/**
 * @file url_pattern.h
 * @brief URLPattern API implementation.
 *
 * This header provides the URLPattern API as specified by the WHATWG URL
 * Pattern Standard. URLPattern allows matching URLs against patterns with
 * wildcards and named groups, similar to how regular expressions match strings.
 *
 * @see https://urlpattern.spec.whatwg.org/
 * @see https://developer.mozilla.org/en-US/docs/Web/API/URL_Pattern_API
 */
#ifndef ADA_URL_PATTERN_H
#define ADA_URL_PATTERN_H

#include "ada/implementation.h"
#include "ada/expected.h"
#include "ada/parser.h"
#include "ada/url_pattern_init.h"

#include <ostream>
#include <string>
#include <string_view>
#include <unordered_map>
#include <variant>
#include <vector>

#if ADA_TESTING
#include <iostream>
#endif  // ADA_TESTING

#if ADA_INCLUDE_URL_PATTERN
namespace ada {

enum class url_pattern_part_type : uint8_t {
  // The part represents a simple fixed text string.
  FIXED_TEXT,
  // The part represents a matching group with a custom regular expression.
  REGEXP,
  // The part represents a matching group that matches code points up to the
  // next separator code point. This is typically used for a named group like
  // ":foo" that does not have a custom regular expression.
  SEGMENT_WILDCARD,
  // The part represents a matching group that greedily matches all code points.
  // This is typically used for the "*" wildcard matching group.
  FULL_WILDCARD,
};

// Pattern type for fast-path matching optimization.
// This allows skipping expensive regex evaluation for common simple patterns.
enum class url_pattern_component_type : uint8_t {
  // Pattern is "^$" - only matches empty string
  EMPTY,
  // Pattern is "^<literal>$" - exact string match (no regex needed)
  EXACT_MATCH,
  // Pattern is "^(.*)$" - matches anything (full wildcard)
  FULL_WILDCARD,
  // Pattern requires actual regex evaluation
  REGEXP,
};

enum class url_pattern_part_modifier : uint8_t {
  // The part does not have a modifier.
  none,
  // The part has an optional modifier indicated by the U+003F (?) code point.
  optional,
  // The part has a "zero or more" modifier indicated by the U+002A (*) code
  // point.
  zero_or_more,
  // The part has a "one or more" modifier indicated by the U+002B (+) code
  // point.
  one_or_more,
};

// @see https://urlpattern.spec.whatwg.org/#part
class url_pattern_part {
 public:
  url_pattern_part(url_pattern_part_type _type, std::string&& _value,
                   url_pattern_part_modifier _modifier)
      : type(_type), value(std::move(_value)), modifier(_modifier) {}

  url_pattern_part(url_pattern_part_type _type, std::string&& _value,
                   url_pattern_part_modifier _modifier, std::string&& _name,
                   std::string&& _prefix, std::string&& _suffix)
      : type(_type),
        value(std::move(_value)),
        modifier(_modifier),
        name(std::move(_name)),
        prefix(std::move(_prefix)),
        suffix(std::move(_suffix)) {}
  // A part has an associated type, a string, which must be set upon creation.
  url_pattern_part_type type;
  // A part has an associated value, a string, which must be set upon creation.
  std::string value;
  // A part has an associated modifier a string, which must be set upon
  // creation.
  url_pattern_part_modifier modifier;
  // A part has an associated name, a string, initially the empty string.
  std::string name{};
  // A part has an associated prefix, a string, initially the empty string.
  std::string prefix{};
  // A part has an associated suffix, a string, initially the empty string.
  std::string suffix{};

  inline bool is_regexp() const noexcept;
};

// @see https://urlpattern.spec.whatwg.org/#options-header
struct url_pattern_compile_component_options {
  url_pattern_compile_component_options() = default;
  explicit url_pattern_compile_component_options(
      std::optional<char> new_delimiter = std::nullopt,
      std::optional<char> new_prefix = std::nullopt)
      : delimiter(new_delimiter), prefix(new_prefix) {}

  inline std::string_view get_delimiter() const ada_warn_unused;
  inline std::string_view get_prefix() const ada_warn_unused;

  // @see https://urlpattern.spec.whatwg.org/#options-ignore-case
  bool ignore_case = false;

  static url_pattern_compile_component_options DEFAULT;
  static url_pattern_compile_component_options HOSTNAME;
  static url_pattern_compile_component_options PATHNAME;

 private:
  // @see https://urlpattern.spec.whatwg.org/#options-delimiter-code-point
  std::optional<char> delimiter{};
  // @see https://urlpattern.spec.whatwg.org/#options-prefix-code-point
  std::optional<char> prefix{};
};

