// Copyright 2026 Google LLC

//

// 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

//

//     https://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.

#ifndef THIRD_PARTY_CEL_CPP_PARSER_INTERNAL_LEXER_H_

#define THIRD_PARTY_CEL_CPP_PARSER_INTERNAL_LEXER_H_

#include <cstddef>

#include <cstdint>

#include <limits>

#include <optional>

#include <string>

#include <string_view>

#include <utility>

#include "absl/base/attributes.h"

#include "absl/base/optimization.h"

#include "absl/functional/function_ref.h"

#include "absl/log/absl_check.h"

#include "absl/strings/ascii.h"

#include "common/source.h"

namespace cel::parser_internal {

enum class TokenType {

  kError = 0,

  kEnd,

  kWhitespace,

  kComment,

  // Keywords

  kNull,

  kFalse,

  kTrue,

  kIn,

  kReservedWord,

  // Literals

  kInt,

  kUint,

  kFloat,

  kString,

  kBytes,

  // Identifiers (standard bare identifiers and backtick-quoted identifiers).

  // Note: The lexer does not validate whether a quoted/escaped identifier is

  // source-legal or permitted in its syntactic context. Because

  // non-source-legal identifiers are used internally in macros and functions,

  // the parser must strictly validate the characters inside quoted identifiers

  // and verify that they only appear where permitted (e.g., field selections

  // and struct field specifiers).

  kIdent,

  // Delimiters

  kLeftBracket,   // [

  kRightBracket,  // ]

  kLeftBrace,     // {

  kRightBrace,    // }

  kLeftParen,     // (

  kRightParen,    // )

  // Operators

  kDot,               // .

  kComma,             // ,

  kMinus,             // -

  kPlus,              // +

  kAsterisk,          // *

  kSlash,             // /

  kPercent,           // %

  kQuestion,          // ?

  kColon,             // :

  kExclamation,       // !

  kEqual,             // =

  kEqualEqual,        // ==

  kExclamationEqual,  // !=

  kLess,              // <

  kLessEqual,         // <=

  kGreater,           // >

  kGreaterEqual,      // >=

  kLogicalAnd,        // &&

  kLogicalOr,         // ||

};

ABSL_ATTRIBUTE_PURE_FUNCTION std::string_view TokenTypeToString(TokenType type);

struct Token final {

  TokenType type = TokenType::kError;

  int32_t start = 0;

  int32_t end = 0;

};

struct LexerError final {

  int32_t start = 0;

  int32_t end = 0;

  std::string message;

};

// Lexer performs fast tokenization of CEL expression source code.

//

// Responsibilities & Parser Expectations:

// This lexer is designed for speed and does not perform comprehensive semantic

// or syntax validation:

//

// 1. String and Bytes Literal Escape Sequences:

//    - For standard single- and double-quoted literals ("..." and '...'), the

//      lexer recognizes backslash ('\') only to determine whether the closing

//      delimiter ('"' or '\'') is escaped (e.g., \" and \' do not terminate the

//      literal, whereas \\" terminates it because the backslash is escaped).

//    - For triple-quoted literals ("""...""" and '''...''') and raw literals

//      (r"...", r'''...'''), backslashes and escape sequences are not processed

//      when locating the closing delimiter.

//    - The lexer does NOT validate, decode, or check the syntax of any escape

//      sequences (e.g., \n, \r, \t, \xHH, \uHHHH, \U00HHHHHH, \0, or invalid

//      escapes like \q). All characters and backslashes within the literal

//      boundaries are preserved verbatim in the token's text span.

//    - The parser/caller is strictly responsible for validating and unescaping

//      all escape sequences and reporting syntax errors for invalid escape

//      sequences when converting string and bytes tokens during AST

//      construction.

//

// 2. Numeric Literals:

//    - Performs only general bounds and format matching for integers and

//      floating-point numeric literals. The lexer expects the parser to perform

//      final validation and numeric conversion when building the AST.

class Lexer final {

 public:

  explicit Lexer(const cel::Source& source)

      : content_(source.content()), position_(0) {

    ABSL_DCHECK_LE(content_.size(), static_cast<SourcePosition>(

                                        std::numeric_limits<int32_t>::max()));

  }

  struct Position final {

    int32_t position = 0;

    bool at_end = false;

    bool done = false;

    LexerError error;

  };

  Lexer(const Lexer&) = delete;

  Lexer(Lexer&&) = delete;

  Lexer& operator=(const Lexer&) = delete;

  Lexer& operator=(Lexer&&) = delete;

  // Scans and returns the next token from the source.

  [[nodiscard]] ABSL_ATTRIBUTE_NOINLINE Token Lex();

  // Inspect the error from the last call to `Lex()` that returned an error

  // token. The reference is not guaranteed to be valid after further calls to

  // `Lex`.

