toBeLessThanOrEqual(chunk.text.length);
expect(style.length).toBeGreaterThan(0);
}
}
// Verify styles exist somewhere
const allStyles = chunks.flatMap((c) => c.styles.map((s) => s.style));
expect(allStyles).toContain("BOLD");
expect(allStyles).toContain("ITALIC");
});
it("multiple links near chunk boundary all get properly chunked", () => {
const limit = 50;
const markdown =
"[first](https://first.com/long/path) [second](https://second.com/another/path) [third](https://third.com)";
const chunks = markdownToSignalTextChunks(markdown, limit);
// All chunks should respect limit
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
// All link labels should appear somewhere
const combined = chunks.map((c) => c.text).join("");
expect(combined).toContain("first");
expect(combined).toContain("second");
expect(combined).toContain("third");
});
it("preserves spoiler style through link expansion and chunking", () => {
const limit = 40;
const markdown =
"||secret content|| and [link](https://example.com/path) with more text to chunk";
const chunks = markdownToSignalTextChunks(markdown, limit);
// All chunks should respect limit
for (const chunk of chunks) {
expect(chunk.text.length).toBeLessThanOrEqual(limit);
}
// Spoiler style should exist and be valid
const chunkWithSpoiler = chunks.find((c) => c.styles.some((s) => s.style === "SPOILER"));
expect(chunkWithSpoiler).toBeDefined();
const spoilerStyle = chunkWithSpoiler!.styles.find((s) => s.style === "SPOILER");
expect(spoilerStyle).toBeDefined();
expect(spoilerStyle!.start).toBeGreaterThanOrEqual(0);
expect(spoilerStyle!.start + spoilerStyle!.length).toBeLessThanOrEqual(
chunkWithSpoiler!.text.length,
);
});
});
});