26.10.2013 13:17, Pavel Levshin:
Is it declared somewhere which integer size rsyslog does support?
There is a inconsistence between internal number representation, which
is signed long long, and JSON-c integers, which is signed int only. So
you can use integers in range -2^63 .. 2^63-1 for RainerScript
expressions, but variables can hold no more than -2^31 .. 2^31-1.
This appears to be easy to fix. Json-c supports int64 from and uses them
internally in any case.
Note that this patch requires json-c >= 0.10.
Ommongodb seems to be prepared for int64, but I haven't tested it. Other
modules, notably mmcount, left unchanged.
After the patch:
set $!min = -2147483648;
set $!max = 2147483647;
set $!under = -2147483648 - 1;
set $!over = 2147483647 + 1;
set $!under_s = cstr(-2147483648 - 1);
set $!over_s = cstr(2147483647 + 1);
"min": -2147483648, "max": 2147483647, "under": -2147483649, "over":
2147483648, "under_s": "-2147483650", "over_s": "2147483649"
set $!under_sj = "-1000000000000";
set $!over_sj = "1000000000000";
set $!under_sjn = cstr(cnum($!under_sj));
set $!over_sjn = cstr(cnum($!over_sj));
set $!under_sn = cstr(cnum("-1000000000000"));
set $!over_sn = cstr(cnum("-1000000000000"));
"under_sj": "-1000000000000", "over_sj": "1000000000000", "under_sjn":
"-1000000000000", "over_sjn": "1000000000000", "under_sn":
"-1000000000000", "over_sn": "-1000000000000"
--
Pavel Levshin
>From 4d7951f4b080656f6af155b5a396d13e8a60c799 Mon Sep 17 00:00:00 2001
From: Pavel Levshin <[email protected]>
Date: Sat, 26 Oct 2013 14:57:37 +0400
Subject: [PATCH] Make use of int64 json numbers
---
configure.ac | 4 ++--
grammar/rainerscript.c | 2 +-
plugins/ommongodb/ommongodb.c | 3 +--
runtime/msg.c | 4 ++--
4 files changed, 6 insertions(+), 7 deletions(-)
diff --git a/configure.ac b/configure.ac
index 2f23147..108520d 100644
--- a/configure.ac
+++ b/configure.ac
@@ -33,8 +33,8 @@ PKG_PROG_PKG_CONFIG
# modules we require
PKG_CHECK_MODULES(LIBESTR, libestr >= 0.1.8)
-PKG_CHECK_MODULES([JSON_C], [json],, [
- PKG_CHECK_MODULES([JSON_C], [json-c])
+PKG_CHECK_MODULES([JSON_C], [json >= 0.10],, [
+ PKG_CHECK_MODULES([JSON_C], [json-c >= 0.10])
])
case "${host}" in
diff --git a/grammar/rainerscript.c b/grammar/rainerscript.c
index 37f3509..6dc6eb1 100644
--- a/grammar/rainerscript.c
+++ b/grammar/rainerscript.c
@@ -1192,7 +1192,7 @@ var2Number(struct var *r, int *bSuccess)
n = es_str2num(r->d.estr, bSuccess);
} else {
if(r->datatype == 'J') {
- n = (r->d.json == NULL) ? 0 :
json_object_get_int(r->d.json);
+ n = (r->d.json == NULL) ? 0 :
json_object_get_int64(r->d.json);
} else {
n = r->d.n;
}
diff --git a/plugins/ommongodb/ommongodb.c b/plugins/ommongodb/ommongodb.c
index 78781c0..afd2d35 100644
--- a/plugins/ommongodb/ommongodb.c
+++ b/plugins/ommongodb/ommongodb.c
@@ -317,8 +317,7 @@ BSONAppendJSONObject(bson *doc, const gchar *name, struct
json_object *json)
case json_type_int: {
int64_t i;
- /* FIXME: the future version will have get_int64 */
- i = json_object_get_int(json);
+ i = json_object_get_int64(json);
if (i >= INT32_MIN && i <= INT32_MAX)
return bson_append_int32(doc, name, i);
else
diff --git a/runtime/msg.c b/runtime/msg.c
index d173644..7a8045b 100644
--- a/runtime/msg.c
+++ b/runtime/msg.c
@@ -4049,7 +4049,7 @@ jsonDeepCopy(struct json_object *src)
dst = json_object_new_double(json_object_get_double(src));
break;
case json_type_int:
- dst = json_object_new_int(json_object_get_int(src));
+ dst = json_object_new_int64(json_object_get_int64(src));
break;
case json_type_string:
dst = json_object_new_string(json_object_get_string(src));
@@ -4092,7 +4092,7 @@ msgSetJSONFromVar(msg_t *pMsg, uchar *varname, struct var
*v)
free(cstr);
break;
case 'N':/* number (integer) */
- json = json_object_new_int((int) v->d.n);
+ json = json_object_new_int64(v->d.n);
break;
case 'J':/* native JSON */
json = jsonDeepCopy(v->d.json);
--
1.7.9.5
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? Follow https://twitter.com/rgerhards
NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of
sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE
THAT.
