blob: 5853c82af8e7650f44149a62a99d4591d21531f4 [file] [log] [blame] [edit]
// Copyright 2020 The Tint Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include "src/reader/wgsl/lexer.h"
#include <cmath>
#include <cstring>
#include <limits>
#include <utility>
#include "src/debug.h"
namespace tint {
namespace reader {
namespace wgsl {
namespace {
bool is_whitespace(char c) {
return std::isspace(c);
}
uint32_t dec_value(char c) {
if (c >= '0' && c <= '9') {
return static_cast<uint32_t>(c - '0');
}
return 0;
}
uint32_t hex_value(char c) {
if (c >= '0' && c <= '9') {
return static_cast<uint32_t>(c - '0');
}
if (c >= 'a' && c <= 'f') {
return 0xA + static_cast<uint32_t>(c - 'a');
}
if (c >= 'A' && c <= 'F') {
return 0xA + static_cast<uint32_t>(c - 'A');
}
return 0;
}
} // namespace
Lexer::Lexer(const std::string& file_path, const Source::FileContent* content)
: file_path_(file_path),
content_(content),
len_(static_cast<uint32_t>(content->data.size())),
location_{1, 1} {}
Lexer::~Lexer() = default;
Token Lexer::next() {
skip_whitespace();
skip_comments();
if (is_eof()) {
return {Token::Type::kEOF, begin_source()};
}
auto t = try_hex_float();
if (!t.IsUninitialized()) {
return t;
}
t = try_hex_integer();
if (!t.IsUninitialized()) {
return t;
}
t = try_float();
if (!t.IsUninitialized()) {
return t;
}
t = try_integer();
if (!t.IsUninitialized()) {
return t;
}
t = try_punctuation();
if (!t.IsUninitialized()) {
return t;
}
t = try_ident();
if (!t.IsUninitialized()) {
return t;
}
return {Token::Type::kError, begin_source(), "invalid character found"};
}
Source Lexer::begin_source() const {
Source src{};
src.file_path = file_path_;
src.file_content = content_;
src.range.begin = location_;
src.range.end = location_;
return src;
}
void Lexer::end_source(Source& src) const {
src.range.end = location_;
}
bool Lexer::is_eof() const {
return pos_ >= len_;
}
bool Lexer::is_alpha(char ch) const {
return std::isalpha(ch) || ch == '_';
}
bool Lexer::is_digit(char ch) const {
return std::isdigit(ch);
}
bool Lexer::is_alphanum(char ch) const {
return is_alpha(ch) || is_digit(ch);
}
bool Lexer::is_hex(char ch) const {
return std::isxdigit(ch);
}
bool Lexer::matches(size_t pos, const std::string& substr) {
if (pos >= len_)
return false;
return content_->data.substr(pos, substr.size()) == substr;
}
void Lexer::skip_whitespace() {
for (;;) {
auto pos = pos_;
while (!is_eof() && is_whitespace(content_->data[pos_])) {
if (matches(pos_, "\n")) {
pos_++;
location_.line++;
location_.column = 1;
continue;
}
pos_++;
location_.column++;
}
skip_comments();
// If the cursor didn't advance we didn't remove any whitespace
// so we're done.
if (pos == pos_)
break;
}
}
void Lexer::skip_comments() {
if (matches(pos_, "//")) {
// Line comment: ignore everything until the end of line.
while (!is_eof() && !matches(pos_, "\n")) {
pos_++;
location_.column++;
}
return;
}
if (matches(pos_, "/*")) {
// Block comment: ignore everything until the closing '*/' token.
pos_ += 2;
location_.column += 2;
int depth = 1;
while (!is_eof() && depth > 0) {
if (matches(pos_, "/*")) {
// Start of block comment: increase nesting depth.
pos_ += 2;
location_.column += 2;
depth++;
} else if (matches(pos_, "*/")) {
// End of block comment: decrease nesting depth.
pos_ += 2;
location_.column += 2;
depth--;
} else if (matches(pos_, "\n")) {
// Newline: skip and update source location.
