human-readable: scale progress rates by default (#1138)

Author: steadytao

progress.c

@@ -26,6 +26,7 @@
 extern int am_server;
 extern int flist_eof;
 extern int quiet;
+extern int human_readable;
 extern int need_unsorted_flist;
 extern int output_needs_newline;
 extern int stdout_format_has_i;
@@ -69,9 +70,10 @@ static unsigned long msdiff(struct timeval *t1, struct timeval *t2)
 static void rprint_progress(OFF_T ofs, OFF_T size, struct timeval *now, int is_last)
 {
 	char rembuf[64], eol[128];
+	const char *rate_str;
 	unsigned long diff;
 	double rate, remain;
-	int pct;
+	int pct, rate_human_level;
 
 	if (is_last) {
 		int len = snprintf(eol, sizeof eol,
@@ -115,9 +117,13 @@ static void rprint_progress(OFF_T ofs, OFF_T size, struct timeval *now, int is_l
 
 	output_needs_newline = 0;
 	pct = ofs == size ? 100 : (int) (100.0 * ofs / size);
+	/* Keep progress rates scaled while byte counts retain the level-1 default. */
+	rate_human_level = human_readable == 1 ? 2 : human_readable;
+	rate_str = rate >= INT64_MAX ? ">MAX "
+		: do_big_num((int64)rate, rate_human_level, NULL);
 	rput_progress();
 	rprintf(FCLIENT, "%15sB %3d%% %7sB/s %s%s", human_num(ofs), pct,
-		rate >= INT64_MAX ? ">MAX " : human_num((int64)rate), rembuf, eol);
+		rate_str, rembuf, eol);
 	if (!is_last && !quiet) {
 		output_needs_newline = 1;
 		rflush(FCLIENT);

rsync.1.md

@@ -3613,7 +3613,9 @@ sign) if you want the local shell to expand it.
 
     The default is human-readable level 1.  Each `-h` option increases the
     level by one.  You can take the level down to 0 (to output numbers as pure
-    digits) by specifying the `--no-human-readable` (`--no-h`) option.
+    digits) by specifying the `--no-human-readable` (`--no-h`) option.  Progress
+    rates are an exception: the default and `-h` use level 2, `--no-h` uses
+    level 0 and `-hh` uses level 3.
 
     The unit letters that are appended in levels 2 and 3 are: `K` (kilo), `M`
     (mega), `G` (giga), `T` (tera), `P` (peta) or `E` (exa). For example, a
@@ -3821,13 +3823,17 @@ sign) if you want the local shell to expand it.
     While rsync is transferring a regular file, it updates a progress line that
     looks like this:
 
-    >     782,448B  63% 113,295B/s    0:00:04
+    >     782,448B  63%    113KB/s    0:00:04
 
     In this example, the receiver has reconstructed 782,448 bytes or 63% of the
-    sender's file, which is being reconstructed at a rate of 113,295 bytes
+    sender's file, which is being reconstructed at a rate of 113 kilobytes
     per second, and the transfer will finish in 4 seconds if the current rate
     is maintained until the end.
 
+    Progress rates are scaled in decimal units by default.  `--no-h` keeps the
+    rate in bytes per second, while `-hh` uses binary units such as `KiB/s` and
+    `MiB/s`.
+
     These statistics can be misleading if rsync's delta-transfer algorithm is
     in use.  For example, if the sender's file consists of the basis file
     followed by additional data, the reported rate will probably drop
@@ -3838,10 +3844,10 @@ sign) if you want the local shell to expand it.
     When the file transfer finishes, rsync replaces the progress line with a
     summary line that looks like this:
 
-    >     1,238,099B 100% 149,893B/s    0:00:08  (xfr#5, to-chk=169/396)
+    >     1,238,099B 100%    150KB/s    0:00:08  (xfr#5, to-chk=169/396)
 
     In this example, the file was 1,238,099 bytes long in total, the average
-    rate of transfer for the whole file was 149,893 bytes per second over the
+    rate of transfer for the whole file was 150 kilobytes per second over the
     8 seconds that it took to complete, it was the 5th transfer of a
     regular file during the current rsync session, and there are 169 more files
     for the receiver to check (to see if they are up-to-date or not) remaining

testsuite/output-options_test.py

@@ -95,7 +95,7 @@ def out(*args, want_rc=0, env=None, text=True):
 # --- --progress shows a percentage ------------------------------------------
 rmtree(TODIR)
 p = out('-a', '--progress', f'{src}/', f'{TODIR}/')
-progress_re = r'[\d.,]+B\s+100%\s+[\d.,]+B/s\s+\d+:\d\d:\d\d'
+progress_re = r'[\d.,]+B\s+100%\s+[\d.,]+(?:[KMGTPE]i?)?B/s\s+\d+:\d\d:\d\d'
 if not re.search(progress_re, p.stdout):
     test_fail(f"--progress did not show the expected final shape:\n{p.stdout}")
 
@@ -122,6 +122,17 @@ def out(*args, want_rc=0, env=None, text=True):
 if re.search(suffix_re, plain):
     test_fail(f"--stats without -h unexpectedly used a unit suffix:\n{plain}")
 
+for label, options, progress_re in (
+    ('default', (), r'50[,.]000B\s+100%\s+[\d.,]+[KMGTPE]B/s'),
+    ('--no-h', ('--no-h',), r'50000B\s+100%\s+\d+B/s'),
+    ('-h', ('-h',), r'50\.0KB\s+100%\s+[\d.,]+[KMGTPE]B/s'),
+    ('-hh', ('-hh',), r'48\.8KiB\s+100%\s+[\d.,]+[KMGTPE]iB/s'),
+):
+    rmtree(TODIR)
+    progress = out('-a', '--progress', *options, f'{src}/', f'{TODIR}/').stdout
+    if not re.search(progress_re, progress):
+        test_fail(f"{label} progress output has the wrong units:\n{progress}")
+
 rmtree(TODIR)
 human_binary = out('-a', '-hh', '--stats', f'{src}/', f'{TODIR}/').stdout
 if 'Total file size: 48.8Ki bytes' not in human_binary: