node-api: implement external strings

Introduce APIs that allow for the creation of JavaScript strings without
copying the underlying native string into the engine. The APIs fall back
to regular string creation if the engine's external string APIs are
unavailable. In this case, an optional boolean out-parameter indicates
that the string was copied, and the optional finalizer is called if
given.

PR-URL: #48339
Fixes: #48198
Reviewed-By: Daeyeon Jeong <daeyeon.dev@gmail.com>
Signed-off-by: Gabriel Schulhof <gabrielschulhof@gmail.com>

benchmark/napi/string/.gitignore

@@ -0,0 +1 @@
+build/

benchmark/napi/string/binding.c

@@ -0,0 +1,56 @@
+#include <assert.h>
+#define NAPI_EXPERIMENTAL
+#include <node_api.h>
+
+#define NAPI_CALL(call)                                                        \
+  do {                                                                         \
+    napi_status status = call;                                                 \
+    assert(status == napi_ok && #call " failed");                              \
+  } while (0);
+
+#define EXPORT_FUNC(env, exports, name, func)                                  \
+  do {                                                                         \
+    napi_value js_func;                                                        \
+    NAPI_CALL(napi_create_function(                                            \
+        (env), (name), NAPI_AUTO_LENGTH, (func), NULL, &js_func));             \
+    NAPI_CALL(napi_set_named_property((env), (exports), (name), js_func));     \
+  } while (0);
+
+const char* one_byte_string = "The Quick Brown Fox Jumped Over The Lazy Dog.";
+const char16_t* two_byte_string =
+    u"The Quick Brown Fox Jumped Over The Lazy Dog.";
+
+#define DECLARE_BINDING(CapName, lowercase_name, var_name)                     \
+  static napi_value CreateString##CapName(napi_env env,                        \
+                                          napi_callback_info info) {           \
+    size_t argc = 4;                                                           \
+    napi_value argv[4];                                                        \
+    uint32_t n;                                                                \
+    uint32_t index;                                                            \
+    napi_handle_scope scope;                                                   \
+    napi_value js_string;                                                      \
+                                                                               \
+    NAPI_CALL(napi_get_cb_info(env, info, &argc, argv, NULL, NULL));           \
+    NAPI_CALL(napi_get_value_uint32(env, argv[0], &n));                        \
+    NAPI_CALL(napi_open_handle_scope(env, &scope));                            \
+    NAPI_CALL(napi_call_function(env, argv[1], argv[2], 0, NULL, NULL));       \
+    for (index = 0; index < n; index++) {                                      \
+      NAPI_CALL(napi_create_string_##lowercase_name(                           \
+          env, (var_name), NAPI_AUTO_LENGTH, &js_string));                     \
+    }                                                                          \
+    NAPI_CALL(napi_call_function(env, argv[1], argv[3], 1, &argv[0], NULL));   \
+    NAPI_CALL(napi_close_handle_scope(env, scope));                            \
+                                                                               \
+    return NULL;                                                               \
+  }
+
+DECLARE_BINDING(Latin1, latin1, one_byte_string)
+DECLARE_BINDING(Utf8, utf8, one_byte_string)
+DECLARE_BINDING(Utf16, utf16, two_byte_string)
+
+NAPI_MODULE_INIT() {
+  EXPORT_FUNC(env, exports, "createStringLatin1", CreateStringLatin1);
+  EXPORT_FUNC(env, exports, "createStringUtf8", CreateStringUtf8);
+  EXPORT_FUNC(env, exports, "createStringUtf16", CreateStringUtf16);
+  return exports;
+}

benchmark/napi/string/binding.gyp

@@ -0,0 +1,8 @@
+{
+  'targets': [
+    {
+      'target_name': 'binding',
+      'sources': [ 'binding.c' ]
+    }
+  ]
+}

