Merge pull request #308 from cuviper/extract_if

Add map and set `extract_if`

Author: cuviper

src/map.rs

@@ -16,7 +16,8 @@ mod tests;
 pub use self::core::raw_entry_v1::{self, RawEntryApiV1};
 pub use self::core::{Entry, IndexedEntry, OccupiedEntry, VacantEntry};
 pub use self::iter::{
-    Drain, IntoIter, IntoKeys, IntoValues, Iter, IterMut, IterMut2, Keys, Splice, Values, ValuesMut,
+    Drain, ExtractIf, IntoIter, IntoKeys, IntoValues, Iter, IterMut, IterMut2, Keys, Splice,
+    Values, ValuesMut,
 };
 pub use self::mutable::MutableEntryKey;
 pub use self::mutable::MutableKeys;
@@ -36,7 +37,7 @@ use alloc::vec::Vec;
 #[cfg(feature = "std")]
 use std::collections::hash_map::RandomState;
 
-use self::core::IndexMapCore;
+pub(crate) use self::core::{ExtractCore, IndexMapCore};
 use crate::util::{third, try_simplify_range};
 use crate::{Bucket, Equivalent, GetDisjointMutError, HashValue, TryReserveError};
 
@@ -303,6 +304,55 @@ impl<K, V, S> IndexMap<K, V, S> {
         Drain::new(self.core.drain(range))
     }
 
+    /// Creates an iterator which uses a closure to determine if an element should be removed,
+    /// for all elements in the given range.
+    ///
+    /// If the closure returns true, the element is removed from the map and yielded.
+    /// If the closure returns false, or panics, the element remains in the map and will not be
+    /// yielded.
+    ///
+    /// Note that `extract_if` lets you mutate every value in the filter closure, regardless of
+    /// whether you choose to keep or remove it.
+    ///
+    /// The range may be any type that implements [`RangeBounds<usize>`],
+    /// including all of the `std::ops::Range*` types, or even a tuple pair of
+    /// `Bound` start and end values. To check the entire map, use `RangeFull`
+    /// like `map.extract_if(.., predicate)`.
+    ///
+    /// If the returned `ExtractIf` is not exhausted, e.g. because it is dropped without iterating
+    /// or the iteration short-circuits, then the remaining elements will be retained.
+    /// Use [`retain`] with a negated predicate if you do not need the returned iterator.
+    ///
+    /// [`retain`]: IndexMap::retain
+    ///
+    /// ***Panics*** if the starting point is greater than the end point or if
+    /// the end point is greater than the length of the map.
+    ///
+    /// # Examples
+    ///
+    /// Splitting a map into even and odd keys, reusing the original map:
+    ///
+    /// ```
+    /// use indexmap::IndexMap;
+    ///
+    /// let mut map: IndexMap<i32, i32> = (0..8).map(|x| (x, x)).collect();
+    /// let extracted: IndexMap<i32, i32> = map.extract_if(.., |k, _v| k % 2 == 0).collect();
+    ///
+    /// let evens = extracted.keys().copied().collect::<Vec<_>>();
+    /// let odds = map.keys().copied().collect::<Vec<_>>();
+    ///
+    /// assert_eq!(evens, vec![0, 2, 4, 6]);
+    /// assert_eq!(odds, vec![1, 3, 5, 7]);
+    /// ```
+    #[track_caller]
+    pub fn extract_if<F, R>(&mut self, range: R, pred: F) -> ExtractIf<'_, K, V, F>
+    where
+        F: FnMut(&K, &mut V) -> bool,
+        R: RangeBounds<usize>,
+    {
+        ExtractIf::new(&mut self.core, range, pred)
+    }
+
     /// Splits the collection into two at the given index.
     ///
     /// Returns a newly allocated map containing the elements in the range

src/map/core.rs

@@ -8,6 +8,7 @@
 //! However, we should probably not let this show in the public API or docs.
 
 mod entry;
+mod extract;
 
 pub mod raw_entry_v1;
 
@@ -23,6 +24,7 @@ type Indices = hash_table::HashTable<usize>;
 type Entries<K, V> = Vec<Bucket<K, V>>;
 
 pub use entry::{Entry, IndexedEntry, OccupiedEntry, VacantEntry};
+pub(crate) use extract::ExtractCore;
 
 /// Core of the map that does not depend on S
 #[derive(Debug)]
@@ -157,6 +159,7 @@ impl<K, V> IndexMapCore<K, V> {
 
