Method Parse

cpp/src/arrow/util/uri.cc:247–317

const std::string& Uri::ToString() const { return impl_->string_rep_; }

Status Uri::Parse(const std::string& uri_string) {
  impl_->Reset();

  const auto& s = impl_->KeepString(uri_string);
  impl_->string_rep_ = s;
  const char* error_pos;
  int retval =
      uriParseSingleUriExA(&impl_->uri_, s.data(), s.data() + s.size(), &error_pos);
  if (retval != URI_SUCCESS) {
    if (retval == URI_ERROR_SYNTAX) {
      return Status::Invalid("Cannot parse URI: '", uri_string,
                             "' due to syntax error at character '", *error_pos,
                             "' (position ", error_pos - s.data(), ")");
    } else {
      return Status::Invalid("Cannot parse URI: '", uri_string, "'");
    }
  }

  const auto scheme = TextRangeToView(impl_->uri_.scheme);
  if (scheme.empty()) {
    return Status::Invalid("URI has empty scheme: '", uri_string, "'");
  }
  impl_->is_file_uri_ = (scheme == "file");

  // Gather path segments
  auto path_seg = impl_->uri_.pathHead;
  while (path_seg != nullptr) {
    impl_->path_segments_.push_back(TextRangeToView(path_seg->text));
    path_seg = path_seg->next;
  }

  // Decide whether URI path is absolute
  impl_->is_absolute_path_ = false;
  if (impl_->uri_.absolutePath == URI_TRUE) {
    impl_->is_absolute_path_ = true;
  } else if (has_host() && impl_->path_segments_.size() > 0) {
    // When there's a host (even empty), uriparser considers the path relative.
    // Several URI parsers for Python all consider it absolute, though.
    // For example, the path for "file:///tmp/foo" is "/tmp/foo", not "tmp/foo".
    // Similarly, the path for "file://localhost/" is "/".
    // However, the path for "file://localhost" is "".
    impl_->is_absolute_path_ = true;
  }
#ifdef _WIN32
  // There's an exception on Windows: "file:/C:foo/bar" is relative.
  if (impl_->is_file_uri_ && impl_->path_segments_.size() > 0) {
    const auto& first_seg = impl_->path_segments_[0];
    if (IsDriveSpec(first_seg) && (first_seg.length() >= 3 && first_seg[2] != '/')) {
      impl_->is_absolute_path_ = false;
    }
  }
#endif

  if (impl_->is_file_uri_ && !impl_->is_absolute_path_) {
    return Status::Invalid("File URI cannot be relative: '", uri_string, "'");
  }

  // Parse port number