implement sync formdata parser (#2911)

* implement sync formdata parsser

* move utf8decodebytes

* fixup

* fixup! off-by-one

* fixup

* fixup

* fix bugs, fix lint

* mark wpts as passing

* this one fails on node 18

* apply suggestions

* fixup! body can end with CRLF

Author: KhafraDev

lib/web/fetch/body.js

@@ -1,6 +1,5 @@
 'use strict'
 
-const Busboy = require('@fastify/busboy')
 const util = require('../../core/util')
 const {
   ReadableStreamFrom,
@@ -9,23 +8,20 @@ const {
   readableStreamClose,
   createDeferredPromise,
   fullyReadBody,
-  extractMimeType
+  extractMimeType,
+  utf8DecodeBytes
 } = require('./util')
 const { FormData } = require('./formdata')
 const { kState } = require('./symbols')
 const { webidl } = require('./webidl')
-const { Blob, File: NativeFile } = require('node:buffer')
+const { Blob } = require('node:buffer')
 const assert = require('node:assert')
 const { isErrored } = require('../../core/util')
 const { isArrayBuffer } = require('node:util/types')
-const { File: UndiciFile } = require('./file')
 const { serializeAMimeType } = require('./data-url')
-const { Readable } = require('node:stream')
+const { multipartFormDataParser } = require('./formdata-parser')
 
-/** @type {globalThis['File']} */
-const File = NativeFile ?? UndiciFile
 const textEncoder = new TextEncoder()
-const textDecoder = new TextDecoder()
 
 // https://fetch.spec.whatwg.org/#concept-bodyinit-extract
 function extractBody (object, keepalive = false) {
@@ -338,116 +334,56 @@ function bodyMixinMethods (instance) {
       return consumeBody(this, parseJSONFromBytes, instance)
     },
 
-    async formData () {
-      webidl.brandCheck(this, instance)
-
-      throwIfAborted(this[kState])
-
-      // 1. Let mimeType be the result of get the MIME type with this.
-      const mimeType = bodyMimeType(this)
-
-      // If mimeType’s essence is "multipart/form-data", then:
-      if (mimeType !== null && mimeType.essence === 'multipart/form-data') {
-        const responseFormData = new FormData()
-
-        let busboy
-
-        try {
-          busboy = new Busboy({
-            headers: {
-              'content-type': serializeAMimeType(mimeType)
-            },
-            preservePath: true
-          })
-        } catch (err) {
-          throw new DOMException(`${err}`, 'AbortError')
-        }
-
-        busboy.on('field', (name, value) => {
-          responseFormData.append(name, value)
-        })
-        busboy.on('file', (name, value, filename, encoding, mimeType) => {
-          const chunks = []
-
-          if (encoding === 'base64' || encoding.toLowerCase() === 'base64') {
-            let base64chunk = ''
-
-            value.on('data', (chunk) => {
-              base64chunk += chunk.toString().replace(/[\r\n]/gm, '')
-
-              const end = base64chunk.length - base64chunk.length % 4
-              chunks.push(Buffer.from(base64chunk.slice(0, end), 'base64'))
-
-              base64chunk = base64chunk.slice(end)
-            })
-            value.on('end', () => {
-              chunks.push(Buffer.from(base64chunk, 'base64'))
-              responseFormData.append(name, new File(chunks, filename, { type: mimeType }))
-            })
-          } else {
-            value.on('data', (chunk) => {
-              chunks.push(chunk)
-            })
-            value.on('end', () => {
-              responseFormData.append(name, new File(chunks, filename, { type: mimeType }))
-            })
-          }
-        })
-
-        const busboyResolve = new Promise((resolve, reject) => {
-          busboy.on('finish', resolve)
-          busboy.on('error', (err) => reject(new TypeError(err)))
-        })
-
-        if (this.body !== null) {
-          Readable.from(this[kState].body.stream).pipe(busboy)
-        }
+    formData () {
+      // The formData() method steps are to return the result of running
+      // consume body with this and the following step given a byte sequence bytes:
+      return consumeBody(this, (value) => {
+        // 1. Let mimeType be the result of get the MIME type with this.
+        const mimeType = bodyMimeType(this)
+
+        // 2. If mimeType is non-null, then switch on mimeType’s essence and run
+        //    the corresponding steps:
+        if (mimeType !== null) {
+          switch (mimeType.essence) {
+            case 'multipart/form-data': {
+              // 1. ... [long step]
+              const parsed = multipartFormDataParser(value, mimeType)
+
+              // 2. If that fails for some reason, then throw a TypeError.
+              if (parsed === 'failure') {
+                throw new TypeError('Failed to parse body as FormData.')
+              }
+
+              // 3. Return a new FormData object, appending each entry,
+              //    resulting from the parsing operation, to its entry list.
+              const fd = new FormData()
+              fd[kState] = parsed
+
+              return fd
+            }
+            case 'application/x-www-form-urlencoded': {
+              // 1. Let entries be the result of parsing bytes.
+              const entries = new URLSearchParams(value.toString())
 