pos_++;
location_.line++;
location_.column = 1;
} else {
// Anything else: skip and update source location.
pos_++;
location_.column++;
}
}
}
}
Token Lexer::try_float() {
auto start = pos_;
auto end = pos_;
auto source = begin_source();
if (matches(end, "-")) {
end++;
}
while (end < len_ && is_digit(content_->data[end])) {
end++;
}
if (end >= len_ || !matches(end, ".")) {
return {};
}
end++;
while (end < len_ && is_digit(content_->data[end])) {
end++;
}
// Parse the exponent if one exists
if (end < len_ && matches(end, "e")) {
end++;
if (end < len_ && (matches(end, "+") || matches(end, "-"))) {
end++;
}
auto exp_start = end;
while (end < len_ && isdigit(content_->data[end])) {
end++;
}
// Must have an exponent
if (exp_start == end)
return {};
}
auto str = content_->data.substr(start, end - start);
if (str == "." || str == "-.")
return {};
pos_ = end;
location_.column += (end - start);
end_source(source);
auto res = strtod(content_->data.c_str() + start, nullptr);
// This handles if the number is a really small in the exponent
if (res > 0 && res < static_cast<double>(std::numeric_limits<float>::min())) {
return {Token::Type::kError, source, "f32 (" + str + " too small"};
}
// This handles if the number is really large negative number
if (res < static_cast<double>(std::numeric_limits<float>::lowest())) {
return {Token::Type::kError, source, "f32 (" + str + ") too small"};
}
if (res > static_cast<double>(std::numeric_limits<float>::max())) {
return {Token::Type::kError, source, "f32 (" + str + ") too large"};
}
return {source, static_cast<float>(res)};
}
Token Lexer::try_hex_float() {
constexpr uint32_t kTotalBits = 32;
constexpr uint32_t kTotalMsb = kTotalBits - 1;
constexpr uint32_t kMantissaBits = 23;
constexpr uint32_t kMantissaMsb = kMantissaBits - 1;
constexpr uint32_t kMantissaShiftRight = kTotalBits - kMantissaBits;
constexpr int32_t kExponentBias = 127;
constexpr int32_t kExponentMax = 255;
constexpr uint32_t kExponentBits = 8;
constexpr uint32_t kExponentMask = (1 << kExponentBits) - 1;
constexpr uint32_t kExponentLeftShift = kMantissaBits;
constexpr uint32_t kSignBit = 31;
auto start = pos_;
auto end = pos_;
auto source = begin_source();
// clang-format off
// -?0x([0-9a-fA-F]*.?[0-9a-fA-F]+ | [0-9a-fA-F]+.[0-9a-fA-F]*)(p|P)(+|-)?[0-9]+ // NOLINT
// clang-format on
// -?
int32_t sign_bit = 0;
if (matches(end, "-")) {
sign_bit = 1;
end++;
}
// 0x
if (matches(end, "0x")) {
end += 2;
} else {
return {};
}
uint32_t mantissa = 0;
int32_t exponent = 0;
// `set_next_mantissa_bit_to` sets next `mantissa` bit starting from msb to
// lsb to value 1 if `set` is true, 0 otherwise
uint32_t mantissa_next_bit = kTotalMsb;
auto set_next_mantissa_bit_to = [&](bool set, bool integer_part) -> bool {
// If adding bits for the integer part, we can overflow whether we set the
// bit or not. For the fractional part, we can only overflow when setting
// the bit.
const bool check_overflow = integer_part || set;
if (check_overflow && (mantissa_next_bit > kTotalMsb)) {
return false; // Overflowed mantissa
}
if (set) {
mantissa |= (1 << mantissa_next_bit);
}
--mantissa_next_bit;
return true;
};
// Collect integer range (if any)
auto integer_range = std::make_pair(end, end);
while (end < len_ && is_hex(content_->data[end])) {
integer_range.second = ++end;
}
// .?
if (matches(end, ".")) {
end++;
}
// Collect fractional range (if any)
auto fractional_range = std::make_pair(end, end);
while (end < len_ && is_hex(content_->data[end])) {
fractional_range.second = ++end;
}
// Must have at least an integer or fractional part
if ((integer_range.first == integer_range.second) &&
(fractional_range.first == fractional_range.second)) {
return {};
}
// (p|P)
if (matches(end, "p") || matches(end, "P")) {
end++;
} else {
return {};
}
// At this point, we know for sure our token is a hex float value.
