Give the json_valid() function an optional second argument that determines

what is meant by "valid".

FossilOrigin-Name: a4e19ad43dac81e7655ec03ff69bb99d1d02b0c227034c90fb41415fd4793fe3

diff --git a/manifest b/manifest
index 17d296c97f4b2f0d68b2ea2edb9c00444042bcd9..188b5f527747f03468c97a23549470eabce51c78 100644 (file)
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C Convert\sthe\sjson_tree()\sand\sjson_each()\svirtual\stables\sover\sto\suse\sJSONB\shas\ntheir\sinternal\srepresentation.
-D 2023-11-27T12:36:29.017
+C Give\sthe\sjson_valid()\sfunction\san\soptional\ssecond\sargument\sthat\sdetermines\nwhat\sis\smeant\sby\s"valid".
+D 2023-11-27T15:57:11.976
 F .fossil-settings/empty-dirs dbb81e8fc0401ac46a1491ab34a7f2c7c0452f2f06b54ebb845d024ca8283ef1
 F .fossil-settings/ignore-glob 35175cdfcf539b2318cb04a9901442804be81cd677d8b889fcc9149c21f239ea
 F LICENSE.md df5091916dbb40e6e9686186587125e1b2ff51f022cc334e886c19a0e9982724
@@ -688,7 +688,7 @@ F src/hash.h 3340ab6e1d13e725571d7cee6d3e3135f0779a7d8e76a9ce0a85971fa3953c51
 F src/hwtime.h f9c2dfb84dce7acf95ce6d289e46f5f9d3d1afd328e53da8f8e9008e3b3caae6
 F src/in-operator.md 10cd8f4bcd225a32518407c2fb2484089112fd71
 F src/insert.c 3f0a94082d978bbdd33c38fefea15346c6c6bffb70bc645a71dc0f1f87dd3276
-F src/json.c efd9ec1cb0f0cbee3e43ebaba69469b3d3f44ee623b53b1ce72361da42e05fbe
+F src/json.c 6c1de0d7802a15669a78008125b409b67aa23eefcacfc52436d713419d09be9e
 F src/legacy.c d7874bc885906868cd51e6c2156698f2754f02d9eee1bae2d687323c3ca8e5aa
 F src/loadext.c 7432c944ff197046d67a1207790a1b13eec4548c85a9457eb0896bb3641dfb36
 F src/main.c 1b89f3de98d1b59fec5bac1d66d6ece21f703821b8eaa0d53d9604c35309f6f9
@@ -1325,7 +1325,7 @@ F test/json/json-generator.tcl dc0dd0f393800c98658fc4c47eaa6af29d4e17527380cd286
 F test/json/json-q1-b.txt 606818a5fba6d9e418c9f4ea7d8418af026775042dad81439b72447a147a462c
 F test/json/json-q1.txt 65f9d1cdcc4cffa9823fb73ed936aae5658700cd001fde448f68bfb91c807307
 F test/json/json-speed-check.sh b060a9a6c696c0a807d8929400fa11bd7113edc58b0d66b9795f424f8d0db326 x
-F test/json101.test 19f9abc77c33151bd7fdb7850fb473cb0154f229b863b740590ad9de807a032d
+F test/json101.test 643936557daeb2eeeb839e9d87a682c7d7c6f6aee66045eedd8613cb8471d8d8
 F test/json102.test 557a46e16df1aa9bdbc4076a71a45814ea0e7503d6621d87d42a8c04cbc2b0ef
 F test/json103.test 53df87f83a4e5fa0c0a56eb29ff6c94055c6eb919f33316d62161a8880112dbe
 F test/json104.test 1b844a70cddcfa2e4cd81a5db0657b2e61e7f00868310f24f56a9ba0114348c1
@@ -2145,9 +2145,8 @@ F vsixtest/vsixtest.tcl 6a9a6ab600c25a91a7acc6293828957a386a8a93
 F vsixtest/vsixtest.vcxproj.data 2ed517e100c66dc455b492e1a33350c1b20fbcdc
 F vsixtest/vsixtest.vcxproj.filters 37e51ffedcdb064aad6ff33b6148725226cd608e
 F vsixtest/vsixtest_TemporaryKey.pfx e5b1b036facdb453873e7084e1cae9102ccc67a0
-P ab2644aacf4757a51cf62e05cff6711a0a3605d60502a3dd310887df1b993545 b5a5660ca22437640c9bf32c44d92c76a7293dafcbaf4fa6a4c171128d64871d
-R 6545b38a1b858ef8dae597ee252577f2
-T +closed b5a5660ca22437640c9bf32c44d92c76a7293dafcbaf4fa6a4c171128d64871d
+P ec18caa3f7790b780dde66c1ccbb6eb09d2f1507629cc45955fc1b08380b4017
+R ae458da46f597856d9232f7ad8297973
 U drh
-Z 9c0c7326e55f087530f0f53e8712d89c
+Z 6c8b34340e171672e2ebcad369756a23
 # Remove this line to create a well-formed Fossil manifest.