  [[nodiscard]] const LexerError& GetError() const

      ABSL_ATTRIBUTE_LIFETIME_BOUND {

    return error_;

  }

  [[nodiscard]] int32_t GetPosition() const { return position_; }

  [[nodiscard]] Position SavePosition() const {

    return Position{position_, at_end_, done_, error_};

  }

  void RestorePosition(const Position& position) {

    position_ = position.position;

    at_end_ = position.at_end;

    done_ = position.done;

    error_ = position.error;

  }

 private:

  [[nodiscard]] bool Match(char32_t c) const {

    return position_ < content_.size() && content_.at(position_) == c;

  }

  [[nodiscard]] bool MatchIgnoreCase(char32_t c) const {

    if (position_ >= content_.size()) return false;

    char32_t cp = content_.at(position_);

    return cp <= 0x7f && c <= 0x7f &&

           absl::ascii_tolower(static_cast<char>(cp)) ==

               absl::ascii_tolower(static_cast<char>(c));

  }

  void Advance(size_t n) {

    ABSL_DCHECK_LE(n, static_cast<size_t>(content_.size() - position_));

    position_ += static_cast<int32_t>(n);

  }

  void AdvanceProcessingNewLines(int32_t end_position) {

    ABSL_DCHECK_LE(end_position, content_.size());

    ABSL_DCHECK_GE(end_position, position_);

    Advance(static_cast<size_t>(end_position - position_));

  }

  [[nodiscard]] Token MakeToken(TokenType type, int32_t start, int32_t end) {

    if (ABSL_PREDICT_FALSE(at_end_)) {

      AtEndTokenCreated();

    }

    return Token{.type = type, .start = start, .end = end};

  }

  [[nodiscard]] Token SetError(int32_t start, int32_t end,

                               std::string message) {

    error_ =

        LexerError{.start = start, .end = end, .message = std::move(message)};

    return Token{.type = TokenType::kError, .start = start, .end = end};

  }

  void AtEndTokenCreated() { done_ = true; }

  // Consumes characters up to and including the first occurrence of character

  // `c` without interpreting backslashes as escapes. Returns true if `c` was

  // found and consumed; false if end of input was reached.

  [[nodiscard]] bool ConsumeUntilAfter(char32_t c);

  // Consumes characters up to and including the first occurrence of substring

  // `s` without interpreting backslashes as escapes (`s` must not contain

  // newlines). Returns true if `s` was found and consumed; false if end of

  // input was reached.

  [[nodiscard]] bool ConsumeUntilAfterString(std::u32string_view s);

  // Consumes characters up to and including the first occurrence of `c` that is

  // not preceded by an odd number of backslash ('\') escape characters. Returns

  // true if an unescaped `c` was found and consumed; false if reached EOF.

  [[nodiscard]] bool ConsumeUntilAfterUnescaped(char32_t c);

  // Consumes characters up to and including the first occurrence of substring

  // `s` where the first character of `s` is not preceded by an odd number of

  // backslashes. Returns true if an unescaped `s` was found and consumed; false

  // if reached EOF.

  [[nodiscard]] bool ConsumeUntilAfterUnescapedString(std::u32string_view s);

  [[nodiscard]] bool MatchString(std::u32string_view s) const;

  [[nodiscard]] std::optional<char32_t> MatchIf(

      absl::FunctionRef<bool(char32_t)> predicate) const;

  void ConsumeLine();

  void ConsumeWhitespace();

  [[nodiscard]] bool Consume(char32_t c);

  [[nodiscard]] bool ConsumeIgnoreCase(char32_t c);

  [[nodiscard]] bool ConsumeString(std::u32string_view s);

  [[nodiscard]] std::optional<char32_t> ConsumeIf(

      absl::FunctionRef<bool(char32_t)> predicate);

  [[nodiscard]] bool ConsumeDigits();

  [[nodiscard]] bool ConsumeHexDigits();

  [[nodiscard]] TokenType ConsumeIntegralSuffix();

  // Consumes a backtick-quoted identifier (`...`) and returns

  // TokenType::kIdent. The token text preserves the surrounding backticks so

  // the parser can detect quoted identifiers and enforce restrictions on their

  // characters and allowed syntactic locations (such as field selections and

  // struct field specifiers).

  [[nodiscard]] Token ConsumeQuotedIdent();

  [[nodiscard]] Token ConsumeStringLiteral(int32_t start, char32_t quote,

                                           bool is_bytes = false,

                                           bool is_raw = false);

  // Consumes prefixed string and bytes literals.

  // Handles the following prefix sequences (case-insensitive for 'r' and 'b'):

  // - Raw strings: r"...", r'...', r"""...""", r'''...'''

  // - Bytes: b"...", b'...', b"""...""", b'''...'''

  // - Raw bytes: br"...", br'...', br"""...""", br'''...''', rb"...", rb'...',

  // rb"""...""", rb'''...'''

  [[nodiscard]] std::optional<Token> ConsumePrefixedStringLiteral();

  // Consumes a numeric literal token and returns its TokenType (kInt, kUint, or

  // kFloat). Recognizes the following literal formats:

  // - Decimal integers (kInt / kUint): 0, 45U, 123456

  // - Hexadecimal integers (kInt / kUint): 0x1A, 0XFFu, 0x0U

  // - Floating-point numbers (kFloat): .12345, 1.23, 1e6, 1.5e+10, .5e-3

  [[nodiscard]] Token ConsumeNumericLiteral();

  // Consumes an identifier token and checks if it matches any reserved

  // keywords.

  [[nodiscard]] Token ConsumeIdent();

  cel::SourceContentView content_;

  int32_t position_ = 0;

  bool at_end_ = false;

  bool done_ = false;

  LexerError error_;

};

}  // namespace cel::parser_internal

#endif  // THIRD_PARTY_CEL_CPP_PARSER_INTERNAL_LEXER_H_