lib/token: move Ref to token

Ref is a type that allows us to discriminate between different lifetimes
specifically, whether the underlying Token is borrowed from the data, or
borrowed from the scratch space.

This buys us the ability to attempt zero copy deserialization, but fall
back to copying if required.
This commit is contained in:
Paul Stemmet 2021-06-23 17:01:43 +00:00 committed by Paul Stemmet
parent 70d2f3a2e5
commit c4e514f6c2
2 changed files with 90 additions and 12 deletions

View File

@ -3,9 +3,14 @@ use crate::{
error::{ScanError, ScanResult as Result},
scalar::escape::flow_unescape,
},
token::{ScalarStyle, Token},
token::{Ref, ScalarStyle, Token},
};
/// Scans a single or double quoted (flow) scalar returning
/// a Token containing the contents, and the amount read
/// from .base. This function will attempt to borrow from
/// the underlying .base, however it may be required to copy
/// into .scratch and borrow from that lifetime.
pub(super) fn scan_flow_scalar<'b, 'c>(
base: &'b str,
scratch: &'c mut Vec<u8>,
@ -238,17 +243,6 @@ fn set_no_borrow(can_borrow: &mut bool, base: &str, buffer: &str, scratch: &mut
*can_borrow = false
}
/// This allows us to discriminate between a Token with
/// different lifetimes, specifically either a lifetime
/// 'borrow-ed from the underlying data or 'copy-ied from
/// some scratch space provided.
#[derive(Debug, PartialEq)]
pub enum Ref<'borrow, 'copy>
{
Borrow(Token<'borrow>),
Copy(Token<'copy>),
}
#[derive(Debug, PartialEq, Clone, Copy)]
enum EscapeState
{

View File

@ -49,6 +49,90 @@ pub enum Token<'a>
Scalar(Slice<'a>, ScalarStyle),
}
impl<'a> Token<'a>
{
pub fn into_owned(self) -> Token<'static>
{
match self
{
Token::StreamStart(encoding) => Token::StreamStart(encoding),
Token::StreamEnd => Token::StreamEnd,
Token::VersionDirective(major, minor) => Token::VersionDirective(major, minor),
Token::TagDirective(handle, suffix) => Token::TagDirective(
Slice::Owned(handle.into_owned()),
Slice::Owned(suffix.into_owned()),
),
Token::DocumentStart => Token::DocumentStart,
Token::DocumentEnd => Token::DocumentEnd,
Token::BlockSequenceStart => Token::FlowSequenceStart,
Token::BlockMappingStart => Token::FlowMappingStart,
Token::BlockEnd => Token::BlockEnd,
Token::FlowSequenceStart => Token::FlowSequenceStart,
Token::FlowSequenceEnd => Token::FlowSequenceEnd,
Token::FlowMappingStart => Token::FlowMappingStart,
Token::FlowMappingEnd => Token::FlowMappingEnd,
Token::BlockEntry => Token::BlockEntry,
Token::FlowEntry => Token::FlowEntry,
Token::Key => Token::Key,
Token::Value => Token::Value,
Token::Alias(alias) => Token::Alias(Slice::Owned(alias.into_owned())),
Token::Anchor(anchor) => Token::Anchor(Slice::Owned(anchor.into_owned())),
Token::Tag(handle, suffix) => Token::Tag(
Slice::Owned(handle.into_owned()),
Slice::Owned(suffix.into_owned()),
),
Token::Scalar(contents, kind) =>
{
Token::Scalar(Slice::Owned(contents.into_owned()), kind)
},
}
}
}
/// This allows us to discriminate between a Token with
/// different lifetimes, specifically either a lifetime
/// 'borrow-ed from the underlying data or 'copy-ied from
/// some scratch space provided.
#[derive(Debug, PartialEq)]
pub enum Ref<'borrow, 'copy>
{
Borrow(Token<'borrow>),
Copy(Token<'copy>),
}
impl<'b, 'c> Ref<'b, 'c>
{
/// Unifies the lifetimes of the underlying Token,
/// returning one that lives at least as long as
/// 'borrow. Note that this _will_ allocate if a copy
/// needs to be made.
pub fn into_inner(self) -> Token<'b>
{
match self
{
Self::Borrow(t) => t,
Self::Copy(t) => t.into_owned(),
}
}
/// Short hand check if the Ref contains a borrowed
/// Token
pub fn is_borrowed(&self) -> bool
{
match self
{
Self::Borrow(_) => true,
Self::Copy(_) => false,
}
}
/// Short hand check if the Ref contains a copied Token
pub fn is_copied(&self) -> bool
{
!self.is_borrowed()
}
}
#[derive(Debug, PartialEq)]
pub enum StreamEncoding
{