feat: emscripten support (#412)

* emscripten support

* Trying to fix the new CI test

* Trying to add test

* Only preduce wasm.js when building for wasm

* Automating the tests.

* Fix.

* Adding tests

Co-authored-by: Moshe Atlow <[email protected]>

* Reformat

* Fixing ci.

* Removing dead code.

---------

Co-authored-by: Moshe Atlow <[email protected]>

Author: lemire

.github/workflows/emscripten.yml

@@ -0,0 +1,39 @@
+name: emscripten
+
+on:
+  pull_request:
+    types: [opened, synchronize, reopened, ready_for_review]
+    paths-ignore:
+      - '**.md'
+      - 'docs/**'
+  push:
+    branches:
+      - main
+    paths-ignore:
+      - '**.md'
+      - 'docs/**'
+
+permissions:
+  contents: read
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/setup-node@v3
+      - uses: mymindstorm/setup-emsdk@v12
+      - name: Verify
+        run: emcc -v
+      - name: Checkout
+        uses: actions/checkout@v3
+      - name: Configure
+        run: emcmake  cmake -B buildwasm
+      - name: Build
+        run: cmake --build buildwasm
+      - name: Test
+        run: ctest --test-dir buildwasm

CMakeLists.txt

@@ -33,23 +33,27 @@ if((BUILD_TESTING OR ADA_BENCHMARKS) AND (CMAKE_SIZEOF_VOID_P EQUAL 8))
   add_dependency(gtest)
 endif()
 
-if (BUILD_TESTING)
+if (BUILD_TESTING AND NOT EMSCRIPTEN)
   message(STATUS "The tests are enabled.")
   add_subdirectory(tests)
 else()
   if(is_top_project)
     message(STATUS "The tests are disabled.")
   endif()
-endif(BUILD_TESTING)
+endif(BUILD_TESTING AND NOT EMSCRIPTEN)
 
-If(ADA_BENCHMARKS)
+If(ADA_BENCHMARKS AND NOT EMSCRIPTEN)
   message(STATUS "Ada benchmarks enabled.")
   add_subdirectory(benchmarks)
-else(ADA_BENCHMARKS)
+else(ADA_BENCHMARKS AND NOT EMSCRIPTEN)
   if(is_top_project)
     message(STATUS "Ada benchmarks disabled. Set ADA_BENCHMARKS=ON to enable them.")
   endif()
-endif(ADA_BENCHMARKS)
+endif(ADA_BENCHMARKS AND NOT EMSCRIPTEN)
+
+if (BUILD_TESTING AND EMSCRIPTEN)
+  add_subdirectory(tests-wasm)
+endif(BUILD_TESTING AND EMSCRIPTEN)
 
 add_library(ada::ada ALIAS ada)
 
@@ -60,16 +64,10 @@ set_target_properties(
   WINDOWS_EXPORT_ALL_SYMBOLS YES
 )
 
-include (TestBigEndian)
-TEST_BIG_ENDIAN(IS_BIG_ENDIAN)
-if(IS_BIG_ENDIAN)
-  message(STATUS "Big-endian system detected.")
-endif()
-
 include(CMakePackageConfigHelpers)
 include(GNUInstallDirs)
 
-if(NOT ADA_COVERAGE)
+if(NOT ADA_COVERAGE AND NOT EMSCRIPTEN)
   add_subdirectory(singleheader)
   add_subdirectory(tools)
 endif()

src/helpers.cpp

@@ -238,7 +238,7 @@ ada_really_inline size_t find_next_host_delimiter_special(
                         has_zero_byte(xor3) | has_zero_byte(xor4) |
                         has_zero_byte(xor5);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
   if (i < view.size()) {
@@ -256,7 +256,7 @@ ada_really_inline size_t find_next_host_delimiter_special(
                         has_zero_byte(xor3) | has_zero_byte(xor4) |
                         has_zero_byte(xor5);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
   return view.size();
@@ -300,7 +300,7 @@ ada_really_inline size_t find_next_host_delimiter(std::string_view view,
     uint64_t is_match = has_zero_byte(xor1) | has_zero_byte(xor2) |
                         has_zero_byte(xor4) | has_zero_byte(xor5);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
   if (i < view.size()) {
@@ -318,7 +318,7 @@ ada_really_inline size_t find_next_host_delimiter(std::string_view view,
     uint64_t is_match = has_zero_byte(xor1) | has_zero_byte(xor2) |
                         has_zero_byte(xor4) | has_zero_byte(xor5);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
   return view.size();
@@ -625,7 +625,7 @@ find_authority_delimiter_special(std::string_view view) noexcept {
     uint64_t is_match = has_zero_byte(xor1) | has_zero_byte(xor2) |
                         has_zero_byte(xor3) | has_zero_byte(xor4);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
 