-        await busboyResolve
+              // 2. If entries is failure, then throw a TypeError.
 
-        return responseFormData
-      } else if (mimeType !== null && mimeType.essence === 'application/x-www-form-urlencoded') {
-        // Otherwise, if mimeType’s essence is "application/x-www-form-urlencoded", then:
+              // 3. Return a new FormData object whose entry list is entries.
+              const fd = new FormData()
 
-        // 1. Let entries be the result of parsing bytes.
-        let entries
-        try {
-          let text = ''
-          // application/x-www-form-urlencoded parser will keep the BOM.
-          // https://url.spec.whatwg.org/#concept-urlencoded-parser
-          // Note that streaming decoder is stateful and cannot be reused
-          const stream = this[kState].body.stream.pipeThrough(new TextDecoderStream('utf-8', { ignoreBOM: true }))
+              for (const [name, value] of entries) {
+                fd.append(name, value)
+              }
 
-          for await (const chunk of stream) {
-            text += chunk
+              return fd
+            }
           }
-
-          entries = new URLSearchParams(text)
-        } catch (err) {
-          // istanbul ignore next: Unclear when new URLSearchParams can fail on a string.
-          // 2. If entries is failure, then throw a TypeError.
-          throw new TypeError(err)
         }
 
-        // 3. Return a new FormData object whose entries are entries.
-        const formData = new FormData()
-        for (const [name, value] of entries) {
-          formData.append(name, value)
-        }
-        return formData
-      } else {
-        // Wait a tick before checking if the request has been aborted.
-        // Otherwise, a TypeError can be thrown when an AbortError should.
-        await Promise.resolve()
-
-        throwIfAborted(this[kState])
-
-        // Otherwise, throw a TypeError.
-        throw webidl.errors.exception({
-          header: `${instance.name}.formData`,
-          message: 'Could not parse content as FormData.'
-        })
-      }
+        // 3. Throw a TypeError.
+        throw new TypeError(
+          'Content-Type was not one of "multipart/form-data" or "application/x-www-form-urlencoded".'
+        )
+      }, instance)
     }
   }
 
@@ -516,32 +452,6 @@ function bodyUnusable (body) {
   return body != null && (body.stream.locked || util.isDisturbed(body.stream))
 }
 
-/**
- * @see https://encoding.spec.whatwg.org/#utf-8-decode
- * @param {Buffer} buffer
- */
-function utf8DecodeBytes (buffer) {
-  if (buffer.length === 0) {
-    return ''
-  }
-
-  // 1. Let buffer be the result of peeking three bytes from
-  //    ioQueue, converted to a byte sequence.
-
-  // 2. If buffer is 0xEF 0xBB 0xBF, then read three
-  //    bytes from ioQueue. (Do nothing with those bytes.)
-  if (buffer[0] === 0xEF && buffer[1] === 0xBB && buffer[2] === 0xBF) {
-    buffer = buffer.subarray(3)
-  }
-
-  // 3. Process a queue with an instance of UTF-8’s
-  //    decoder, ioQueue, output, and "replacement".
-  const output = textDecoder.decode(buffer)
-
-  // 4. Return output.
-  return output
-}
-
 /**
  * @see https://infra.spec.whatwg.org/#parse-json-bytes-to-a-javascript-value
  * @param {Uint8Array} bytes

lib/web/fetch/data-url.js

@@ -628,7 +628,6 @@ function removeASCIIWhitespace (str, leading = true, trailing = true) {
 }
 
 /**
- *
  * @param {string} str
  * @param {boolean} leading
  * @param {boolean} trailing
@@ -738,5 +737,7 @@ module.exports = {
   collectAnHTTPQuotedString,
   serializeAMimeType,
   removeChars,
-  minimizeSupportedMimeType
+  minimizeSupportedMimeType,
+  HTTP_TOKEN_CODEPOINTS,
+  isomorphicDecode
 }