- Added
MovableWriter
trait so that all writers can take things that implement that, allowing them to take FileHandles, or FileDescriptors as well as anything else that can be written to. - Fixed a bug in BufferedReader.read_until that would stop early if a line with just a newline (or whatever the "until" character was was hit)
- Update to Mojo 25.1.0
- Rename FileReader as BufferedReader
- Added buffered.BufferedWriter
- Added delimited.DelimWriter/DelimReader
- Added stats.ReservoirSampler
- Added cli.OptParser
- Added utils.ir.dump_ir from @soraros
- Removed benchmark module
- Added tests across the board
- Added docs across the board
- Added GH Pages docs with modo and help from @mlange-42 - https://extramojo.github.io/ExtraMojo/
- Many small bugfixes
- Added BufferedWriter in io.buffered
- Pinned to
max == 24.6
for CI
- Moved find_chr_next_occurance to bstr.memchr and renamed it memchr and added a memchr_wide
- This also includes some small perf improvements to memchr
- Increase file reader buffer size
- Switch to memcpy internally for BufferedReader to copy bytes from buffer to line reader
- Added context manager to BufferedReader
- Do aligned reads for find next chr and to ascii lower/upper
- Added MIT and Unlincense files.
- Fixed bug in find_next_chr that didn't use the passed in start on small inputs.
- Fixed bug in BufferedReader which wasn't deallocating its buffer on deletion.
- Improved the to_ascii_lowercase perf and added a corresponding to_ascii_uppercase.
- Added more tests for all the bstr methods to better cover the different code paths based on SIMD register sizes.
- Move BufferedReader from being backed by a tensor buffer to using a raw buffer with less copies.
- Added
bstr
package with basic support for split, find, and lowercasing of byte strings (more to come!) - Added Span api for regex's
- Deleted a lot of the Tensor related helper code