Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
78 changes: 62 additions & 16 deletions html5ever/src/tokenizer/mod.rs
Original file line number Diff line number Diff line change
Expand Up @@ -707,10 +707,6 @@ macro_rules! peek ( ($me:expr, $input:expr) => (
unwrap_or_return!($me.peek($input), ProcessResult::Suspend)
));

macro_rules! pop_except_from ( ($me:expr, $input:expr, $set:expr) => (
unwrap_or_return!($me.pop_except_from($input, $set), ProcessResult::Suspend)
));

macro_rules! eat ( ($me:expr, $input:expr, $pat:expr) => (
unwrap_or_return!($me.eat($input, $pat, u8::eq_ignore_ascii_case), ProcessResult::Suspend)
));
Expand Down Expand Up @@ -798,7 +794,13 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ rcdata-state
states::RawData(Rcdata) => loop {
match pop_except_from!(self, input, small_char_set!('\r' '\0' '&' '<' '\n')) {
let Some(set_result) =
self.pop_except_from(input, small_char_set!('\r' '\0' '&' '<' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\0') => {
self.bad_char_error();
self.emit_char('\u{fffd}');
Expand All @@ -812,7 +814,13 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ rawtext-state
states::RawData(Rawtext) => loop {
match pop_except_from!(self, input, small_char_set!('\r' '\0' '<' '\n')) {
let Some(set_result) =
self.pop_except_from(input, small_char_set!('\r' '\0' '<' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\0') => {
self.bad_char_error();
self.emit_char('\u{fffd}');
Expand All @@ -825,7 +833,13 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ script-data-state
states::RawData(ScriptData) => loop {
match pop_except_from!(self, input, small_char_set!('\r' '\0' '<' '\n')) {
let Some(set_result) =
self.pop_except_from(input, small_char_set!('\r' '\0' '<' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\0') => {
self.bad_char_error();
self.emit_char('\u{fffd}');
Expand All @@ -838,7 +852,13 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ script-data-escaped-state
states::RawData(ScriptDataEscaped(Escaped)) => loop {
match pop_except_from!(self, input, small_char_set!('\r' '\0' '-' '<' '\n')) {
let Some(set_result) =
self.pop_except_from(input, small_char_set!('\r' '\0' '-' '<' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\0') => {
self.bad_char_error();
self.emit_char('\u{fffd}');
Expand All @@ -857,7 +877,13 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ script-data-double-escaped-state
states::RawData(ScriptDataEscaped(DoubleEscaped)) => loop {
match pop_except_from!(self, input, small_char_set!('\r' '\0' '-' '<' '\n')) {
let Some(set_result) =
self.pop_except_from(input, small_char_set!('\r' '\0' '-' '<' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\0') => {
self.bad_char_error();
self.emit_char('\u{fffd}');
Expand All @@ -877,7 +903,12 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ plaintext-state
states::Plaintext => loop {
match pop_except_from!(self, input, small_char_set!('\r' '\0' '\n')) {
let Some(set_result) = self.pop_except_from(input, small_char_set!('\r' '\0' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\0') => {
self.bad_char_error();
self.emit_char('\u{fffd}');
Expand Down Expand Up @@ -1239,7 +1270,13 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ attribute-value-(double-quoted)-state
states::AttributeValue(DoubleQuoted) => loop {
match pop_except_from!(self, input, small_char_set!('\r' '"' '&' '\0' '\n')) {
let Some(set_result) =
self.pop_except_from(input, small_char_set!('\r' '"' '&' '\0' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('"') => go!(self: to State::AfterAttributeValueQuoted),
FromSet('&') => go!(self: consume_char_ref),
FromSet('\0') => {
Expand All @@ -1253,7 +1290,13 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ attribute-value-(single-quoted)-state
states::AttributeValue(SingleQuoted) => loop {
match pop_except_from!(self, input, small_char_set!('\r' '\'' '&' '\0' '\n')) {
let Some(set_result) =
self.pop_except_from(input, small_char_set!('\r' '\'' '&' '\0' '\n'))
else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\'') => go!(self: to State::AfterAttributeValueQuoted),
FromSet('&') => go!(self: consume_char_ref),
FromSet('\0') => {
Expand All @@ -1267,11 +1310,14 @@ impl<Sink: TokenSink> Tokenizer<Sink> {

