# Changelog ## 0.5.0 – 2019-06-15 * Removed most of reexports from crate root. * Renamed `White` trait to `FromStream` and reorganized modules a little bit. * Removed `reader::BorrowedError`. The intended usage of this struct was avoid allocations on a hot path (in a case errors are expected). You can now disable rendering errors using `WithCheapError` adapter, which should cause the `reader::Error` not to allocate. * Removed `pretty_unwrap` and made the `Debug` display rendered errors by default. * Split the `Error::TooShort` to differentiate between zero input and partial input. * Implemented `FromStream` for `Option` that returns `None` on empty input. * Added `Reader::from_stdin_naive`. This helper allows to avoid explitely calling and locking `io::stdin`, but has some small disadvantages (hence “naive”). * Added rendered errors to `parse_line`. * Added `parse_stdin` helper for parsing whole stdin. * Bumped minimal Rust version to 1.20 ## 0.4.4 – 2017-11-19 * Added a type annotation in a test (needed due to rust/pull/44287) ## 0.4.3 – 2017-05-24 * Added `Skip` and `SkipAll` to handle ignored values. * Added `Zeroed` wrapper for parsing 0-terminated lists. * More consistent API for `WhiteReader`: * Added `continue_` as a synonym to `parse` (to match `continue_line`) * Added `finish` (as a `finish_line` counterpart) * Renamed `unwrap` to `into_inner` * Moved `open` from standalone function to static method. * Added `parse_file` helper. * Ensured that the crate can be pasted into a module and added a binary – now `cargo run` generates a template with `mod whiteread`. * Split crate into multiple modules. * Added separate Error types for Reader, which can `Display` the error location in a nice way. * Fix bug in reading empty vector using `Lengthed`. * More impls for tuples (up to 6) ## 0.3.0 – 2016-06-10 * Added `SplitAsciiWhitespace`, a faster version of `SplitWhitespace`. * Added `open` for opening files as a `WhiteReader`. ## 0.2.0 – 2015-06-04 * Added `Leftovers` error variant returned on excessive input. * `WhiteReader` API changes: * Rename `parse_line` to `line` and added `start_line` and `finish_line`. So now you can *start*, *continue* or *finish* parsing a line, or parse the *line* as a whole. * Renamed the `&str`-returning `get_line` to `next_line`. ## 0.1.0 – 2015-06-03 * Initial release