url.hpp

/*
 * SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION.
 * SPDX-License-Identifier: Apache-2.0
 */
#pragma once
 
#include <optional>
#include <string>
 
#include <curl/curl.h>
 
namespace kvikio::detail {
class CurlUrlHandle {
 private:
  CURLU* _handle{nullptr};
 
 public:
  CurlUrlHandle();
 
  ~CurlUrlHandle() noexcept;
 
  CurlUrlHandle(CurlUrlHandle const&)            = delete;
  CurlUrlHandle& operator=(CurlUrlHandle const&) = delete;
 
  CurlUrlHandle(CurlUrlHandle&& other) noexcept;
  CurlUrlHandle& operator=(CurlUrlHandle&& other) noexcept;
 
  CURLU* get() const;
};
 
class UrlParser {
 public:
  struct UrlComponents {
    std::optional<std::string> scheme;
 
    std::optional<std::string> host;
 
    std::optional<std::string> port;
 
    std::optional<std::string> path;
 
    std::optional<std::string> query;
 
    std::optional<std::string> fragment;
  };
 
  static UrlComponents parse(std::string const& url,
                             std::optional<unsigned int> bitmask_url_flags       = std::nullopt,
                             std::optional<unsigned int> bitmask_component_flags = std::nullopt);
 
  static std::optional<std::string> extract_component(
    CurlUrlHandle const& handle,
    CURLUPart part,
    std::optional<unsigned int> bitmask_component_flags = std::nullopt,
    std::optional<CURLUcode> allowed_err_code           = std::nullopt);
 
  static std::optional<std::string> extract_component(
    std::string const& url,
    CURLUPart part,
    std::optional<unsigned int> bitmask_url_flags       = std::nullopt,
    std::optional<unsigned int> bitmask_component_flags = std::nullopt,
    std::optional<CURLUcode> allowed_err_code           = std::nullopt);
};
 
class UrlBuilder {
 private:
  CurlUrlHandle _handle;
 
  UrlBuilder& set_component(CURLUPart part,
                            char const* value,
                            std::optional<unsigned int> flags = std::nullopt);
 
 public:
  explicit UrlBuilder();
 
  explicit UrlBuilder(std::string const& url,
                      std::optional<unsigned int> bitmask_url_flags = std::nullopt);
 
  explicit UrlBuilder(UrlParser::UrlComponents const& components,
                      std::optional<unsigned int> bitmask_url_flags = std::nullopt);
 
  UrlBuilder& set_scheme(std::optional<std::string> const& scheme);
 
  UrlBuilder& set_host(std::optional<std::string> const& host);
 
  UrlBuilder& set_port(std::optional<std::string> const& port);
 
  UrlBuilder& set_path(std::optional<std::string> const& path);
 
  UrlBuilder& set_query(std::optional<std::string> const& query);
 
  UrlBuilder& set_fragment(std::optional<std::string> const& fragment);
 
  std::string build(std::optional<unsigned int> bitmask_component_flags = std::nullopt) const;
 
  static std::string build_manually(UrlParser::UrlComponents const& components);
};
 
class UrlEncoder {
 public:
  static constexpr char aws_special_chars[] = {
    '!',    '*',    '\'',   '(',    ')',    '&',    '$',    '@',    '=',    ';',    ':',    '+',
    ' ',    ',',    '?',    '\x00', '\x01', '\x02', '\x03', '\x04', '\x05', '\x06', '\x07', '\x08',
    '\x09', '\x0A', '\x0B', '\x0C', '\x0D', '\x0E', '\x0F', '\x10', '\x11', '\x12', '\x13', '\x14',
    '\x15', '\x16', '\x17', '\x18', '\x19', '\x1A', '\x1B', '\x1C', '\x1D', '\x1E', '\x1F', '\x7F'};
 
  static std::string encode_path(std::string_view path,
                                 std::string_view chars_to_encode = std::string_view{
                                   aws_special_chars, sizeof(aws_special_chars)});
};
 
}  // namespace kvikio::detail