From d32a73320e4bcf9e19daa8fe145a90bb25aff6a1 Mon Sep 17 00:00:00 2001
From: onur-ozkan <work@onurozkan.dev>
Date: Wed, 11 Oct 2023 17:02:03 +0300
Subject: [PATCH] optimize file read in `Config::verify`

`Config::verify` refactored to improve the efficiency and
memory usage of file hashing.

Signed-off-by: onur-ozkan <work@onurozkan.dev>
---
 src/bootstrap/download.rs | 37 +++++++++++++++++++++++++++----------
 1 file changed, 27 insertions(+), 10 deletions(-)

diff --git a/src/bootstrap/download.rs b/src/bootstrap/download.rs
index 8e9614ec89a0a..2653793da17d4 100644
--- a/src/bootstrap/download.rs
+++ b/src/bootstrap/download.rs
@@ -2,7 +2,7 @@ use std::{
     env,
     ffi::{OsStr, OsString},
     fs::{self, File},
-    io::{BufRead, BufReader, BufWriter, ErrorKind, Write},
+    io::{BufRead, BufReader, BufWriter, ErrorKind, Read, Write},
     path::{Path, PathBuf},
     process::{Command, Stdio},
 };
@@ -324,21 +324,38 @@ impl Config {
         use sha2::Digest;
 
         self.verbose(&format!("verifying {}", path.display()));
+
+        if self.dry_run() {
+            return false;
+        }
+
         let mut hasher = sha2::Sha256::new();
-        // FIXME: this is ok for rustfmt (4.1 MB large at time of writing), but it seems memory-intensive for rustc and larger components.
-        // Consider using streaming IO instead?
-        let contents = if self.dry_run() { vec![] } else { t!(fs::read(path)) };
-        hasher.update(&contents);
-        let found = hex::encode(hasher.finalize().as_slice());
-        let verified = found == expected;
-        if !verified && !self.dry_run() {
+
+        let file = t!(File::open(path));
+        let mut reader = BufReader::new(file);
+        let mut buffer = [0; 4096]; // read in chunks of 4KB
+
+        loop {
+            let read_len = t!(reader.read(&mut buffer));
+            // break if EOF
+            if read_len == 0 {
+                break;
+            }
+            hasher.update(&buffer[0..read_len]);
+        }
+
+        let checksum = hex::encode(hasher.finalize().as_slice());
+        let verified = checksum == expected;
+
+        if !verified {
             println!(
                 "invalid checksum: \n\
-                found:    {found}\n\
+                found:    {checksum}\n\
                 expected: {expected}",
             );
         }
-        return verified;
+
+        verified
     }
 }