davidl/tolaria
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: strip list markers from snippets + bump cache version for full rescan

Browse Source 
Snippet extraction was including raw list markers (* , - , + , 1. ) in
the preview text, producing ugly leading spaces. Notes whose snippets
were cached before this fix showed stale/incorrect previews.

- Add strip_list_marker() in both Rust and TS to remove bullet/ordered
  list prefixes before snippet assembly
- Trim final snippet to remove leading/trailing whitespace
- Bump CACHE_VERSION 6 → 7 to force full rescan on next vault load,
  ensuring all entries get clean snippets
- Add Rust + Vitest tests for list marker stripping
- Update Playwright smoke test with stricter snippet assertions

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Test
2026-03-16 02:59:43 +01:00
parent 8684b36bba
commit ee556aa004
5 changed files with 140 additions and 53 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src-tauri/src/vault/cache.rs
View File

@@ -8,7 +8,7 @@ use super::{parse_md_file, scan_vault, VaultEntry};
// --- Vault Cache ---
/// Bump this when VaultEntry fields change to force a full rescan.
const CACHE_VERSION: u32 = 6;
const CACHE_VERSION: u32 = 7;
#[derive(Debug, Serialize, Deserialize)]
struct VaultCache {

75
src-tauri/src/vault/parsing.rs
View File

@@ -40,6 +40,24 @@ fn is_snippet_line(line: &str) -> bool {
!t.is_empty() && !t.starts_with('#') && !t.starts_with("```") && !t.starts_with("---")
}
/// Strip leading list markers (*, -, +, 1.) from a line.
fn strip_list_marker(line: &str) -> &str {
let t = line.trim_start();
// Unordered: "* ", "- ", "+ "
for prefix in &["* ", "- ", "+ "] {
if let Some(rest) = t.strip_prefix(prefix) {
return rest;
}
}
// Ordered: "1. ", "2. ", etc.
if let Some(dot_pos) = t.find(". ") {
if dot_pos <= 3 && t[..dot_pos].chars().all(|c| c.is_ascii_digit()) {
return &t[dot_pos + 2..];
}
}
t
}
/// Truncate a string to `max_len` bytes at a valid UTF-8 boundary, appending "...".
fn truncate_with_ellipsis(s: &str, max_len: usize) -> String {
if s.len() <= max_len {
@@ -71,9 +89,15 @@ pub(super) fn extract_snippet(content: &str) -> String {
let clean: String = body
.lines()
.filter(|line| is_snippet_line(line))
.map(strip_list_marker)
.collect::<Vec<&str>>()
.join(" ");
truncate_with_ellipsis(&strip_markdown_chars(&clean), 160)
let stripped = strip_markdown_chars(&clean);
let trimmed = stripped.trim();
if trimmed.is_empty() {
return String::new();
}
truncate_with_ellipsis(trimmed, 160)
}
fn without_h1_line(s: &str) -> Option<&str> {
@@ -332,6 +356,55 @@ mod tests {
assert_eq!(snippet, "Content after rule.");
}
// --- strip_list_marker tests ---
#[test]
fn test_strip_list_marker_unordered() {
assert_eq!(strip_list_marker("* Item one"), "Item one");
assert_eq!(strip_list_marker("- Item two"), "Item two");
assert_eq!(strip_list_marker("+ Item three"), "Item three");
}
#[test]
fn test_strip_list_marker_ordered() {
assert_eq!(strip_list_marker("1. First item"), "First item");
assert_eq!(strip_list_marker("10. Tenth item"), "Tenth item");
assert_eq!(strip_list_marker("99. Large number"), "Large number");
}
#[test]
fn test_strip_list_marker_preserves_non_list() {
assert_eq!(strip_list_marker("Regular text"), "Regular text");
assert_eq!(strip_list_marker(" Indented text"), "Indented text");
}
#[test]
fn test_extract_snippet_strips_list_markers() {
let content =
"---\ntype: Project\n---\n# My Project\n\n* First bullet\n* Second bullet\n- Dash item";
let snippet = extract_snippet(content);
assert_eq!(snippet, "First bullet Second bullet Dash item");
}
#[test]
fn test_extract_snippet_mixed_headings_and_bullets() {
let content = "---\ntype: Project\nstatus: Active\n---\n# Migrate newsletter to Beehiiv\n\n### 1) Newsletter is 100% on Beehiiv\n\n* Migration is successful\n\n### 2) Open rate is >27%\n\n* No regressions on open rate";
let snippet = extract_snippet(content);
assert!(
snippet.starts_with("Migration is successful"),
"snippet should start with first bullet content, got: {}",
snippet
);
assert!(snippet.contains("No regressions on open rate"));
}
#[test]
fn test_extract_snippet_ordered_list() {
let content = "# Title\n\n1. First step\n2. Second step\n3. Third step";
let snippet = extract_snippet(content);
assert_eq!(snippet, "First step Second step Third step");
}
// --- count_body_words tests ---
#[test]

