RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-07-30 01:37:36 +00:00
Code Issues Activity Actions

AK: Serialize URL hosts with 'concept-host-serializer'

Browse source 
In order to follow spec text to achieve this, we need to change the
underlying representation of a host in AK::URL to deserialized format.
Before this, we were parsing the host and then immediately serializing
it again.

Making that change resulted in a whole bunch of fallout.

After this change, callers can access the serialized data through
this concept-host-serializer. The functional end result of this
change is that IPv6 hosts are now correctly serialized to be
surrounded with '[' and ']'.
This commit is contained in:
Shannon Booth 2023-07-27 21:40:41 +12:00 • committed by Andreas Kling
parent 768f070b86
commit 8751be09f9
36 changed files with 175 additions and 143 deletions
Download patch file Download diff file Expand all files Collapse all files

31
AK/URL.cpp
View file

@ -101,12 +101,18 @@ void URL::set_password(DeprecatedString password, ApplyPercentEncoding apply_per
m_valid = compute_validity();
}
void URL::set_host(DeprecatedString host)
void URL::set_host(Host host)
{
m_host = move(host);
m_valid = compute_validity();
}
// https://url.spec.whatwg.org/#concept-host-serializer
ErrorOr<String> URL::serialized_host() const
{
return URLParser::serialize_host(m_host);
}
void URL::set_port(Optional<u16> port)
{
if (port == default_port_for_scheme(m_scheme)) {
@ -157,7 +163,7 @@ bool URL::cannot_have_a_username_or_password_or_port() const
{
// A URL cannot have a username/password/port if its host is null or the empty string, or its scheme is "file".
// FIXME: The spec does not mention anything to do with 'cannot be a base URL'.
return m_host.is_null() || m_host.is_empty() || m_cannot_be_a_base_url || m_scheme == "file"sv;
return m_host.has<Empty>() || m_host == String {} || m_cannot_be_a_base_url || m_scheme == "file"sv;
}
// FIXME: This is by no means complete.
@ -192,7 +198,7 @@ bool URL::compute_validity() const
}
// NOTE: A file URL's host should be the empty string for localhost, not null.
if (m_scheme == "file" && m_host.is_null())
if (m_scheme == "file" && m_host.has<Empty>())
return false;
return true;
@ -227,7 +233,7 @@ URL URL::create_with_file_scheme(DeprecatedString const& path, DeprecatedString
url.set_scheme("file");
// NOTE: If the hostname is localhost (or null, which implies localhost), it should be set to the empty string.
// This is because a file URL always needs a non-null hostname.
url.set_host(hostname.is_null() || hostname == "localhost" ? DeprecatedString::empty() : hostname);
url.set_host(hostname.is_null() || hostname == "localhost" ? String {} : String::from_deprecated_string(hostname).release_value_but_fixme_should_propagate_errors());
url.set_paths(lexical_path.parts());
if (path.ends_with('/'))
url.append_slash();
@ -243,7 +249,8 @@ URL URL::create_with_help_scheme(DeprecatedString const& path, DeprecatedString
url.set_scheme("help");
// NOTE: If the hostname is localhost (or null, which implies localhost), it should be set to the empty string.
// This is because a file URL always needs a non-null hostname.
url.set_host(hostname.is_null() || hostname == "localhost" ? DeprecatedString::empty() : hostname);
url.set_host(hostname.is_null() || hostname == "localhost" ? String {} : String::from_deprecated_string(hostname).release_value_but_fixme_should_propagate_errors());
url.set_paths(lexical_path.parts());
if (path.ends_with('/'))
url.append_slash();
@ -309,7 +316,7 @@ DeprecatedString URL::serialize(ExcludeFragment exclude_fragment) const
output.append(':');
// 2. If url’s host is non-null:
if (!m_host.is_null()) {
if (!m_host.has<Empty>()) {
// 1. Append "//" to output.
output.append("//"sv);
@ -329,7 +336,7 @@ DeprecatedString URL::serialize(ExcludeFragment exclude_fragment) const
}
// 3. Append url’s host, serialized, to output.
output.append(m_host);
output.append(serialized_host().release_value_but_fixme_should_propagate_errors());
// 4. If url’s port is non-null, append U+003A (:) followed by url’s port, serialized, to output.
if (m_port.has_value())
@ -342,7 +349,7 @@ DeprecatedString URL::serialize(ExcludeFragment exclude_fragment) const
if (cannot_be_a_base_url()) {
output.append(m_paths[0]);
} else {
if (m_host.is_null() && m_paths.size() > 1 && m_paths[0].is_empty())
if (m_host.has<Empty>() && m_paths.size() > 1 && m_paths[0].is_empty())
output.append("/."sv);
for (auto& segment : m_paths) {
output.append('/');
@ -379,9 +386,9 @@ DeprecatedString URL::serialize_for_display() const
builder.append(m_scheme);
builder.append(':');
if (!m_host.is_null()) {
if (!m_host.has<Empty>()) {
builder.append("//"sv);
builder.append(m_host);
builder.append(serialized_host().release_value_but_fixme_should_propagate_errors());
if (m_port.has_value())
builder.appendff(":{}", *m_port);
}
@ -389,7 +396,7 @@ DeprecatedString URL::serialize_for_display() const
if (cannot_be_a_base_url()) {
builder.append(m_paths[0]);
} else {
if (m_host.is_null() && m_paths.size() > 1 && m_paths[0].is_empty())
if (m_host.has<Empty>() && m_paths.size() > 1 && m_paths[0].is_empty())
builder.append("/."sv);
for (auto& segment : m_paths) {
builder.append('/');
@ -437,7 +444,7 @@ DeprecatedString URL::serialize_origin() const
StringBuilder builder;
builder.append(m_scheme);
builder.append("://"sv);
builder.append(m_host);
builder.append(serialized_host().release_value_but_fixme_should_propagate_errors());
if (m_port.has_value())
builder.appendff(":{}", *m_port);
return builder.to_deprecated_string();