@@ -640,7 +640,7 @@ find_authority_delimiter_special(std::string_view view) noexcept {
     uint64_t is_match = has_zero_byte(xor1) | has_zero_byte(xor2) |
                         has_zero_byte(xor3) | has_zero_byte(xor4);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
 
@@ -673,7 +673,7 @@ find_authority_delimiter(std::string_view view) noexcept {
     uint64_t is_match =
         has_zero_byte(xor1) | has_zero_byte(xor2) | has_zero_byte(xor3);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
 
@@ -687,7 +687,7 @@ find_authority_delimiter(std::string_view view) noexcept {
     uint64_t is_match =
         has_zero_byte(xor1) | has_zero_byte(xor2) | has_zero_byte(xor3);
     if (is_match) {
-      return i + index_of_first_set_byte(is_match);
+      return size_t(i + index_of_first_set_byte(is_match));
     }
   }
 

src/parser.cpp

@@ -36,8 +36,7 @@ result_type parse_url(std::string_view user_input,
 
   // We refuse to parse URL strings that exceed 4GB. Such strings are almost
   // surely the result of a bug or are otherwise a security concern.
-  if (user_input.size() >=
-      std::string_view::size_type(std::numeric_limits<uint32_t>::max)) {
+  if (user_input.size() > std::numeric_limits<uint32_t>::max()) {
     url.is_valid = false;
   }
   // Going forward, user_input.size() is in [0,

tests-wasm/CMakeLists.txt

@@ -0,0 +1,11 @@
+link_libraries(ada)
+add_executable(wasm wasm.cpp)
+set_target_properties(wasm PROPERTIES LINK_FLAGS    "-Os -s WASM=1 -s ENVIRONMENT=node -s EXPORT_NAME=loadWASM -s MODULARIZE=1 --bind --no-entry")
+configure_file(test.js.in test.js)
+
+find_program(NODEJS_BINARY NAMES node nodejs)
+if(NODEJS_BINARY)
+  add_test(NAME wasmtest
+         COMMAND "${NODEJS_BINARY}" "${CMAKE_CURRENT_BINARY_DIR}/test.js")
+endif(NODEJS_BINARY)
+

tests-wasm/test.js.in

@@ -0,0 +1,35 @@
+const child_process = require('child_process');
+const assert = require('assert');
+const test = require('node:test');
+const wasm = require('${CMAKE_CURRENT_BINARY_DIR}/wasm');
+
+function toJS(obj) {
+  const result = {};
+  for (const key of Object.keys(obj.__proto__)) {
+       result[key] = typeof obj[key] === "object" ? toJS(obj[key]) : obj[key];
+  }
+  return result;
+}
+
+const expected = {
+  "result": "success",
+  "href": "https://google.com/?q=Yagiz#Nizipli",
+  "type": 2,
+  "components": {
+    "protocol_end": 6,
+    "username_end": 8,
+    "host_start": 8,
+    "host_end": 18,
+    "port": 4294967295,
+    "pathname_start": 18,
+    "search_start": 19,
+    "hash_start": 27
+  }
+};
+
+test('wasm', async () => {
+  const { parse } = await wasm();
+  console.log();
+  assert.deepStrictEqual(toJS(parse('https://google.com/?q=Yagiz#Nizipli')), expected);
+});
+

tests-wasm/wasm.cpp

@@ -0,0 +1,45 @@
+#include "ada.h"
+#include <emscripten/emscripten.h>
+#include <emscripten/bind.h>
+
+using namespace emscripten;
+
+struct parse_result {
+  std::string result;
+  std::string href;
+  uint32_t type;
+  ada::url_components components;
+};
+
+parse_result parse(const std::string &input) {
+  auto out = ada::parse<ada::url_aggregator>(input);
+  parse_result result;
+  if (!out.has_value()) {
+    result.result = "fail";
+  } else {
+    result.result = "success";
+    result.href = std::string(out->get_href());
+    result.type = out->type;
+    result.components = out->get_components();
+  }
+  return result;
+}
+
+EMSCRIPTEN_BINDINGS(url_components) {
+  class_<parse_result>("Result")
+      .property("result", &parse_result::result)
+      .property("href", &parse_result::href)
+      .property("type", &parse_result::type)
+      .property("components", &parse_result::components);
+  class_<ada::url_components>("URLComponents")
+      .property("protocol_end", &ada::url_components::protocol_end)
+      .property("username_end", &ada::url_components::username_end)
+      .property("host_start", &ada::url_components::host_start)
+      .property("host_end", &ada::url_components::host_end)
+      .property("port", &ada::url_components::port)
+      .property("pathname_start", &ada::url_components::pathname_start)
+      .property("search_start", &ada::url_components::search_start)
+      .property("hash_start", &ada::url_components::hash_start);
+
+  function("parse", &parse);
+}