Make binary detection work similar to vanilla git

Main change: Don't treat chars > 128 as non-printable (common in UTF-8 files)

Signed-off-by: Sven Strickroth <email@cs-ware.de>

CHANGELOG.md

@@ -3,6 +3,9 @@ v0.22 + 1
 
 ### Changes or improvements
 
+* Updated binary identification in CRLF filtering to avoid false positives in
+  UTF-8 files.
+
 * Rename and copy detection is enabled for small files.
 
 ### API additions

src/buf_text.c

@@ -191,7 +191,10 @@ bool git_buf_text_is_binary(const git_buf *buf)
 	while (scan < end) {
 		unsigned char c = *scan++;
 
-		if (c > 0x1F && c < 0x7F)
+		/* Printable characters are those above SPACE (0x1F) excluding DEL,
+		 * and including BS, ESC and FF.
+		 */
+		if ((c > 0x1F && c != 127) || c == '\b' || c == '\033' || c == '\014')
 			printable++;
 		else if (c == '\0')
 			return true;

tests/core/buffer.c

@@ -830,7 +830,7 @@ void test_core_buffer__classify_with_utf8(void)
 	cl_assert(!git_buf_text_contains_nul(&b));
 
 	b.ptr = data1; b.size = b.asize = data1len;
-	cl_assert(git_buf_text_is_binary(&b));
+	cl_assert(!git_buf_text_is_binary(&b));
 	cl_assert(!git_buf_text_contains_nul(&b));
 
 	b.ptr = data2; b.size = b.asize = data2len;