// Parse integer part
// [0-9a-fA-F]*
bool has_zero_integer = true;
bool leading_bit_seen = false;
for (auto i = integer_range.first; i < integer_range.second; ++i) {
const auto nibble = hex_value(content_->data[i]);
if (nibble != 0) {
has_zero_integer = false;
}
for (int32_t bit = 3; bit >= 0; --bit) {
auto v = 1 & (nibble >> bit);
// Skip leading 0s and the first 1
if (leading_bit_seen) {
if (!set_next_mantissa_bit_to(v != 0, true)) {
return {Token::Type::kError, source,
"mantissa is too large for hex float"};
}
++exponent;
} else {
if (v == 1) {
leading_bit_seen = true;
}
}
}
}
// Parse fractional part
// [0-9a-fA-F]*
leading_bit_seen = false;
for (auto i = fractional_range.first; i < fractional_range.second; ++i) {
auto nibble = hex_value(content_->data[i]);
for (int32_t bit = 3; bit >= 0; --bit) {
auto v = 1 & (nibble >> bit);
if (v == 1) {
leading_bit_seen = true;
}
// If integer part is 0 (denorm), we only start writing bits to the
// mantissa once we have a non-zero fractional bit. While the fractional
// values are 0, we adjust the exponent to avoid overflowing `mantissa`.
if (has_zero_integer && !leading_bit_seen) {
--exponent;
} else {
if (!set_next_mantissa_bit_to(v != 0, false)) {
return {Token::Type::kError, source,
"mantissa is too large for hex float"};
}
}
}
}
// (+|-)?
int32_t exponent_sign = 1;
if (matches(end, "+")) {
end++;
} else if (matches(end, "-")) {
exponent_sign = -1;
end++;
}
// Parse exponent from input
// [0-9]+
bool has_exponent = false;
uint32_t input_exponent = 0;
while (end < len_ && isdigit(content_->data[end])) {
has_exponent = true;
auto prev_exponent = input_exponent;
input_exponent = (input_exponent * 10) + dec_value(content_->data[end]);
if (prev_exponent > input_exponent) {
return {Token::Type::kError, source,
"exponent is too large for hex float"};
}
end++;
}
if (!has_exponent) {
return {Token::Type::kError, source,
"expected an exponent value for hex float"};
}
pos_ = end;
location_.column += (end - start);
end_source(source);
// Compute exponent so far
exponent = exponent + (input_exponent * exponent_sign);
// Determine if value is zero
// Note: it's not enough to check mantissa == 0 as we drop initial bit from
// integer part.
bool is_zero = has_zero_integer && mantissa == 0;
TINT_ASSERT(Reader, !is_zero || mantissa == 0);
if (is_zero) {
// If value is zero, then ignore the exponent and produce a zero
exponent = 0;
} else {
// Bias exponent if non-zero
// After this, if exponent is <= 0, our value is a denormal
exponent += kExponentBias;
// Denormal uses biased exponent of -126, not -127
if (has_zero_integer) {
mantissa <<= 1;
--exponent;
}
}
// Shift mantissa to occupy the low 23 bits
mantissa >>= kMantissaShiftRight;
// If denormal, shift mantissa until our exponent is zero
if (!is_zero) {
// Denorm has exponent 0 and non-zero mantissa. We set the top bit here,
// then shift the mantissa to make exponent zero.
