Remove std::wstring usage in CJS_Util.
Use WideString instead. Clean up modified methods along the way.
Bug: pdfium:488
Change-Id: I9f461e6e5bae337ae01ebe16708484b1f04c0ea4
Reviewed-on: https://pdfium-review.googlesource.com/c/pdfium/+/63630
Commit-Queue: Lei Zhang <thestig@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
diff --git a/fxjs/cjs_util.cpp b/fxjs/cjs_util.cpp
index b390975..9a85c8a 100644
--- a/fxjs/cjs_util.cpp
+++ b/fxjs/cjs_util.cpp
@@ -11,7 +11,6 @@
#include <algorithm>
#include <cmath>
#include <cwctype>
-#include <string>
#include <vector>
#include "build/build_config.h"
@@ -98,64 +97,64 @@
CJS_Result CJS_Util::printf(CJS_Runtime* pRuntime,
const std::vector<v8::Local<v8::Value>>& params) {
- const size_t iSize = params.size();
- if (iSize < 1)
+ const size_t num_params = params.size();
+ if (num_params < 1)
return CJS_Result::Failure(JSMessage::kParamError);
- std::wstring unsafe_fmt_string(pRuntime->ToWideString(params[0]).c_str());
- std::vector<std::wstring> unsafe_conversion_specifiers;
- int iOffset = 0;
- int iOffend = 0;
- unsafe_fmt_string.insert(unsafe_fmt_string.begin(), L'S');
- while (iOffset != -1) {
- iOffend = unsafe_fmt_string.find(L"%", iOffset + 1);
- std::wstring strSub;
- if (iOffend == -1)
- strSub = unsafe_fmt_string.substr(iOffset);
- else
- strSub = unsafe_fmt_string.substr(iOffset, iOffend - iOffset);
- unsafe_conversion_specifiers.push_back(strSub);
- iOffset = iOffend;
+ // Use 'S' as a sentinel to ensure we always have some text before the first
+ // format specifier.
+ WideString unsafe_fmt_string = L'S' + pRuntime->ToWideString(params[0]);
+ std::vector<WideString> unsafe_conversion_specifiers;
+
+ {
+ size_t offset = 0;
+ while (true) {
+ Optional<size_t> offset_end = unsafe_fmt_string.Find(L"%", offset + 1);
+ if (!offset_end.has_value()) {
+ unsafe_conversion_specifiers.push_back(
+ unsafe_fmt_string.Right(unsafe_fmt_string.GetLength() - offset));
+ break;
+ }
+
+ unsafe_conversion_specifiers.push_back(
+ unsafe_fmt_string.Mid(offset, offset_end.value() - offset));
+ offset = offset_end.value();
+ }
}
- std::wstring c_strResult;
- for (size_t iIndex = 0; iIndex < unsafe_conversion_specifiers.size();
- ++iIndex) {
- std::wstring c_strFormat = unsafe_conversion_specifiers[iIndex];
- if (iIndex == 0) {
- c_strResult = c_strFormat;
+ WideString result = unsafe_conversion_specifiers[0];
+ for (size_t i = 1; i < unsafe_conversion_specifiers.size(); ++i) {
+ WideString fmt = unsafe_conversion_specifiers[i];
+ if (i >= num_params) {
+ result += fmt;
continue;
}
- if (iIndex >= iSize) {
- c_strResult += c_strFormat;
- continue;
- }
-
- WideString strSegment;
- switch (ParseDataType(&c_strFormat)) {
+ WideString segment;
+ switch (ParseDataType(&fmt)) {
case UTIL_INT:
- strSegment = WideString::Format(c_strFormat.c_str(),
- pRuntime->ToInt32(params[iIndex]));
+ segment = WideString::Format(fmt.c_str(), pRuntime->ToInt32(params[i]));
break;
case UTIL_DOUBLE:
- strSegment = WideString::Format(c_strFormat.c_str(),
- pRuntime->ToDouble(params[iIndex]));
+ segment =
+ WideString::Format(fmt.c_str(), pRuntime->ToDouble(params[i]));
break;
case UTIL_STRING:
- strSegment =
- WideString::Format(c_strFormat.c_str(),
- pRuntime->ToWideString(params[iIndex]).c_str());
+ segment = WideString::Format(fmt.c_str(),
+ pRuntime->ToWideString(params[i]).c_str());
break;
default:
- strSegment = WideString::Format(L"%ls", c_strFormat.c_str());
+ segment = WideString::Format(L"%ls", fmt.c_str());
break;
}
- c_strResult += strSegment.c_str();
+ result += segment;
}
- c_strResult.erase(c_strResult.begin());
- return CJS_Result::Success(pRuntime->NewString(c_strResult.c_str()));
+ // Remove the 'S' sentinel introduced earlier.
