else
if u:get_user() then
table.insert(res, u)
- elseif u:is_subject() then
+ elseif u:is_subject() or u:is_phished() then
table.insert(res, u)
end
end
* @return {boolean},{rspamd_task|error} status + new task or error message
*/
LUA_FUNCTION_DEF (task, load_from_file);
+/***
+ * @function rspamd_task.load_from_string(message[, cfg])
+ * Loads a message from specific file
+ * @return {boolean},{rspamd_task|error} status + new task or error message
+ */
+LUA_FUNCTION_DEF (task, load_from_string);
LUA_FUNCTION_DEF (task, get_message);
/***
static const struct luaL_reg tasklib_f[] = {
LUA_INTERFACE_DEF (task, load_from_file),
+ LUA_INTERFACE_DEF (task, load_from_string),
{NULL, NULL}
};
}
}
-static int
+static gint
lua_task_load_from_file (lua_State * L)
{
struct rspamd_task *task = NULL, **ptask;
return 2;
}
+static gint
+lua_task_load_from_string (lua_State * L)
+{
+ struct rspamd_task *task = NULL, **ptask;
+ const gchar *str_message = luaL_checkstring (L, 1), *err = NULL;
+ gsize message_len = lua_strlen(L, 1);
+ struct rspamd_config *cfg = NULL;
+ gboolean res = FALSE;
+
+ if (str_message) {
+
+ if (lua_type (L, 2) == LUA_TUSERDATA) {
+ gpointer p;
+ p = rspamd_lua_check_udata_maybe (L, 2, "rspamd{config}");
+
+ if (p) {
+ cfg = *(struct rspamd_config **)p;
+ }
+ }
+
+ task = rspamd_task_new (NULL, cfg, NULL, NULL);
+ task->msg.begin = str_message;
+ task->msg.len = message_len;
+ rspamd_mempool_add_destructor (task->task_pool,
+ lua_task_unmap_dtor, task);
+ res = TRUE;
+ }
+ else {
+ return luaL_error (L, "invalid arguments");
+ }
+
+ lua_pushboolean (L, res);
+
+ if (res) {
+ ptask = lua_newuserdata (L, sizeof (*ptask));
+ *ptask = task;
+ rspamd_lua_setclass (L, "rspamd{task}", -1);
+ }
+ else {
+ if (err) {
+ lua_pushstring (L, err);
+ }
+ else {
+ lua_pushnil (L);
+ }
+ }
+
+ return 2;
+}
+
static int
lua_task_get_mempool (lua_State * L)
{
*/
LUA_FUNCTION_DEF (util, load_rspamd_config);
/***
- * @function util.config_from_ucl(any)
+ * @function util.config_from_ucl(any, string)
* Load rspamd config from ucl reperesented by any lua table
* @return {confg} new configuration object suitable for access
*/
return 1;
}
+static gint
+parse_config_options(const char *str_options)
+{
+ gint ret = 0;
+ gchar **vec;
+ const gchar *str;
+ guint i, l;
+
+ vec = g_strsplit_set (str_options, ",;", -1);
+ if (vec) {
+ l = g_strv_length (vec);
+ for (i = 0; i < l; i ++) {
+ str = vec[i];
+
+ if (g_ascii_strcasecmp (str, "INIT_URL") == 0) {
+ ret |= RSPAMD_CONFIG_INIT_URL;
+ } else if (g_ascii_strcasecmp (str, "INIT_LIBS") == 0) {
+ ret |= RSPAMD_CONFIG_INIT_LIBS;
+ } else if (g_ascii_strcasecmp (str, "INIT_SYMCACHE") == 0) {
+ ret |= RSPAMD_CONFIG_INIT_SYMCACHE;
+ } else if (g_ascii_strcasecmp (str, "INIT_VALIDATE") == 0) {
+ ret |= RSPAMD_CONFIG_INIT_VALIDATE;
+ } else if (g_ascii_strcasecmp (str, "INIT_NO_TLD") == 0) {
+ ret |= RSPAMD_CONFIG_INIT_NO_TLD;
+ } else if (g_ascii_strcasecmp (str, "INIT_PRELOAD_MAPS") == 0) {
+ ret |= RSPAMD_CONFIG_INIT_PRELOAD_MAPS;
+ } else {
+ msg_warn ("bad type: %s", str);
+ }
+ }
+
+ g_strfreev (vec);
+ }
+
+ return ret;
+}
+
static gint
lua_util_config_from_ucl (lua_State *L)
{
- struct rspamd_config *cfg, **pcfg;
+ struct rspamd_config *cfg = NULL, **pcfg;
struct rspamd_rcl_section *top;
GError *err = NULL;
ucl_object_t *obj;
+ const char *str_options = NULL;
+ gint int_options = 0;
+
obj = ucl_object_lua_import (L, 1);
+ if (lua_gettop (L) == 2) {
+ if (lua_type (L, 2) == LUA_TSTRING) {
+ str_options = lua_tostring (L, 2);
+ int_options = parse_config_options(str_options);
+ }
+ else {
+ msg_err_config ("config_from_ucl: second parameter is expected to be string");
+ ucl_object_unref (obj);
+ lua_pushnil (L);
+ }
+ }
if (obj) {
- cfg = g_malloc0 (sizeof (struct rspamd_config));
cfg = rspamd_config_new (RSPAMD_CONFIG_INIT_SKIP_LUA);
cfg->lua_state = L;
lua_pushnil (L);
}
else {
- rspamd_config_post_load (cfg, 0);
+ rspamd_config_post_load (cfg, int_options);
pcfg = lua_newuserdata (L, sizeof (struct rspamd_config *));
rspamd_lua_setclass (L, "rspamd{config}", -1);
*pcfg = cfg;
local url = require "rspamd_url"
