fix: handle multi-line metadata blocks in HOOK.md frontmatter
The frontmatter parser was using a simple line-by-line regex that only captured single-line key-value pairs. This meant multi-line metadata blocks (as used by bundled hooks) were not parsed correctly. Changes: - Add extractMultiLineValue() to handle indented continuation lines - Use JSON5 instead of JSON.parse() to support trailing commas - Add comprehensive test coverage for frontmatter parsing Fixes #1113
This commit is contained in:
Sebastian Slight
committed by
Peter Steinberger
Peter Steinberger
parent
1c4297d8b5
commit
35a1d81518
259
src/hooks/frontmatter.test.ts
Normal file
259
src/hooks/frontmatter.test.ts
Normal file
@@ -0,0 +1,259 @@
|
||||
import { describe, expect, it } from "vitest";
|
||||
import { parseFrontmatter, resolveClawdbotMetadata } from "./frontmatter.js";
|
||||
|
||||
describe("parseFrontmatter", () => {
|
||||
it("parses single-line key-value pairs", () => {
|
||||
const content = `---
|
||||
name: test-hook
|
||||
description: "A test hook"
|
||||
homepage: https://example.com
|
||||
---
|
||||
|
||||
# Test Hook
|
||||
`;
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("test-hook");
|
||||
expect(result.description).toBe("A test hook");
|
||||
expect(result.homepage).toBe("https://example.com");
|
||||
});
|
||||
|
||||
it("handles missing frontmatter", () => {
|
||||
const content = "# Just a markdown file";
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result).toEqual({});
|
||||
});
|
||||
|
||||
it("handles unclosed frontmatter", () => {
|
||||
const content = `---
|
||||
name: broken
|
||||
`;
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result).toEqual({});
|
||||
});
|
||||
|
||||
it("parses multi-line metadata block with indented JSON", () => {
|
||||
const content = `---
|
||||
name: session-memory
|
||||
description: "Save session context"
|
||||
metadata:
|
||||
{
|
||||
"clawdbot": {
|
||||
"emoji": "💾",
|
||||
"events": ["command:new"]
|
||||
}
|
||||
}
|
||||
---
|
||||
|
||||
# Session Memory Hook
|
||||
`;
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("session-memory");
|
||||
expect(result.description).toBe("Save session context");
|
||||
expect(result.metadata).toBeDefined();
|
||||
expect(typeof result.metadata).toBe("string");
|
||||
|
||||
// Verify the metadata is valid JSON
|
||||
const parsed = JSON.parse(result.metadata as string);
|
||||
expect(parsed.clawdbot.emoji).toBe("💾");
|
||||
expect(parsed.clawdbot.events).toEqual(["command:new"]);
|
||||
});
|
||||
|
||||
it("parses multi-line metadata with complex nested structure", () => {
|
||||
const content = `---
|
||||
name: command-logger
|
||||
description: "Log all command events"
|
||||
metadata:
|
||||
{
|
||||
"clawdbot":
|
||||
{
|
||||
"emoji": "📝",
|
||||
"events": ["command"],
|
||||
"requires": { "config": ["workspace.dir"] },
|
||||
"install": [{ "id": "bundled", "kind": "bundled", "label": "Bundled" }]
|
||||
}
|
||||
}
|
||||
---
|
||||
`;
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("command-logger");
|
||||
expect(result.metadata).toBeDefined();
|
||||
|
||||
