avoid writing chars to the atom table in several places

This commit is contained in:
Mark Thom
2022-04-16 19:57:26 -06:00
parent b448681872
commit 2d5fe21b05
4 changed files with 31 additions and 160 deletions

View File

@@ -364,6 +364,18 @@ pub enum AtomOrString {
}
impl AtomOrString {
#[inline]
pub fn as_atom(&self, atom_tbl: &mut AtomTable) -> Atom {
match self {
&AtomOrString::Atom(atom) => {
atom
}
AtomOrString::String(string) => {
atom_tbl.build_with(&string)
}
}
}
#[inline]
pub fn as_str(&self) -> &str {
match self {

View File

@@ -737,8 +737,7 @@ impl<'a, Outputter: HCValueOutputter> HCPrinter<'a, Outputter> {
}
self.state_stack.push(TokenOrRedirect::RightCurly);
self.state_stack
.push(TokenOrRedirect::FunctorRedirect(max_depth));
self.state_stack.push(TokenOrRedirect::FunctorRedirect(max_depth));
self.state_stack.push(TokenOrRedirect::LeftCurly);
true

View File

@@ -703,7 +703,7 @@ impl MachineState {
pub(crate) fn parse_number_from_string(
&mut self,
mut string: String,
string: &str,
indices: &IndexStore,
stub_gen: impl Fn() -> FunctorStub,
) -> CallResult {
@@ -725,10 +725,13 @@ impl MachineState {
}
}
string.push('.');
let mut dot_buf: [u8; '.'.len_utf8()] = [0u8];
'.'.encode_utf8(&mut dot_buf);
let stream = CharReader::new(std::io::Cursor::new(string));
let mut parser = Parser::new(stream, self);
let cursor = std::io::Cursor::new(string);
let iter = std::io::Read::chain(cursor, std::io::Cursor::new(dot_buf));
let mut parser = Parser::new(CharReader::new(iter), self);
match parser.read_term(&CompositeOpDir::new(&indices.op_dir, None)) {
Err(err) => {
@@ -816,7 +819,7 @@ impl MachineState {
}
}
(HeapCellValueTag::Char, c) => {
Some(AtomOrString::Atom(self.atom_tbl.build_with(&c.to_string())))
Some(AtomOrString::String(c.to_string()))
}
_ => {
if value.is_constant() {
@@ -1292,20 +1295,20 @@ impl Machine {
let a2 = self.machine_st.store(self.machine_st.deref(self.machine_st.registers[2]));
if let Some(str_like) = self.machine_st.value_to_str_like(a2) {
let atom = match str_like {
let atom_cell = match str_like {
AtomOrString::Atom(atom) => {
if atom == atom!("[]") {
atom_as_cell!(if atom == atom!("[]") {
self.machine_st.atom_tbl.build_with("")
} else {
atom
}
})
}
AtomOrString::String(string) => {
self.machine_st.atom_tbl.build_with(&string)
atom_as_cell!(self.machine_st.atom_tbl.build_with(&string))
}
};
self.machine_st.bind(a1.as_var().unwrap(), atom_as_cell!(atom));
self.machine_st.bind(a1.as_var().unwrap(), atom_cell);
return;
}
@@ -1423,7 +1426,7 @@ impl Machine {
if let Some(atom_or_string) = self.machine_st.value_to_str_like(a1) {
self.machine_st.parse_number_from_string(
atom_or_string.to_string(),
atom_or_string.as_str(),
&self.indices,
stub_gen,
)?;
@@ -1887,7 +1890,7 @@ impl Machine {
}
Ok(addrs) => {
let string = self.machine_st.codes_to_string(addrs.into_iter(), stub_gen)?;
self.machine_st.parse_number_from_string(string, &self.indices, stub_gen)?;
self.machine_st.parse_number_from_string(string.as_str(), &self.indices, stub_gen)?;
}
}
@@ -3317,14 +3320,7 @@ impl Machine {
Err(e) => return Err(e)
};
if let Some(address_sink) = self.machine_st.value_to_str_like(address_sink) {
let address_string = match address_sink {
AtomOrString::Atom(atom) => {
String::from(atom.as_str())
}
AtomOrString::String(string) => {
String::from(string.as_str())
}
};
let address_string = address_sink.as_str(); //to_string();
let address: Uri = address_string.parse().unwrap();
let stream = self.runtime.block_on(async {
@@ -3408,15 +3404,8 @@ impl Machine {
let options = self.machine_st.to_stream_options(alias, eof_action, reposition, stream_type);
let src_sink = self.machine_st.store(self.machine_st.deref(self.machine_st.registers[1]));
if let Some(str_like) = self.machine_st.value_to_str_like(src_sink) {
let file_spec = match str_like {
AtomOrString::Atom(atom) => {
atom
}
AtomOrString::String(string) => {
self.machine_st.atom_tbl.build_with(&string)
}
};
if let Some(file_spec) = self.machine_st.value_to_str_like(src_sink) {
let file_spec = file_spec.as_atom(&mut self.machine_st.atom_tbl);
let mut stream = self.machine_st.stream_from_file_spec(
file_spec,

