From ed8836fe48d70ec0c8581c5dfa3c4ff03e6a3d9e Mon Sep 17 00:00:00 2001
From: "Jorge C. Leitao" <jorgecarleitao@gmail.com>
Date: Tue, 2 Nov 2021 20:08:45 +0000
Subject: [PATCH] Revert "Made null_count not cached. (#563)"

This reverts commit f435f1f597980b2569907dc99b0bc498fd3f45e0.
---
 src/bitmap/immutable.rs | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/src/bitmap/immutable.rs b/src/bitmap/immutable.rs
index c530ff93c24..8a98c2caa97 100644
--- a/src/bitmap/immutable.rs
+++ b/src/bitmap/immutable.rs
@@ -20,6 +20,8 @@ pub struct Bitmap {
     // both are measured in bits. They are used to bound the bitmap to a region of Bytes.
     offset: usize,
     length: usize,
+    // this is a cache: it must be computed on initialization
+    null_count: usize,
 }
 
 impl std::fmt::Debug for Bitmap {
@@ -66,10 +68,12 @@ impl Bitmap {
     #[inline]
     pub(crate) fn from_bytes(bytes: Bytes<u8>, length: usize) -> Self {
         assert!(length <= bytes.len() * 8);
+        let null_count = count_zeros(&bytes, 0, length);
         Self {
             length,
             offset: 0,
             bytes: Arc::new(bytes),
+            null_count,
         }
     }
 
@@ -99,7 +103,7 @@ impl Bitmap {
     /// Returns the number of unset bits on this [`Bitmap`].
     #[inline]
     pub fn null_count(&self) -> usize {
-        count_zeros(&self.bytes, self.offset, self.length)
+        self.null_count
     }
 
     /// Slices `self`, offseting by `offset` and truncating up to `length` bits.
@@ -117,6 +121,17 @@ impl Bitmap {
     /// The caller must ensure that `self.offset + offset + length <= self.len()`
     #[inline]
     pub unsafe fn slice_unchecked(mut self, offset: usize, length: usize) -> Self {
+        // count the smallest chunk
+        if length < self.length / 2 {
+            // count the null values in the slice
+            self.null_count = count_zeros(&self.bytes, offset, length);
+        } else {
+            // subtract the null count of the chunks we slice off
+            let start_end = self.offset + offset + length;
+            let head_count = count_zeros(&self.bytes, self.offset, offset);
+            let tail_count = count_zeros(&self.bytes, start_end, self.length - length - offset);
+            self.null_count -= head_count + tail_count;
+        }
         self.offset += offset;
         self.length = length;
         self