parse.cc

Go to the documentation of this file.

#include <fuzzer/FuzzedDataProvider.h>
 
#include <cassert>
#include <cstdio>
#include <memory>
#include <string>
 
#include "ada.cpp"
#include "ada.h"
 
bool is_valid_utf8_string(const char *buf, size_t len) {
  const uint8_t *data = reinterpret_cast<const uint8_t *>(buf);
  uint64_t pos = 0;
  uint32_t code_point = 0;
  while (pos < len) {
    uint64_t next_pos = pos + 16;
    if (next_pos <= len) {  // if it is safe to read 16 more bytes, check that
                            // they are ascii
      uint64_t v1;
      std::memcpy(&v1, data + pos, sizeof(uint64_t));
      uint64_t v2;
      std::memcpy(&v2, data + pos + sizeof(uint64_t), sizeof(uint64_t));
      uint64_t v{v1 | v2};
      if ((v & 0x8080808080808080) == 0) {
        pos = next_pos;
        continue;
      }
    }
    unsigned char byte = data[pos];
    while (byte < 0b10000000) {
      if (++pos == len) {
        return true;
      }
      byte = data[pos];
    }
 
    if ((byte & 0b11100000) == 0b11000000) {
      next_pos = pos + 2;
      if (next_pos > len) {
        return false;
      }
      if ((data[pos + 1] & 0b11000000) != 0b10000000) {
        return false;
      }
      code_point = (byte & 0b00011111) << 6 | (data[pos + 1] & 0b00111111);
      if ((code_point < 0x80) || (0x7ff < code_point)) {
        return false;
      }
    } else if ((byte & 0b11110000) == 0b11100000) {
      next_pos = pos + 3;
      if (next_pos > len) {
        return false;
      }
      if ((data[pos + 1] & 0b11000000) != 0b10000000) {
        return false;
      }
      if ((data[pos + 2] & 0b11000000) != 0b10000000) {
        return false;
      }
      code_point = (byte & 0b00001111) << 12 |
                   (data[pos + 1] & 0b00111111) << 6 |
                   (data[pos + 2] & 0b00111111);
      if ((code_point < 0x800) || (0xffff < code_point) ||
          (0xd7ff < code_point && code_point < 0xe000)) {
        return false;
      }
    } else if ((byte & 0b11111000) == 0b11110000) {  // 0b11110000
      next_pos = pos + 4;
      if (next_pos > len) {
        return false;
      }
      if ((data[pos + 1] & 0b11000000) != 0b10000000) {
        return false;
      }
      if ((data[pos + 2] & 0b11000000) != 0b10000000) {
        return false;
      }
      if ((data[pos + 3] & 0b11000000) != 0b10000000) {
        return false;
      }
      code_point =
          (byte & 0b00000111) << 18 | (data[pos + 1] & 0b00111111) << 12 |
          (data[pos + 2] & 0b00111111) << 6 | (data[pos + 3] & 0b00111111);
      if (code_point <= 0xffff || 0x10ffff < code_point) {
        return false;
      }
    } else {
      return false;
    }
    pos = next_pos;
  }
  return true;
}
 
// Exercise all getters and boolean predicates on ada::url
static void exercise_url_predicates(const ada::url &u) {
  volatile size_t length = 0;
  length += u.get_href().size();
  length += u.get_origin().size();
  length += u.get_protocol().size();
  length += u.get_username().size();
  length += u.get_password().size();
  length += u.get_host().size();
  length += u.get_hostname().size();
  length += u.get_pathname().size();
  length += u.get_search().size();
  length += u.get_hash().size();
  length += u.get_port().size();
  length += u.to_string().size();
  length += u.get_pathname_length();
  (void)u.has_valid_domain();
  (void)u.has_credentials();
  (void)u.has_empty_hostname();
  (void)u.has_hostname();
  (void)u.has_port();
  (void)u.has_hash();
  (void)u.has_search();
  (void)u.get_components();
}
 
// Exercise all getters and boolean predicates on ada::url_aggregator
static void exercise_aggregator_predicates(const ada::url_aggregator &u) {
  volatile size_t length = 0;
  length += u.get_href().size();
  length += u.get_origin().size();
  length += u.get_protocol().size();
  length += u.get_username().size();
  length += u.get_password().size();
  length += u.get_host().size();
  length += u.get_hostname().size();
  length += u.get_pathname().size();
  length += u.get_search().size();
  length += u.get_hash().size();
  length += u.get_port().size();
  length += u.to_string().size();
  length += u.get_pathname_length();
  (void)u.has_valid_domain();
  (void)u.has_credentials();
  (void)u.has_empty_hostname();
  (void)u.has_hostname();
  (void)u.has_non_empty_username();
  (void)u.has_non_empty_password();
  (void)u.has_password();
  (void)u.has_port();
  (void)u.has_hash();
  (void)u.has_search();
  (void)u.get_components();
  volatile bool is_valid = u.validate();
  (void)is_valid;
  (void)u.to_diagram();
}
 
extern "C" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) {
  FuzzedDataProvider fdp(data, size);
  std::string source = fdp.ConsumeRandomLengthString(256);
  std::string base = fdp.ConsumeRandomLengthString(256);
 
