Parse HTTP date in map requests.

3 files changed, 294 insertions, 6 deletions

diff --git a/src/map.c b/src/map.c
index 0950d5682..ea7f0adbb 100644
--- a/src/map.c
+++ b/src/map.c
@@ -305,7 +305,7 @@ read_http_common (struct rspamd_map *map, struct http_map_data *data, struct htt
 {
 	gchar                         *remain, *pos;
 	ssize_t                         r;
-	gchar                           *te;
+	gchar                           *te, *date;
 	
 	if ((r = read (fd, data->read_buf + data->rlen, sizeof (data->read_buf) - data->rlen)) > 0) {
 		r += data->rlen;
@@ -346,6 +346,14 @@ read_http_common (struct rspamd_map *map, struct http_map_data *data, struct htt
 					data->chunked = -1;
 				}
 			}
+			/* Check for date */
+			date = g_hash_table_lookup (reply->headers, "Date");
+			if (date != NULL) {
+				data->last_checked = parse_http_date (date, -1);
+			}
+			else {
+				data->last_checked = (time_t)-1;
+			}
 
 			if (data->chunked > 0) {
 				return read_http_chunked (data->read_buf, r, map, data, cbdata);
@@ -402,7 +410,9 @@ read_http_sync (struct rspamd_map *map, struct http_map_data *data)
 
 	map->fin_callback (map->pool, &cbdata);
 	*map->user_data = cbdata.cur_data;
-	data->last_checked = time (NULL);
+	if (data->last_checked == (time_t)-1) {
+		data->last_checked = time (NULL);
+	}
 
 	g_hash_table_destroy (repl->headers);
 	g_free (repl);
@@ -846,14 +856,18 @@ http_async_callback (gint fd, short what, void *ud)
 			if (!read_http_common (cbd->map, cbd->data, cbd->reply, &cbd->cbdata, cbd->fd)) {
 				/* Handle Not-Modified in a special way */
 				if (cbd->reply->code == 304) {
-					cbd->data->last_checked = time (NULL);
+					if (cbd->data->last_checked == (time_t)-1) {
+						cbd->data->last_checked = time (NULL);
+					}
 					msg_info ("data is not modified for server %s", cbd->data->host);
 				}
 				else if (cbd->cbdata.cur_data != NULL) {
 					/* Destroy old data and start reading request data */
 					cbd->map->fin_callback (cbd->map->pool, &cbd->cbdata);
 					*cbd->map->user_data = cbd->cbdata.cur_data;
-					cbd->data->last_checked = time (NULL);
+					if (cbd->data->last_checked == (time_t)-1) {
+						cbd->data->last_checked = time (NULL);
+					}
 				}
 				if (cbd->state == 1 && cbd->reply->code == 200) {
 					/* Write to log that data is modified */
diff --git a/src/util.c b/src/util.c
index 7757c7ad0..441ce007a 100644
--- a/src/util.c
+++ b/src/util.c
@@ -38,8 +38,6 @@
 /* Default connect timeout for sync sockets */
 #define CONNECT_TIMEOUT 3
 
-static gchar* rspamd_sprintf_num (gchar *buf, gchar *last, guint64 ui64, gchar zero, guint hexadecimal, guint width);
-
 gint
 make_socket_nonblocking (gint fd)
 {
@@ -1860,5 +1858,273 @@ parse_ipmask_v4 (const char *line, struct in_addr *ina, int *mask)
 }
 
