summaryrefslogtreecommitdiffstats
path: root/testing/mozbase/rust/mozprofile/src/prefreader.rs
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--testing/mozbase/rust/mozprofile/src/prefreader.rs1064
1 files changed, 1064 insertions, 0 deletions
diff --git a/testing/mozbase/rust/mozprofile/src/prefreader.rs b/testing/mozbase/rust/mozprofile/src/prefreader.rs
new file mode 100644
index 0000000000..cff3edb617
--- /dev/null
+++ b/testing/mozbase/rust/mozprofile/src/prefreader.rs
@@ -0,0 +1,1064 @@
+/* This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
+
+use crate::preferences::{Pref, PrefValue, Preferences};
+use std::borrow::Borrow;
+use std::borrow::Cow;
+use std::char;
+use std::error::Error;
+use std::fmt;
+use std::io::{self, Write};
+use std::iter::Iterator;
+use std::mem;
+use std::str;
+
+impl PrefReaderError {
+ fn new(message: String, position: Position, parent: Option<Box<dyn Error>>) -> PrefReaderError {
+ PrefReaderError {
+ message,
+ position,
+ parent,
+ }
+ }
+}
+
+impl fmt::Display for PrefReaderError {
+ fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+ write!(
+ f,
+ "{} at line {}, column {}",
+ self.message, self.position.line, self.position.column
+ )
+ }
+}
+
+impl Error for PrefReaderError {
+ fn description(&self) -> &str {
+ &self.message
+ }
+
+ fn cause(&self) -> Option<&dyn Error> {
+ self.parent.as_deref()
+ }
+}
+
+impl From<io::Error> for PrefReaderError {
+ fn from(err: io::Error) -> PrefReaderError {
+ PrefReaderError::new("IOError".into(), Position::new(), Some(err.into()))
+ }
+}
+
+#[derive(Copy, Clone, Debug, PartialEq)]
+enum TokenizerState {
+ Junk,
+ CommentStart,
+ CommentLine,
+ CommentBlock,
+ FunctionName,
+ AfterFunctionName,
+ FunctionArgs,
+ FunctionArg,
+ DoubleQuotedString,
+ SingleQuotedString,
+ Number,
+ Bool,
+ AfterFunctionArg,
+ AfterFunction,
+ Error,
+}
+
+#[derive(Copy, Clone, Debug, Default, PartialEq)]
+pub struct Position {
+ line: u32,
+ column: u32,
+}
+
+impl Position {
+ pub fn new() -> Position {
+ Position { line: 1, column: 0 }
+ }
+}
+
+#[derive(Copy, Clone, Debug, PartialEq)]
+pub enum TokenType {
+ None,
+ PrefFunction,
+ UserPrefFunction,
+ StickyPrefFunction,
+ CommentBlock,
+ CommentLine,
+ CommentBashLine,
+ Paren,
+ Semicolon,
+ Comma,
+ String,
+ Int,
+ Bool,
+ Error,
+}
+
+#[derive(Debug, PartialEq)]
+pub enum PrefToken<'a> {
+ PrefFunction(Position),
+ UserPrefFunction(Position),
+ StickyPrefFunction(Position),
+ CommentBlock(Cow<'a, str>, Position),
+ CommentLine(Cow<'a, str>, Position),
+ CommentBashLine(Cow<'a, str>, Position),
+ Paren(char, Position),
+ Semicolon(Position),
+ Comma(Position),
+ String(Cow<'a, str>, Position),
+ Int(i64, Position),
+ Bool(bool, Position),
+ Error(String, Position),
+}
+
+impl<'a> PrefToken<'a> {
+ fn position(&self) -> Position {
+ match *self {
+ PrefToken::PrefFunction(position) => position,
+ PrefToken::UserPrefFunction(position) => position,
+ PrefToken::StickyPrefFunction(position) => position,
+ PrefToken::CommentBlock(_, position) => position,
