git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH v4] Refactor recv_sideband()

From
Junio C Hamano <gitster@pobox.com>
Date
Jun 29, 2016, 03:00 UTC
Message-ID
<xmqqshvw66oc.fsf@gitster.mtv.corp.google.com>
In-Reply-To
<xmqq8txp6icn.fsf@gitster.mtv.corp.google.com>
Junio C Hamano <gitster@pobox.com> writes:
> It's just that if you take the latter, then the conditional after
> the loop exits (i.e. the last transmission was an incomplete line)
> cannot be "is outbuf empty?", as your base state is "has PREFIX and
> can never be empty".  I was working back from that if statement.
Let's try this again.  How does this look?
In this version:
 - "outbuf" is where we keep the (possibly partial) data collected
   to be eventually shown;
 - output of pending (possibly partial) data is handled by a helper
   function drain().  It is responsible for prepending of the
   PREFIX, which is treated purely as a cosmetic thing.  It also is
   responsible for completing an incomplete line at the end of the
   transmission (e.g. flushing of the buffered input upon reception of
   the emergency exit packet).
 - locally generated errors go directly to fprintf(stderr),
   bypassing outbuf (hence drain()).
 sideband.c | 43 +++++++++++++++++++++++++------------------
 1 file changed, 25 insertions(+), 18 deletions(-)
diff --git a/sideband.c b/sideband.c
index 226a8c2..6873137 100644
--- a/sideband.c
+++ b/sideband.c
@@ -18,6 +18,16 @@
 #define ANSI_SUFFIX "\033[K"
 #define DUMB_SUFFIX "        "
 
+static void drain(struct strbuf *outbuf)
+{
+	if (!outbuf->len)
+		return;
+	strbuf_splice(outbuf, 0, 0, PREFIX, strlen(PREFIX));
+	strbuf_complete_line(outbuf);
+	fwrite(outbuf->buf, 1, outbuf->len, stderr);
+	strbuf_reset(outbuf);
+}
+
 int recv_sideband(const char *me, int in_stream, int out)
 {
 	const char *term, *suffix;
@@ -26,20 +36,21 @@ int recv_sideband(const char *me, int in_stream, int out)
 	const char *b, *brk;
 	int retval = 0;
 
-	strbuf_addf(&outbuf, "%s", PREFIX);
 	term = getenv("TERM");
 	if (isatty(2) && term && strcmp(term, "dumb"))
 		suffix = ANSI_SUFFIX;
 	else
 		suffix = DUMB_SUFFIX;
 
-	while (retval == 0) {
+	while (!retval) {
 		int band, len;
 		len = packet_read(in_stream, NULL, NULL, buf, LARGE_PACKET_MAX, 0);
 		if (len == 0)
 			break;
 		if (len < 1) {
-			fprintf(stderr, "%s: protocol error: no band designator\n", me);
+			drain(&outbuf);
+			fprintf(stderr, "%s: protocol error: no band designator\n",
+				me);
 			retval = SIDEBAND_PROTOCOL_ERROR;
 			break;
 		}
@@ -48,7 +59,8 @@ int recv_sideband(const char *me, int in_stream, int out)
 		len--;
 		switch (band) {
 		case 3:
-			fprintf(stderr, "%s%s\n", PREFIX, buf + 1);
+			drain(&outbuf);
+			strbuf_addf(&outbuf, "%s\n", buf + 1);
 			retval = SIDEBAND_REMOTE_ERROR;
 			break;
 		case 2:
@@ -58,13 +70,12 @@ int recv_sideband(const char *me, int in_stream, int out)
 			 * Append a suffix to each nonempty line to clear the
 			 * end of the screen line.
 			 *
-			 * The output is accumulated in a buffer and each line
-			 * is printed to stderr using fprintf() with a single
-			 * conversion specifier. This is a "best effort"
-			 * approach to supporting both inter-process atomicity
-			 * (single conversion specifiers are likely to end up
-			 * in single atomic write() system calls) and the ANSI
-			 * control code emulation under Windows.
+			 * The output is accumulated in a buffer and
+			 * each line is printed to stderr using
+			 * fwrite(3).  This is a "best effort"
+			 * approach to support inter-process atomicity
+			 * (single fwrite(3) call is likely to end up
+			 * in single atomic write() system calls).
 			 */
 			while ((brk = strpbrk(b, "\n\r"))) {
 				int linelen = brk - b;
@@ -75,11 +86,7 @@ int recv_sideband(const char *me, int in_stream, int out)
 				} else {
 					strbuf_addf(&outbuf, "%c", *brk);
 				}
-				fprintf(stderr, "%.*s", (int)outbuf.len,
-					outbuf.buf);
-				strbuf_reset(&outbuf);
-				strbuf_addf(&outbuf, "%s", PREFIX);
-
+				drain(&outbuf);
 				b = brk + 1;
 			}
 
@@ -90,6 +97,7 @@ int recv_sideband(const char *me, int in_stream, int out)
 			write_or_die(out, buf + 1, len);
 			break;
 		default:
+			drain(&outbuf);
 			fprintf(stderr, "%s: protocol error: bad band #%d\n",
 				me, band);
 			retval = SIDEBAND_PROTOCOL_ERROR;
@@ -97,8 +105,7 @@ int recv_sideband(const char *me, int in_stream, int out)
 		}
 	}
 
-	if (outbuf.len > 0)
-		fprintf(stderr, "%.*s", (int)outbuf.len, outbuf.buf);
+	drain(&outbuf);
 	strbuf_release(&outbuf);
 	return retval;
 }
Previous: Junio C HamanoNext: Nicolas Pitre
Message 45 of 60 in “Refactor recv_sideband()”
  1. Refactor recv_sideband()Lukas Fleischer, Jun 13, 2016
  2. Nicolas PitreJun 13, 2016
  3. Johannes SchindelinJun 14, 2016
  4. Nicolas PitreJun 14, 2016
  5. Johannes SchindelinJun 14, 2016
  6. Nicolas PitreJun 14, 2016
  7. Refactor recv_sideband()Lukas Fleischer, Jun 14, 2016
  8. Lukas FleischerJun 14, 2016
  9. Junio C HamanoJun 14, 2016
  10. Jeff KingJun 15, 2016
  11. Jeff KingJun 24, 2016
  12. Johannes SchindelinJun 24, 2016
  13. Jeff KingJun 24, 2016
  14. Junio C HamanoJun 24, 2016
  15. Lukas FleischerJun 27, 2016
  16. Junio C HamanoJun 27, 2016
  17. Jeff KingJun 27, 2016
  18. Junio C HamanoJun 27, 2016
  19. Lukas FleischerJun 27, 2016
  20. Nicolas PitreJun 27, 2016
  21. Lukas FleischerJun 28, 2016
  22. Junio C HamanoJun 28, 2016
  23. Johannes SchindelinJun 28, 2016
  24. Johannes SchindelinJun 28, 2016
  25. Junio C HamanoJun 28, 2016
  26. Johannes SchindelinJun 28, 2016
  27. Dennis KaarsemakerJun 24, 2016
  28. Refactor recv_sideband()Lukas Fleischer, Jun 22, 2016
  29. Nicolas PitreJun 22, 2016
  30. Nicolas PitreJun 22, 2016
  31. Lukas FleischerJun 23, 2016
  32. Nicolas PitreJun 23, 2016
  33. Refactor recv_sideband()Lukas Fleischer, Jun 28, 2016
  34. Junio C HamanoJun 28, 2016
  35. Junio C HamanoJun 28, 2016
  36. Nicolas PitreJun 28, 2016
  37. Junio C HamanoJun 28, 2016
  38. Nicolas PitreJun 28, 2016
  39. Junio C HamanoJun 28, 2016
  40. Nicolas PitreJun 28, 2016
  41. Junio C HamanoJun 28, 2016
  42. Nicolas PitreJun 28, 2016
  43. Junio C HamanoJun 28, 2016
  44. Junio C HamanoJun 28, 2016
  45. Junio C HamanoJun 29, 2016
  46. Nicolas PitreJun 29, 2016
  47. Nicolas PitreJun 29, 2016
  48. Junio C HamanoJun 29, 2016
  49. Lukas FleischerJun 30, 2016
  50. Junio C HamanoJul 1, 2016
  51. Nicolas PitreJul 5, 2016
  52. Junio C HamanoJul 6, 2016
  53. Nicolas PitreJul 7, 2016
  54. Nicolas PitreJun 14, 2016
  55. Nicolas PitreJun 14, 2016
  56. Junio C HamanoJun 14, 2016
  57. Lukas FleischerJun 14, 2016
  58. Junio C HamanoJun 14, 2016
  59. Nicolas PitreJun 14, 2016
  60. Lukas FleischerJun 19, 2016

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.