if (exponent <= 0) {
mantissa >>= 1;
mantissa |= (1 << kMantissaMsb);
}
while (exponent < 0) {
mantissa >>= 1;
++exponent;
// If underflow, clamp to zero
if (mantissa == 0) {
exponent = 0;
}
}
}
if (exponent > kExponentMax) {
// Overflow: set to infinity
exponent = kExponentMax;
mantissa = 0;
} else if (exponent == kExponentMax && mantissa != 0) {
// NaN: set to infinity
mantissa = 0;
}
// Combine sign, mantissa, and exponent
uint32_t result_u32 = sign_bit << kSignBit;
result_u32 |= mantissa;
result_u32 |= (exponent & kExponentMask) << kExponentLeftShift;
// Reinterpret as float and return
float result;
std::memcpy(&result, &result_u32, sizeof(result));
return {source, static_cast<float>(result)};
}
Token Lexer::build_token_from_int_if_possible(Source source,
size_t start,
size_t end,
int32_t base) {
auto res = strtoll(content_->data.c_str() + start, nullptr, base);
if (matches(pos_, "u")) {
if (static_cast<uint64_t>(res) >
static_cast<uint64_t>(std::numeric_limits<uint32_t>::max())) {
return {
Token::Type::kError, source,
"u32 (" + content_->data.substr(start, end - start) + ") too large"};
}
pos_ += 1;
location_.column += 1;
end_source(source);
return {source, static_cast<uint32_t>(res)};
}
if (res < static_cast<int64_t>(std::numeric_limits<int32_t>::min())) {
return {
Token::Type::kError, source,
"i32 (" + content_->data.substr(start, end - start) + ") too small"};
}
if (res > static_cast<int64_t>(std::numeric_limits<int32_t>::max())) {
return {
Token::Type::kError, source,
"i32 (" + content_->data.substr(start, end - start) + ") too large"};
}
end_source(source);
return {source, static_cast<int32_t>(res)};
}
Token Lexer::try_hex_integer() {
auto start = pos_;
auto end = pos_;
auto source = begin_source();
if (matches(end, "-")) {
end++;
}
if (!matches(end, "0x")) {
return Token();
}
end += 2;
while (!is_eof() && is_hex(content_->data[end])) {
end += 1;
}
pos_ = end;
location_.column += (end - start);
return build_token_from_int_if_possible(source, start, end, 16);
}
Token Lexer::try_integer() {
auto start = pos_;
auto end = start;
auto source = begin_source();
if (matches(end, "-")) {
end++;
}
if (end >= len_ || !is_digit(content_->data[end])) {
return {};
}
auto first = end;
while (end < len_ && is_digit(content_->data[end])) {
end++;
}
// If the first digit is a zero this must only be zero as leading zeros
// are not allowed.
if (content_->data[first] == '0' && (end - first != 1))
return {};
pos_ = end;
location_.column += (end - start);
return build_token_from_int_if_possible(source, start, end, 10);
}
Token Lexer::try_ident() {
// Must begin with an a-zA-Z_
if (!is_alpha(content_->data[pos_])) {
return {};
}
auto source = begin_source();
auto s = pos_;
while (!is_eof() && is_alphanum(content_->data[pos_])) {
pos_++;
location_.column++;
}
auto str = content_->data.substr(s, pos_ - s);
end_source(source);
auto t = check_keyword(source, str);
if (!t.IsUninitialized()) {
return t;
}
return {Token::Type::kIdentifier, source, str};
}
Token Lexer::try_punctuation() {
auto source = begin_source();
auto type = Token::Type::kUninitialized;
if (matches(pos_, "[[")) {
type = Token::Type::kAttrLeft;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "]]")) {
type = Token::Type::kAttrRight;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "(")) {
type = Token::Type::kParenLeft;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, ")")) {
type = Token::Type::kParenRight;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "[")) {
type = Token::Type::kBracketLeft;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "]")) {
type = Token::Type::kBracketRight;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "{")) {
