From 5495be2784547be3d9cde09177186199087c7977 Mon Sep 17 00:00:00 2001
From: wangmm0220 <wangmm0220@gmail.com>
Date: Tue, 19 Oct 2021 13:21:47 +0800
Subject: [PATCH] TD-6129<feature> add taos json key hash free function

---
 src/client/src/tscUtil.c        |  8 +++++++-
 src/util/inc/tutil.h            |  1 +
 src/util/src/tutil.c            | 11 +++++++++++
 tests/pytest/stable/json_tag.py |  9 +++++++++
 4 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/src/client/src/tscUtil.c b/src/client/src/tscUtil.c
index efe52ac8dc..f963a1e6a2 100644
--- a/src/client/src/tscUtil.c
+++ b/src/client/src/tscUtil.c
@@ -5285,6 +5285,7 @@ end:
 }
 
 int parseJsontoTagData(char* json, SKVRowBuilder* kvRowBuilder, char* errMsg, int16_t startColId){
+  // set json NULL data
   uint8_t nullTypeVal[CHAR_BYTES + VARSTR_HEADER_SIZE + CHAR_BYTES] = {0};
   uint8_t jsonNULL = TSDB_DATA_JSON_NULL;
   int jsonIndex = startColId + 1;
@@ -5302,6 +5303,7 @@ int parseJsontoTagData(char* json, SKVRowBuilder* kvRowBuilder, char* errMsg, in
   *(uint8_t*)(varDataVal(nullTypeVal + CHAR_BYTES)) = jsonNotNull;
   tdAddColToKVRow(kvRowBuilder, jsonIndex++, TSDB_DATA_TYPE_NCHAR, &nullTypeVal, true);   // add json type
 
+  // set json real data
   cJSON *root = cJSON_Parse(json);
   if (root == NULL){
     tscError("json parse error");
@@ -5329,7 +5331,11 @@ int parseJsontoTagData(char* json, SKVRowBuilder* kvRowBuilder, char* errMsg, in
     if (item->type == cJSON_NULL){
       continue;
     }
-
+    if(!isValidateTag(jsonKey)){
+      tscError("json key not validate");
+      retCode =  tscSQLSyntaxErrMsg(errMsg, "json key not validate", NULL);
+      goto end;
+    }
     if(strlen(jsonKey) > TSDB_MAX_JSON_KEY_LEN){
       tscError("json key too long error");
       retCode =  tscSQLSyntaxErrMsg(errMsg, "json key too long, more than 256", NULL);
diff --git a/src/util/inc/tutil.h b/src/util/inc/tutil.h
index 43382a4856..3eb675f569 100644
--- a/src/util/inc/tutil.h
+++ b/src/util/inc/tutil.h
@@ -46,6 +46,7 @@ int  taosCheckVersion(char *input_client_version, char *input_server_version, in
 char *   taosIpStr(uint32_t ipInt);
 uint32_t ip2uint(const char *const ip_addr);
 void jsonKeyMd5(void *pMsg, int msgLen, void *pKey);
+bool isValidateTag(char *input);
 
 static FORCE_INLINE void taosEncryptPass(uint8_t *inBuf, size_t inLen, char *target) {
   MD5_CTX context;
diff --git a/src/util/src/tutil.c b/src/util/src/tutil.c
index 2def2be8a4..441ee74e16 100644
--- a/src/util/src/tutil.c
+++ b/src/util/src/tutil.c
@@ -466,6 +466,17 @@ void jsonKeyMd5(void *pMsg, int msgLen, void *pKey) {
   memcpy(pKey, context.digest, sizeof(context.digest));
 }
 
+bool isValidateTag(char *input) {
+  if (!input) return false;
+  int len = strlen(input);
+  if (len == 0) return false;
+  if (input[0] != '_' || isalpha(input[0]) == 0) return false;
+  for (int i = 1; i < len; ++i) {
+    if (input[0] != '_' || isalnum(input[0]) == 0) return false;
+  }
+  return true;
+}
+
 FORCE_INLINE float taos_align_get_float(const char* pBuf) {
 #if __STDC_VERSION__ >= 201112L
   static_assert(sizeof(float) == sizeof(uint32_t), "sizeof(float) must equal to sizeof(uint32_t)");
diff --git a/tests/pytest/stable/json_tag.py b/tests/pytest/stable/json_tag.py
index 7a469647d0..27be0b7eac 100644
--- a/tests/pytest/stable/json_tag.py
+++ b/tests/pytest/stable/json_tag.py
@@ -33,6 +33,9 @@ class TDTestCase:
         tdSql.execute("create table if not exists db_json_tag_test.jsons1(ts timestamp, dataInt int, dataBool bool, dataStr nchar(50)) tags(jtag json)")
         tdSql.execute("CREATE TABLE if not exists db_json_tag_test.jsons1_1 using db_json_tag_test.jsons1 tags('{\"loc\":\"fff\",\"id\":5}')")
         tdSql.error("CREATE TABLE if not exists db_json_tag_test.jsons1_3 using db_json_tag_test.jsons1 tags(3333)")
+        tdSql.error("CREATE TABLE if not exists db_json_tag_test.jsons1_3 using db_json_tag_test.jsons1 tags('{\"1loc\":\"fff\",\";id\":5}')")
+        tdSql.error("CREATE TABLE if not exists db_json_tag_test.jsons1_3 using db_json_tag_test.jsons1 tags('{\"。loc\":\"fff\",\"fsd\":5}')")
+        tdSql.error("CREATE TABLE if not exists db_json_tag_test.jsons1_3 using db_json_tag_test.jsons1 tags('{\"试试\":\"fff\",\";id\":5}')")
         tdSql.execute("insert into db_json_tag_test.jsons1_2 using db_json_tag_test.jsons1 tags('{\"num\":5,\"location\":\"beijing\"}') values (now, 2, true, 'json2')")
         tdSql.error("insert into db_json_tag_test.jsons1_4 using db_json_tag_test.jsons1 tags(3)")
         tdSql.execute("insert into db_json_tag_test.jsons1_1 values(now, 1, false, 'json1')")
@@ -228,6 +231,12 @@ class TDTestCase:
         tdSql.query("select distinct jtag from db_json_tag_test.jsons1")
         tdSql.checkRows(6)
 
+        tdSql.query("select distinct jtag->'location' from db_json_tag_test.jsons1")
+        tdSql.checkRows(3)
+
+        # test chinese
+        tdSql.execute("CREATE TABLE if not exists db_json_tag_test.jsons1_11 using db_json_tag_test.jsons1 tags('{\"k1\":\"中国\",\"k2\":\"是是是\"}')")
+
     def stop(self):
         tdSql.close()
         tdLog.success("%s successfully executed" % __file__)
-- 
GitLab