7
AK/URL.h
View file

@ -79,7 +79,8 @@ public:
DeprecatedString const& scheme() const { return m_scheme; }
DeprecatedString username(ApplyPercentDecoding = ApplyPercentDecoding::Yes) const;
DeprecatedString password(ApplyPercentDecoding = ApplyPercentDecoding::Yes) const;
DeprecatedString const& host() const { return m_host; }
Host const& host() const { return m_host; }
ErrorOr<String> serialized_host() const;
DeprecatedString basename(ApplyPercentDecoding = ApplyPercentDecoding::Yes) const;
DeprecatedString query(ApplyPercentDecoding = ApplyPercentDecoding::No) const;
DeprecatedString fragment(ApplyPercentDecoding = ApplyPercentDecoding::Yes) const;
@ -101,7 +102,7 @@ public:
void set_scheme(DeprecatedString);
void set_username(DeprecatedString, ApplyPercentEncoding = ApplyPercentEncoding::Yes);
void set_password(DeprecatedString, ApplyPercentEncoding = ApplyPercentEncoding::Yes);
void set_host(DeprecatedString);
void set_host(Host);
void set_port(Optional<u16>);
void set_paths(Vector<DeprecatedString>, ApplyPercentEncoding = ApplyPercentEncoding::Yes);
void set_query(DeprecatedString, ApplyPercentEncoding = ApplyPercentEncoding::Yes);
@ -178,7 +179,7 @@ private:
DeprecatedString m_password;
// A URL’s host is null or a host. It is initially null.
DeprecatedString m_host;
Host m_host;
// A URL’s port is either null or a 16-bit unsigned integer that identifies a networking port. It is initially null.
Optional<u16> m_port;