type = Token::Type::kBraceLeft;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "}")) {
type = Token::Type::kBraceRight;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "&&")) {
type = Token::Type::kAndAnd;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "&")) {
type = Token::Type::kAnd;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "/")) {
type = Token::Type::kForwardSlash;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "!=")) {
type = Token::Type::kNotEqual;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "!")) {
type = Token::Type::kBang;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, ":")) {
type = Token::Type::kColon;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, ",")) {
type = Token::Type::kComma;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "==")) {
type = Token::Type::kEqualEqual;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "=")) {
type = Token::Type::kEqual;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, ">=")) {
type = Token::Type::kGreaterThanEqual;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, ">>")) {
type = Token::Type::kShiftRight;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, ">")) {
type = Token::Type::kGreaterThan;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "<=")) {
type = Token::Type::kLessThanEqual;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "<<")) {
type = Token::Type::kShiftLeft;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "<")) {
type = Token::Type::kLessThan;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "%")) {
type = Token::Type::kMod;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "->")) {
type = Token::Type::kArrow;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "--")) {
type = Token::Type::kMinusMinus;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "-")) {
type = Token::Type::kMinus;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, ".")) {
type = Token::Type::kPeriod;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "++")) {
type = Token::Type::kPlusPlus;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "+")) {
type = Token::Type::kPlus;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "||")) {
type = Token::Type::kOrOr;
pos_ += 2;
location_.column += 2;
} else if (matches(pos_, "|")) {
type = Token::Type::kOr;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, ";")) {
type = Token::Type::kSemicolon;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "*")) {
type = Token::Type::kStar;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "~")) {
type = Token::Type::kTilde;
pos_ += 1;
location_.column += 1;
} else if (matches(pos_, "^")) {
type = Token::Type::kXor;
pos_ += 1;
location_.column += 1;
}
end_source(source);
return {type, source};
}
Token Lexer::check_keyword(const Source& source, const std::string& str) {
if (str == "array")
return {Token::Type::kArray, source, "array"};
if (str == "atomic")
return {Token::Type::kAtomic, source, "atomic"};
if (str == "bitcast")
return {Token::Type::kBitcast, source, "bitcast"};
if (str == "bool")
return {Token::Type::kBool, source, "bool"};
if (str == "break")
return {Token::Type::kBreak, source, "break"};
if (str == "case")
return {Token::Type::kCase, source, "case"};
if (str == "continue")
return {Token::Type::kContinue, source, "continue"};
if (str == "continuing")
return {Token::Type::kContinuing, source, "continuing"};
if (str == "discard")
return {Token::Type::kDiscard, source, "discard"};
if (str == "default")
return {Token::Type::kDefault, source, "default"};
if (str == "else")
return {Token::Type::kElse, source, "else"};
if (str == "elseif")
return {Token::Type::kElseIf, source, "elseif"};
if (str == "f32")
return {Token::Type::kF32, source, "f32"};
if (str == "fallthrough")