View File

@@ -246,33 +246,6 @@ impl<R: Read> CharRead for CharReader<R> {
}
}
/*
impl<R: Seek> BufReader<R> {
/// Seeks relative to the current position. If the new position lies within the buffer,
/// the buffer will not be flushed, allowing for more efficient seeks.
/// This method does not return the location of the underlying reader, so the caller
/// must track this information themselves if it is required.
#[stable(feature = "bufreader_seek_relative", since = "1.53.0")]
pub fn seek_relative(&mut self, offset: i64) -> io::Result<()> {
let pos = self.pos as u64;
if offset < 0 {
if let Some(new_pos) = pos.checked_sub((-offset) as u64) {
self.pos = new_pos as usize;
return Ok(());
}
} else {
if let Some(new_pos) = pos.checked_add(offset as u64) {
if new_pos <= self.cap as u64 {
self.pos = new_pos as usize;
return Ok(());
}
}
}
self.seek(SeekFrom::Current(offset)).map(drop)
}
}
*/
impl<R: Read> Read for CharReader<R> {
fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> {
// // If we don't have any buffered data and we're doing a massive read
@@ -377,108 +350,6 @@ where
}
}
/*
#[stable(feature = "rust1", since = "1.0.0")]
impl<R: Seek> Seek for BufReader<R> {
/// Seek to an offset, in bytes, in the underlying reader.
///
/// The position used for seeking with [`SeekFrom::Current`]`(_)` is the
/// position the underlying reader would be at if the `BufReader<R>` had no
/// internal buffer.
///
/// Seeking always discards the internal buffer, even if the seek position
/// would otherwise fall within it. This guarantees that calling
/// [`BufReader::into_inner()`] immediately after a seek yields the underlying reader
/// at the same position.
///
/// To seek without discarding the internal buffer, use [`BufReader::seek_relative`].
///
/// See [`std::io::Seek`] for more details.
///
/// Note: In the edge case where you're seeking with [`SeekFrom::Current`]`(n)`
/// where `n` minus the internal buffer length overflows an `i64`, two
/// seeks will be performed instead of one. If the second seek returns
/// [`Err`], the underlying reader will be left at the same position it would
/// have if you called `seek` with [`SeekFrom::Current`]`(0)`.
///
/// [`std::io::Seek`]: Seek
fn seek(&mut self, pos: SeekFrom) -> io::Result<u64> {
let result: u64;
if let SeekFrom::Current(n) = pos {
let remainder = (self.cap - self.pos) as i64;
// it should be safe to assume that remainder fits within an i64 as the alternative
// means we managed to allocate 8 exbibytes and that's absurd.
// But it's not out of the realm of possibility for some weird underlying reader to
// support seeking by i64::MIN so we need to handle underflow when subtracting
// remainder.
if let Some(offset) = n.checked_sub(remainder) {
result = self.inner.seek(SeekFrom::Current(offset))?;
} else {
// seek backwards by our remainder, and then by the offset
self.inner.seek(SeekFrom::Current(-remainder))?;
self.discard_buffer();
result = self.inner.seek(SeekFrom::Current(n))?;
}
} else {
// Seeking with Start/End doesn't care about our buffer length.
result = self.inner.seek(pos)?;
}
self.discard_buffer();
Ok(result)
}
/// Returns the current seek position from the start of the stream.
///
/// The value returned is equivalent to `self.seek(SeekFrom::Current(0))`
/// but does not flush the internal buffer. Due to this optimization the
/// function does not guarantee that calling `.into_inner()` immediately
/// afterwards will yield the underlying reader at the same position. Use
/// [`BufReader::seek`] instead if you require that guarantee.
///
/// # Panics
///
/// This function will panic if the position of the inner reader is smaller
/// than the amount of buffered data. That can happen if the inner reader
/// has an incorrect implementation of [`Seek::stream_position`], or if the
/// position has gone out of sync due to calling [`Seek::seek`] directly on
/// the underlying reader.
///
/// # Example
///
/// ```no_run
/// use std::{
/// io::{self, BufRead, BufReader, Seek},
/// fs::File,
/// };
///
/// fn main() -> io::Result<()> {
/// let mut f = BufReader::new(File::open("foo.txt")?);
///
/// let before = f.stream_position()?;
/// f.read_line(&mut String::new())?;
/// let after = f.stream_position()?;
///
/// println!("The first line was {} bytes long", after - before);
/// Ok(())
/// }
/// ```
fn stream_position(&mut self) -> io::Result<u64> {
let remainder = (self.cap - self.pos) as u64;
self.inner.stream_position().map(|pos| {
pos.checked_sub(remainder).expect(
"overflow when subtracting remaining buffer size from inner stream position",
)
})
}
}
*/
/*
impl<T> SizeHint for CharReader<T> {
fn lower_bound(&self) -> usize {
self.buffer().len()
}
}
*/
#[cfg(test)]
mod tests {