Merge #120

120: clean up and prepare 0.7 release r=crepererum a=crepererum



Co-authored-by: Marco Neumann <marco@crepererum.net>

Author: bors[bot]

CHANGELOG.md

@@ -1,5 +1,26 @@
 # Changelog
 
+## 0.7 --- Misc Improvements & Fixes
+
+### All
+- support `?Sized` where possible
+- implement `Send` where possible
+
+### Filters
+- proper union support
+
+### HyperLogLog
+- fix potential out-of-bounds access (#74)
+- improve performance
+
+### T-Digest
+- more scale functions
+
+### Dependencies
+- replace `Void` with `Infallible`
+- update dependencies
+- make all dependencies optional
+
 ## 0.6 --- T-Digest, Lossy Counter
 
 ### T-Digest

README.md

@@ -1,6 +1,7 @@
 # pdatastructs
 
-Simple probabilistic data structures
+A collection of data structures that are based probability theory and therefore only provide the correct answer if a
+certain probability. In exchange they have a better runtime and memory complexity compared to traditional data structures.
 
 [![Build Status](https://github.com/crepererum/pdatastructs.rs/workflows/CI/badge.svg)](https://github.com/crepererum/pdatastructs.rs/actions?query=workflow%3ACI)
 [![Crates.io](https://img.shields.io/crates/v/pdatastructs.svg)](https://crates.io/crates/pdatastructs)
@@ -25,8 +26,8 @@ The following data structures are implemented:
 
 Licensed under either of these:
 
- * Apache License, Version 2.0 ([LICENSE-APACHE](LICENSE-APACHE) or https://www.apache.org/licenses/LICENSE-2.0)
- * MIT License ([LICENSE-MIT](LICENSE-MIT) or https://opensource.org/licenses/MIT)
+ * Apache License, Version 2.0 ([LICENSE-APACHE](LICENSE-APACHE) or <https://www.apache.org/licenses/LICENSE-2.0>)
+ * MIT License ([LICENSE-MIT](LICENSE-MIT) or <https://opensource.org/licenses/MIT>)
 
 ### Contributing
 

src/countminsketch.rs

@@ -203,7 +203,7 @@ where
 
         let w = (f64::consts::E / epsilon).ceil() as usize;
         let d = (1. / delta).ln().ceil() as usize;
-        CountMinSketch::with_params_and_hasher(w, d, buildhasher)
+        Self::with_params_and_hasher(w, d, buildhasher)
     }
 
     /// Get number of columns of internal counter table.
@@ -302,7 +302,7 @@ impl<T> fmt::Debug for CountMinSketch<T>
 where
     T: Hash + ?Sized,
 {
-    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         write!(f, "CountMinSketch {{ w: {}, d: {} }}", self.w, self.d)
     }
 }

src/filters/bloomfilter.rs

@@ -220,7 +220,7 @@ where
         let ln2 = (2f64).ln();
         let m = (-((n as f64) * p.ln()) / (ln2 * ln2)) as usize;
 
-        BloomFilter::with_params_and_hash(m, k, buildhasher)
+        Self::with_params_and_hash(m, k, buildhasher)
     }
 
     /// Get `k` (number of hash functions).
@@ -318,7 +318,7 @@ impl<T> fmt::Debug for BloomFilter<T>
 where
     T: Hash + ?Sized,
 {
-    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         write!(f, "BloomFilter {{ m: {}, k: {} }}", self.bs.len(), self.k)
     }
 }

src/filters/cuckoofilter.rs

@@ -14,7 +14,7 @@ const MAX_NUM_KICKS: usize = 500; // mentioned in paper
 
 /// Error struct used to signal that a `CuckooFilter` is full, i.e. that a value cannot be inserted
 /// because the implementation was unable to find a free bucket.
-#[derive(Debug)]
+#[derive(Debug, Clone, Copy)]
 pub struct CuckooFilterFull;
 
 /// A CuckooFilter is a set-like data structure, that keeps track of elements it has seen without
@@ -587,7 +587,7 @@ where
     R: Rng,
     B: BuildHasher + Clone + Eq,
 {
-    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         write!(
             f,
             "CuckooFilter {{ bucketsize: {}, n_buckets: {} }}",

src/filters/quotientfilter.rs

@@ -11,7 +11,7 @@ use crate::filters::Filter;
 use crate::helpers::all_zero_intvector;
 