  // volatile forces the compiler to store the results without undue
  // optimizations
  volatile size_t length = 0;
 
  auto parse_url = ada::parse<ada::url>(source);
  auto parse_url_aggregator = ada::parse<ada::url_aggregator>(source);
 
  if (is_valid_utf8_string(source.data(), source.length())) {
    if (parse_url.has_value() ^ parse_url_aggregator.has_value()) {
      printf("Source used to parse: %s", source.c_str());
      abort();
    }
  }
 
  if (parse_url) {
    length += parse_url->get_href().size();
    length += parse_url->get_origin().size();
  }
 
  if (parse_url_aggregator) {
    length += parse_url_aggregator->get_href().size();
    length += parse_url_aggregator->get_origin().size();
 
    volatile bool is_parse_url_aggregator_output_valid = false;
    is_parse_url_aggregator_output_valid = parse_url_aggregator->validate();
 
    assert(parse_url->get_protocol() == parse_url_aggregator->get_protocol());
    assert(parse_url->get_href() == parse_url_aggregator->get_href());
    assert(std::string(parse_url->get_hostname()) ==
           std::string(parse_url_aggregator->get_hostname()));
    assert(std::string(parse_url->get_pathname()) ==
           std::string(parse_url_aggregator->get_pathname()));
    assert(std::string(parse_url->get_search()) ==
           std::string(parse_url_aggregator->get_search()));
    assert(std::string(parse_url->get_hash()) ==
           std::string(parse_url_aggregator->get_hash()));
    assert(std::string(parse_url->get_port()) ==
           std::string(parse_url_aggregator->get_port()));
    assert(parse_url->get_username() ==
           std::string(parse_url_aggregator->get_username()));
    assert(parse_url->get_password() ==
           std::string(parse_url_aggregator->get_password()));
    assert(std::string(parse_url->get_host()) ==
           std::string(parse_url_aggregator->get_host()));
 
    // Exercise all predicates on both types
    exercise_url_predicates(*parse_url);
    exercise_aggregator_predicates(*parse_url_aggregator);
 
    // Test set_href consistency
    parse_url->set_href(source);
    parse_url_aggregator->set_href(source);
    assert(parse_url->get_href() == parse_url_aggregator->get_href());
  }

  if (parse_url) {
    // Copy constructor
    ada::url copied_url = *parse_url;
    assert(copied_url.get_href() == parse_url->get_href());
 
    // Copy assignment
    ada::url assigned_url;
    assigned_url = *parse_url;
    assert(assigned_url.get_href() == parse_url->get_href());
 
    // Move constructor
    ada::url moved_url = std::move(copied_url);
    assert(moved_url.get_href() == parse_url->get_href());
  }
 
  if (parse_url_aggregator) {
    // Copy constructor
    ada::url_aggregator copied_agg = *parse_url_aggregator;
    assert(std::string(copied_agg.get_href()) ==
           std::string(parse_url_aggregator->get_href()));
 
    // Copy assignment
    ada::url_aggregator assigned_agg;
    assigned_agg = *parse_url_aggregator;
    assert(std::string(assigned_agg.get_href()) ==
           std::string(parse_url_aggregator->get_href()));
 
    // Move constructor
    ada::url_aggregator moved_agg = std::move(copied_agg);
    assert(std::string(moved_agg.get_href()) ==
           std::string(parse_url_aggregator->get_href()));
 
    // Move assignment
    ada::url_aggregator move_assigned_agg;
    move_assigned_agg = std::move(assigned_agg);
    assert(std::string(move_assigned_agg.get_href()) ==
           std::string(parse_url_aggregator->get_href()));
  }

  auto out_url = ada::parse<ada::url>("https://www.ada-url.com");
 
  if (out_url) {
    out_url->set_protocol(source);
    out_url->set_username(source);
    out_url->set_password(source);
    out_url->set_hostname(source);
    out_url->set_host(source);
    out_url->set_pathname(source);
    out_url->set_search(source);
    out_url->set_hash(source);
    out_url->set_port(source);
 
    // getters
    length += out_url->get_protocol().size();
    length += out_url->get_username().size();
    length += out_url->get_password().size();
    length += out_url->get_hostname().size();
    length += out_url->get_host().size();
    length += out_url->get_pathname().size();
    length += out_url->get_search().size();
    length += out_url->get_hash().size();
    length += out_url->get_origin().size();
    length += out_url->get_port().size();
    length += out_url->get_pathname_length();
 
    length += out_url->to_string().size();
 
