url,buffer: implement URL.createObjectURL

Signed-off-by: James M Snell <[email protected]>

PR-URL: #39693
Reviewed-By: Anna Henningsen <[email protected]>
Reviewed-By: Benjamin Gruenbaum <[email protected]>
Reviewed-By: Bradley Farias <[email protected]>
Reviewed-By: Stephen Belanger <[email protected]>

Author: jasnell

doc/api/buffer.md

@@ -4948,6 +4948,20 @@ added: v3.0.0
 
 An alias for [`buffer.constants.MAX_STRING_LENGTH`][].
 
+### `buffer.resolveObjectURL(id)`
+<!-- YAML
+added: REPLACEME
+-->
+
+> Stability: 1 - Experimental
+
+* `id` {string} A `'blob:nodedata:...` URL string returned by a prior call to
+  `URL.createObjectURL()`.
+* Returns: {Blob}
+
+Resolves a `'blob:nodedata:...'` an associated {Blob} object registered using
+a prior call to `URL.createObjectURL()`.
+
 ### `buffer.transcode(source, fromEnc, toEnc)`
 <!-- YAML
 added: v7.1.0

doc/api/url.md

@@ -608,6 +608,53 @@ console.log(JSON.stringify(myURLs));
 // Prints ["https://www.example.com/","https://test.example.org/"]
 ```
 
+#### `URL.createObjectURL(blob)`
+<!-- YAML
+added: REPLACEME
+-->
+
+> Stability: 1 - Experimental
+
+* `blob` {Blob}
+* Returns: {string}
+
+Creates a `'blob:nodedata:...'` URL string that represents the given {Blob}
+object and can be used to retrieve the `Blob` later.
+
+```js
+const {
+  Blob,
+  resolveObjectURL,
+} = require('buffer');
+
+const blob = new Blob(['hello']);
+const id = URL.createObjectURL(blob);
+
+// later...
+
+const otherBlob = resolveObjectURL(id);
+console.log(otherBlob.size);
+```
+
+The data stored by the registered {Blob} will be retained in memory until
+`URL.revokeObjectURL()` is called to remove it.
+
+`Blob` objects are registered within the current thread. If using Worker
+Threads, `Blob` objects registered within one Worker will not be available
+to other workers or the main thread.
+
+#### `URL.revokeObjectURL(id)`
+<!-- YAML
+added: REPLACEME
+-->
+
+> Stability: 1 - Experimental
+
+* `id` {string} A `'blob:nodedata:...` URL string returned by a prior call to
+  `URL.createObjectURL()`.
+
+Removes the stored {Blob} identified by the given ID.
+
 ### Class: `URLSearchParams`
 <!-- YAML
 added:

lib/buffer.js

@@ -120,6 +120,7 @@ const {
 
 const {
   Blob,
+  resolveObjectURL,
 } = require('internal/blob');
 
 FastBuffer.prototype.constructor = Buffer;
@@ -1239,6 +1240,7 @@ function atob(input) {
 
 module.exports = {
   Blob,
+  resolveObjectURL,
   Buffer,
   SlowBuffer,
   transcode,

lib/internal/blob.js

@@ -7,10 +7,11 @@ const {
   ObjectDefineProperty,
   PromiseResolve,
   PromiseReject,
-  PromisePrototypeFinally,
+  SafePromisePrototypeFinally,
   ReflectConstruct,
   RegExpPrototypeTest,
   StringPrototypeToLowerCase,
+  StringPrototypeSplit,
   Symbol,
   SymbolIterator,
   SymbolToStringTag,
@@ -20,7 +21,8 @@ const {
 const {
   createBlob: _createBlob,
   FixedSizeBlobCopyJob,
-} = internalBinding('buffer');
+  getDataObject,
+} = internalBinding('blob');
 
 const { TextDecoder } = require('internal/encoding');
 
@@ -57,26 +59,37 @@ const {
 } = require('internal/validators');
 
 const kHandle = Symbol('kHandle');
+const kState = Symbol('kState');
 const kType = Symbol('kType');
 const kLength = Symbol('kLength');
 const kArrayBufferPromise = Symbol('kArrayBufferPromise');
 
+const kMaxChunkSize = 65536;
+
 const disallowedTypeCharacters = /[^\u{0020}-\u{007E}]/u;
 
