[CHORE] drop unused deps and add machete CI check (#2520)

* removes unneeded deps for daft crates
* adds check for them in the future in CI via cargo machete

.github/workflows/python-package.yml

@@ -675,6 +675,20 @@ jobs:
       if: ${{ contains(needs.*.result, 'failure') }}
       run: exit 1
 
+  check-unused-cargo-deps:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+    - uses: moonrepo/setup-rust@v1
+      with:
+        cache: false
+        channel: stable
+
+    - name: Install Machete
+      run: cargo +stable install cargo-machete
+    - name: Run Machete
+      run: cargo machete --with-metadata
+
   doctests:
     runs-on: ubuntu-latest
     continue-on-error: true

Cargo.toml

@@ -59,6 +59,9 @@ name = "daft"
 publish = false
 version = "0.2.0-dev0"
 
+[package.metadata.cargo-machete]
+ignored = ["lzma_sys"]
+
 [patch.crates-io]
 arrow2 = {path = "src/arrow2"}
 parquet2 = {path = "src/parquet2"}

src/arrow2/Cargo.toml

@@ -247,5 +247,8 @@ version = "0.17.4"
 [package.metadata.cargo-all-features]
 allowlist = ["compute", "compute_sort", "compute_hash", "compute_nullif"]
 
+[package.metadata.cargo-machete]
+ignored = ["arrow_array", "arrow_buffer", "avro_rs", "criterion", "crossbeam_channel", "flate2", "getrandom", "rustc_version", "sample_arrow2", "sample_std", "sample_test", "tokio", "tokio_util"]
+
 [target.wasm32-unknown-unknown.dependencies]
 getrandom = {version = "0.2", features = ["js"]}

src/common/daft-config/Cargo.toml

@@ -1,7 +1,6 @@
 [dependencies]
 bincode = {workspace = true}
 common-io-config = {path = "../io-config", default-features = false}
-lazy_static = {workspace = true}
 pyo3 = {workspace = true, optional = true}
 serde = {workspace = true}
 

src/common/treenode/Cargo.toml

@@ -1,6 +1,5 @@
 [dependencies]
 common-error = {path = "../error", default-features = false}
-pyo3 = {workspace = true, optional = true}
 
 [features]
 python = ["common-error/python"]

src/daft-core/Cargo.toml

@@ -23,7 +23,6 @@ common-daft-config = {path = "../common/daft-config", default-features = false}
 common-error = {path = "../common/error", default-features = false}
 daft-minhash = {path = "../daft-minhash", default-features = false}
 daft-sketch = {path = "../daft-sketch", default-features = false}
-dyn-clone = "1.0.17"
 fastrand = "2.1.0"
 fnv = "1.0.7"
 html-escape = {workspace = true}
@@ -40,7 +39,6 @@ ndarray = "0.15.6"
 num-derive = {workspace = true}
 num-traits = {workspace = true}
 pyo3 = {workspace = true, optional = true}
-rand = {workspace = true}
 regex = {workspace = true}
 serde = {workspace = true}
 serde_json = {workspace = true}

src/daft-core/src/utils/mod.rs

@@ -3,6 +3,8 @@ pub mod display_table;
 pub mod hashable_float_wrapper;
 pub mod supertype;
 
+pub use bincode;
+
 #[macro_export]
 macro_rules! impl_binary_trait_by_reference {
     ($ty:ty, $trait:ident, $fname:ident) => {
@@ -26,7 +28,10 @@ macro_rules! impl_bincode_py_state_serialization {
                     Self::type_object(py)
                         .getattr("_from_serialized")?
                         .to_object(py),
-                    (PyBytes::new(py, &bincode::serialize(&self).unwrap()).to_object(py),)
+                    (
+                        PyBytes::new(py, &$crate::utils::bincode::serialize(&self).unwrap())
+                            .to_object(py),
+                    )
                         .to_object(py),
                 ))
             }
@@ -35,7 +40,7 @@ macro_rules! impl_bincode_py_state_serialization {
             pub fn _from_serialized(py: Python, serialized: PyObject) -> PyResult<Self> {
                 serialized
                     .extract::<&PyBytes>(py)
-                    .map(|s| bincode::deserialize(s.as_bytes()).unwrap())
+                    .map(|s| $crate::utils::bincode::deserialize(s.as_bytes()).unwrap())
             }
         }
     };

src/daft-csv/Cargo.toml

@@ -3,10 +3,6 @@ arrow2 = {workspace = true, features = ["io_csv", "io_csv_async"]}
 async-compat = {workspace = true}
 async-compression = {workspace = true}
 async-stream = {workspace = true}
-bincode = {workspace = true}
-bytes = {workspace = true}
-chrono = {workspace = true}
-chrono-tz = {workspace = true}
 common-error = {path = "../common/error", default-features = false}
 csv-async = "1.3.0"
 daft-compression = {path = "../daft-compression", default-features = false}
@@ -16,15 +12,11 @@ daft-dsl = {path = "../daft-dsl", default-features = false}
 daft-io = {path = "../daft-io", default-features = false}
 daft-table = {path = "../daft-table", default-features = false}
 futures = {workspace = true}
-lexical-core = {version = "0.8"}
-log = {workspace = true}
 pyo3 = {workspace = true, optional = true}
 rayon = {workspace = true}
 serde = {workspace = true}
-simdutf8 = "0.1.3"
 snafu = {workspace = true}
 tokio = {workspace = true}
-tokio-stream = {workspace = true}
 tokio-util = {workspace = true}
 url = {workspace = true}
 

src/daft-decoding/Cargo.toml

@@ -1,14 +1,10 @@
 [dependencies]
 arrow2 = {workspace = true, features = ["io_csv", "io_csv_async"]}
-async-compression = {workspace = true}
 atoi_simd = "0.15.5"
 chrono = {workspace = true}
-chrono-tz = {workspace = true}
 csv-async = "1.2.6"
 fast-float = "0.2.0"
 simdutf8 = "0.1.3"
-tokio = {workspace = true}
-url = {workspace = true}
 
 [package]
 edition = {workspace = true}