     #[inline]
     pub(crate) fn len(&self) -> usize {
+        debug_assert_eq!(self.entries.len(), self.indices.len());
         self.indices.len()
     }
 

src/map/core/extract.rs

@@ -0,0 +1,108 @@
+#![allow(unsafe_code)]
+
+use super::{Bucket, IndexMapCore};
+use crate::util::simplify_range;
+
+use core::ops::RangeBounds;
+
+impl<K, V> IndexMapCore<K, V> {
+    #[track_caller]
+    pub(crate) fn extract<R>(&mut self, range: R) -> ExtractCore<'_, K, V>
+    where
+        R: RangeBounds<usize>,
+    {
+        let range = simplify_range(range, self.entries.len());
+
+        // SAFETY: We must have consistent lengths to start, so that's a hard assertion.
+        // Then the worst `set_len` can do is leak items if `ExtractCore` doesn't drop.
+        assert_eq!(self.entries.len(), self.indices.len());
+        unsafe {
+            self.entries.set_len(range.start);
+        }
+        ExtractCore {
+            map: self,
+            new_len: range.start,
+            current: range.start,
+            end: range.end,
+        }
+    }
+}
+
+pub(crate) struct ExtractCore<'a, K, V> {
+    map: &'a mut IndexMapCore<K, V>,
+    new_len: usize,
+    current: usize,
+    end: usize,
+}
+
+impl<K, V> Drop for ExtractCore<'_, K, V> {
+    fn drop(&mut self) {
+        let old_len = self.map.indices.len();
+        let mut new_len = self.new_len;
+
+        debug_assert!(new_len <= self.current);
+        debug_assert!(self.current <= self.end);
+        debug_assert!(self.current <= old_len);
+        debug_assert!(old_len <= self.map.entries.capacity());
+
+        // SAFETY: We assume `new_len` and `current` were correctly maintained by the iterator.
+        // So `entries[new_len..current]` were extracted, but the rest before and after are valid.
+        unsafe {
+            if new_len == self.current {
+                // Nothing was extracted, so any remaining items can be left in place.
+                new_len = old_len;
+            } else if self.current < old_len {
+                // Need to shift the remaining items down.
+                let tail_len = old_len - self.current;
+                let base = self.map.entries.as_mut_ptr();
+                let src = base.add(self.current);
+                let dest = base.add(new_len);
+                src.copy_to(dest, tail_len);
+                new_len += tail_len;
+            }
+            self.map.entries.set_len(new_len);
+        }
+
+        if new_len != old_len {
+            // We don't keep track of *which* items were extracted, so reindex everything.
+            self.map.rebuild_hash_table();
+        }
+    }
+}
+
+impl<K, V> ExtractCore<'_, K, V> {
+    pub(crate) fn extract_if<F>(&mut self, mut pred: F) -> Option<Bucket<K, V>>
+    where
+        F: FnMut(&mut Bucket<K, V>) -> bool,
+    {
+        debug_assert!(self.end <= self.map.entries.capacity());
+
+        let base = self.map.entries.as_mut_ptr();
+        while self.current < self.end {
+            // SAFETY: We're maintaining both indices within bounds of the original entries, so
+            // 0..new_len and current..indices.len() are always valid items for our Drop to keep.
+            unsafe {
+                let item = base.add(self.current);
+                if pred(&mut *item) {
+                    // Extract it!
+                    self.current += 1;
+                    return Some(item.read());
+                } else {
+                    // Keep it, shifting it down if needed.
+                    if self.new_len != self.current {
+                        debug_assert!(self.new_len < self.current);
+                        let dest = base.add(self.new_len);
+                        item.copy_to_nonoverlapping(dest, 1);
+                    }
+                    self.current += 1;
+                    self.new_len += 1;
+                }
+            }
+        }
+        None
+    }
+
+    pub(crate) fn remaining(&self) -> usize {
+        self.end - self.current
+    }
+}

src/map/iter.rs

@@ -1,5 +1,4 @@
-use super::core::IndexMapCore;
-use super::{Bucket, IndexMap, Slice};
+use super::{Bucket, ExtractCore, IndexMap, IndexMapCore, Slice};
 
 use alloc::vec::{self, Vec};
 use core::fmt;
@@ -774,3 +773,58 @@ where
             .finish()
     }
 }
+
+/// An extracting iterator for `IndexMap`.
+///
+/// This `struct` is created by [`IndexMap::extract_if()`].
+/// See its documentation for more.
+pub struct ExtractIf<'a, K, V, F> {
+    inner: ExtractCore<'a, K, V>,
+    pred: F,
+}
+
+impl<K, V, F> ExtractIf<'_, K, V, F> {
+    #[track_caller]
+    pub(super) fn new<R>(core: &mut IndexMapCore<K, V>, range: R, pred: F) -> ExtractIf<'_, K, V, F>
+    where
+        R: RangeBounds<usize>,
+        F: FnMut(&K, &mut V) -> bool,
+    {
+        ExtractIf {
+            inner: core.extract(range),
+            pred,
+        }
+    }
+}
+
+impl<K, V, F> Iterator for ExtractIf<'_, K, V, F>
+where
+    F: FnMut(&K, &mut V) -> bool,
+{
+    type Item = (K, V);
+
+    fn next(&mut self) -> Option<Self::Item> {
+        self.inner
+            .extract_if(|bucket| {
+                let (key, value) = bucket.ref_mut();
+                (self.pred)(key, value)
+            })
+            .map(Bucket::key_value)
+    }
+
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        (0, Some(self.inner.remaining()))
+    }
+}
+
+impl<K, V, F> FusedIterator for ExtractIf<'_, K, V, F> where F: FnMut(&K, &mut V) -> bool {}
+
+impl<K, V, F> fmt::Debug for ExtractIf<'_, K, V, F>
+where
+    K: fmt::Debug,
+    V: fmt::Debug,
+{
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        f.debug_struct("ExtractIf").finish_non_exhaustive()
+    }
+}

