Files
uv/Cargo.toml
T
Ibraheem Ahmed ffd18cc75d Implement marker trees using algebraic decision diagrams (#5898)
## Summary

This PR rewrites the `MarkerTree` type to use algebraic decision
diagrams (ADD). This has many benefits:
- The diagram is canonical for a given marker function. It is impossible
to create two functionally equivalent marker trees that don't refer to
the same underlying ADD. This also means that any trivially true or
unsatisfiable markers are represented by the same constants.
- The diagram can handle complex operations (conjunction/disjunction) in
polynomial time, as well as constant-time negation.
- The diagram can be converted to a simplified DNF form for user-facing
output.

The new representation gives us a lot more confidence in our marker
operations and simplification, which is proving to be very important
(see https://github.com/astral-sh/uv/pull/5733 and
https://github.com/astral-sh/uv/pull/5163).

Unfortunately, it is not easy to split this PR into multiple commits
because it is a large rewrite of the `marker` module. I'd suggest
reading through the `marker/algebra.rs`, `marker/simplify.rs`, and
`marker/tree.rs` files for the new implementation, as well as the
updated snapshots to verify how the new simplification rules work in
practice. However, a few other things were changed:
- [We now use release-only comparisons for `python_full_version`, where
we previously only did for
`python_version`](https://github.com/astral-sh/uv/blob/ibraheem/canonical-markers/crates/pep508-rs/src/marker/algebra.rs#L522).
I'm unsure how marker operations should work in the presence of
pre-release versions if we decide that this is incorrect.
- [Meaningless marker expressions are now
ignored](https://github.com/astral-sh/uv/blob/ibraheem/canonical-markers/crates/pep508-rs/src/marker/parse.rs#L502).
This means that a marker such as `'x' == 'x'` will always evaluate to
`true` (as if the expression did not exist), whereas we previously
treated this as always `false`. It's negation however, remains `false`.
- [Unsatisfiable markers are written as `python_version <
'0'`](https://github.com/astral-sh/uv/blob/ibraheem/canonical-markers/crates/pep508-rs/src/marker/tree.rs#L1329).
- The `PubGrubSpecifier` type has been moved to the new `uv-pubgrub`
crate, shared by `pep508-rs` and `uv-resolver`. `pep508-rs` also depends
on the `pubgrub` crate for the `Range` type, we probably want to move
`pubgrub::Range` into a separate crate to break this, but I don't think
that should block this PR (cc @konstin).

There is still some remaining work here that I decided to leave for now
for the sake of unblocking some of the related work on the resolver.
- We still use `Option<MarkerTree>` throughout uv, which is unnecessary
now that `MarkerTree::TRUE` is canonical.
- The `MarkerTree` type is now interned globally and can potentially
implement `Copy`. However, it's unclear if we want to add more
information to marker trees that would make it `!Copy`. For example, we
may wish to attach extra and requires-python environment information to
avoid simplifying after construction.
- We don't currently combine `python_full_version` and `python_version`
markers.
- I also have not spent too much time investigating performance and
there is probably some low-hanging fruit. Many of the test cases I did
run actually saw large performance improvements due to the markers being
simplified internally, reducing the stress on the old `normalize`
routine, especially for the extremely large markers seen in
`transformers` and other projects.

Resolves https://github.com/astral-sh/uv/issues/5660,
https://github.com/astral-sh/uv/issues/5179.
2024-08-09 13:40:02 -04:00