 /*
+ * Obtained from nginx
+ * Copyright (C) Igor Sysoev
+ */
+static guint mday[] = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 };
+
+time_t
+parse_http_date (const gchar *header, gsize len)
+{
+	const gchar *p, *end;
+	gint month;
+	guint day, year, hour, min, sec;
+	guint64 time;
+	enum {
+		no = 0, rfc822, /* Tue, 10 Nov 2002 23:50:13   */
+		rfc850, /* Tuesday, 10-Dec-02 23:50:13 */
+		isoc /* Tue Dec 10 23:50:13 2002    */
+	} fmt;
+
+	fmt = 0;
+	if (len > 0) {
+		end = header + len;
+	}
+	else {
+		end = header + strlen (header);
+	}
+
+#if (NGX_SUPPRESS_WARN)
+	day = 32;
+	year = 2038;
+#endif
+
+	for (p = header; p < end; p++) {
+		if (*p == ',') {
+			break;
+		}
+
+		if (*p == ' ') {
+			fmt = isoc;
+			break;
+		}
+	}
+
+	for (p++; p < end; p++)
+		if (*p != ' ') {
+			break;
+		}
+
+	if (end - p < 18) {
+		return (time_t)-1;
+	}
+
+	if (fmt != isoc) {
+		if (*p < '0' || *p > '9' || *(p + 1) < '0' || *(p + 1) > '9') {
+			return (time_t)-1;
+		}
+
+		day = (*p - '0') * 10 + *(p + 1) - '0';
+		p += 2;
+
+		if (*p == ' ') {
+			if (end - p < 18) {
+				return (time_t)-1;
+			}
+			fmt = rfc822;
+
+		}
+		else if (*p == '-') {
+			fmt = rfc850;
+
+		}
+		else {
+			return (time_t)-1;
+		}
+
+		p++;
+	}
+
+	switch (*p) {
+
+	case 'J':
+		month = *(p + 1) == 'a' ? 0 : *(p + 2) == 'n' ? 5 : 6;
+		break;
+
+	case 'F':
+		month = 1;
+		break;
+
+	case 'M':
+		month = *(p + 2) == 'r' ? 2 : 4;
+		break;
+
+	case 'A':
+		month = *(p + 1) == 'p' ? 3 : 7;
+		break;
+
+	case 'S':
+		month = 8;
+		break;
+
+	case 'O':
+		month = 9;
+		break;
+
+	case 'N':
+		month = 10;
+		break;
+
+	case 'D':
+		month = 11;
+		break;
+
+	default:
+		return (time_t)-1;
+	}
+
+	p += 3;
+
+	if ((fmt == rfc822 && *p != ' ') || (fmt == rfc850 && *p != '-')) {
+		return (time_t)-1;
+	}
+
+	p++;
+
+	if (fmt == rfc822) {
+		if (*p < '0' || *p > '9' || *(p + 1) < '0' || *(p + 1) > '9'
+		|| *(p + 2) < '0' || *(p + 2) > '9' || *(p + 3) < '0'
+		|| *(p + 3) > '9') {
+			return (time_t)-1;
+		}
+
+		year = (*p - '0') * 1000 + (*(p + 1) - '0') * 100
+				+ (*(p + 2) - '0') * 10 + *(p + 3) - '0';
+		p += 4;
+
+	}
+	else if (fmt == rfc850) {
+		if (*p < '0' || *p > '9' || *(p + 1) < '0' || *(p + 1) > '9') {
+			return (time_t)-1;
+		}
+
+		year = (*p - '0') * 10 + *(p + 1) - '0';
+		year += (year < 70) ? 2000 : 1900;
+		p += 2;
+	}
+
+	if (fmt == isoc) {
+		if (*p == ' ') {
+			p++;
+		}
+
+		if (*p < '0' || *p > '9') {
+			return (time_t)-1;
+		}
+
+		day = *p++ - '0';
+
+		if (*p != ' ') {
+			if (*p < '0' || *p > '9') {
+				return (time_t)-1;
+			}
+
+			day = day * 10 + *p++ - '0';
+		}
+
+		if (end - p < 14) {
+			return (time_t)-1;
+		}
+	}
+
+	if (*p++ != ' ') {
+		return (time_t)-1;
+	}
+
+	if (*p < '0' || *p > '9' || *(p + 1) < '0' || *(p + 1) > '9') {
+		return (time_t)-1;
+	}
+
+	hour = (*p - '0') * 10 + *(p + 1) - '0';
+	p += 2;
+
+	if (*p++ != ':') {
+		return (time_t)-1;
+	}
+
+	if (*p < '0' || *p > '9' || *(p + 1) < '0' || *(p + 1) > '9') {
+		return (time_t)-1;
+	}
+
+	min = (*p - '0') * 10 + *(p + 1) - '0';
+	p += 2;
+
+	if (*p++ != ':') {
+		return (time_t)-1;
+	}
+
+	if (*p < '0' || *p > '9' || *(p + 1) < '0' || *(p + 1) > '9') {
+		return (time_t)-1;
+	}
+
+	sec = (*p - '0') * 10 + *(p + 1) - '0';
+
+	if (fmt == isoc) {
+		p += 2;
+
+		if (*p++ != ' ') {
+			return (time_t)-1;
+		}
+
+		if (*p < '0' || *p > '9' || *(p + 1) < '0' || *(p + 1) > '9'
+		|| *(p + 2) < '0' || *(p + 2) > '9' || *(p + 3) < '0'
+		|| *(p + 3) > '9') {
+			return (time_t)-1;
+		}
+
+		year = (*p - '0') * 1000 + (*(p + 1) - '0') * 100
+				+ (*(p + 2) - '0') * 10 + *(p + 3) - '0';
+	}
+
+	if (hour > 23 || min > 59 || sec > 59) {
+		return (time_t)-1;
+	}
+
+	if (day == 29 && month == 1) {
+		if ((year & 3) || ((year % 100 == 0) && (year % 400) != 0)) {
+			return (time_t)-1;
+		}
+
+	}
+	else if (day > mday[month]) {
+		return (time_t)-1;
+	}
+
+	/*
+	 * shift new year to March 1 and start months from 1 (not 0),
+	 * it is needed for Gauss' formula
+	 */
+
+	if (--month <= 0) {
+		month += 12;
+		year -= 1;
+	}
+
+	/* Gauss' formula for Gregorian days since March 1, 1 BC */
+
+	time = (guint64) (
+			/* days in years including leap years since March 1, 1 BC */
+
+			365 * year + year / 4 - year / 100 + year / 400
+
+			/* days before the month */
+
+			+ 367 * month / 12 - 30
+
+			/* days before the day */
+
+			+ day - 1
+
+			/*
+			 * 719527 days were between March 1, 1 BC and March 1, 1970,
+			 * 31 and 28 days were in January and February 1970
+			 */
+
+			- 719527 + 31 + 28) * 86400 + hour * 3600 + min * 60 + sec;
+
+	return (time_t) time;
+}
+
+/*
  * vi:ts=4
  */
diff --git a/src/util.h b/src/util.h
index 228fb7445..b1e07c538 100644
--- a/src/util.h
+++ b/src/util.h
@@ -400,4 +400,12 @@ gpointer rspamd_str_pool_copy (gconstpointer data, gpointer ud);
  */
 gboolean parse_ipmask_v4 (const char *line, struct in_addr *ina, int *mask);
 
+/**
+ * Parse HTTP date header and return it as time_t
+ * @param header HTTP date header
+ * @param len length of header
+ * @return time_t or (time_t)-1 in case of error
+ */
+time_t parse_http_date (const gchar *header, gsize len);
+
 #endif