+ PrefToken::CommentLine(_, position) => position,
+ PrefToken::CommentBashLine(_, position) => position,
+ PrefToken::Paren(_, position) => position,
+ PrefToken::Semicolon(position) => position,
+ PrefToken::Comma(position) => position,
+ PrefToken::String(_, position) => position,
+ PrefToken::Int(_, position) => position,
+ PrefToken::Bool(_, position) => position,
+ PrefToken::Error(_, position) => position,
+ }
+ }
+}
+
+#[derive(Debug)]
+pub struct PrefReaderError {
+ message: String,
+ position: Position,
+ parent: Option<Box<dyn Error>>,
+}
+
+struct TokenData<'a> {
+ token_type: TokenType,
+ complete: bool,
+ position: Position,
+ data: Cow<'a, str>,
+ start_pos: usize,
+}
+
+impl<'a> TokenData<'a> {
+ fn new(token_type: TokenType, position: Position, start_pos: usize) -> TokenData<'a> {
+ TokenData {
+ token_type,
+ complete: false,
+ position,
+ data: Cow::Borrowed(""),
+ start_pos,
+ }
+ }
+
+ fn start(&mut self, tokenizer: &PrefTokenizer, token_type: TokenType) {
+ self.token_type = token_type;
+ self.position = tokenizer.position;
+ self.start_pos = tokenizer.pos;
+ }
+
+ fn end(&mut self, buf: &'a [u8], end_pos: usize) -> Result<(), PrefReaderError> {
+ self.complete = true;
+ self.add_slice_to_token(buf, end_pos)
+ }
+
+ fn add_slice_to_token(&mut self, buf: &'a [u8], end_pos: usize) -> Result<(), PrefReaderError> {
+ let data = match str::from_utf8(&buf[self.start_pos..end_pos]) {
+ Ok(x) => x,
+ Err(_) => {
+ return Err(PrefReaderError::new(
+ "Could not convert string to utf8".into(),
+ self.position,
+ None,
+ ));
+ }
+ };
+ if self.data != "" {
+ self.data.to_mut().push_str(data)
+ } else {
+ self.data = Cow::Borrowed(data)
+ };
+ Ok(())
+ }
+
+ fn push_char(&mut self, tokenizer: &PrefTokenizer, data: char) {
+ self.data.to_mut().push(data);
+ self.start_pos = tokenizer.pos + 1;
+ }
+}
+
+pub struct PrefTokenizer<'a> {
+ data: &'a [u8],
+ pos: usize,
+ cur: Option<char>,
+ position: Position,
+ state: TokenizerState,
+ next_state: Option<TokenizerState>,
+}
+
+impl<'a> PrefTokenizer<'a> {
+ pub fn new(data: &'a [u8]) -> PrefTokenizer<'a> {
+ PrefTokenizer {
+ data,
+ pos: 0,
+ cur: None,
+ position: Position::new(),
+ state: TokenizerState::Junk,
+ next_state: Some(TokenizerState::FunctionName),
+ }
+ }
+
+ fn make_token(&mut self, token_data: TokenData<'a>) -> PrefToken<'a> {
+ let buf = token_data.data;
+ let position = token_data.position;
+ // Note: the panic! here are for cases where the invalid input is regarded as
+ // a bug in the caller. In cases where `make_token` can legitimately be called
+ // with invalid data we must instead return a PrefToken::Error
+ match token_data.token_type {
+ TokenType::None => panic!("Got a token without a type"),
+ TokenType::PrefFunction => PrefToken::PrefFunction(position),
+ TokenType::UserPrefFunction => PrefToken::UserPrefFunction(position),
+ TokenType::StickyPrefFunction => PrefToken::StickyPrefFunction(position),
+ TokenType::CommentBlock => PrefToken::CommentBlock(buf, position),