 let Buffer;
 let ReadableStream;
+let URL;
+
+
+// Yes, lazy loading is annoying but because of circular
+// references between the url, internal/blob, and buffer
+// modules, lazy loading here makes sure that things work.
+
+function lazyURL(id) {
+  URL ??= require('internal/url').URL;
+  return new URL(id);
+}
 
 function lazyBuffer() {
-  if (Buffer === undefined)
-    Buffer = require('buffer').Buffer;
+  Buffer ??= require('buffer').Buffer;
   return Buffer;
 }
 
 function lazyReadableStream(options) {
-  if (ReadableStream === undefined) {
-    ReadableStream =
-      require('internal/webstreams/readablestream').ReadableStream;
-  }
+  ReadableStream ??=
+    require('internal/webstreams/readablestream').ReadableStream;
   return new ReadableStream(options);
 }
 
@@ -232,9 +245,9 @@ class Blob {
       return PromiseReject(new ERR_INVALID_THIS('Blob'));
 
     // If there's already a promise in flight for the content,
-    // reuse it, but only once. After the cached promise resolves
-    // it will be cleared, allowing it to be garbage collected
-    // as soon as possible.
+    // reuse it, but only while it's in flight. After the cached
+    // promise resolves it will be cleared, allowing it to be
+    // garbage collected as soon as possible.
     if (this[kArrayBufferPromise])
       return this[kArrayBufferPromise];
 
@@ -260,15 +273,14 @@ class Blob {
       resolve(ab);
     };
     this[kArrayBufferPromise] =
-    PromisePrototypeFinally(
+    SafePromisePrototypeFinally(
       promise,
       () => this[kArrayBufferPromise] = undefined);
 
     return this[kArrayBufferPromise];
   }
 
   /**
-   *
    * @returns {Promise<string>}
    */
   async text() {
@@ -288,10 +300,20 @@ class Blob {
 
     const self = this;
     return new lazyReadableStream({
-      async start(controller) {
-        const ab = await self.arrayBuffer();
-        controller.enqueue(new Uint8Array(ab));
-        controller.close();
+      async start() {
+        this[kState] = await self.arrayBuffer();
+      },
+
+      pull(controller) {
+        if (this[kState].byteLength <= kMaxChunkSize) {
+          controller.enqueue(new Uint8Array(this[kState]));
+          controller.close();
+          this[kState] = undefined;
+        } else {
+          const slice = this[kState].slice(0, kMaxChunkSize);
+          this[kState] = this[kState].slice(kMaxChunkSize);
+          controller.enqueue(new Uint8Array(slice));
+        }
       }
     });
   }
@@ -315,9 +337,47 @@ ObjectDefineProperty(Blob.prototype, SymbolToStringTag, {
   value: 'Blob',
 });
 
+function resolveObjectURL(url) {
+  url = `${url}`;
+  try {
+    const parsed = new lazyURL(url);
+
+    const split = StringPrototypeSplit(parsed.pathname, ':');
+
+    if (split.length !== 2)
+      return;
+
+    const {
+      0: base,
+      1: id,
+    } = split;
+
+    if (base !== 'nodedata')
+      return;
+
+    const ret = getDataObject(id);
+
+    if (ret === undefined)
+      return;
+
+    const {
+      0: handle,
+      1: length,
+      2: type,
+    } = ret;
+
+    if (handle !== undefined)
+      return createBlob(handle, length, type);
+  } catch {
+    // If there's an error, it's ignored and nothing is returned
+  }
+}
+
 module.exports = {
   Blob,
   ClonedBlob,
   createBlob,
   isBlob,
+  kHandle,
+  resolveObjectURL,
 };

