Examples

This package presents 2 entry points to the regex engine: Rure, an OO wrapper of the underlying Rust API, and a drop-in replacement for the stdlib re module (compile, search, match, findall, finditer, RegexObject and MatchObject).

The Rure interface exposes the “pay for what you use” API, enabling you to request the minimum computation you need: does the text match (is_match), where does it match (find, find_iter), and where are the submatches (captures, captures_iter).

The drop-in replacement should be as simple as import rure as re, and using the API as documented in the Python documentation ( https://docs.python.org/3/library/re.html , https://docs.python.org/2/library/re.html). The flags supported by re are automatically translated to those supported by rure. Note that the rure engine is more strict than re, and will reject expressions that contain unnecessary escapes, or use features not supported by the engine.

One important note regarding this shim: the Rust engine operates on byte offsets in the given search text, while Python operates on Unicode code points. The Rust engine guarantees returning offsets that correspond to valid UTF8 segments. By default, the MatchObject that is returned by this library will decode the captured text. The offsets returned by start, end, and span, however, are byte offsets and not character offsets. Using them with the string attribute is safe, so you can do:

>>> email = u"tony@tiremove_thisger.net"
>>> m = re.search(u"remove_this", email)
>>> m.string[:m.start()] + m.string[m.end():].decode('utf8')
u'tony@tiger.net'

This package also includes an is_match(pattern, string, flags=0) function (and corresponding method on RegexObject), that only returns a boolean.

Performance

It’s fast. Its core matching engine is a lazy DFA, which is what GNU grep and RE2 use. Like GNU grep, this regex engine can detect multi byte literals in the regex and will use fast literal string searching to quickly skip through the input to find possible match locations.

All memory usage is bounded and all searching takes linear time with respect to the input string.

For more details, see the PERFORMANCE guide: https://github.com/rust-lang-nursery/regex/blob/master/PERFORMANCE.md

Missing

There are a few things missing from this package that are present in the Rust API. There’s no particular (known) reason why they don’t, they just haven’t been implemented yet.

• Splitting a string by a regex.

• Replacing regex matches in a string with some other text.

Install

Binary wheels are provided for MacOS. The specific versions of the Rust compiler, rure and regex crates will be available in the changelog.

Installing from a source tarball requires manually building the Rust rure crate and pointing at the built directory. If you are wanting to take advantage of a modern CPU, it’s likely that you’ll want to build the regex crate with SSE3 and SIMD. To do so, you will need to update the regex/regex-capi/Cargo.toml to include the simd-accel feature: regex = { version = “0.2.2”, path = “..”, features=[“simd-accel”] }.

• git clone https://github.com/rust-lang-nursery/regex

• cargo build –release –manifest-path /path/to/regex/regex-capi/Cargo.toml

  • To build with SSE3: RUSTFLAGS=”-C target-feature=+ssse3” cargo build –release –features simd-accel

• RURE_DIR=/path/to/regex/regex-capi python setup.py bdist_wheel

• pip install rure –no-index -f ./dist

0.2.1 (2019-04-07)

• Update build pipeline to support multiple Python versions

• Binary wheels compiled against:

  • Rust: 1.33.0

  • regex: 1.0

  • rure (regex-capi): 0.2.1

0.2.0 (2018-03-04)

• Add support for RegexSet

0.1.2 (2017-10-09)

• First release on PyPI

• Binary wheels compiled against:

  • Rust: 1.20.0

  • regex: 0.2.2

  • rure (regex-capi): 0.2.0