From bc63ec57e7f965796479f68f6f687187c089bd40 Mon Sep 17 00:00:00 2001
From: Jeremy Evans <code@jeremyevans.net>
Date: Wed, 28 Oct 2020 10:58:28 -0700
Subject: [PATCH] [ruby/zlib] Allow Zlib.crc32 and .adler32 to accept IO
 instance

This reads from the IO in 8192 byte chunks, so you don't need to
have the entire string in memory.

Fixes #16

https://github.com/ruby/zlib/commit/ba9793c550
---
 ext/zlib/zlib.c        | 19 ++++++++++++++++---
 test/zlib/test_zlib.rb | 26 ++++++++++++++++++++++++++
 2 files changed, 42 insertions(+), 3 deletions(-)

diff --git a/ext/zlib/zlib.c b/ext/zlib/zlib.c
index c540d4c5a1..fa2112f6b8 100644
--- a/ext/zlib/zlib.c
+++ b/ext/zlib/zlib.c
@@ -56,7 +56,7 @@ max_uint(long n)
 #define MAX_UINT(n) (uInt)(n)
 #endif
 
-static ID id_dictionaries;
+static ID id_dictionaries, id_read;
 
 /*--------- Prototypes --------*/
 
@@ -407,6 +407,15 @@ do_checksum(int argc, VALUE *argv, uLong (*func)(uLong, const Bytef*, uInt))
     if (NIL_P(str)) {
 	sum = func(sum, Z_NULL, 0);
     }
+    else if (rb_obj_is_kind_of(str, rb_cIO)) {
+        VALUE buf;
+        VALUE buflen = INT2NUM(8192);
+
+        while (!NIL_P(buf = rb_funcall(str, id_read, 1, buflen))) {
+            StringValue(buf);
+            sum = checksum_long(func, sum, (Bytef*)RSTRING_PTR(buf), RSTRING_LEN(buf));
+        }
+    }
     else {
 	StringValue(str);
 	sum = checksum_long(func, sum, (Bytef*)RSTRING_PTR(str), RSTRING_LEN(str));
@@ -422,6 +431,8 @@ do_checksum(int argc, VALUE *argv, uLong (*func)(uLong, const Bytef*, uInt))
  * Calculates Adler-32 checksum for +string+, and returns updated value of
  * +adler+. If +string+ is omitted, it returns the Adler-32 initial value. If
  * +adler+ is omitted, it assumes that the initial value is given to +adler+.
+ * If +string+ is an IO instance, reads from the IO until the IO returns nil
+ * and returns Adler-32 of all read data.
  *
  * Example usage:
  *
@@ -466,7 +477,9 @@ rb_zlib_adler32_combine(VALUE klass, VALUE adler1, VALUE adler2, VALUE len2)
  *
  * Calculates CRC checksum for +string+, and returns updated value of +crc+. If
  * +string+ is omitted, it returns the CRC initial value. If +crc+ is omitted, it
- * assumes that the initial value is given to +crc+.
+ * assumes that the initial value is given to +crc+. If +string+ is an IO instance,
+ * reads from the IO until the IO returns nil and returns CRC checksum of all read
+ * data.
  *
  * FIXME: expression.
  */
@@ -2198,7 +2211,7 @@ rb_inflate_set_dictionary(VALUE obj, VALUE dic)
 #define OS_CODE  OS_UNIX
 #endif
 
-static ID id_write, id_read, id_readpartial, id_flush, id_seek, id_close, id_path, id_input;
+static ID id_write, id_readpartial, id_flush, id_seek, id_close, id_path, id_input;
 static VALUE cGzError, cNoFooter, cCRCError, cLengthError;
 
 
diff --git a/test/zlib/test_zlib.rb b/test/zlib/test_zlib.rb
index c58eafe112..c72fe76858 100644
--- a/test/zlib/test_zlib.rb
+++ b/test/zlib/test_zlib.rb
@@ -1145,6 +1145,19 @@ if defined? Zlib
       assert_equal(0x02820145, Zlib.adler32("foo"))
       assert_equal(0x02820145, Zlib.adler32("o", Zlib.adler32("fo")))
       assert_equal(0x8a62c964, Zlib.adler32("abc\x01\x02\x03" * 10000))
+      Tempfile.create("test_zlib_gzip_file_to_io") {|t|
+        File.binwrite(t.path, "foo")
+        t.rewind
+        assert_equal(0x02820145, Zlib.adler32(t))
+
+        t.rewind
+        crc = Zlib.adler32(t.read(2))
+        assert_equal(0x02820145, Zlib.adler32(t, crc))
+
+        File.binwrite(t.path, "abc\x01\x02\x03" * 10000)
+        t.rewind
+        assert_equal(0x8a62c964, Zlib.adler32(t))
+      }
     end
 
     def test_adler32_combine
@@ -1167,6 +1180,19 @@ if defined? Zlib
       assert_equal(0x8c736521, Zlib.crc32("foo"))
       assert_equal(0x8c736521, Zlib.crc32("o", Zlib.crc32("fo")))
       assert_equal(0x07f0d68f, Zlib.crc32("abc\x01\x02\x03" * 10000))
+      Tempfile.create("test_zlib_gzip_file_to_io") {|t|
+        File.binwrite(t.path, "foo")
+        t.rewind
+        assert_equal(0x8c736521, Zlib.crc32(t))
+
+        t.rewind
+        crc = Zlib.crc32(t.read(2))
+        assert_equal(0x8c736521, Zlib.crc32(t, crc))
+
+        File.binwrite(t.path, "abc\x01\x02\x03" * 10000)
+        t.rewind
+        assert_equal(0x07f0d68f, Zlib.crc32(t))
+      }
     end
 
     def test_crc32_combine