// The default options is an options struct with delimiter code point set to
// the empty string and prefix code point set to the empty string.
inline url_pattern_compile_component_options
    url_pattern_compile_component_options::DEFAULT(std::nullopt, std::nullopt);

// The hostname options is an options struct with delimiter code point set
// "." and prefix code point set to the empty string.
inline url_pattern_compile_component_options
    url_pattern_compile_component_options::HOSTNAME('.', std::nullopt);

// The pathname options is an options struct with delimiter code point set
// "/" and prefix code point set to "/".
inline url_pattern_compile_component_options
    url_pattern_compile_component_options::PATHNAME('/', '/');

// A struct providing the URLPattern matching results for a single
// URL component. The URLPatternComponentResult is only ever used
// as a member attribute of a URLPatternResult struct. The
// URLPatternComponentResult API is defined as part of the URLPattern
// specification.
struct url_pattern_component_result {
  std::string input;
  std::unordered_map<std::string, std::optional<std::string>> groups;

  bool operator==(const url_pattern_component_result&) const;

#if ADA_TESTING
  friend void PrintTo(const url_pattern_component_result& result,
                      std::ostream* os) {
    *os << "input: '" << result.input << "', group: ";
    for (const auto& group : result.groups) {
      *os << "(" << group.first << ", " << group.second.value_or("undefined")
          << ") ";
    }
  }
#endif  // ADA_TESTING
};

template <url_pattern_regex::regex_concept regex_provider>
class url_pattern_component {
 public:
  url_pattern_component() = default;

  // This function explicitly takes a std::string because it is moved.
  // To avoid unnecessary copy, move each value while calling the constructor.
  url_pattern_component(std::string&& new_pattern,
                        typename regex_provider::regex_type&& new_regexp,
                        std::vector<std::string>&& new_group_name_list,
                        bool new_has_regexp_groups,
                        url_pattern_component_type new_type,
                        std::string&& new_exact_match_value = {})
      : regexp(std::move(new_regexp)),
        pattern(std::move(new_pattern)),
        group_name_list(std::move(new_group_name_list)),
        exact_match_value(std::move(new_exact_match_value)),
        has_regexp_groups(new_has_regexp_groups),
        type(new_type) {}

  // @see https://urlpattern.spec.whatwg.org/#compile-a-component
  template <url_pattern_encoding_callback F>
  static tl::expected<url_pattern_component, errors> compile(
      std::string_view input, F& encoding_callback,
      url_pattern_compile_component_options& options);

  // @see https://urlpattern.spec.whatwg.org/#create-a-component-match-result
  url_pattern_component_result create_component_match_result(
      std::string&& input,
      std::vector<std::optional<std::string>>&& exec_result);

  // Fast path test that returns true/false without constructing result groups.
  // Uses cached pattern type to skip regex evaluation for simple patterns.
  bool fast_test(std::string_view input) const noexcept;

  // Fast path match that returns capture groups without regex for simple
  // patterns. Returns nullopt if pattern doesn't match, otherwise returns
  // capture groups.
  std::optional<std::vector<std::optional<std::string>>> fast_match(
      std::string_view input) const;

#if ADA_TESTING
  friend void PrintTo(const url_pattern_component& component,
                      std::ostream* os) {
    *os << "pattern: '" << component.pattern
        << "', has_regexp_groups: " << component.has_regexp_groups
        << "group_name_list: ";
    for (const auto& name : component.group_name_list) {
      *os << name << ", ";
    }
  }
#endif  // ADA_TESTING

  typename regex_provider::regex_type regexp{};
  std::string pattern{};
  std::vector<std::string> group_name_list{};
  // For EXACT_MATCH type: the literal string to compare against
  std::string exact_match_value{};
  bool has_regexp_groups = false;
  // Cached pattern type for fast-path optimization
  url_pattern_component_type type = url_pattern_component_type::REGEXP;
};

// A URLPattern input can be either a string or a URLPatternInit object.
// If it is a string, it must be a valid UTF-8 string.
using url_pattern_input = std::variant<std::string_view, url_pattern_init>;

// A struct providing the URLPattern matching results for all
// components of a URL. The URLPatternResult API is defined as
// part of the URLPattern specification.
struct url_pattern_result {
  std::vector<url_pattern_input> inputs;
  url_pattern_component_result protocol;
  url_pattern_component_result username;
  url_pattern_component_result password;
  url_pattern_component_result hostname;
  url_pattern_component_result port;
  url_pattern_component_result pathname;
  url_pattern_component_result search;
  url_pattern_component_result hash;
};

struct url_pattern_options {
  bool ignore_case = false;