local logger = require "rspamd_logger"
local ffi = require "ffi"
+ local rspamd_util = require "rspamd_util"
+ local rspamd_task = require "rspamd_task"
ffi.cdef[[
void rspamd_url_init (const char *tld_file);
esld_limit = 2,
need_emails = true,
prefix = 'p'
+ },
+ {
+ input = {"abc@a.google.com", "b.google.com", "c.google.com", "a.net", "bb.net", "a.bb.net", "b.bb.net"},
+ expect = {"abc@a.google.com", "a.bb.net", "b.google.com", "a.net", "bb.net", "abc@a.google.com"},
+ filter = nil,
+ limit = 9999,
+ esld_limit = 2,
+ need_emails = true,
+ prefix = 'p'
}
}
+ local function prepare_actual_result(actual)
+ return fun.totable(fun.map(
+ function(u) return u:get_raw():gsub('^%w+://', '') end,
+ actual
+ ))
+ end
+
local pool = mpool.create()
for i,c in ipairs(cases) do
local function prepare_url_list(c)
return fun.totable(fun.map(
- function (u) return url.create(pool, u) end,
- c.input or url_list
- ))
- end
+ function (u) return url.create(pool, u) end,
+ c.input or url_list
+ ))
+ end
test("extract_specific_urls, backward compatibility case #" .. i, function()
task_object.urls = prepare_url_list(c)
end
local actual = util.extract_specific_urls(task_object, c.limit, c.need_emails, c.filter, c.prefix)
- local actual_result = fun.totable(fun.map(
- function(u) return u:get_host() end,
- actual
- ))
+ local actual_result = prepare_actual_result(actual)
--[[
local s = logger.slog("%1 =?= %2", c.expect, actual_result)
prefix = c.prefix,
})
- local actual_result = fun.totable(fun.map(
- function(u) return u:get_host() end,
- actual
- ))
+ local actual_result = prepare_actual_result(actual)
--[[
local s = logger.slog("case[%1] %2 =?= %3", i, c.expect, actual_result)
end)
end
+
+--[[ ******************* kinda functional *************************************** ]]
+ local test_dir = string.gsub(debug.getinfo(1).source, "^@(.+/)[^/]+$", "%1")
+ local tld_file = string.format('%s/%s', test_dir, "test_tld.dat")
+
+ local config = {
+ options = {
+ filters = {'spf', 'dkim', 'regexp'},
+ url_tld = tld_file,
+ dns = {
+ nameserver = {'8.8.8.8'}
+ },
+ },
+ logging = {
+ type = 'console',
+ level = 'debug'
+ },
+ metric = {
+ name = 'default',
+ actions = {
+ reject = 100500,
+ },
+ unknown_weight = 1
+ }
+ }
+
+ test("extract_specific_urls - from email", function()
+ local cfg = rspamd_util.config_from_ucl(config, "INIT_URL,INIT_LIBS,INIT_SYMCACHE,INIT_VALIDATE,INIT_PRELOAD_MAPS")
+ assert_not_nil(cfg)
+
+ local msg = [[
+From: <>
+To: <nobody@example.com>
+Subject: test
+Content-Type: multipart/alternative;
+ boundary="_000_6be055295eab48a5af7ad4022f33e2d0_"
+
+--_000_6be055295eab48a5af7ad4022f33e2d0_
+Content-Type: text/plain; charset="utf-8"
+Content-Transfer-Encoding: base64
+
+Hello world
+
+
+--_000_6be055295eab48a5af7ad4022f33e2d0_
+Content-Type: text/html; charset="utf-8"
+
+<html><body>
+<a href="http://example.net">http://example.net</a>
+<a href="http://example1.net">http://example1.net</a>
+<a href="http://example2.net">http://example2.net</a>
+<a href="http://example3.net">http://example3.net</a>
+<a href="http://example4.net">http://example4.net</a>
+<a href="http://domain1.com">http://domain1.com</a>
+<a href="http://domain2.com">http://domain2.com</a>
+<a href="http://domain3.com">http://domain3.com</a>
+<a href="http://domain4.com">http://domain4.com</a>
+<a href="http://domain5.com">http://domain5.com</a>
+<a href="http://domain.com">http://example.net/</a>
+</html>
+]]
+ local expect = {"example.net", "domain.com"}
+ local res,task = rspamd_task.load_from_string(msg, rspamd_config)
+
+ if not res then
+ assert_true(false, "failed to load message")
+ end
+
+ if not task:process_message() then
+ assert_true(false, "failed to process message")
+ end
+
+ local actual = util.extract_specific_urls({
+ task = task,
+ limit = 2,
+ esld_limit = 2,
+ })
+
+ local actual_result = prepare_actual_result(actual)
+
+ --[[
+ local s = logger.slog("case[%1] %2 =?= %3", i, expect, actual_result)
+ print(s) --]]
+
+ assert_equal("domain.com", actual_result[1], "checking that first url is the one with highest suspiciousness level")
+
+ end)
end)
\ No newline at end of file