benchmark/napi/string/index.js

@@ -0,0 +1,19 @@
+'use strict';
+const common = require('../../common.js');
+
+let binding;
+try {
+  binding = require(`./build/${common.buildType}/binding`);
+} catch {
+  console.error(`${__filename}: Binding failed to load`);
+  process.exit(0);
+}
+
+const bench = common.createBenchmark(main, {
+  n: [1e5, 1e6, 1e7],
+  stringType: ['Latin1', 'Utf8', 'Utf16'],
+});
+
+function main({ n, stringType }) {
+  binding[`createString${stringType}`](n, bench, bench.start, bench.end);
+}

doc/api/n-api.md

@@ -801,7 +801,7 @@ napiVersion: 1
 
 Function pointer type for add-on provided functions that allow the user to be
 notified when externally-owned data is ready to be cleaned up because the
-object with which it was associated with, has been garbage-collected. The user
+object with which it was associated with has been garbage-collected. The user
 must provide a function satisfying the following signature which would get
 called upon the object's collection. Currently, `napi_finalize` can be used for
 finding out when objects that have external data are collected.
@@ -819,6 +819,11 @@ Since these functions may be called while the JavaScript engine is in a state
 where it cannot execute JavaScript code, some Node-API calls may return
 `napi_pending_exception` even when there is no exception pending.
 
+In the case of [`node_api_create_external_string_latin1`][] and
+[`node_api_create_external_string_utf16`][] the `env` parameter may be null,
+because external strings can be collected during the latter part of environment
+shutdown.
+
 Change History:
 
 * experimental (`NAPI_EXPERIMENTAL` is defined):
@@ -2882,6 +2887,56 @@ string. The native string is copied.
 The JavaScript `string` type is described in
 [Section 6.1.4][] of the ECMAScript Language Specification.
 
+#### `node_api_create_external_string_latin1`
+
+<!-- YAML
+added: REPLACEME
+-->
+
+> Stability: 1 - Experimental
+
+```c
+napi_status
+node_api_create_external_string_latin1(napi_env env,
+                                       char* str,
+                                       size_t length,
+                                       napi_finalize finalize_callback,
+                                       void* finalize_hint,
+                                       napi_value* result,
+                                       bool* copied);
+```
+
+* `[in] env`: The environment that the API is invoked under.
+* `[in] str`: Character buffer representing an ISO-8859-1-encoded string.
+* `[in] length`: The length of the string in bytes, or `NAPI_AUTO_LENGTH` if it
+  is null-terminated.
+* `[in] finalize_callback`: The function to call when the string is being
+  collected. The function will be called with the following parameters:
+  * `[in] env`: The environment in which the add-on is running. This value
+    may be null if the string is being collected as part of the termination
+    of the worker or the main Node.js instance.
+  * `[in] data`: This is the value `str` as a `void*` pointer.
+  * `[in] finalize_hint`: This is the value `finalize_hint` that was given
+    to the API.
+    [`napi_finalize`][] provides more details.
+    This parameter is optional. Passing a null value means that the add-on
+    doesn't need to be notified when the corresponding JavaScript string is
+    collected.
+* `[in] finalize_hint`: Optional hint to pass to the finalize callback during
+  collection.
+* `[out] result`: A `napi_value` representing a JavaScript `string`.
+* `[out] copied`: Whether the string was copied. If it was, the finalizer will
+  already have been invoked to destroy `str`.
+
+Returns `napi_ok` if the API succeeded.
+
+This API creates a JavaScript `string` value from an ISO-8859-1-encoded C
+string. The native string may not be copied and must thus exist for the entire
+life cycle of the JavaScript value.
+
+The JavaScript `string` type is described in
+[Section 6.1.4][] of the ECMAScript Language Specification.
+
 #### `napi_create_string_utf16`
 
 <!-- YAML
@@ -2910,6 +2965,56 @@ The native string is copied.
 The JavaScript `string` type is described in
 [Section 6.1.4][] of the ECMAScript Language Specification.
 