//§ attribute-value-(unquoted)-state
states::AttributeValue(Unquoted) => loop {
match pop_except_from!(
self,
let Some(set_result) = self.pop_except_from(
input,
small_char_set!('\r' '\t' '\n' '\x0C' ' ' '&' '>' '\0')
) {
small_char_set!('\r' '\t' '\n' '\x0C' ' ' '&' '>' '\0'),
) else {
return ProcessResult::Suspend;
};

match set_result {
FromSet('\t') | FromSet('\n') | FromSet('\x0C') | FromSet(' ') => {
go!(self: to State::BeforeAttributeName)
},
Expand Down
39 changes: 28 additions & 11 deletions xml5ever/src/tokenizer/mod.rs
Original file line number Diff line number Diff line change
Expand Up @@ -627,13 +627,6 @@ macro_rules! get_char ( ($me:expr, $input:expr) => {{
character
}});

macro_rules! pop_except_from ( ($me:expr, $input:expr, $set:expr) => {{
let Some(popped_element) = $me.pop_except_from($input, $set) else {
return ProcessResult::Done;
};
popped_element
}});

macro_rules! eat ( ($me:expr, $input:expr, $pat:expr) => {{
let Some(value) = $me.eat($input, $pat) else {
return ProcessResult::Done;
Expand Down Expand Up @@ -664,7 +657,13 @@ impl<Sink: TokenSink> XmlTokenizer<Sink> {
match self.state.get() {
//§ data-state
XmlState::Data => loop {
match pop_except_from!(self, input, small_char_set!('\r' '&' '<')) {
let Some(popped_element) =
self.pop_except_from(input, small_char_set!('\r' '&' '<'))
else {
return ProcessResult::Done;
};

match popped_element {
FromSet('&') => go!(self: consume_char_ref),
FromSet('<') => go!(self: to XmlState::TagState),
FromSet(c) => go!(self: emit c),
Expand Down Expand Up @@ -925,7 +924,13 @@ impl<Sink: TokenSink> XmlTokenizer<Sink> {
},
//§ tag-attribute-value-double-quoted-state
XmlState::TagAttrValue(DoubleQuoted) => loop {
match pop_except_from!(self, input, small_char_set!('\n' '"' '&')) {
let Some(popped_element) =
self.pop_except_from(input, small_char_set!('\n' '"' '&'))
else {
return ProcessResult::Done;
};

match popped_element {
FromSet('"') => go!(self: to XmlState::TagAttrNameBefore),
FromSet('&') => go!(self: consume_char_ref '"' ),
FromSet(c) => go!(self: push_value c),
Expand All @@ -934,7 +939,13 @@ impl<Sink: TokenSink> XmlTokenizer<Sink> {
},
//§ tag-attribute-value-single-quoted-state
XmlState::TagAttrValue(SingleQuoted) => loop {
match pop_except_from!(self, input, small_char_set!('\n' '\'' '&')) {
let Some(popped_element) =
self.pop_except_from(input, small_char_set!('\n' '\'' '&'))
else {
return ProcessResult::Done;
};

match popped_element {
FromSet('\'') => go!(self: to XmlState::TagAttrNameBefore),
FromSet('&') => go!(self: consume_char_ref '\''),
FromSet(c) => go!(self: push_value c),
Expand All @@ -943,7 +954,13 @@ impl<Sink: TokenSink> XmlTokenizer<Sink> {
},
//§ tag-attribute-value-double-quoted-state
XmlState::TagAttrValue(Unquoted) => loop {
match pop_except_from!(self, input, small_char_set!('\n' '\t' ' ' '&' '>')) {
let Some(popped_element) =
self.pop_except_from(input, small_char_set!('\n' '\t' ' ' '&' '>'))
else {
return ProcessResult::Done;
};

match popped_element {
FromSet('\t') | FromSet('\n') | FromSet(' ') => {
go!(self: to XmlState::TagAttrNameBefore)
},
Expand Down
Loading