+ DCHECK_EQ(L'S', result[0]);
+ auto result_view = result.AsStringView();
+ return CJS_Result::Success(
+ pRuntime->NewString(result_view.Right(result_view.GetLength() - 1)));
}
CJS_Result CJS_Util::printd(CJS_Runtime* pRuntime,
@@ -275,6 +274,7 @@
.AsStringView()));
}
+// static
WideString CJS_Util::StringPrintx(const WideString& wsFormat,
const WideString& wsSource) {
WideString wsResult;
@@ -394,18 +394,15 @@
return CJS_Result::Success(pRuntime->NewString(wStr.AsStringView()));
}
-// Ensure that sFormat contains at most one well-understood printf formatting
-// directive which is safe to use with a single argument, and return the type
-// of argument expected, or -1 otherwise. If -1 is returned, it is NOT safe
-// to use sFormat with printf() and it must be copied byte-by-byte.
-int CJS_Util::ParseDataType(std::wstring* sFormat) {
+// static
+int CJS_Util::ParseDataType(WideString* sFormat) {
enum State { BEFORE, FLAGS, WIDTH, PRECISION, SPECIFIER, AFTER };
int result = -1;
State state = BEFORE;
size_t precision_digits = 0;
size_t i = 0;
- while (i < sFormat->length()) {
+ while (i < sFormat->GetLength()) {
wchar_t c = (*sFormat)[i];
switch (state) {
case BEFORE:
@@ -454,7 +451,7 @@
// Map s to S since we always deal internally with wchar_t strings.
// TODO(tsepez): Probably 100% borked. %S is not a standard
// conversion.
- (*sFormat)[i] = L'S';
+ sFormat->SetAt(i, L'S');
result = UTIL_STRING;
} else {
return -1;
diff --git a/fxjs/cjs_util.h b/fxjs/cjs_util.h
index 3339ce7..74e96cc 100644
--- a/fxjs/cjs_util.h
+++ b/fxjs/cjs_util.h
@@ -7,9 +7,9 @@
#ifndef FXJS_CJS_UTIL_H_
#define FXJS_CJS_UTIL_H_
-#include <string>
#include <vector>
+#include "core/fxcrt/widestring.h"
#include "fxjs/cjs_object.h"
#include "fxjs/js_define.h"
@@ -26,8 +26,16 @@
CJS_Util(v8::Local<v8::Object> pObject, CJS_Runtime* pRuntime);
~CJS_Util() override;
+ // Ensure that |sFormat| contains at most one well-understood printf
+ // formatting directive which is safe to use with a single argument, and
+ // return the type of argument expected, or -1 otherwise. If -1 is returned,
+ // it is NOT safe to use |sFormat| with printf() and it must be copied
+ // byte-by-byte.
+ //
// Exposed for testing.
- static int ParseDataType(std::wstring* sFormat);
+ static int ParseDataType(WideString* sFormat);
+
+ // Exposed for testing.
static WideString StringPrintx(const WideString& cFormat,
const WideString& cSource);
diff --git a/fxjs/cjs_util_unittest.cpp b/fxjs/cjs_util_unittest.cpp
index 5a1fa0f..d405746 100644
--- a/fxjs/cjs_util_unittest.cpp
+++ b/fxjs/cjs_util_unittest.cpp
@@ -105,7 +105,7 @@
};
for (size_t i = 0; i < FX_ArraySize(cases); i++) {
- std::wstring input(cases[i].input_string);
+ WideString input(cases[i].input_string);
EXPECT_EQ(cases[i].expected, CJS_Util::ParseDataType(&input))
<< cases[i].input_string;
}
diff --git a/testing/fuzzers/pdf_cjs_util_fuzzer.cc b/testing/fuzzers/pdf_cjs_util_fuzzer.cc
index a773244..5ccb65b 100644
--- a/testing/fuzzers/pdf_cjs_util_fuzzer.cc
+++ b/testing/fuzzers/pdf_cjs_util_fuzzer.cc
@@ -2,8 +2,6 @@
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
-#include <string>
-
#include "core/fxcrt/widestring.h"
#include "fxjs/cjs_util.h"
@@ -12,8 +10,7 @@
size_t short_size = size / sizeof(unsigned short);
if (short_size > 1) {
WideString input = WideString::FromUTF16LE(short_data, short_size);
- std::wstring winput(input.c_str(), input.GetLength());
- CJS_Util::ParseDataType(&winput);
+ CJS_Util::ParseDataType(&input);
}
if (short_size > 2) {
size_t short_len1 = short_size / 2;