src/set.rs

@@ -8,7 +8,7 @@ mod slice;
 mod tests;
 
 pub use self::iter::{
-    Difference, Drain, Intersection, IntoIter, Iter, Splice, SymmetricDifference, Union,
+    Difference, Drain, ExtractIf, Intersection, IntoIter, Iter, Splice, SymmetricDifference, Union,
 };
 pub use self::mutable::MutableValues;
 pub use self::slice::Slice;
@@ -249,6 +249,52 @@ impl<T, S> IndexSet<T, S> {
         Drain::new(self.map.core.drain(range))
     }
 
+    /// Creates an iterator which uses a closure to determine if a value should be removed,
+    /// for all values in the given range.
+    ///
+    /// If the closure returns true, then the value is removed and yielded.
+    /// If the closure returns false, the value will remain in the list and will not be yielded
+    /// by the iterator.
+    ///
+    /// The range may be any type that implements [`RangeBounds<usize>`],
+    /// including all of the `std::ops::Range*` types, or even a tuple pair of
+    /// `Bound` start and end values. To check the entire set, use `RangeFull`
+    /// like `set.extract_if(.., predicate)`.
+    ///
+    /// If the returned `ExtractIf` is not exhausted, e.g. because it is dropped without iterating
+    /// or the iteration short-circuits, then the remaining elements will be retained.
+    /// Use [`retain`] with a negated predicate if you do not need the returned iterator.
+    ///
+    /// [`retain`]: IndexSet::retain
+    ///
+    /// ***Panics*** if the starting point is greater than the end point or if
+    /// the end point is greater than the length of the set.
+    ///
+    /// # Examples
+    ///
+    /// Splitting a set into even and odd values, reusing the original set:
+    ///
+    /// ```
+    /// use indexmap::IndexSet;
+    ///
+    /// let mut set: IndexSet<i32> = (0..8).collect();
+    /// let extracted: IndexSet<i32> = set.extract_if(.., |v| v % 2 == 0).collect();
+    ///
+    /// let evens = extracted.into_iter().collect::<Vec<_>>();
+    /// let odds = set.into_iter().collect::<Vec<_>>();
+    ///
+    /// assert_eq!(evens, vec![0, 2, 4, 6]);
+    /// assert_eq!(odds, vec![1, 3, 5, 7]);
+    /// ```
+    #[track_caller]
+    pub fn extract_if<F, R>(&mut self, range: R, pred: F) -> ExtractIf<'_, T, F>
+    where
+        F: FnMut(&T) -> bool,
+        R: RangeBounds<usize>,
+    {
+        ExtractIf::new(&mut self.map.core, range, pred)
+    }
+
     /// Splits the collection into two at the given index.
     ///
     /// Returns a newly allocated set containing the elements in the range

src/set/iter.rs

@@ -1,3 +1,5 @@
+use crate::map::{ExtractCore, IndexMapCore};
+
 use super::{Bucket, IndexSet, Slice};
 
 use alloc::vec::{self, Vec};
@@ -626,3 +628,54 @@ impl<I: fmt::Debug> fmt::Debug for UnitValue<I> {
         fmt::Debug::fmt(&self.0, f)
     }
 }
+
+/// An extracting iterator for `IndexSet`.
+///
+/// This `struct` is created by [`IndexSet::extract_if()`].
+/// See its documentation for more.
+pub struct ExtractIf<'a, T, F> {
+    inner: ExtractCore<'a, T, ()>,
+    pred: F,
+}
+
+impl<T, F> ExtractIf<'_, T, F> {
+    #[track_caller]
+    pub(super) fn new<R>(core: &mut IndexMapCore<T, ()>, range: R, pred: F) -> ExtractIf<'_, T, F>
+    where
+        R: RangeBounds<usize>,
+        F: FnMut(&T) -> bool,
+    {
+        ExtractIf {
+            inner: core.extract(range),
+            pred,
+        }
+    }
+}
+
+impl<T, F> Iterator for ExtractIf<'_, T, F>
+where
+    F: FnMut(&T) -> bool,
+{
+    type Item = T;
+
+    fn next(&mut self) -> Option<Self::Item> {
+        self.inner
+            .extract_if(|bucket| (self.pred)(bucket.key_ref()))
+            .map(Bucket::key)
+    }
+
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        (0, Some(self.inner.remaining()))
+    }
+}
+
+impl<T, F> FusedIterator for ExtractIf<'_, T, F> where F: FnMut(&T) -> bool {}
+
+impl<T, F> fmt::Debug for ExtractIf<'_, T, F>
+where
+    T: fmt::Debug,
+{
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        f.debug_struct("ExtractIf").finish_non_exhaustive()
+    }
+}