Skip to contentSkip to conversation

Subject: RFC Migrate from RFC2822 to ISO8601 for time representations

PATCH Olivier Le Thanh Duong 3 participants 4 messages Oct 18, 2009 – Oct 22, 2009

lore
Share & tools
lore
#0 lore raw

Hi,

the format used to represent time in ofono is currently RFC2822 which is complex and a bit deprecated. I think it would be better to move to ISO8601 which is a well used and supported [1] format and simpler to parse.

RFC2822 date example: Mon, 05 Oct 2009 19:31:32 +0200 ISO8601 date example: 2009-10-05T19:31:26+0200

The proposed patch makes this transition.

Any comments?

Regards,
Olivier

[1] for example Glib has a g_time_val_from_iso8601 which understand timezone compared to strptime which doesn't support %z

From 16359eccfcb4a7a5444e6f87729008192edefc5e Mon Sep 17 00:00:00 2001
From: Olivier Le Thanh Duong <[email protected]>
Date: Mon, 5 Oct 2009 19:10:25 +0200
Subject: [PATCH] Migrate from RFC2822 to ISO8601 for time representations

---
  plugins/example_history.c |    6 +++---
  src/sms.c                 |    4 ++--
  src/voicecall.c           |    2 +-
  unit/test-sms.c           |    4 ++--
  4 files changed, 8 insertions(+), 8 deletions(-)
plugins/example_history.c +0 -0
index 384c88d..d923a8e 100644
@@ -68,11 +68,11 @@ static void example_history_call_ended(struct

ofono_history_context *context,
else
ofono_debug("From: %s", from);

- strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", localtime(&start));
+ strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", localtime(&start));
buf[127] = '\0';
ofono_debug("StartTime: %s", buf);

- strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", localtime(&end));
+ strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", localtime(&end));
buf[127] = '\0';
ofono_debug("EndTime: %s", buf);
}
@@ -96,7 +96,7 @@ static void example_history_call_missed(struct ofono_history_context *context,
from = phone_number_to_string(&call->phone_number);

ofono_debug("From: %s", from);
- strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", localtime(&when));
+ strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", localtime(&when));
buf[127] = '\0';
ofono_debug("When: %s", buf);
}

src/sms.c +0 -0
index 141f288..e02ac85 100644
@@ -467,11 +467,11 @@ static void dispatch_text_message(struct ofono_sms

*sms,

  	ts = sms_scts_to_time(scts, &remote);

- strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", localtime(&ts));
+ strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", localtime(&ts));
buf[127] = '\0';
ofono_dbus_dict_append(&dict, "LocalSentTime", DBUS_TYPE_STRING, &str);

- strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", &remote);
+ strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", &remote);
buf[127] = '\0';
ofono_dbus_dict_append(&dict, "SentTime", DBUS_TYPE_STRING, &str);

src/voicecall.c +1 -1
index eff6321..3f70f9e 100644
@@ -164,7 +164,7 @@ static const char *time_to_str(const time_t *t)
  {
  	static char buf[128];
 
-	strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", localtime(t));
+	strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", localtime(t));
  	buf[127] = '\0';
 
  	return buf;
unit/test-sms.c +0 -0
index ac4f1ff..9c1528c 100644
@@ -59,12 +59,12 @@ static void print_scts(struct sms_scts *scts, const

char *prefix)

  	ts = sms_scts_to_time(scts, &remote);

- strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", localtime(&ts));
+ strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", localtime(&ts));
buf[127] = '\0';

  	g_print("local time: %s\n", buf);

- strftime(buf, 127, "%a, %d %b %Y %H:%M:%S %z", &remote);
+ strftime(buf, 127, "%Y-%m-%dT%H:%M:%S%z", &remote);
buf[127] = '\0';

g_print("remote time: %s\n", buf);

Signature · 1 line
1.6.3.3
#1 lore raw

Hi Olivier,

Quoted 3 lines

the format used to represent time in ofono is currently RFC2822 which is complex and a bit deprecated. I think it would be better to move to ISO8601 which is a well used and supported [1] format and simpler to parse.

So I don't have a problem with the patch, since RFC822 does indeed have issues with i18n. However, this will break existing applications which will need to be updated to the new format.

If no one objects in the next 24 hours or so I will apply this one.

Regards,
-Denis

#2 lore raw

Hi Denis,

Quoted 9 lines
Quoted 3 lines

the format used to represent time in ofono is currently RFC2822 which is complex and a bit deprecated. I think it would be better to move to ISO8601 which is a well used and supported [1] format and simpler to parse.

So I don't have a problem with the patch, since RFC822 does indeed have issues with i18n. However, this will break existing applications which will need to be updated to the new format.

If no one objects in the next 24 hours or so I will apply this one.

go ahead and apply the patch. Lets break the apps now. I am perfectly fine with this patch and it makes sense.

Regards

Marcel