#if ADA_TESTING
  friend void PrintTo(const url_pattern_options& options, std::ostream* os) {
    *os << "ignore_case: '" << options.ignore_case;
  }
#endif  // ADA_TESTING
};

/**
 * @brief URL pattern matching class implementing the URLPattern API.
 *
 * URLPattern provides a way to match URLs against patterns with wildcards
 * and named capture groups. It's useful for routing, URL-based dispatching,
 * and URL validation.
 *
 * Pattern syntax supports:
 * - Literal text matching
 * - Named groups: `:name` (matches up to the next separator)
 * - Wildcards: `*` (matches everything)
 * - Custom regex: `(pattern)`
 * - Optional segments: `:name?`
 * - Repeated segments: `:name+`, `:name*`
 *
 * @tparam regex_provider The regex implementation to use for pattern matching.
 *         Must satisfy the url_pattern_regex::regex_concept.
 *
 * @note All string inputs must be valid UTF-8.
 *
 * @see https://urlpattern.spec.whatwg.org/
 */
template <url_pattern_regex::regex_concept regex_provider>
class url_pattern {
 public:
  url_pattern() = default;

  /**
   * If non-null, base_url must pointer at a valid UTF-8 string.
   * @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-exec
   */
  result<std::optional<url_pattern_result>> exec(
      const url_pattern_input& input,
      const std::string_view* base_url = nullptr);

  /**
   * If non-null, base_url must pointer at a valid UTF-8 string.
   * @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-test
   */
  result<bool> test(const url_pattern_input& input,
                    const std::string_view* base_url = nullptr);

  /**
   * @see https://urlpattern.spec.whatwg.org/#url-pattern-match
   * This function expects a valid UTF-8 string if input is a string.
   */
  result<std::optional<url_pattern_result>> match(
      const url_pattern_input& input,
      const std::string_view* base_url_string = nullptr);

  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-protocol
  [[nodiscard]] std::string_view get_protocol() const ada_lifetime_bound;
  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-username
  [[nodiscard]] std::string_view get_username() const ada_lifetime_bound;
  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-password
  [[nodiscard]] std::string_view get_password() const ada_lifetime_bound;
  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-hostname
  [[nodiscard]] std::string_view get_hostname() const ada_lifetime_bound;
  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-port
  [[nodiscard]] std::string_view get_port() const ada_lifetime_bound;
  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-pathname
  [[nodiscard]] std::string_view get_pathname() const ada_lifetime_bound;
  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-search
  [[nodiscard]] std::string_view get_search() const ada_lifetime_bound;
  // @see https://urlpattern.spec.whatwg.org/#dom-urlpattern-hash
  [[nodiscard]] std::string_view get_hash() const ada_lifetime_bound;

  // If ignoreCase is true, the JavaScript regular expression created for each
  // pattern must use the `vi` flag. Otherwise, they must use the `v` flag.
  [[nodiscard]] bool ignore_case() const;

  // @see https://urlpattern.spec.whatwg.org/#url-pattern-has-regexp-groups
  [[nodiscard]] bool has_regexp_groups() const;

  // Helper to test all components at once. Returns true if all match.
  [[nodiscard]] bool test_components(
      std::string_view protocol, std::string_view username,
      std::string_view password, std::string_view hostname,
      std::string_view port, std::string_view pathname, std::string_view search,
      std::string_view hash) const;

#if ADA_TESTING
  friend void PrintTo(const url_pattern& c, std::ostream* os) {
    *os << "protocol_component: '" << c.get_protocol() << ", ";
    *os << "username_component: '" << c.get_username() << ", ";
    *os << "password_component: '" << c.get_password() << ", ";
    *os << "hostname_component: '" << c.get_hostname() << ", ";
    *os << "port_component: '" << c.get_port() << ", ";
    *os << "pathname_component: '" << c.get_pathname() << ", ";
    *os << "search_component: '" << c.get_search() << ", ";
    *os << "hash_component: '" << c.get_hash();
  }
#endif  // ADA_TESTING

  template <url_pattern_regex::regex_concept P>
  friend tl::expected<url_pattern<P>, errors> parser::parse_url_pattern_impl(
      std::variant<std::string_view, url_pattern_init>&& input,
      const std::string_view* base_url, const url_pattern_options* options);

  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> protocol_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> username_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> password_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> hostname_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> port_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> pathname_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> search_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  url_pattern_component<regex_provider> hash_component{};
  /**
   * @private
   * We can not make this private due to a LLVM bug.
   * Ref: https://github.com/ada-url/ada/pull/859
   */
  bool ignore_case_ = false;
};
}  // namespace ada
#endif  // ADA_INCLUDE_URL_PATTERN
#endif