StringLowering: Escape the JSON in the custom section

src/passes/StringLowering.cpp

@@ -37,7 +37,7 @@
 #include "ir/type-updating.h"
 #include "ir/utils.h"
 #include "pass.h"
-#include "support/json.h"
+#include "support/string.h"
 #include "wasm-builder.h"
 #include "wasm.h"
 
@@ -205,8 +205,9 @@ struct StringLowering : public StringGathering {
 
  void makeImports(Module* module) {
  Index importIndex = 0;
- json::Value stringArray;
- stringArray.setArray();
+ std::stringstream json;
+ json << '[';
+ bool first = true;
  std::vector<Name> importedStrings;
  for (auto& global : module->globals) {
  if (global->init) {
@@ -216,16 +217,19 @@ struct StringLowering : public StringGathering {
  importIndex++;
  global->init = nullptr;
 
- auto str = json::Value::make(std::string(c->string.str).c_str());
- stringArray.push_back(str);
+ if (first) {
+ first = false;
+ } else {
+ json << ',';
+ }
+ String::printEscapedJSON(json, c->string.str);
  }
  }
  }
 
  // Add a custom section with the JSON.
- std::stringstream stream;
- stringArray.stringify(stream);
- auto str = stream.str();
+ json << ']';
+ auto str = json.str();
  auto vec = std::vector<char>(str.begin(), str.end());
  module->customSections.emplace_back(
  CustomSection{"string.consts", std::move(vec)});

src/support/json.cpp

@@ -15,13 +15,13 @@
  */
 
 #include "support/json.h"
+#include "support/string.h"
 
 namespace json {
 
 void Value::stringify(std::ostream& os, bool pretty) {
  if (isString()) {
- // TODO: escaping
- os << '"' << getCString() << '"';
+ wasm::String::printEscapedJSON(os, getCString());
  } else if (isArray()) {
  os << '[';
  auto first = true;

src/support/string.cpp

@@ -106,7 +106,7 @@ std::string trim(const std::string& input) {
  return input.substr(0, size);
 }
 
-std::ostream& printEscaped(std::ostream& os, std::string_view str) {
+std::ostream& printEscaped(std::ostream& os, const std::string_view str) {
  os << '"';
  for (unsigned char c : str) {
  switch (c) {
@@ -140,4 +140,76 @@ std::ostream& printEscaped(std::ostream& os, std::string_view str) {
  return os << '"';
 }
 
+std::ostream& printEscapedJSON(std::ostream& os, const std::string_view str) {
+ os << '"';
+ for (size_t i = 0; i < str.size(); i++) {
+ int u0 = str[i];
+ switch (u0) {
+ case '\t':
+ os << "\\t";
+ continue;
+ case '\n':
+ os << "\\n";
+ continue;
+ case '\r':
+ os << "\\r";
+ continue;
+ case '"':
+ os << "\\\"";
+ continue;
+ case '\'':
+ os << "'";
+ continue;
+ case '\\':
+ os << "\\\\";
+ continue;
+ default: {
+ auto uEscape = [&](uint32_t v) {
+ os << std::hex;
+ os << "\\u";
+ os << ((v >> 24) & 0xff);
+ os << ((v >> 16) & 0xff);
+ os << ((v >> 8) & 0xff);
+ os << (v & 0xff);
+ os << std::dec;
+ };
+
+ // Based off of
+ // https://github.com/emscripten-core/emscripten/blob/59e6b8f1262d75585d8416b728e8cbb3db176fe2/src/library_strings.js#L72-L91
+ if (!(u0 & 0x80)) {
+ if (u0 >= 32 && u0 < 127) {
+ os << char(u0);
+ } else {
+ uEscape(u0);
+ }
+ continue;
+ }
+ i++;
+ int u1 = str[i] & 63;
+ if ((u0 & 0xE0) == 0xC0) {
+ uEscape((((u0 & 31) << 6) | u1));
+ continue;
+ }
+ i++;
+ int u2 = str[i] & 63;
+ if ((u0 & 0xF0) == 0xE0) {
+ u0 = ((u0 & 15) << 12) | (u1 << 6) | u2;
+ } else {
+ i++;
+ u0 = ((u0 & 7) << 18) | (u1 << 12) | (u2 << 6) | (str[i] & 63);
+ }
+
+ if (u0 < 0x10000) {
+ uEscape(u0);
+ } else {
+ auto ch = u0 - 0x10000;
+ uEscape(0xD800 | (ch >> 10));
+ uEscape(0xDC00 | (ch & 0x3FF));
+ }
+ }
+ }
+ }
+ return os << '"';
+}
+
 } // namespace wasm::String

src/support/string.h

@@ -75,7 +75,9 @@ inline bool isNumber(const std::string& str) {
  return !str.empty() && std::all_of(str.begin(), str.end(), ::isdigit);
 }
 
-std::ostream& printEscaped(std::ostream& os, std::string_view str);
+std::ostream& printEscaped(std::ostream& os, const std::string_view str);
+
+std::ostream& printEscapedJSON(std::ostream& os, const std::string_view str);
 
 } // namespace wasm::String
 

test/lit/passes/string-lowering.js

@@ -0,0 +1,8 @@
+var filename = process.argv[2];
+var module = new WebAssembly.Module(require('fs').readFileSync(filename));
+var sections = WebAssembly.Module.customSections(module, 'string.consts');
+var array = new Uint8Array(sections[0]);
+var string = new TextDecoder('utf-8').decode(array);
+console.log("string:", string);
+var json = JSON.parse(string);
+console.log("JSON:", json);

test/lit/passes/string-lowering.wast

@@ -2,8 +2,6 @@
 ;; operations are tested in string-gathering.wast (which is auto-updated, unlike
 ;; this which is manual).
 
-;; RUN: foreach %s %t wasm-opt --string-lowering -all -S -o - | filecheck %s
-
 (module
  (func $consts
  (drop
@@ -15,9 +13,26 @@
  (drop
  (string.const "foo")
  )
+ (drop
+ (string.const "needs\tescaping\00.")
+ )
  )
 )
 
-;; The custom section should contain foo and bar, and foo only once.
-;; CHECK: custom section "string.consts", size 13, contents: "[\"bar\",\"foo\"]"
+;; The custom section should contain foo and bar, and foo only once, and the
+;; string with \t should be escaped.
+;;
+;; RUN: wasm-opt %s --string-lowering -all -S -o - | filecheck %s
+;;
+;; CHECK: custom section "string.consts", size 38, contents: "[\"bar\",\"foo\",\"needs\\tescaping\\u0000.\"]"
+
+;; The custom section should parse OK using JSON.parse from node.
+;; (Note we run --remove-unused-module-elements to remove externref-using
+;; imports, which require a newer version of node.)
+;;
+;; RUN: wasm-opt %s --string-lowering --remove-unused-module-elements -all -o %t.wasm
+;; RUN: node %S/string-lowering.js %t.wasm | filecheck %s --check-prefix=CHECK-JS
+;;
+;; CHECK-JS: string: ["bar","foo","needs\tescaping\u0000."]
+;; CHECK-JS: JSON: [ 'bar', 'foo', 'needs\tescaping\u0000.' ]