fix(metascraper): log rule values without a string form

packages/metascraper/src/get-data.js

@@ -3,6 +3,11 @@
 const debug = require('debug-logfmt')('metascraper:get-data')
 const { findRule, has } = require('@metascraper/helpers')
 
+const PRINTABLE_TYPES = new Set(['string', 'number', 'boolean', 'bigint'])
+
+const formatValue = value =>
+  value === null || PRINTABLE_TYPES.has(typeof value) ? value : typeof value
+
 const getData = async ({ rules, ...props }) => {
   const data = {}
 
@@ -26,7 +31,9 @@ const getData = async ({ rules, ...props }) => {
       }
 
       duration(
-        `${propName}=${normalizedValue} rules=${innerRules.length} status=${status}`
+        `${propName}=${formatValue(normalizedValue)} rules=${
+          innerRules.length
+        } status=${status}`
       )
       data[propName] = normalizedValue
     })

packages/metascraper/test/unit/index.js

@@ -1,8 +1,32 @@
 'use strict'
 
+const path = require('path')
+
 const test = require('ava').default
 
 const createMetascraper = require('../..')
+const debug = require(require.resolve('debug', {
+  paths: [path.dirname(require.resolve('debug-logfmt'))]
+}))
+
+const withGetDataLogs = async fn => {
+  const previous = debug.disable()
+  debug.enable('metascraper:get-data')
+  const chunks = []
+  const write = process.stderr.write
+  process.stderr.write = (chunk, encoding, cb) => {
+    chunks.push(Buffer.isBuffer(chunk) ? chunk.toString() : String(chunk))
+    return write.call(process.stderr, chunk, encoding, cb)
+  }
+  try {
+    await fn()
+    return chunks.join('')
+  } finally {
+    process.stderr.write = write
+    debug.disable()
+    if (previous) debug.enable(previous)
+  }
+}
 
 test('returns all rules fields by default', async t => {
   const rulesNames = [
@@ -86,3 +110,87 @@ test('avoid to run rules by `omitPropNames`', async t => {
     rulesNames.filter(field => field !== 'author')
   )
 })
+
+test('a rule value without a string form is returned as is', async t => {
+  const location = Object.assign(Object.create(null), {
+    href: 'https://example.com/'
+  })
+  const cyclic = { name: 'cyclic' }
+  cyclic.self = cyclic
+  const customInspect = {
+    name: 'custom',
+    [Symbol.for('nodejs.util.inspect.custom')]: () => {
+      throw new Error('custom')
+    }
+  }
+  const revoked = Proxy.revocable({}, {})
+  revoked.revoke()
+  const symbol = Symbol('value')
+  const bareFunction = Object.setPrototypeOf(() => {}, null)
+
+  const metascraper = createMetascraper([
+    {
+      location: [() => location],
+      cyclic: [() => cyclic],
+      customInspect: [() => customInspect],
+      revoked: [() => revoked.proxy],
+      symbol: [() => symbol],
+      bareFunction: [() => bareFunction],
+      title: [() => 'title']
+    }
+  ])
+  const metadata = await metascraper({ url: 'https://example.com' })
+
+  t.is(metadata.location, location)
+  t.is(metadata.cyclic, cyclic)
+  t.is(metadata.customInspect, customInspect)
+  t.is(metadata.revoked, null)
+  t.is(metadata.symbol, symbol)
+  t.is(metadata.bareFunction, bareFunction)
+  t.is(metadata.title, 'title')
+})
+
+test.serial(
+  'logs a primitive rule value and the type of any other',
+  async t => {
+    const location = Object.assign(Object.create(null), {
+      href: 'https://example.com/'
+    })
+    const symbol = Symbol('value')
+    const bareFunction = Object.setPrototypeOf(() => {}, null)
+    const metascraper = createMetascraper([
+      {
+        label: [() => 'hello'],
+        count: [() => 0],
+        flag: [() => false],
+        big: [() => 10n],
+        empty: [() => null],
+        location: [() => location],
+        symbol: [() => symbol],
+        bareFunction: [() => bareFunction]
+      }
+    ])
+
+    let metadata
+    const log = await withGetDataLogs(async () => {
+      metadata = await metascraper({ url: 'https://example.com' })
+    })
+
+    t.is(metadata.label, 'hello')
+    t.is(metadata.count, 0)
+    t.is(metadata.flag, false)
+    t.is(metadata.big, 10n)
+    t.is(metadata.empty, null)
+    t.is(metadata.location, location)
+    t.is(metadata.symbol, symbol)
+    t.is(metadata.bareFunction, bareFunction)
+    t.regex(log, /label=hello rules=/)
+    t.regex(log, /count=0 rules=/)
+    t.regex(log, /flag=false rules=/)
+    t.regex(log, /big=10 rules=/)
+    t.regex(log, /empty=null rules=/)
+    t.regex(log, /location=object rules=/)
+    t.regex(log, /symbol=symbol rules=/)
+    t.regex(log, /bareFunction=function rules=/)
+  }
+)