return {Token::Type::kFallthrough, source, "fallthrough"};
if (str == "false")
return {Token::Type::kFalse, source, "false"};
if (str == "fn")
return {Token::Type::kFn, source, "fn"};
if (str == "for")
return {Token::Type::kFor, source, "for"};
if (str == "bgra8unorm")
return {Token::Type::kFormatBgra8Unorm, source, "bgra8unorm"};
if (str == "bgra8unorm_srgb")
return {Token::Type::kFormatBgra8UnormSrgb, source, "bgra8unorm_srgb"};
if (str == "r16float")
return {Token::Type::kFormatR16Float, source, "r16float"};
if (str == "r16sint")
return {Token::Type::kFormatR16Sint, source, "r16sint"};
if (str == "r16uint")
return {Token::Type::kFormatR16Uint, source, "r16uint"};
if (str == "r32float")
return {Token::Type::kFormatR32Float, source, "r32float"};
if (str == "r32sint")
return {Token::Type::kFormatR32Sint, source, "r32sint"};
if (str == "r32uint")
return {Token::Type::kFormatR32Uint, source, "r32uint"};
if (str == "r8sint")
return {Token::Type::kFormatR8Sint, source, "r8sint"};
if (str == "r8snorm")
return {Token::Type::kFormatR8Snorm, source, "r8snorm"};
if (str == "r8uint")
return {Token::Type::kFormatR8Uint, source, "r8uint"};
if (str == "r8unorm")
return {Token::Type::kFormatR8Unorm, source, "r8unorm"};
if (str == "rg11b10float")
return {Token::Type::kFormatRg11B10Float, source, "rg11b10float"};
if (str == "rg16float")
return {Token::Type::kFormatRg16Float, source, "rg16float"};
if (str == "rg16sint")
return {Token::Type::kFormatRg16Sint, source, "rg16sint"};
if (str == "rg16uint")
return {Token::Type::kFormatRg16Uint, source, "rg16uint"};
if (str == "rg32float")
return {Token::Type::kFormatRg32Float, source, "rg32float"};
if (str == "rg32sint")
return {Token::Type::kFormatRg32Sint, source, "rg32sint"};
if (str == "rg32uint")
return {Token::Type::kFormatRg32Uint, source, "rg32uint"};
if (str == "rg8sint")
return {Token::Type::kFormatRg8Sint, source, "rg8sint"};
if (str == "rg8snorm")
return {Token::Type::kFormatRg8Snorm, source, "rg8snorm"};
if (str == "rg8uint")
return {Token::Type::kFormatRg8Uint, source, "rg8uint"};
if (str == "rg8unorm")
return {Token::Type::kFormatRg8Unorm, source, "rg8unorm"};
if (str == "rgb10a2unorm")
return {Token::Type::kFormatRgb10A2Unorm, source, "rgb10a2unorm"};
if (str == "rgba16float")
return {Token::Type::kFormatRgba16Float, source, "rgba16float"};
if (str == "rgba16sint")
return {Token::Type::kFormatRgba16Sint, source, "rgba16sint"};
if (str == "rgba16uint")
return {Token::Type::kFormatRgba16Uint, source, "rgba16uint"};
if (str == "rgba32float")
return {Token::Type::kFormatRgba32Float, source, "rgba32float"};
if (str == "rgba32sint")
return {Token::Type::kFormatRgba32Sint, source, "rgba32sint"};
if (str == "rgba32uint")
return {Token::Type::kFormatRgba32Uint, source, "rgba32uint"};
if (str == "rgba8sint")
return {Token::Type::kFormatRgba8Sint, source, "rgba8sint"};
if (str == "rgba8snorm")
return {Token::Type::kFormatRgba8Snorm, source, "rgba8snorm"};
if (str == "rgba8uint")
return {Token::Type::kFormatRgba8Uint, source, "rgba8uint"};
if (str == "rgba8unorm")
return {Token::Type::kFormatRgba8Unorm, source, "rgba8unorm"};
if (str == "rgba8unorm_srgb")
return {Token::Type::kFormatRgba8UnormSrgb, source, "rgba8unorm_srgb"};
if (str == "function")
return {Token::Type::kFunction, source, "function"};
if (str == "i32")
return {Token::Type::kI32, source, "i32"};
if (str == "if")
return {Token::Type::kIf, source, "if"};
if (str == "image")
return {Token::Type::kImage, source, "image"};
if (str == "import")
return {Token::Type::kImport, source, "import"};
if (str == "let")
return {Token::Type::kLet, source, "let"};
if (str == "loop")
return {Token::Type::kLoop, source, "loop"};
if (str == "mat2x2")