+ TokenType::CommentLine => PrefToken::CommentLine(buf, position),
+ TokenType::CommentBashLine => PrefToken::CommentBashLine(buf, position),
+ TokenType::Paren => {
+ if buf.len() != 1 {
+ panic!("Expected a buffer of length one");
+ }
+ PrefToken::Paren(buf.chars().next().unwrap(), position)
+ }
+ TokenType::Semicolon => PrefToken::Semicolon(position),
+ TokenType::Comma => PrefToken::Comma(position),
+ TokenType::String => PrefToken::String(buf, position),
+ TokenType::Int => {
+ return match buf.parse::<i64>() {
+ Ok(value) => PrefToken::Int(value, position),
+ Err(_) => PrefToken::Error(format!("Expected integer, got {}", buf), position),
+ }
+ }
+ TokenType::Bool => {
+ let value = match buf.borrow() {
+ "true" => true,
+ "false" => false,
+ x => panic!("Boolean wasn't 'true' or 'false' (was {})", x),
+ };
+ PrefToken::Bool(value, position)
+ }
+ TokenType::Error => panic!("make_token can't construct errors"),
+ }
+ }
+
+ fn get_char(&mut self) -> Option<char> {
+ if self.pos + 1 >= self.data.len() {
+ self.cur = None;
+ return None;
+ };
+ if self.cur.is_some() {
+ self.pos += 1;
+ }
+ let c = self.data[self.pos] as char;
+ if self.cur == Some('\n') {
+ self.position.line += 1;
+ self.position.column = 0;
+ } else if self.cur.is_some() {
+ self.position.column += 1;
+ };
+ self.cur = Some(c);
+ self.cur
+ }
+
+ fn unget_char(&mut self) -> Option<char> {
+ if self.pos == 0 {
+ self.position.column = 0;
+ self.cur = None
+ } else {
+ self.pos -= 1;
+ let c = self.data[self.pos] as char;
+ if c == '\n' {
+ self.position.line -= 1;
+ let mut col_pos = self.pos;
+ while col_pos > 0 {
+ col_pos -= 1;
+ if self.data[col_pos] as char == '\n' {
+ break;
+ }
+ }
+ self.position.column = (self.pos - col_pos) as u32;
+ } else {
+ self.position.column -= 1;
+ }
+ self.cur = Some(c);
+ }
+ self.cur
+ }
+
+ fn is_space(c: char) -> bool {
+ matches!(c, ' ' | '\t' | '\r' | '\n')
+ }
+
+ fn skip_whitespace(&mut self) -> Option<char> {
+ while let Some(c) = self.cur {
+ if PrefTokenizer::is_space(c) {
+ self.get_char();
+ } else {
+ break;
+ };
+ }
+ self.cur
+ }
+
+ fn consume_escape(&mut self, token_data: &mut TokenData<'a>) -> Result<(), PrefReaderError> {
+ let pos = self.pos;
+ let escaped = self.read_escape()?;
+ if let Some(escape_char) = escaped {
+ token_data.add_slice_to_token(self.data, pos)?;
+ token_data.push_char(self, escape_char);
+ };
+ Ok(())
+ }
+
+ fn read_escape(&mut self) -> Result<Option<char>, PrefReaderError> {
+ let escape_char = match self.get_char() {
+ Some('u') => self.read_hex_escape(4, true)?,
+ Some('x') => self.read_hex_escape(2, true)?,
+ Some('\\') => '\\' as u32,
+ Some('"') => '"' as u32,
+ Some('\'') => '\'' as u32,
+ Some('r') => '\r' as u32,
+ Some('n') => '\n' as u32,
+ Some(_) => return Ok(None),
+ None => {
+ return Err(PrefReaderError::new(
+ "EOF in character escape".into(),
+ self.position,
+ None,
+ ))
+ }
+ };
+ Ok(Some(char::from_u32(escape_char).ok_or_else(|| {
+ PrefReaderError::new(
+ "Invalid codepoint decoded from escape".into(),
+ self.position,
+ None,
+ )
+ })?))