317 lines
12 KiB
TOML

[workspace]
members = ["crates/*"]
exclude = [
"scripts",
# Needs nightly
"crates/uv-trampoline",
]
resolver = "2"
[workspace.package]
edition = "2021"
rust-version = "1.80"
homepage = "https://pypi.org/project/uv/"
documentation = "https://pypi.org/project/uv/"
repository = "https://github.com/astral-sh/uv"
authors = ["uv"]
license = "MIT OR Apache-2.0"
[workspace.dependencies]
cache-key = { path = "crates/cache-key" }
distribution-filename = { path = "crates/distribution-filename" }
distribution-types = { path = "crates/distribution-types" }
install-wheel-rs = { path = "crates/install-wheel-rs", default-features = false }
once-map = { path = "crates/once-map" }
pep440_rs = { path = "crates/pep440-rs" }
pep508_rs = { path = "crates/pep508-rs", features = ["non-pep508-extensions"] }
platform-tags = { path = "crates/platform-tags" }
pypi-types = { path = "crates/pypi-types" }
requirements-txt = { path = "crates/requirements-txt" }
uv-auth = { path = "crates/uv-auth" }
uv-build = { path = "crates/uv-build" }
uv-cache = { path = "crates/uv-cache" }
uv-cli = { path = "crates/uv-cli" }
uv-client = { path = "crates/uv-client" }
uv-configuration = { path = "crates/uv-configuration" }
uv-dispatch = { path = "crates/uv-dispatch" }
uv-distribution = { path = "crates/uv-distribution" }
uv-extract = { path = "crates/uv-extract" }
uv-fs = { path = "crates/uv-fs" }
uv-git = { path = "crates/uv-git" }
uv-installer = { path = "crates/uv-installer" }
uv-macros = { path = "crates/uv-macros" }
uv-normalize = { path = "crates/uv-normalize" }
uv-options-metadata = { path = "crates/uv-options-metadata" }
uv-python = { path = "crates/uv-python" }
uv-requirements = { path = "crates/uv-requirements" }
uv-resolver = { path = "crates/uv-resolver" }
uv-pubgrub = { path = "crates/uv-pubgrub" }
uv-scripts = { path = "crates/uv-scripts" }
uv-settings = { path = "crates/uv-settings" }
uv-shell = { path = "crates/uv-shell" }
uv-state = { path = "crates/uv-state" }
uv-tool = { path = "crates/uv-tool" }
uv-types = { path = "crates/uv-types" }
uv-version = { path = "crates/uv-version" }
uv-virtualenv = { path = "crates/uv-virtualenv" }
uv-warnings = { path = "crates/uv-warnings" }
uv-workspace = { path = "crates/uv-workspace" }
anstream = { version = "0.6.13" }
anyhow = { version = "1.0.80" }
async-channel = { version = "2.2.0" }
async-compression = { version = "0.4.6" }
async-trait = { version = "0.1.78" }
async_http_range_reader = { version = "0.8.0" }
async_zip = { git = "https://github.com/charliermarsh/rs-async-zip", rev = "011b24604fa7bc223daaad7712c0694bac8f0a87", features = ["deflate"] }
axoupdater = { version = "0.7.0", default-features = false }
backoff = { version = "0.4.0" }
base64 = { version = "0.22.0" }
boxcar = { version = "0.2.5" }
cachedir = { version = "0.3.1" }
cargo-util = { version = "0.2.8" }
chrono = { version = "0.4.31" }
clap = { version = "4.5.9" }
clap_complete_command = { version = "0.6.0" }
configparser = { version = "3.0.4" }
console = { version = "0.15.8", default-features = false }
csv = { version = "1.3.0" }
ctrlc = { version = "3.4.4" }
dashmap = { version = "6.0.0" }
data-encoding = { version = "2.5.0" }
derivative = { version = "2.2.0" }
directories = { version = "5.0.1" }
dirs-sys = { version = "0.4.1" }
dunce = { version = "1.0.4" }
either = { version = "1.12.0" }
encoding_rs_io = { version = "0.1.7" }
flate2 = { version = "1.0.28", default-features = false }
fs-err = { version = "2.11.0" }
fs2 = { version = "0.4.3" }
futures = { version = "0.3.30" }
glob = { version = "0.3.1" }
hex = { version = "0.4.3" }
home = { version = "0.5.9" }
html-escape = { version = "0.2.13" }
http = { version = "1.1.0" }
indexmap = { version = "2.2.5" }
indicatif = { version = "0.17.7" }
indoc = { version = "2.0.4" }
itertools = { version = "0.13.0" }
junction = { version = "1.0.0" }
mailparse = { version = "0.15.0" }
md-5 = { version = "0.10.6" }
memchr = { version = "2.7.4" }
miette = { version = "7.2.0" }
nanoid = { version = "0.4.0" }