    // boolean predicates after setters
    (void)out_url->has_valid_domain();
    (void)out_url->has_credentials();
    (void)out_url->has_empty_hostname();
    (void)out_url->has_hostname();
    (void)out_url->has_port();
    (void)out_url->has_hash();
    (void)out_url->has_search();
    (void)out_url->get_components();
  }

  auto out_aggregator =
      ada::parse<ada::url_aggregator>("https://www.ada-url.com");
 
  if (out_aggregator) {
    out_aggregator->set_protocol(source);
    out_aggregator->set_username(source);
    out_aggregator->set_password(source);
    out_aggregator->set_hostname(source);
    out_aggregator->set_host(source);
    out_aggregator->set_pathname(source);
    out_aggregator->set_search(source);
    out_aggregator->set_hash(source);
    out_aggregator->set_port(source);
 
    // getters
    length += out_aggregator->get_protocol().size();
    length += out_aggregator->get_username().size();
    length += out_aggregator->get_password().size();
    length += out_aggregator->get_hostname().size();
    length += out_aggregator->get_host().size();
    length += out_aggregator->get_pathname().size();
    length += out_aggregator->get_search().size();
    length += out_aggregator->get_hash().size();
    length += out_aggregator->get_origin().size();
    length += out_aggregator->get_port().size();
    length += out_aggregator->get_pathname_length();
 
    length += out_aggregator->to_string().size();
 
    volatile bool is_output_valid = false;
    is_output_valid = out_aggregator->validate();
 
    (void)out_aggregator->to_diagram();
 
    // boolean predicates after setters
    (void)out_aggregator->has_valid_domain();
    (void)out_aggregator->has_credentials();
    (void)out_aggregator->has_empty_hostname();
    (void)out_aggregator->has_hostname();
    (void)out_aggregator->has_non_empty_username();
    (void)out_aggregator->has_non_empty_password();
    (void)out_aggregator->has_password();
    (void)out_aggregator->has_port();
    (void)out_aggregator->has_hash();
    (void)out_aggregator->has_search();
    (void)out_aggregator->get_components();
 
    // clear methods
    out_aggregator->clear_port();
    out_aggregator->clear_search();
    out_aggregator->clear_hash();
  }

  auto base_url = ada::parse<ada::url>(base);
  auto base_agg = ada::parse<ada::url_aggregator>(base);
 
  if (base_url) {
    auto result = ada::parse<ada::url>(source, &*base_url);
    if (result) {
      length += result->get_href().size();
      length += result->get_origin().size();
      exercise_url_predicates(*result);
    }
  }
 
  if (base_agg) {
    auto result = ada::parse<ada::url_aggregator>(source, &*base_agg);
    if (result) {
      length += result->get_href().size();
      length += result->get_origin().size();
      exercise_aggregator_predicates(*result);
    }
  }
 
  // Cross-type consistency: relative URL parsing with a base should agree
  // between url and url_aggregator representations for valid UTF-8 inputs.
  if (is_valid_utf8_string(source.data(), source.length()) &&
      is_valid_utf8_string(base.data(), base.length()) && base_url &&
      base_agg) {
    auto res_url = ada::parse<ada::url>(source, &*base_url);
    auto res_agg = ada::parse<ada::url_aggregator>(source, &*base_agg);
    if (res_url.has_value() ^ res_agg.has_value()) {
      printf("Relative parse inconsistency for source=%s base=%s\n",
             source.c_str(), base.c_str());
      abort();
    }
    if (res_url && res_agg) {
      if (res_url->get_href() != std::string(res_agg->get_href())) {
        printf("Relative parse href mismatch for source=%s base=%s\n",
               source.c_str(), base.c_str());
        abort();
      }
    }
  }

  if (base_agg) {
    auto level1 = ada::parse<ada::url_aggregator>(source, &*base_agg);
    if (level1) {
      std::string input2 = fdp.ConsumeRandomLengthString(128);
      auto level2 = ada::parse<ada::url_aggregator>(input2, &*level1);
      if (level2) {
        length += level2->get_href().size();
        volatile bool v = level2->validate();
        (void)v;
      }
    }
  }

  {
    auto known_base =
        ada::parse<ada::url_aggregator>("https://example.com/a/b/c?query#hash");
    if (known_base) {
      auto result = ada::parse<ada::url_aggregator>(source, &*known_base);
      if (result) {
        length += result->get_href().size();
        exercise_aggregator_predicates(*result);
      }
    }
  }

  length += ada::href_from_file(source).size();

  bool is_valid = ada::checkers::verify_dns_length(source);
 
  (void)is_valid;
 
  return 0;
}  // extern "C" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) {