return {Token::Type::kMat2x2, source, "mat2x2"};
if (str == "mat2x3")
return {Token::Type::kMat2x3, source, "mat2x3"};
if (str == "mat2x4")
return {Token::Type::kMat2x4, source, "mat2x4"};
if (str == "mat3x2")
return {Token::Type::kMat3x2, source, "mat3x2"};
if (str == "mat3x3")
return {Token::Type::kMat3x3, source, "mat3x3"};
if (str == "mat3x4")
return {Token::Type::kMat3x4, source, "mat3x4"};
if (str == "mat4x2")
return {Token::Type::kMat4x2, source, "mat4x2"};
if (str == "mat4x3")
return {Token::Type::kMat4x3, source, "mat4x3"};
if (str == "mat4x4")
return {Token::Type::kMat4x4, source, "mat4x4"};
if (str == "private")
return {Token::Type::kPrivate, source, "private"};
if (str == "ptr")
return {Token::Type::kPtr, source, "ptr"};
if (str == "return")
return {Token::Type::kReturn, source, "return"};
if (str == "sampler")
return {Token::Type::kSampler, source, "sampler"};
if (str == "sampler_comparison")
return {Token::Type::kComparisonSampler, source, "sampler_comparison"};
if (str == "storage_buffer" || str == "storage")
return {Token::Type::kStorage, source, "storage"};
if (str == "struct")
return {Token::Type::kStruct, source, "struct"};
if (str == "switch")
return {Token::Type::kSwitch, source, "switch"};
if (str == "texture_1d")
return {Token::Type::kTextureSampled1d, source, "texture_1d"};
if (str == "texture_2d")
return {Token::Type::kTextureSampled2d, source, "texture_2d"};
if (str == "texture_2d_array")
return {Token::Type::kTextureSampled2dArray, source, "texture_2d_array"};
if (str == "texture_3d")
return {Token::Type::kTextureSampled3d, source, "texture_3d"};
if (str == "texture_cube")
return {Token::Type::kTextureSampledCube, source, "texture_cube"};
if (str == "texture_cube_array") {
return {Token::Type::kTextureSampledCubeArray, source,
"texture_cube_array"};
}
if (str == "texture_depth_2d")
return {Token::Type::kTextureDepth2d, source, "texture_depth_2d"};
if (str == "texture_depth_2d_array") {
return {Token::Type::kTextureDepth2dArray, source,
"texture_depth_2d_array"};
}
if (str == "texture_depth_cube")
return {Token::Type::kTextureDepthCube, source, "texture_depth_cube"};
if (str == "texture_depth_cube_array") {
return {Token::Type::kTextureDepthCubeArray, source,
"texture_depth_cube_array"};
}
if (str == "texture_depth_multisampled_2d") {
return {Token::Type::kTextureDepthMultisampled2d, source,
"texture_depth_multisampled_2d"};
}
if (str == "texture_external") {
return {Token::Type::kTextureExternal, source, "texture_external"};
}
if (str == "texture_multisampled_2d") {
return {Token::Type::kTextureMultisampled2d, source,
"texture_multisampled_2d"};
}
if (str == "texture_storage_1d") {
return {Token::Type::kTextureStorage1d, source, "texture_storage_1d"};
}
if (str == "texture_storage_2d") {
return {Token::Type::kTextureStorage2d, source, "texture_storage_2d"};
}
if (str == "texture_storage_2d_array") {
return {Token::Type::kTextureStorage2dArray, source,
"texture_storage_2d_array"};
}
if (str == "texture_storage_3d") {
return {Token::Type::kTextureStorage3d, source, "texture_storage_3d"};
}
if (str == "true")
return {Token::Type::kTrue, source, "true"};
if (str == "type")
return {Token::Type::kType, source, "type"};
if (str == "u32")
return {Token::Type::kU32, source, "u32"};
if (str == "uniform")
return {Token::Type::kUniform, source, "uniform"};
if (str == "var")
return {Token::Type::kVar, source, "var"};
if (str == "vec2")
return {Token::Type::kVec2, source, "vec2"};
if (str == "vec3")
return {Token::Type::kVec3, source, "vec3"};
if (str == "vec4")
return {Token::Type::kVec4, source, "vec4"};
if (str == "workgroup")
return {Token::Type::kWorkgroup, source, "workgroup"};
return {};
}
} // namespace wgsl
} // namespace reader
} // namespace tint