 /// Error that signals that the QuotientFilter is full.
-#[derive(Debug)]
+#[derive(Debug, Clone, Copy)]
 pub struct QuotientFilterFull;
 
 /// Internal results for scanning the quotientfilter.
@@ -282,7 +282,7 @@ where
     /// and a default hasher.
     pub fn with_params(bits_quotient: usize, bits_remainder: usize) -> Self {
         let buildhasher = BuildHasherDefault::<DefaultHasher>::default();
-        QuotientFilter::with_params_and_hash(bits_quotient, bits_remainder, buildhasher)
+        Self::with_params_and_hash(bits_quotient, bits_remainder, buildhasher)
     }
 }
 

src/hash_utils.rs

@@ -105,7 +105,7 @@ where
     ///
     /// This call is CPU-expensive and may only be used if the resulting iterator will really be
     /// used.
-    pub fn iter_for<T>(&self, obj: &T) -> HashIter<B>
+    pub fn iter_for<T>(&self, obj: &T) -> HashIter<'_, B>
     where
         T: Hash + ?Sized,
     {
@@ -141,7 +141,7 @@ where
 #[derive(Debug)]
 pub struct HashIter<'a, B>
 where
-    B: 'a + BuildHasher,
+    B: BuildHasher,
 {
     builder: &'a HashIterBuilder<B>,
     h1: u64,
@@ -191,7 +191,7 @@ where
 }
 
 /// BuildHasher that takes a seed.
-#[derive(Clone, Debug, PartialEq, Eq)]
+#[derive(Clone, Copy, Debug, PartialEq, Eq)]
 pub struct BuildHasherSeeded {
     seed: usize,
 }

src/hyperloglog.rs

@@ -332,7 +332,7 @@ impl<T> fmt::Debug for HyperLogLog<T>
 where
     T: Hash + ?Sized,
 {
-    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         write!(f, "HyperLogLog {{ b: {} }}", self.b)
     }
 }

src/lib.rs

@@ -1,23 +1,30 @@
-//! A collection of data structures that are based probability theory and therefore only provide
-//! the correct answer if a certain probability. In exchange they have a better runtime and memory
-//! complexity compared to traditional data structures.
-
-#![deny(const_err)]
-#![deny(anonymous_parameters)]
-#![deny(bare_trait_objects)]
-#![deny(dead_code)]
-#![deny(illegal_floating_point_literal_pattern)]
-#![deny(missing_debug_implementations)]
-#![deny(missing_docs)]
-#![deny(non_camel_case_types)]
-#![deny(non_snake_case)]
-#![deny(non_upper_case_globals)]
-#![deny(unknown_lints)]
-#![deny(unreachable_code)]
-#![deny(unreachable_patterns)]
-#![deny(unreachable_pub)]
-#![deny(unsafe_code)]
-#![deny(unused_extern_crates)]
+#![doc = concat!(include_str!("../README.md"), "\n\n", include_str!("../CHANGELOG.md"))]
+#![deny(
+    anonymous_parameters,
+    bare_trait_objects,
+    clippy::clone_on_ref_ptr,
+    clippy::explicit_iter_loop,
+    clippy::future_not_send,
+    clippy::use_self,
+    const_err,
+    dead_code,
+    illegal_floating_point_literal_pattern,
+    missing_copy_implementations,
+    missing_debug_implementations,
+    missing_docs,
+    non_camel_case_types,
+    non_snake_case,
+    non_upper_case_globals,
+    rust_2018_idioms,
+    rustdoc::bare_urls,
+    rustdoc::broken_intra_doc_links,
+    unknown_lints,
+    unreachable_code,
+    unreachable_patterns,
+    unreachable_pub,
+    unsafe_code,
+    unused_extern_crates
+)]
 
 #[cfg(feature = "num-traits")]
 pub use num_traits;

src/reservoirsampling.rs

@@ -148,7 +148,7 @@ impl<T, R> fmt::Debug for ReservoirSampling<T, R>
 where
     R: Rng,
 {
-    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         write!(f, "ReservoirSampling {{ k: {} }}", self.k)
     }
 }