const parsed = JSON.parse(result.metadata as string);
|
||||
expect(parsed.clawdbot.emoji).toBe("📝");
|
||||
expect(parsed.clawdbot.events).toEqual(["command"]);
|
||||
expect(parsed.clawdbot.requires.config).toEqual(["workspace.dir"]);
|
||||
expect(parsed.clawdbot.install[0].kind).toBe("bundled");
|
||||
});
|
||||
|
||||
it("handles single-line metadata (inline JSON)", () => {
|
||||
const content = `---
|
||||
name: simple-hook
|
||||
metadata: {"clawdbot": {"events": ["test"]}}
|
||||
---
|
||||
`;
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("simple-hook");
|
||||
expect(result.metadata).toBe('{"clawdbot": {"events": ["test"]}}');
|
||||
});
|
||||
|
||||
it("handles mixed single-line and multi-line values", () => {
|
||||
const content = `---
|
||||
name: mixed-hook
|
||||
description: "A hook with mixed values"
|
||||
homepage: https://example.com
|
||||
metadata:
|
||||
{
|
||||
"clawdbot": {
|
||||
"events": ["command:new"]
|
||||
}
|
||||
}
|
||||
enabled: true
|
||||
---
|
||||
`;
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("mixed-hook");
|
||||
expect(result.description).toBe("A hook with mixed values");
|
||||
expect(result.homepage).toBe("https://example.com");
|
||||
expect(result.metadata).toBeDefined();
|
||||
expect(result.enabled).toBe("true");
|
||||
});
|
||||
|
||||
it("strips surrounding quotes from values", () => {
|
||||
const content = `---
|
||||
name: "quoted-name"
|
||||
description: 'single-quoted'
|
||||
---
|
||||
`;
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("quoted-name");
|
||||
expect(result.description).toBe("single-quoted");
|
||||
});
|
||||
|
||||
it("handles CRLF line endings", () => {
|
||||
const content = "---\r\nname: test\r\ndescription: crlf\r\n---\r\n";
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("test");
|
||||
expect(result.description).toBe("crlf");
|
||||
});
|
||||
|
||||
it("handles CR line endings", () => {
|
||||
const content = "---\rname: test\rdescription: cr\r---\r";
|
||||
const result = parseFrontmatter(content);
|
||||
expect(result.name).toBe("test");
|
||||
expect(result.description).toBe("cr");
|
||||
});
|
||||
});
|
||||
|
||||
describe("resolveClawdbotMetadata", () => {
|
||||
it("extracts clawdbot metadata from parsed frontmatter", () => {
|
||||
const frontmatter = {
|
||||
name: "test-hook",
|
||||
metadata: JSON.stringify({
|
||||
clawdbot: {
|
||||
emoji: "🔥",
|
||||
events: ["command:new", "command:reset"],
|
||||
requires: {
|
||||
config: ["workspace.dir"],
|
||||
bins: ["git"],
|
||||
},
|
||||
},
|
||||
}),
|
||||
};
|
||||
|
||||
const result = resolveClawdbotMetadata(frontmatter);
|
||||
expect(result).toBeDefined();
|
||||
expect(result?.emoji).toBe("🔥");
|
||||
expect(result?.events).toEqual(["command:new", "command:reset"]);
|
||||
expect(result?.requires?.config).toEqual(["workspace.dir"]);
|
||||
expect(result?.requires?.bins).toEqual(["git"]);
|
||||
});
|
||||
|
||||
it("returns undefined when metadata is missing", () => {
|
||||
const frontmatter = { name: "no-metadata" };
|
||||
const result = resolveClawdbotMetadata(frontmatter);
|
||||
expect(result).toBeUndefined();