26
src/utils/wikilinks.test.ts
View File

@@ -565,11 +565,29 @@ describe('extractSnippet', () => {
expect(snippet).toContain('Ship the minimum viable product')
})
it('includes list items in snippet', () => {
const content = '# Title\n\n- First item\n- Second item\n- Third item'
it('strips list markers from bullet items', () => {
const content = '# Title\n\n* First bullet\n* Second bullet\n- Dash item'
const snippet = extractSnippet(content)
expect(snippet).toContain('First item')
expect(snippet).toContain('Second item')
expect(snippet).toBe('First bullet Second bullet Dash item')
})
it('strips ordered list markers', () => {
const content = '# Title\n\n1. First step\n2. Second step\n3. Third step'
const snippet = extractSnippet(content)
expect(snippet).toBe('First step Second step Third step')
})
it('handles mixed headings and bullet lists (real-world format)', () => {
const content = '---\ntype: Project\nstatus: Active\n---\n# Migrate newsletter\n\n### 1) Goal one\n\n* Migration is successful\n\n### 2) Goal two\n\n* No regressions on open rate'
const snippet = extractSnippet(content)
expect(snippet).toMatch(/^Migration is successful/)
expect(snippet).toContain('No regressions on open rate')
})
it('trims leading/trailing whitespace from snippet', () => {
const content = '# Title\n\n Some text with spaces \n'
const snippet = extractSnippet(content)
expect(snippet).toBe('Some text with spaces')
})
it('includes code content lines (not fences) in snippet', () => {

18
src/utils/wikilinks.ts
View File

@@ -160,6 +160,19 @@ function isSnippetLine(line: string): boolean {
return t !== '' && !t.startsWith('#') && !t.startsWith('```') && !t.startsWith('---')
}
/** Strip leading list markers (*, -, +, 1.) from a line. */
function stripListMarker(line: string): string {
const t = line.trimStart()
for (const prefix of ['* ', '- ', '+ ']) {
if (t.startsWith(prefix)) return t.slice(prefix.length)
}
const dotPos = t.indexOf('. ')
if (dotPos >= 1 && dotPos <= 3 && /^\d+$/.test(t.slice(0, dotPos))) {
return t.slice(dotPos + 2)
}
return t
}
/** Remove the first H1 heading line, allowing leading blank lines. */
function removeH1Line(body: string): string {
const lines = body.split('\n')
@@ -204,8 +217,9 @@ function stripMarkdownChars(s: string): string {
export function extractSnippet(content: string): string {
const [, body] = splitFrontmatter(content)
const withoutH1 = removeH1Line(body)
const clean = withoutH1.split('\n').filter(isSnippetLine).join(' ')
const stripped = stripMarkdownChars(clean)
const clean = withoutH1.split('\n').filter(isSnippetLine).map(stripListMarker).join(' ')
const stripped = stripMarkdownChars(clean).trim()
if (!stripped) return ''
if (stripped.length <= 160) return stripped
return stripped.slice(0, 160) + '...'
}