+ }
+
+ fn read_hex_escape(&mut self, hex_chars: isize, first: bool) -> Result<u32, PrefReaderError> {
+ let mut value = 0;
+ for _ in 0..hex_chars {
+ match self.get_char() {
+ Some(x) => {
+ value <<= 4;
+ match x {
+ '0'..='9' => value += x as u32 - '0' as u32,
+ 'a'..='f' => value += x as u32 - 'a' as u32,
+ 'A'..='F' => value += x as u32 - 'A' as u32,
+ _ => {
+ return Err(PrefReaderError::new(
+ "Unexpected character in escape".into(),
+ self.position,
+ None,
+ ))
+ }
+ }
+ }
+ None => {
+ return Err(PrefReaderError::new(
+ "Unexpected EOF in escape".into(),
+ self.position,
+ None,
+ ))
+ }
+ }
+ }
+ if first && (0xD800..=0xDBFF).contains(&value) {
+ // First part of a surrogate pair
+ if self.get_char() != Some('\\') || self.get_char() != Some('u') {
+ return Err(PrefReaderError::new(
+ "Lone high surrogate in surrogate pair".into(),
+ self.position,
+ None,
+ ));
+ }
+ self.unget_char();
+ let high_surrogate = value;
+ let low_surrogate = self.read_hex_escape(4, false)?;
+ let high_value = (high_surrogate - 0xD800) << 10;
+ let low_value = low_surrogate - 0xDC00;
+ value = high_value + low_value + 0x10000;
+ } else if first && (0xDC00..=0xDFFF).contains(&value) {
+ return Err(PrefReaderError::new(
+ "Lone low surrogate".into(),
+ self.position,
+ None,
+ ));
+ } else if !first && !(0xDC00..=0xDFFF).contains(&value) {
+ return Err(PrefReaderError::new(
+ "Invalid low surrogate in surrogate pair".into(),
+ self.position,
+ None,
+ ));
+ }
+ Ok(value)
+ }
+
+ fn get_match(&mut self, target: &str, separators: &str) -> bool {
+ let initial_pos = self.pos;
+ let mut matched = true;
+ for c in target.chars() {
+ if self.cur == Some(c) {
+ self.get_char();
+ } else {
+ matched = false;
+ break;
+ }
+ }
+
+ if !matched {
+ for _ in 0..(self.pos - initial_pos) {
+ self.unget_char();
+ }
+ } else {
+ // Check that the next character is whitespace or a separator
+ if let Some(c) = self.cur {
+ if !(PrefTokenizer::is_space(c) || separators.contains(c) || c == '/') {
+ matched = false;
+ }
+ self.unget_char();
+ }
+ // Otherwise the token was followed by EOF. That's a valid match, but
+ // will presumably cause a parse error later.
+ }
+
+ matched
+ }
+
+ fn next_token(&mut self) -> Result<Option<TokenData<'a>>, PrefReaderError> {
+ let mut token_data = TokenData::new(TokenType::None, Position::new(), 0);
+
+ loop {
+ let mut c = match self.get_char() {
+ Some(x) => x,
+ None => return Ok(None),
+ };
+
+ self.state = match self.state {
+ TokenizerState::Junk => {
+ c = match self.skip_whitespace() {
+ Some(x) => x,
+ None => return Ok(None),
+ };
+ match c {
+ '/' => TokenizerState::CommentStart,
+ '#' => {
+ token_data.start(self, TokenType::CommentBashLine);
+ token_data.start_pos = self.pos + 1;
+ TokenizerState::CommentLine
+ }
+ _ => {
+ self.unget_char();
+ let next = match self.next_state {
+ Some(x) => x,
+ None => {
+ return Err(PrefReaderError::new(
+ "In Junk state without a next state defined".into(),
+ self.position,
+ None,
+ ))
+ }
+ };
+ self.next_state = None;
+ next
+ }
+ }
+ }
+ TokenizerState::CommentStart => match c {
+ '*' => {
+ token_data.start(self, TokenType::CommentBlock);
+ token_data.start_pos = self.pos + 1;
+ TokenizerState::CommentBlock
+ }
+ '/' => {
+ token_data.start(self, TokenType::CommentLine);
+ token_data.start_pos = self.pos + 1;
+ TokenizerState::CommentLine
