Rollup merge of rust-lang#81136 - Xavientois:io_reader_size_hint, r=c…

…ramertj

Improved IO Bytes Size Hint

After trying to implement better `size_hint()` return values for `File` in [this PR](rust-lang#81044) and changing to implementing it for `BufReader` in [this PR](rust-lang#81052), I have arrived at this implementation that provides tighter bounds for the `Bytes` iterator of various readers including `BufReader`, `Empty`, and `Chain`.

Unfortunately, for `BufReader`, the size_hint only improves after calling `fill_buffer` due to it using the contents of the buffer for the hint. Nevertheless, the the tighter bounds  should result in better pre-allocation of space to handle the contents of the `Bytes` iterator.

Closes rust-lang#81052

library/std/src/io/buffered/bufreader.rs

@@ -1,6 +1,8 @@
 use crate::cmp;
 use crate::fmt;
-use crate::io::{self, BufRead, Initializer, IoSliceMut, Read, Seek, SeekFrom, DEFAULT_BUF_SIZE};
+use crate::io::{
+    self, BufRead, Initializer, IoSliceMut, Read, Seek, SeekFrom, SizeHint, DEFAULT_BUF_SIZE,
+};
 
 /// The `BufReader<R>` struct adds buffering to any reader.
 ///
@@ -435,3 +437,9 @@ impl<R: Seek> Seek for BufReader<R> {
         })
     }
 }
+
+impl<T> SizeHint for BufReader<T> {
+    fn lower_bound(&self) -> usize {
+        self.buffer().len()
+    }
+}

library/std/src/io/mod.rs

@@ -2238,6 +2238,19 @@ impl<T: BufRead, U: BufRead> BufRead for Chain<T, U> {
     }
 }
 
+impl<T, U> SizeHint for Chain<T, U> {
+    fn lower_bound(&self) -> usize {
+        SizeHint::lower_bound(&self.first) + SizeHint::lower_bound(&self.second)
+    }
+
+    fn upper_bound(&self) -> Option<usize> {
+        match (SizeHint::upper_bound(&self.first), SizeHint::upper_bound(&self.second)) {
+            (Some(first), Some(second)) => Some(first + second),
+            _ => None,
+        }
+    }
+}
+
 /// Reader adaptor which limits the bytes read from an underlying reader.
 ///
 /// This struct is generally created by calling [`take`] on a reader.
@@ -2464,6 +2477,30 @@ impl<R: Read> Iterator for Bytes<R> {
             };
         }
     }
+
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        SizeHint::size_hint(&self.inner)
+    }
+}
+
+trait SizeHint {
+    fn lower_bound(&self) -> usize;
+
+    fn upper_bound(&self) -> Option<usize>;
+
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        (self.lower_bound(), self.upper_bound())
+    }
+}
+
+impl<T> SizeHint for T {
+    default fn lower_bound(&self) -> usize {
+        0
+    }
+
+    default fn upper_bound(&self) -> Option<usize> {
+        None
+    }
 }
 
 /// An iterator over the contents of an instance of `BufRead` split on a

library/std/src/io/tests.rs

@@ -1,7 +1,7 @@
 use super::{repeat, Cursor, SeekFrom};
 use crate::cmp::{self, min};
 use crate::io::{self, IoSlice, IoSliceMut};
-use crate::io::{BufRead, Read, Seek, Write};
+use crate::io::{BufRead, BufReader, Read, Seek, Write};
 use crate::ops::Deref;
 
 #[test]
@@ -198,6 +198,53 @@ fn chain_bufread() {
     cmp_bufread(chain1, chain2, &testdata[..]);
 }
 
+#[test]
+fn bufreader_size_hint() {
+    let testdata = b"ABCDEFGHIJKL";
+    let mut buf_reader = BufReader::new(&testdata[..]);
+    assert_eq!(buf_reader.buffer().len(), 0);
+
+    let buffer_length = testdata.len();
+    buf_reader.fill_buf().unwrap();
+
+    // Check that size hint matches buffer contents
+    let mut buffered_bytes = buf_reader.bytes();
+    let (lower_bound, _upper_bound) = buffered_bytes.size_hint();
+    assert_eq!(lower_bound, buffer_length);
+
+    // Check that size hint matches buffer contents after advancing
+    buffered_bytes.next().unwrap().unwrap();
+    let (lower_bound, _upper_bound) = buffered_bytes.size_hint();
+    assert_eq!(lower_bound, buffer_length - 1);
+}
+
+#[test]
+fn empty_size_hint() {
+    let size_hint = io::empty().bytes().size_hint();
+    assert_eq!(size_hint, (0, Some(0)));
+}
+
+#[test]
+fn chain_empty_size_hint() {
+    let chain = io::empty().chain(io::empty());
+    let size_hint = chain.bytes().size_hint();
+    assert_eq!(size_hint, (0, Some(0)));
+}
+
+#[test]
+fn chain_size_hint() {
+    let testdata = b"ABCDEFGHIJKL";
+    let mut buf_reader_1 = BufReader::new(&testdata[..6]);
+    let mut buf_reader_2 = BufReader::new(&testdata[6..]);
+
+    buf_reader_1.fill_buf().unwrap();
+    buf_reader_2.fill_buf().unwrap();
+
+    let chain = buf_reader_1.chain(buf_reader_2);
+    let size_hint = chain.bytes().size_hint();
+    assert_eq!(size_hint, (testdata.len(), None));
+}
+
 #[test]
 fn chain_zero_length_read_is_not_eof() {
     let a = b"A";

library/std/src/io/util.rs

@@ -4,7 +4,9 @@
 mod tests;
 
 use crate::fmt;
-use crate::io::{self, BufRead, Initializer, IoSlice, IoSliceMut, Read, Seek, SeekFrom, Write};
+use crate::io::{
+    self, BufRead, Initializer, IoSlice, IoSliceMut, Read, Seek, SeekFrom, SizeHint, Write,
+};
 
 /// A reader which is always at EOF.
 ///
@@ -80,6 +82,12 @@ impl fmt::Debug for Empty {
     }
 }
 
+impl SizeHint for Empty {
+    fn upper_bound(&self) -> Option<usize> {
+        Some(0)
+    }
+}
+
 /// A reader which yields one byte over and over and over and over and over and...
 ///
 /// This struct is generally created by calling [`repeat()`]. Please