lib/internal/url.js

@@ -42,17 +42,20 @@ const {
 
 const { getConstructorOf, removeColors } = require('internal/util');
 const {
-  ERR_ARG_NOT_ITERABLE,
-  ERR_INVALID_ARG_TYPE,
-  ERR_INVALID_ARG_VALUE,
-  ERR_INVALID_FILE_URL_HOST,
-  ERR_INVALID_FILE_URL_PATH,
-  ERR_INVALID_THIS,
-  ERR_INVALID_TUPLE,
-  ERR_INVALID_URL,
-  ERR_INVALID_URL_SCHEME,
-  ERR_MISSING_ARGS
-} = require('internal/errors').codes;
+  codes: {
+    ERR_ARG_NOT_ITERABLE,
+    ERR_INVALID_ARG_TYPE,
+    ERR_INVALID_ARG_VALUE,
+    ERR_INVALID_FILE_URL_HOST,
+    ERR_INVALID_FILE_URL_PATH,
+    ERR_INVALID_THIS,
+    ERR_INVALID_TUPLE,
+    ERR_INVALID_URL,
+    ERR_INVALID_URL_SCHEME,
+    ERR_MISSING_ARGS,
+    ERR_NO_CRYPTO,
+  },
+} = require('internal/errors');
 const {
   CHAR_AMPERSAND,
   CHAR_BACKWARD_SLASH,
@@ -100,6 +103,11 @@ const {
   kSchemeStart
 } = internalBinding('url');
 
+const {
+  storeDataObject,
+  revokeDataObject,
+} = internalBinding('blob');
+
 const context = Symbol('context');
 const cannotBeBase = Symbol('cannot-be-base');
 const cannotHaveUsernamePasswordPort =
@@ -108,6 +116,24 @@ const special = Symbol('special');
 const searchParams = Symbol('query');
 const kFormat = Symbol('format');
 
+let blob;
+let cryptoRandom;
+
+function lazyBlob() {
+  blob ??= require('internal/blob');
+  return blob;
+}
+
+function lazyCryptoRandom() {
+  try {
+    cryptoRandom ??= require('internal/crypto/random');
+  } catch {
+    // If Node.js built without crypto support, we'll fall
+    // through here and handle it later.
+  }
+  return cryptoRandom;
+}
+
 // https://tc39.github.io/ecma262/#sec-%iteratorprototype%-object
 const IteratorPrototype = ObjectGetPrototypeOf(
   ObjectGetPrototypeOf([][SymbolIterator]())
@@ -930,6 +956,37 @@ class URL {
   toJSON() {
     return this[kFormat]({});
   }
+
+  static createObjectURL(obj) {
+    const cryptoRandom = lazyCryptoRandom();
+    if (cryptoRandom === undefined)
+      throw new ERR_NO_CRYPTO();
+
+    // Yes, lazy loading is annoying but because of circular
+    // references between the url, internal/blob, and buffer
+    // modules, lazy loading here makes sure that things work.
+    const blob = lazyBlob();
+    if (!blob.isBlob(obj))
+      throw new ERR_INVALID_ARG_TYPE('obj', 'Blob', obj);
+
+    const id = cryptoRandom.randomUUID();
+
+    storeDataObject(id, obj[blob.kHandle], obj.size, obj.type);
+
+    return `blob:nodedata:${id}`;
+  }
+
+  static revokeObjectURL(url) {
+    url = `${url}`;
+    try {
+      const parsed = new URL(url);
+      const split = StringPrototypeSplit(parsed.pathname, ':');
+      if (split.length === 2)
+        revokeDataObject(split[1]);
+    } catch {
+      // If there's an error, it's ignored.
+    }
+  }
 }
 
 ObjectDefineProperties(URL.prototype, {

src/node_binding.cc

@@ -40,6 +40,7 @@
 // __attribute__((constructor)) like mechanism in GCC.
 #define NODE_BUILTIN_STANDARD_MODULES(V)                                       \
   V(async_wrap)                                                                \
+  V(blob)                                                                      \
   V(block_list)                                                                \
   V(buffer)                                                                    \
   V(cares_wrap)                                                                \