+ }
+ _ => {
+ return Err(PrefReaderError::new(
+ "Invalid character after /".into(),
+ self.position,
+ None,
+ ))
+ }
+ },
+ TokenizerState::CommentLine => match c {
+ '\n' => {
+ token_data.end(self.data, self.pos)?;
+ TokenizerState::Junk
+ }
+ _ => TokenizerState::CommentLine,
+ },
+ TokenizerState::CommentBlock => match c {
+ '*' => {
+ if self.get_char() == Some('/') {
+ token_data.end(self.data, self.pos - 1)?;
+ TokenizerState::Junk
+ } else {
+ TokenizerState::CommentBlock
+ }
+ }
+ _ => TokenizerState::CommentBlock,
+ },
+ TokenizerState::FunctionName => {
+ let position = self.position;
+ let start_pos = self.pos;
+ match c {
+ 'u' => {
+ if self.get_match("user_pref", "(") {
+ token_data.start(self, TokenType::UserPrefFunction);
+ }
+ }
+ 's' => {
+ if self.get_match("sticky_pref", "(") {
+ token_data.start(self, TokenType::StickyPrefFunction);
+ }
+ }
+ 'p' => {
+ if self.get_match("pref", "(") {
+ token_data.start(self, TokenType::PrefFunction);
+ }
+ }
+ _ => {}
+ };
+ if token_data.token_type == TokenType::None {
+ // We didn't match anything
+ return Err(PrefReaderError::new(
+ "Expected a pref function name".into(),
+ position,
+ None,
+ ));
+ } else {
+ token_data.start_pos = start_pos;
+ token_data.position = position;
+ token_data.end(self.data, self.pos + 1)?;
+ self.next_state = Some(TokenizerState::AfterFunctionName);
+ TokenizerState::Junk
+ }
+ }
+ TokenizerState::AfterFunctionName => match c {
+ '(' => {
+ self.next_state = Some(TokenizerState::FunctionArgs);
+ token_data.start(self, TokenType::Paren);
+ token_data.end(self.data, self.pos + 1)?;
+ self.next_state = Some(TokenizerState::FunctionArgs);
+ TokenizerState::Junk
+ }
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected an opening paren".into(),
+ self.position,
+ None,
+ ))
+ }
+ },
+ TokenizerState::FunctionArgs => match c {
+ ')' => {
+ token_data.start(self, TokenType::Paren);
+ token_data.end(self.data, self.pos + 1)?;
+ self.next_state = Some(TokenizerState::AfterFunction);
+ TokenizerState::Junk
+ }
+ _ => {
+ self.unget_char();
+ TokenizerState::FunctionArg
+ }
+ },
+ TokenizerState::FunctionArg => match c {
+ '"' => {
+ token_data.start(self, TokenType::String);
+ token_data.start_pos = self.pos + 1;
+ TokenizerState::DoubleQuotedString
+ }
+ '\'' => {
+ token_data.start(self, TokenType::String);
+ token_data.start_pos = self.pos + 1;
+ TokenizerState::SingleQuotedString
+ }
+ 't' | 'f' => {
+ self.unget_char();
+ TokenizerState::Bool
+ }
+ '0'..='9' | '-' | '+' => {
+ token_data.start(self, TokenType::Int);
+ TokenizerState::Number
+ }
+ _ => {
+ return Err(PrefReaderError::new(
+ "Invalid character at start of function argument".into(),
+ self.position,
+ None,
+ ))
+ }
+ },
+ TokenizerState::DoubleQuotedString => match c {
+ '"' => {
+ token_data.end(self.data, self.pos)?;
+ self.next_state = Some(TokenizerState::AfterFunctionArg);
+ TokenizerState::Junk
+ }
+ '\n' => {
+ return Err(PrefReaderError::new(
+ "EOL in double quoted string".into(),
+ self.position,
+ None,
+ ))
+ }
+ '\\' => {
+ self.consume_escape(&mut token_data)?;
+ TokenizerState::DoubleQuotedString
+ }
+ _ => TokenizerState::DoubleQuotedString,
+ },
+ TokenizerState::SingleQuotedString => match c {
+ '\'' => {
+ token_data.end(self.data, self.pos)?;
+ self.next_state = Some(TokenizerState::AfterFunctionArg);