|
||||
});
|
||||
|
||||
it("returns undefined when clawdbot key is missing", () => {
|
||||
const frontmatter = {
|
||||
metadata: JSON.stringify({ other: "data" }),
|
||||
};
|
||||
const result = resolveClawdbotMetadata(frontmatter);
|
||||
expect(result).toBeUndefined();
|
||||
});
|
||||
|
||||
it("returns undefined for invalid JSON", () => {
|
||||
const frontmatter = {
|
||||
metadata: "not valid json {",
|
||||
};
|
||||
const result = resolveClawdbotMetadata(frontmatter);
|
||||
expect(result).toBeUndefined();
|
||||
});
|
||||
|
||||
it("handles install specs", () => {
|
||||
const frontmatter = {
|
||||
metadata: JSON.stringify({
|
||||
clawdbot: {
|
||||
events: ["command"],
|
||||
install: [
|
||||
{ id: "bundled", kind: "bundled", label: "Bundled with Clawdbot" },
|
||||
{ id: "npm", kind: "npm", package: "@clawdbot/hook" },
|
||||
],
|
||||
},
|
||||
}),
|
||||
};
|
||||
|
||||
const result = resolveClawdbotMetadata(frontmatter);
|
||||
expect(result?.install).toHaveLength(2);
|
||||
expect(result?.install?.[0].kind).toBe("bundled");
|
||||
expect(result?.install?.[1].kind).toBe("npm");
|
||||
expect(result?.install?.[1].package).toBe("@clawdbot/hook");
|
||||
});
|
||||
|
||||
it("handles os restrictions", () => {
|
||||
const frontmatter = {
|
||||
metadata: JSON.stringify({
|
||||
clawdbot: {
|
||||
events: ["command"],
|
||||
os: ["darwin", "linux"],
|
||||
},
|
||||
}),
|
||||
};
|
||||
|
||||
const result = resolveClawdbotMetadata(frontmatter);
|
||||
expect(result?.os).toEqual(["darwin", "linux"]);
|
||||
});
|
||||
|
||||
it("parses real session-memory HOOK.md format", () => {
|
||||
// This is the actual format used in the bundled hooks
|
||||
const content = `---
|
||||
name: session-memory
|
||||
description: "Save session context to memory when /new command is issued"
|
||||
homepage: https://docs.clawd.bot/hooks#session-memory
|
||||
metadata:
|
||||
{
|
||||
"clawdbot":
|
||||
{
|
||||
"emoji": "💾",
|
||||
"events": ["command:new"],
|
||||
"requires": { "config": ["workspace.dir"] },
|
||||
"install": [{ "id": "bundled", "kind": "bundled", "label": "Bundled with Clawdbot" }],
|
||||
},
|
||||
}
|
||||
---
|
||||
|
||||
# Session Memory Hook
|
||||
`;
|
||||
|
||||
const frontmatter = parseFrontmatter(content);
|
||||
expect(frontmatter.name).toBe("session-memory");
|
||||
expect(frontmatter.metadata).toBeDefined();
|
||||
|
||||
const clawdbot = resolveClawdbotMetadata(frontmatter);
|
||||
expect(clawdbot).toBeDefined();
|
||||
expect(clawdbot?.emoji).toBe("💾");
|
||||
expect(clawdbot?.events).toEqual(["command:new"]);
|
||||
expect(clawdbot?.requires?.config).toEqual(["workspace.dir"]);
|
||||
expect(clawdbot?.install?.[0].kind).toBe("bundled");
|
||||
});
|
||||
});
|
||||
@@ -1,3 +1,4 @@
|
||||
import JSON5 from "json5";
|
||||
import type {
|
||||
ClawdbotHookMetadata,
|
||||
HookEntry,
|
||||
@@ -16,6 +17,48 @@ function stripQuotes(value: string): string {
|
||||
return value;
|
||||
}
|
||||
|
||||
/**
|
||||
* Extract a multi-line block value from frontmatter lines.
|
||||
* Handles indented continuation lines (YAML-style multi-line values).