+#### `node_api_create_external_string_utf16`
+
+<!-- YAML
+added: REPLACEME
+-->
+
+> Stability: 1 - Experimental
+
+```c
+napi_status
+node_api_create_external_string_utf16(napi_env env,
+                                      char16_t* str,
+                                      size_t length,
+                                      napi_finalize finalize_callback,
+                                      void* finalize_hint,
+                                      napi_value* result,
+                                      bool* copied);
+```
+
+* `[in] env`: The environment that the API is invoked under.
+* `[in] str`: Character buffer representing a UTF16-LE-encoded string.
+* `[in] length`: The length of the string in two-byte code units, or
+  `NAPI_AUTO_LENGTH` if it is null-terminated.
+* `[in] finalize_callback`: The function to call when the string is being
+  collected. The function will be called with the following parameters:
+  * `[in] env`: The environment in which the add-on is running. This value
+    may be null if the string is being collected as part of the termination
+    of the worker or the main Node.js instance.
+  * `[in] data`: This is the value `str` as a `void*` pointer.
+  * `[in] finalize_hint`: This is the value `finalize_hint` that was given
+    to the API.
+    [`napi_finalize`][] provides more details.
+    This parameter is optional. Passing a null value means that the add-on
+    doesn't need to be notified when the corresponding JavaScript string is
+    collected.
+* `[in] finalize_hint`: Optional hint to pass to the finalize callback during
+  collection.
+* `[out] result`: A `napi_value` representing a JavaScript `string`.
+* `[out] copied`: Whether the string was copied. If it was, the finalizer will
+  already have been invoked to destroy `str`.
+
+Returns `napi_ok` if the API succeeded.
+
+This API creates a JavaScript `string` value from a UTF16-LE-encoded C string.
+The native string may not be copied and must thus exist for the entire life
+cycle of the JavaScript value.
+
+The JavaScript `string` type is described in
+[Section 6.1.4][] of the ECMAScript Language Specification.
+
 #### `napi_create_string_utf8`
 
 <!-- YAML
@@ -6472,6 +6577,8 @@ the add-on's file name during loading.
 [`napi_wrap`]: #napi_wrap
 [`node-addon-api`]: https://github.com/nodejs/node-addon-api
 [`node_api.h`]: https://github.com/nodejs/node/blob/HEAD/src/node_api.h
+[`node_api_create_external_string_latin1`]: #node_api_create_external_string_latin1
+[`node_api_create_external_string_utf16`]: #node_api_create_external_string_utf16
 [`node_api_create_syntax_error`]: #node_api_create_syntax_error
 [`node_api_throw_syntax_error`]: #node_api_throw_syntax_error
 [`process.release`]: process.md#processrelease

src/js_native_api.h

@@ -92,6 +92,24 @@ NAPI_EXTERN napi_status NAPI_CDECL napi_create_string_utf16(napi_env env,
                                                             const char16_t* str,
                                                             size_t length,
                                                             napi_value* result);
+#ifdef NAPI_EXPERIMENTAL
+NAPI_EXTERN napi_status NAPI_CDECL
+node_api_create_external_string_latin1(napi_env env,
+                                       char* str,
+                                       size_t length,
+                                       napi_finalize finalize_callback,
+                                       void* finalize_hint,
+                                       napi_value* result,
+                                       bool* copied);
+NAPI_EXTERN napi_status NAPI_CDECL
+node_api_create_external_string_utf16(napi_env env,
+                                      char16_t* str,
+                                      size_t length,
+                                      napi_finalize finalize_callback,
+                                      void* finalize_hint,
+                                      napi_value* result,
+                                      bool* copied);
+#endif  // NAPI_EXPERIMENTAL
 NAPI_EXTERN napi_status NAPI_CDECL napi_create_symbol(napi_env env,
                                                       napi_value description,
                                                       napi_value* result);