+ TokenizerState::Junk
+ }
+ '\n' => {
+ return Err(PrefReaderError::new(
+ "EOL in single quoted string".into(),
+ self.position,
+ None,
+ ))
+ }
+ '\\' => {
+ self.consume_escape(&mut token_data)?;
+ TokenizerState::SingleQuotedString
+ }
+ _ => TokenizerState::SingleQuotedString,
+ },
+ TokenizerState::Number => match c {
+ '0'..='9' => TokenizerState::Number,
+ ')' | ',' => {
+ token_data.end(self.data, self.pos)?;
+ self.unget_char();
+ self.next_state = Some(TokenizerState::AfterFunctionArg);
+ TokenizerState::Junk
+ }
+ x if PrefTokenizer::is_space(x) => {
+ token_data.end(self.data, self.pos)?;
+ self.next_state = Some(TokenizerState::AfterFunctionArg);
+ TokenizerState::Junk
+ }
+ _ => {
+ return Err(PrefReaderError::new(
+ "Invalid character in number literal".into(),
+ self.position,
+ None,
+ ))
+ }
+ },
+ TokenizerState::Bool => {
+ let start_pos = self.pos;
+ let position = self.position;
+ match c {
+ 't' => {
+ if self.get_match("true", ",)") {
+ token_data.start(self, TokenType::Bool)
+ }
+ }
+ 'f' => {
+ if self.get_match("false", ",)") {
+ token_data.start(self, TokenType::Bool)
+ }
+ }
+ _ => {}
+ };
+ if token_data.token_type == TokenType::None {
+ return Err(PrefReaderError::new(
+ "Unexpected characters in function argument".into(),
+ position,
+ None,
+ ));
+ } else {
+ token_data.start_pos = start_pos;
+ token_data.position = position;
+ token_data.end(self.data, self.pos + 1)?;
+ self.next_state = Some(TokenizerState::AfterFunctionArg);
+ TokenizerState::Junk
+ }
+ }
+ TokenizerState::AfterFunctionArg => match c {
+ ',' => {
+ token_data.start(self, TokenType::Comma);
+ token_data.end(self.data, self.pos + 1)?;
+ self.next_state = Some(TokenizerState::FunctionArg);
+ TokenizerState::Junk
+ }
+ ')' => {
+ token_data.start(self, TokenType::Paren);
+ token_data.end(self.data, self.pos + 1)?;
+ self.next_state = Some(TokenizerState::AfterFunction);
+ TokenizerState::Junk
+ }
+ _ => {
+ return Err(PrefReaderError::new(
+ "Unexpected character after function argument".into(),
+ self.position,
+ None,
+ ))
+ }
+ },
+ TokenizerState::AfterFunction => match c {
+ ';' => {
+ token_data.start(self, TokenType::Semicolon);
+ token_data.end(self.data, self.pos)?;
+ self.next_state = Some(TokenizerState::FunctionName);
+ TokenizerState::Junk
+ }
+ _ => {
+ return Err(PrefReaderError::new(
+ "Unexpected character after function".into(),
+ self.position,
+ None,
+ ))
+ }
+ },
+ TokenizerState::Error => TokenizerState::Error,
+ };
+ if token_data.complete {
+ return Ok(Some(token_data));
+ }
+ }
+ }
+}
+
+impl<'a> Iterator for PrefTokenizer<'a> {
+ type Item = PrefToken<'a>;
+
+ fn next(&mut self) -> Option<PrefToken<'a>> {
+ if let TokenizerState::Error = self.state {
+ return None;
+ }
+ let token_data = match self.next_token() {
+ Err(e) => {
+ self.state = TokenizerState::Error;
+ return Some(PrefToken::Error(e.message.clone(), e.position));
+ }
+ Ok(Some(token_data)) => token_data,
+ Ok(None) => return None,
+ };
+ let token = self.make_token(token_data);
+ Some(token)
+ }
+}
+
+pub fn tokenize(data: &[u8]) -> PrefTokenizer {
+ PrefTokenizer::new(data)
+}
+
+pub fn serialize_token<T: Write>(token: &PrefToken, output: &mut T) -> Result<(), PrefReaderError> {
+ let mut data_buf = String::new();
+
+ let data = match *token {
+ PrefToken::PrefFunction(_) => "pref",