diff --git a/manifest.uuid b/manifest.uuid
index 189e2ccc7a5c0ec605cd67929b737a57951f8ca9..c418cd97809566d45ccb68e14ab486fe73aee269 100644 (file)
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-ec18caa3f7790b780dde66c1ccbb6eb09d2f1507629cc45955fc1b08380b4017
\ No newline at end of file
+a4e19ad43dac81e7655ec03ff69bb99d1d02b0c227034c90fb41415fd4793fe3
\ No newline at end of file

diff --git a/src/json.c b/src/json.c
index 0e3f6fe5755347f9468bbc0b5a66b2b2de844a47..1107a24f4cfa90eba30ddd18780a15d5e275ad0b 100644 (file)
--- a/src/json.c
+++ b/src/json.c
@@ -5241,15 +5241,60 @@ static void jsonTypeFunc(
 
 /*
 ** json_valid(JSON)
+** json_valid(JSON, FLAGS)
 **
-** Return 1 if the argument is one of:
+** Check the JSON argument to see if it is well-formed.  The FLAGS argument
+** encodes the various constraints on what is meant by "well-formed":
 **
-**    *   A well-formed canonical JSON string according to RFC-8259
-**        (without JSON5 enhancements), or
+**     0x01      Canonical RFC-8259 JSON text
+**     0x02      JSON text with optional JSON-5 extensions
+**     0x04      Superficially appears to be JSONB
+**     0x08      Strictly well-formed JSONB
 **
-**    *   A BLOB that plausibly could be a JSONB value.
+** If the FLAGS argument is omitted, it defaults to 1.  Useful values for
+** FLAGS include:
 **
-** Return 0 otherwise.
+**    1          Strict canonical JSON text
+**    2          JSON text perhaps with JSON-5 extensions
+**    4          Superficially appears to be JSONB
+**    5          Canonical JSON text or superficial JSONB
+**    6          JSON-5 text or superficial JSONB
+**    8          Strict JSONB
+**    9          Canonical JSON text or strict JSONB
+**    10         JSON-5 text or strict JSONB
+**
+** Other flag combinations are redundant.  For example, every canonical
+** JSON text is also well-formed JSON-5 text, so FLAG values 2 and 3
+** are the same.  Similarly, any input that passes a strict JSONB validation
+** will also pass the superficial validation so 12 through 15 are the same
+** as 8 through 11 respectively.
+**
+** This routine runs in linear time to validate text and when doing strict
+** JSONB validation.  Superficial JSONB validation is constant time,
+** assuming the BLOB is already in memory.  The performance advantage
+** of superficial JSONB validation is why that option is provided.
+** Application developers can choose to do fast superficial validation or
+** slower strict validation, according to their specific needs.
+**
+** Only the lower four bits of the FLAGS argument are currently used.
+** Higher bits are reserved for future expansion.   To facilitate
+** compatibility, the current implementation raises an error if any bit
+** in FLAGS is set other than the lower four bits.
+**
+** The original circa 2015 implementation of the JSON routines in
+** SQLite only supported canonical RFC-8259 JSON text and the json_valid()
+** function only accepted one argument.  That is why the default value
+** for the FLAGS argument is 1, since FLAGS=1 causes this routine to only
+** recognize canonical RFC-8259 JSON text as valid.  The extra FLAGS
+** argument was added when the JSON routines were extended to support
+** JSON5-like extensions and binary JSONB stored in BLOBs.