72
tests/smoke/note-list-preview-snippet.spec.ts
View File

@@ -7,69 +7,51 @@ test.describe('Note list preview snippet', () => {
})
test('notes with content show a snippet in the note list', async ({ page }) => {
// The note list should be visible with mock entries that have snippets
// The note list should be visible with entries that have snippets
const noteListContainer = page.locator('[data-testid="note-list-container"]')
await expect(noteListContainer).toBeVisible()
// Check that at least one snippet is rendered (mock entries all have snippets)
// The snippet text lives in a muted-foreground div inside each note item
const snippetElements = page.locator('.text-muted-foreground').filter({
hasText: /\w{10,}/, // at least 10 word-chars → real snippet text
// The snippet text lives in a 12px muted-foreground div inside each note item.
// At least one must contain real text (10+ word-chars) — not just a date.
const snippetElements = noteListContainer.locator('.text-\\[12px\\]').filter({
hasText: /\w{10,}/,
})
const count = await snippetElements.count()
expect(count).toBeGreaterThan(0)
})
test('snippet updates after editing and saving a note', async ({ page }) => {
// Click the first note to open it in the editor
const firstNote = page.locator('[data-testid="note-list-container"]').locator('.cursor-pointer').first()
await firstNote.click()
// Wait for editor to load
const editor = page.locator('[data-testid="editor-container"], .bn-editor, .ProseMirror').first()
await expect(editor).toBeVisible({ timeout: 5000 })
// Type some unique content
const uniqueText = `Snippet test content ${Date.now()}`
await editor.click()
await page.keyboard.press('End')
await page.keyboard.press('Enter')
await page.keyboard.type(uniqueText, { delay: 10 })
// Save with Cmd+S
await page.keyboard.press('Control+s')
// Wait for save to complete
await page.waitForTimeout(500)
// The snippet in the note list should now contain our text (or at least be non-empty)
// The note list item should have a snippet div with content
const noteItem = page.locator('[data-testid="note-list-container"]').locator('.cursor-pointer').first()
const snippetDiv = noteItem.locator('.text-muted-foreground').first()
const snippetText = await snippetDiv.textContent()
expect(snippetText).toBeTruthy()
expect(snippetText!.length).toBeGreaterThan(0)
})
test('snippet is stripped of markdown formatting', async ({ page }) => {
// The mock entry "Kitchen Sink" has bold, italic, code, etc. in its snippet source
// but the extracted snippet should not contain markdown chars like ** or *
test('snippet does not contain raw markdown formatting', async ({ page }) => {
const noteListContainer = page.locator('[data-testid="note-list-container"]')
await expect(noteListContainer).toBeVisible()
// Look for the "Kitchen Sink" note or any note with a snippet
// Verify no raw markdown chars appear in snippet text
const allSnippets = page.locator('[data-testid="note-list-container"] .text-muted-foreground')
const snippetCount = await allSnippets.count()
// Check snippet divs for absence of raw markdown
const snippetDivs = noteListContainer.locator('.text-\\[12px\\]')
const snippetCount = await snippetDivs.count()
for (let i = 0; i < Math.min(snippetCount, 5); i++) {
const text = await allSnippets.nth(i).textContent()
const text = await snippetDivs.nth(i).textContent()
if (text && text.length > 10) {
// Should not contain raw markdown formatting
expect(text).not.toMatch(/\*\*[^*]+\*\*/) // no **bold**
expect(text).not.toContain('```') // no code fences
expect(text).not.toMatch(/\[\[.*\]\]/) // no raw wikilinks
}
}
})
test('snippet does not start with list markers', async ({ page }) => {
const noteListContainer = page.locator('[data-testid="note-list-container"]')
await expect(noteListContainer).toBeVisible()
const snippetDivs = noteListContainer.locator('.text-\\[12px\\]')
const snippetCount = await snippetDivs.count()
for (let i = 0; i < Math.min(snippetCount, 10); i++) {
const text = await snippetDivs.nth(i).textContent()
if (text && text.length > 5) {
// Snippet text should not start with bullet markers
expect(text.trimStart()).not.toMatch(/^[*\-+] /)
expect(text.trimStart()).not.toMatch(/^\d+\. /)
}
}
})
})