+ PrefToken::UserPrefFunction(_) => "user_pref",
+ PrefToken::StickyPrefFunction(_) => "sticky_pref",
+ PrefToken::CommentBlock(ref data, _) => {
+ data_buf.reserve(data.len() + 4);
+ data_buf.push_str("/*");
+ data_buf.push_str(data.borrow());
+ data_buf.push('*');
+ &*data_buf
+ }
+ PrefToken::CommentLine(ref data, _) => {
+ data_buf.reserve(data.len() + 2);
+ data_buf.push_str("//");
+ data_buf.push_str(data.borrow());
+ &*data_buf
+ }
+ PrefToken::CommentBashLine(ref data, _) => {
+ data_buf.reserve(data.len() + 1);
+ data_buf.push('#');
+ data_buf.push_str(data.borrow());
+ &*data_buf
+ }
+ PrefToken::Paren(data, _) => {
+ data_buf.push(data);
+ &*data_buf
+ }
+ PrefToken::Comma(_) => ",",
+ PrefToken::Semicolon(_) => ";\n",
+ PrefToken::String(ref data, _) => {
+ data_buf.reserve(data.len() + 2);
+ data_buf.push('"');
+ data_buf.push_str(escape_quote(data.borrow()).borrow());
+ data_buf.push('"');
+ &*data_buf
+ }
+ PrefToken::Int(data, _) => {
+ data_buf.push_str(&data.to_string());
+ &*data_buf
+ }
+ PrefToken::Bool(data, _) => {
+ if data {
+ "true"
+ } else {
+ "false"
+ }
+ }
+ PrefToken::Error(ref data, pos) => {
+ return Err(PrefReaderError::new(data.clone(), pos, None))
+ }
+ };
+ output.write_all(data.as_bytes())?;
+ Ok(())
+}
+
+pub fn serialize_tokens<'a, I, W>(tokens: I, output: &mut W) -> Result<(), PrefReaderError>
+where
+ I: Iterator<Item = &'a PrefToken<'a>>,
+ W: Write,
+{
+ for token in tokens {
+ serialize_token(token, output)?;
+ }
+ Ok(())
+}
+
+fn escape_quote(data: &str) -> Cow<str> {
+ // Not very efficient…
+ if data.contains('"') || data.contains('\\') {
+ Cow::Owned(data.replace('\\', r#"\\"#).replace('"', r#"\""#))
+ } else {
+ Cow::Borrowed(data)
+ }
+}
+
+#[derive(Debug, PartialEq)]
+enum ParserState {
+ Function,
+ Key,
+ Value,
+}
+
+struct PrefBuilder {
+ key: Option<String>,
+ value: Option<PrefValue>,
+ sticky: bool,
+}
+
+impl PrefBuilder {
+ fn new() -> PrefBuilder {
+ PrefBuilder {
+ key: None,
+ value: None,
+ sticky: false,
+ }
+ }
+}
+
+fn skip_comments<'a>(tokenizer: &mut PrefTokenizer<'a>) -> Option<PrefToken<'a>> {
+ loop {
+ match tokenizer.next() {
+ Some(PrefToken::CommentBashLine(_, _))
+ | Some(PrefToken::CommentBlock(_, _))
+ | Some(PrefToken::CommentLine(_, _)) => {}
+ Some(x) => return Some(x),
+ None => return None,
+ }
+ }
+}
+
+pub fn parse_tokens(tokenizer: &mut PrefTokenizer<'_>) -> Result<Preferences, PrefReaderError> {
+ let mut state = ParserState::Function;
+ let mut current_pref = PrefBuilder::new();
+ let mut rv = Preferences::new();
+
+ loop {
+ // Not just using a for loop here seems strange, but this restricts the
+ // scope of the borrow
+ let token = {
+ match tokenizer.next() {
+ Some(x) => x,
+ None => break,
+ }
+ };
+ // First deal with comments and errors
+ match token {
+ PrefToken::Error(msg, position) => {
+ return Err(PrefReaderError::new(msg, position, None));
+ }
+ PrefToken::CommentBashLine(_, _)
+ | PrefToken::CommentLine(_, _)
+ | PrefToken::CommentBlock(_, _) => continue,
+ _ => {}
+ }
+ state = match state {
+ ParserState::Function => {
+ match token {
+ PrefToken::PrefFunction(_) => {
+ current_pref.sticky = false;
+ }
+ PrefToken::UserPrefFunction(_) => {
+ current_pref.sticky = false;
+ }
+ PrefToken::StickyPrefFunction(_) => {
+ current_pref.sticky = true;
+ }