|
||||
*
|
||||
* @param lines - All lines in the frontmatter block
|
||||
* @param startIndex - Index of the line containing the key
|
||||
* @returns The combined multi-line value and the number of lines consumed
|
||||
*/
|
||||
function extractMultiLineValue(
|
||||
lines: string[],
|
||||
startIndex: number,
|
||||
): { value: string; linesConsumed: number } {
|
||||
const startLine = lines[startIndex];
|
||||
const match = startLine.match(/^([\w-]+):\s*(.*)$/);
|
||||
if (!match) return { value: "", linesConsumed: 1 };
|
||||
|
||||
const inlineValue = match[2].trim();
|
||||
|
||||
// If there's a value on the same line, return it (single-line case)
|
||||
if (inlineValue) {
|
||||
return { value: inlineValue, linesConsumed: 1 };
|
||||
}
|
||||
|
||||
// Multi-line case: collect indented continuation lines
|
||||
const valueLines: string[] = [];
|
||||
let i = startIndex + 1;
|
||||
|
||||
while (i < lines.length) {
|
||||
const line = lines[i];
|
||||
// Stop if we hit a non-indented line (new key or empty line without indent)
|
||||
if (line.length > 0 && !line.startsWith(" ") && !line.startsWith("\t")) {
|
||||
break;
|
||||
}
|
||||
valueLines.push(line);
|
||||
i++;
|
||||
}
|
||||
|
||||
// Join and trim the multi-line value
|
||||
const combined = valueLines.join("\n").trim();
|
||||
return { value: combined, linesConsumed: i - startIndex };
|
||||
}
|
||||
|
||||
export function parseFrontmatter(content: string): ParsedHookFrontmatter {
|
||||
const frontmatter: ParsedHookFrontmatter = {};
|
||||
const normalized = content.replace(/\r\n/g, "\n").replace(/\r/g, "\n");
|
||||
@@ -23,14 +66,47 @@ export function parseFrontmatter(content: string): ParsedHookFrontmatter {
|
||||
const endIndex = normalized.indexOf("\n---", 3);
|
||||
if (endIndex === -1) return frontmatter;
|
||||
const block = normalized.slice(4, endIndex);
|
||||
for (const line of block.split("\n")) {
|
||||
const lines = block.split("\n");
|
||||
|
||||
let i = 0;
|
||||
while (i < lines.length) {
|
||||
const line = lines[i];
|
||||
const match = line.match(/^([\w-]+):\s*(.*)$/);
|
||||
if (!match) continue;
|
||||
if (!match) {
|
||||
i++;
|
||||
continue;
|
||||
}
|
||||
|
||||
const key = match[1];
|
||||
const value = stripQuotes(match[2].trim());
|
||||
if (!key || !value) continue;
|
||||
frontmatter[key] = value;
|
||||
const inlineValue = match[2].trim();
|
||||
|
||||
if (!key) {
|
||||
i++;
|
||||
continue;
|
||||
}
|
||||
|
||||
// Check if this is a multi-line value (no inline value and next line is indented)
|
||||
if (!inlineValue && i + 1 < lines.length) {
|
||||
const nextLine = lines[i + 1];
|
||||
if (nextLine.startsWith(" ") || nextLine.startsWith("\t")) {
|
||||
// Multi-line value
|
||||
const { value, linesConsumed } = extractMultiLineValue(lines, i);
|
||||
if (value) {
|
||||
frontmatter[key] = value;
|
||||
}
|
||||
i += linesConsumed;
|
||||
continue;
|
||||
}
|
||||
}
|
||||
|
||||
// Single-line value
|
||||
const value = stripQuotes(inlineValue);
|
||||
if (value) {
|
||||
frontmatter[key] = value;
|
||||
}
|
||||
i++;
|
||||
}
|
||||
|
||||
return frontmatter;
|
||||
}
|
||||
|
||||
@@ -96,7 +172,8 @@ export function resolveClawdbotMetadata(
|
||||
const raw = getFrontmatterValue(frontmatter, "metadata");
|
||||
if (!raw) return undefined;
|
||||
try {
|
||||
const parsed = JSON.parse(raw) as { clawdbot?: unknown };
|
||||
// Use JSON5 to handle trailing commas and other relaxed JSON syntax
|
||||
const parsed = JSON5.parse(raw) as { clawdbot?: unknown };
|
||||
if (!parsed || typeof parsed !== "object") return undefined;
|
||||
const clawdbot = (parsed as { clawdbot?: unknown }).clawdbot;
|
||||
if (!clawdbot || typeof clawdbot !== "object") return undefined;
|
||||
|
||||
Reference in New Issue
Block a user