owo-colors = { version = "4.0.0" }
path-absolutize = { version = "3.1.1" }
path-slash = { version = "0.2.1" }
pathdiff = { version = "0.2.1" }
petgraph = { version = "0.6.4" }
platform-info = { version = "2.0.2" }
proc-macro2 = { version = "1.0.86" }
pubgrub = { git = "https://github.com/astral-sh/pubgrub", rev = "2fac39371a47e7cb821e510aaa4de25405413d29" }
pyo3 = { version = "0.21.0" }
pyo3-log = { version = "0.10.0" }
quote = { version = "1.0.36" }
rayon = { version = "1.8.0" }
reflink-copy = { version = "0.1.15" }
regex = { version = "1.10.2" }
reqwest = { version = "0.12.3", default-features = false, features = ["json", "gzip", "brotli", "stream", "rustls-tls", "rustls-tls-native-roots"] }
reqwest-middleware = { git = "https://github.com/astral-sh/reqwest-middleware", rev = "21ceec9a5fd2e8d6f71c3ea2999078fecbd13cbe" }
reqwest-retry = { git = "https://github.com/astral-sh/reqwest-middleware", rev = "21ceec9a5fd2e8d6f71c3ea2999078fecbd13cbe" }
rkyv = { version = "0.7.43", features = ["strict", "validation"] }
rmp-serde = { version = "1.1.2" }
rust-netrc = { version = "0.1.1" }
rustc-hash = { version = "2.0.0" }
rustix = { version = "0.38.34", default-features = false, features = ["fs", "std"] }
same-file = { version = "1.0.6" }
schemars = { version = "0.8.16", features = ["url"] }
seahash = { version = "4.1.0" }
serde = { version = "1.0.197", features = ["derive"] }
serde_json = { version = "1.0.114" }
sha2 = { version = "0.10.8" }
smallvec = { version = "1.13.2" }
syn = { version = "2.0.66" }
sys-info = { version = "0.9.1" }
target-lexicon = {version = "0.12.14" }
tempfile = { version = "3.9.0" }
textwrap = { version = "0.16.1" }
thiserror = { version = "1.0.56" }
tl = { version = "0.7.7" }
tokio = { version = "1.35.1", features = ["fs", "io-util", "macros", "process", "signal", "sync"] }
tokio-stream = { version = "0.1.14" }
tokio-tar = { version = "0.3.1" }
tokio-util = { version = "0.7.10", features = ["compat"] }
toml = { version = "0.8.12" }
toml_edit = { version = "0.22.13" }
tracing = { version = "0.1.40" }
tracing-durations-export = { version = "0.2.0", features = ["plot"] }
tracing-subscriber = { version = "0.3.18", features = ["env-filter", "json", "registry"] }
tracing-tree = { version = "0.4.0" }
unicode-width = { version = "0.1.11" }
unscanny = { version = "0.1.0" }
url = { version = "2.5.0" }
urlencoding = { version = "2.1.3" }
walkdir = { version = "2.5.0" }
which = { version = "6.0.0", features = ["regex"] }
windows-sys = { version = "0.52.0", features = ["Win32_Foundation", "Win32_Security", "Win32_Storage_FileSystem", "Win32_System_Ioctl", "Win32_System_IO"] }
winreg = { version = "0.52.0" }
winsafe = { version = "0.0.22", features = ["kernel"] }
wiremock = { version = "0.6.0" }
xz2 = { version = "0.1.7" }
zip = { version = "0.6.6", default-features = false, features = ["deflate"] }
[workspace.metadata.cargo-shear]
ignored = ["flate2", "xz2"]
[patch.crates-io]
# For pyproject-toml
pep440_rs = { path = "crates/pep440-rs" }
pep508_rs = { path = "crates/pep508-rs" }
reqwest-middleware = { git = "https://github.com/astral-sh/reqwest-middleware", rev = "21ceec9a5fd2e8d6f71c3ea2999078fecbd13cbe" }
[workspace.lints.rust]
unsafe_code = "warn"
unreachable_pub = "warn"
# TODO(konsti): Reactivate when check-cfg hits stable rust
# unexpected_cfgs = { level = "warn", check-cfg = ['cfg(codspeed)'] }
[workspace.lints.clippy]
pedantic = { level = "warn", priority = -2 }
# Allowed pedantic lints
char_lit_as_u8 = "allow"
collapsible_else_if = "allow"
collapsible_if = "allow"
implicit_hasher = "allow"
map_unwrap_or = "allow"
match_same_arms = "allow"
missing_errors_doc = "allow"
missing_panics_doc = "allow"
module_name_repetitions = "allow"
must_use_candidate = "allow"
similar_names = "allow"
too_many_lines = "allow"
too_many_arguments = "allow"
# Disallowed restriction lints
print_stdout = "warn"
print_stderr = "warn"
dbg_macro = "warn"
empty_drop = "warn"
empty_structs_with_brackets = "warn"
exit = "warn"
get_unwrap = "warn"
rc_buffer = "warn"