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected pref function".into(),
+ token.position(),
+ None,
+ ));
+ }
+ }
+ let next = skip_comments(tokenizer);
+ match next {
+ Some(PrefToken::Paren('(', _)) => ParserState::Key,
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected open paren".into(),
+ next.map(|x| x.position()).unwrap_or(tokenizer.position),
+ None,
+ ))
+ }
+ }
+ }
+ ParserState::Key => {
+ match token {
+ PrefToken::String(data, _) => current_pref.key = Some(data.into_owned()),
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected string".into(),
+ token.position(),
+ None,
+ ));
+ }
+ }
+ let next = skip_comments(tokenizer);
+ match next {
+ Some(PrefToken::Comma(_)) => ParserState::Value,
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected comma".into(),
+ next.map(|x| x.position()).unwrap_or(tokenizer.position),
+ None,
+ ))
+ }
+ }
+ }
+ ParserState::Value => {
+ match token {
+ PrefToken::String(data, _) => {
+ current_pref.value = Some(PrefValue::String(data.into_owned()))
+ }
+ PrefToken::Int(data, _) => current_pref.value = Some(PrefValue::Int(data)),
+ PrefToken::Bool(data, _) => current_pref.value = Some(PrefValue::Bool(data)),
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected value".into(),
+ token.position(),
+ None,
+ ))
+ }
+ }
+ let next = skip_comments(tokenizer);
+ match next {
+ Some(PrefToken::Paren(')', _)) => {}
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected close paren".into(),
+ next.map(|x| x.position()).unwrap_or(tokenizer.position),
+ None,
+ ))
+ }
+ }
+ let next = skip_comments(tokenizer);
+ match next {
+ Some(PrefToken::Semicolon(_)) | None => {}
+ _ => {
+ return Err(PrefReaderError::new(
+ "Expected semicolon".into(),
+ next.map(|x| x.position()).unwrap_or(tokenizer.position),
+ None,
+ ))
+ }
+ }
+ let key = mem::replace(&mut current_pref.key, None);
+ let value = mem::replace(&mut current_pref.value, None);
+ let pref = if current_pref.sticky {
+ Pref::new_sticky(value.unwrap())
+ } else {
+ Pref::new(value.unwrap())
+ };
+ rv.insert(key.unwrap(), pref);
+ current_pref.sticky = false;
+ ParserState::Function
+ }
+ }
+ }
+ match state {
+ ParserState::Key | ParserState::Value => {
+ return Err(PrefReaderError::new(
+ "EOF in middle of function".into(),
+ tokenizer.position,
+ None,
+ ));
+ }
+ _ => {}
+ }
+ Ok(rv)
+}
+
+pub fn serialize<W: Write>(prefs: &Preferences, output: &mut W) -> io::Result<()> {
+ let mut p: Vec<_> = prefs.iter().collect();
+ p.sort_by(|a, b| a.0.cmp(b.0));
+ for &(key, pref) in &p {
+ let func = if pref.sticky {
+ "sticky_pref("
+ } else {
+ "user_pref("
+ }
+ .as_bytes();
+ output.write_all(func)?;
+ output.write_all(b"\"")?;
+ output.write_all(escape_quote(key).as_bytes())?;
+ output.write_all(b"\"")?;
+ output.write_all(b", ")?;
+ match pref.value {
+ PrefValue::Bool(x) => {
+ output.write_all(if x { b"true" } else { b"false" })?;
+ }
+ PrefValue::Int(x) => {
+ output.write_all(x.to_string().as_bytes())?;
+ }
+ PrefValue::String(ref x) => {
+ output.write_all(b"\"")?;
+ output.write_all(escape_quote(x).as_bytes())?;
+ output.write_all(b"\"")?;
+ }
+ };
+ output.write_all(b");\n")?;
+ }
+ Ok(())
+}
+
+pub fn parse(data: &[u8]) -> Result<Preferences, PrefReaderError> {
+ let mut tokenizer = tokenize(data);
+ parse_tokens(&mut tokenizer)
+}
generated by cgit v1.2.3 (git 2.39.1) at 2024-08-23 09:36:40 +0000