+**
+** Return Values:
+**
+**   *   Raise an error if FLAGS is outside the range of 1 to 15.
+**   *   Return NULL if the input is NULL
+**   *   Return 1 if the input is well-formed.
+**   *   Return 0 if the input is not well-formed.
 */
 static void jsonValidFunc(
   sqlite3_context *ctx,
@@ -5257,26 +5302,47 @@ static void jsonValidFunc(
   sqlite3_value **argv
 ){
   JsonParse *p;          /* The parse */
-  UNUSED_PARAMETER(argc);
-  if( sqlite3_value_type(argv[0])==SQLITE_NULL ){
+  u8 flags = 1;
+  u8 res = 0;
+  if( argc==2 ){
+    i64 f = sqlite3_value_int64(argv[1]);
+    if( f<1 || f>15 ){
+      sqlite3_result_error(ctx, "FLAGS parameter to json_valid() must be"
+                                " between 1 and 15", -1);
+      return;
+    }
+    flags = f & 0x0f;
+  }
+  switch( sqlite3_value_type(argv[0]) ){
+    case SQLITE_NULL: {
 #ifdef SQLITE_LEGACY_JSON_VALID
-    /* Incorrect legacy behavior was to return FALSE for a NULL input */
-    sqlite3_result_int(ctx, 0);
+      /* Incorrect legacy behavior was to return FALSE for a NULL input */
+      sqlite3_result_int(ctx, 0);
 #endif
-    return;
-  }
-  if( jsonFuncArgMightBeBinary(argv[0]) ){
-    sqlite3_result_int(ctx, 1);
-    return;
-  }
-  p = jsonParseCached(ctx, argv[0], 0, 0);
-  if( p==0 || p->oom ){
-    sqlite3_result_error_nomem(ctx);
-    sqlite3_free(p);
-  }else{
-    sqlite3_result_int(ctx, p->nErr==0 && (p->hasNonstd==0 || p->useMod));
-    if( p->nErr ) jsonParseFree(p);
+      return;
+    }
+    case SQLITE_BLOB: {
+      if( (flags & 0x0c)!=0 && jsonFuncArgMightBeBinary(argv[0]) ){
+        /* TO-DO:  strict checking if flags & 0x08 */
+        res = 1;
+      }
+      break;
+    }
+    default: {
+      if( (flags & 0x3)==0 ) break;
+      p = jsonParseCached(ctx, argv[0], 0, 0);
+      if( p==0 || p->oom ){
+        sqlite3_result_error_nomem(ctx);
+        sqlite3_free(p);
+      }else if( p->nErr ){
+        jsonParseFree(p);
+      }else if( (flags & 0x02)!=0 || p->hasNonstd==0 || p->useMod ){
+        res = 1;
+      }
+      break;
+    }
   }
+  sqlite3_result_int(ctx, res);
 }
 
 /*
@@ -6171,6 +6237,7 @@ void sqlite3RegisterJsonFunctions(void){
     JFUNCTION(json_type,          1,1,0, 0,0,0,          jsonTypeFunc),
     JFUNCTION(json_type,          2,1,0, 0,0,0,          jsonTypeFunc),
     JFUNCTION(json_valid,         1,1,0, 0,0,0,          jsonValidFunc),
+    JFUNCTION(json_valid,         2,1,0, 0,0,0,          jsonValidFunc),
 #if SQLITE_DEBUG
     JFUNCTION(json_parse,         1,1,0, 0,0,0,          jsonParseFunc),
     JFUNCTION(json_test1,         1,1,0, 1,0,0,          jsonTest1Func),

diff --git a/test/json101.test b/test/json101.test
index 280d4673938747740a3b049c7c7646288152dc04..ebaf0f3b340df2df1ff17826a4423557fe01930d 100644 (file)
--- a/test/json101.test
+++ b/test/json101.test
@@ -324,7 +324,7 @@ do_execsql_test json101-5.2 {
   SELECT id, json_valid(json), json_type(json), '|' FROM j2 ORDER BY id;
 } {1 1 object | 2 1 object | 3 1 array |}
 do_execsql_test json101-5.2b {
-  SELECT id, json_valid(json), json_type(json), '|' FROM j2b ORDER BY id;
+  SELECT id, json_valid(json,5), json_type(json), '|' FROM j2b ORDER BY id;
 } {1 1 object | 2 1 object | 3 1 array |}
 
 ifcapable !vtab {