rc_mutex = "warn"
rest_pat_in_fully_bound_structs = "warn"
[profile.release]
strip = true
lto = "fat"
# This profile is meant to mimic the `release` profile as closely as
# possible, but using settings that are more beneficial for iterative
# development. That is, the `release` profile is intended for actually
# building the release, where as `profiling` is meant for building `uv`
# for running benchmarks.
#
# The main differences here are to avoid stripping debug information
# and disabling lto. This does result in a mismatch between our release
# configuration and our benchmarking configuration, which is unfortunate.
# But compile times with `lto = true` are completely untenable:
#
# $ cargo b --profile profiling -p uv
# Compiling uv-cli v0.0.1 (/home/andrew/astral/uv/crates/uv-cli)
# Compiling uv v0.2.34 (/home/andrew/astral/uv/crates/uv)
# Finished `profiling` profile [optimized + debuginfo] target(s) in 3m 47s
#
# Using `lto = "thin"` brings a massive improvement, but it's still slow:
#
# $ cargo b --profile profiling -p uv
# Compiling uv v0.2.34 (/home/andrew/astral/uv/crates/uv)
# Finished `profiling` profile [optimized + debuginfo] target(s) in 53.98s
#
# But with `lto = false`:
#
# $ cargo b --profile profiling -p uv
# Compiling uv v0.2.34 (/home/andrew/astral/uv/crates/uv)
# Finished `profiling` profile [optimized + debuginfo] target(s) in 30.09s
#
# We get more reasonable-ish compile times. At least, it's not enough
# time to get up and get a cup of coffee before it completes.
#
# This setup does risk that we are measuring something in benchmarks
# that we are shipping, but in order to make those two the same, we'd
# either need to make compile times way worse for development, or take
# a hit to binary size and a slight hit to runtime performance in our
# release builds.
[profile.profiling]
inherits = "release"
strip = false
debug = "full"
lto = false
[profile.fast-build]
inherits = "dev"
debug = 0
strip = "debuginfo"
# The profile that 'cargo dist' will build with.
[profile.dist]
inherits = "release"
# Config for 'cargo dist'
[workspace.metadata.dist]
# The preferred cargo-dist version to use in CI (Cargo.toml SemVer syntax)
cargo-dist-version = "0.18.0"
# CI backends to support
ci = ["github"]
# The installers to generate for each app
installers = ["shell", "powershell"]
# The archive format to use for windows builds (defaults .zip)
windows-archive = ".zip"
# The archive format to use for non-windows builds (defaults .tar.xz)
unix-archive = ".tar.gz"
# Target platforms to build apps for (Rust target-triple syntax)
targets = [
"aarch64-apple-darwin",
"aarch64-unknown-linux-gnu",
"aarch64-unknown-linux-musl",
"arm-unknown-linux-musleabihf",
"armv7-unknown-linux-gnueabihf",
"armv7-unknown-linux-musleabihf",
"i686-pc-windows-msvc",
"i686-unknown-linux-gnu",
"i686-unknown-linux-musl",
"powerpc64-unknown-linux-gnu",
"powerpc64le-unknown-linux-gnu",
"s390x-unknown-linux-gnu",
"x86_64-apple-darwin",
"x86_64-pc-windows-msvc",
"x86_64-unknown-linux-gnu",
"x86_64-unknown-linux-musl",
]
# Whether to auto-include files like READMEs, LICENSEs, and CHANGELOGs (default true)
auto-includes = false
# Whether cargo-dist should create a GitHub Release or use an existing draft
create-release = true
# Publish jobs to run in CI
pr-run-mode = "skip"
# Whether CI should trigger releases with dispatches instead of tag pushes
dispatch-releases = true
# The stage during which the GitHub Release should be created
github-release = "announce"
# Whether CI should include auto-generated code to build local artifacts
build-local-artifacts = false
# Local artifacts jobs to run in CI
local-artifacts-jobs = ["./build-binaries", "./build-docker"]
# Publish jobs to run in CI
publish-jobs = ["./publish-pypi"]
# Announcement jobs to run in CI
post-announce-jobs = ["./publish-docs"]
# Custom permissions for GitHub Jobs
github-custom-job-permissions = { "build-docker" = { packages = "write", contents = "read" } }
# Whether to install an updater program
install-updater = false