37
AK/URLParser.cpp
View file

@ -32,7 +32,7 @@ static void report_validation_error(SourceLocation const& location = SourceLocat
dbgln_if(URL_PARSER_DEBUG, "URLParser::basic_parse: Validation error! {}", location);
}
static Optional<DeprecatedString> parse_opaque_host(StringView input)
static Optional<URL::Host> parse_opaque_host(StringView input)
{
auto forbidden_host_characters_excluding_percent = "\0\t\n\r #/:<>?@[\\]^|"sv;
for (auto character : forbidden_host_characters_excluding_percent) {
@ -43,7 +43,7 @@ static Optional<DeprecatedString> parse_opaque_host(StringView input)
}
// FIXME: If input contains a code point that is not a URL code point and not U+0025 (%), validation error.
// FIXME: If input contains a U+0025 (%) and the two code points following it are not ASCII hex digits, validation error.
return URL::percent_encode(input, URL::PercentEncodeSet::C0Control);
return String::from_deprecated_string(URL::percent_encode(input, URL::PercentEncodeSet::C0Control)).release_value_but_fixme_should_propagate_errors();
}
struct ParsedIPv4Number {
@ -549,7 +549,7 @@ static bool ends_in_a_number_checker(StringView input)
// https://url.spec.whatwg.org/#concept-host-parser
// NOTE: This is a very bare-bones implementation.
static Optional<DeprecatedString> parse_host(StringView input, bool is_not_special = false)
static Optional<URL::Host> parse_host(StringView input, bool is_not_special = false)
{
// 1. If input starts with U+005B ([), then:
if (input.starts_with('[')) {
@ -563,10 +563,7 @@ static Optional<DeprecatedString> parse_host(StringView input, bool is_not_speci
auto address = parse_ipv6_address(input.substring_view(1, input.length() - 2));
if (!address.has_value())
return {};
StringBuilder output;
serialize_ipv6_address(*address, output);
return output.to_deprecated_string();
return address.release_value();
}
// 2. If isNotSpecial is true, then return the result of opaque-host parsing input.
@ -581,12 +578,16 @@ static Optional<DeprecatedString> parse_host(StringView input, bool is_not_speci
// FIXME: 5. Let asciiDomain be the result of running domain to ASCII on domain.
// FIXME: 6. If asciiDomain is failure, then return failure.
auto& ascii_domain = domain;
auto ascii_domain_or_error = String::from_deprecated_string(domain);
if (ascii_domain_or_error.is_error())
return {};
auto ascii_domain = ascii_domain_or_error.release_value();
// 7. If asciiDomain contains a forbidden domain code point, domain-invalid-code-point validation error, return failure.
auto forbidden_host_characters = "\0\t\n\r #%/:<>?@[\\]^|"sv;
for (auto character : forbidden_host_characters) {
if (ascii_domain.view().contains(character)) {
if (ascii_domain.bytes_as_string_view().contains(character)) {
report_validation_error();
return {};
}
@ -598,11 +599,7 @@ static Optional<DeprecatedString> parse_host(StringView input, bool is_not_speci
if (!ipv4_host.has_value())
return {};
auto result = serialize_ipv4_address(*ipv4_host);
if (result.is_error())
return {};
return result.release_value().to_deprecated_string();
return ipv4_host.release_value();
}
// 9. Return asciiDomain.
@ -880,7 +877,7 @@ URL URLParser::basic_parse(StringView raw_input, Optional<URL> const& base_url,
return *url;
// 4. If url’s scheme is "file" and its host is an empty host, then return.
if (url->scheme() == "file"sv && url->host().is_empty())
if (url->scheme() == "file"sv && url->host() == String {})
return *url;
}
@ -1319,7 +1316,7 @@ URL URLParser::basic_parse(StringView raw_input, Optional<URL> const& base_url,
url->m_scheme = "file";
// 2. Set url’s host to the empty string.
url->m_host = "";
url->m_host = String {};
// 3. If c is U+002F (/) or U+005C (\), then:
if (code_point == '/' || code_point == '\\') {
@ -1422,7 +1419,7 @@ URL URLParser::basic_parse(StringView raw_input, Optional<URL> const& base_url,
// 2. Otherwise, if buffer is the empty string, then:
else if (buffer.is_empty()) {
// 1. Set url’s host to the empty string.
url->m_host = "";
url->m_host = String {};
// 2. If state override is given, then return.
if (state_override.has_value())
@ -1442,8 +1439,8 @@ URL URLParser::basic_parse(StringView raw_input, Optional<URL> const& base_url,
return {};
// 3. If host is "localhost", then set host to the empty string.
if (host.value() == "localhost")
host = "";
if (host.value().has<String>() && host.value().get<String>() == "localhost"sv)
host = String {};
// 4. Set url’s host to host.
url->m_host = host.release_value();
@ -1498,7 +1495,7 @@ URL URLParser::basic_parse(StringView raw_input, Optional<URL> const& base_url,
continue;
}
// 5. Otherwise, if state override is given and url’s host is null, append the empty string to url’s path.
else if (state_override.has_value() && url->host().is_empty()) {
else if (state_override.has_value() && url->host().has<Empty>()) {
url->append_slash();
}
break;