Validity checking of text nodes in JSONB.

FossilOrigin-Name: fa5160687c2f970d407e8af73c246f7cd806bb4ce35f29a79ac534a8646a6c8e

diff --git a/manifest b/manifest
index ada03e02f136a53ec237ad9fada4b9272ccc8bff..91ba64c3dd431b8438ff314cfb9f7dd1095fefa0 100644 (file)
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C The\sjson_error_position()\sfunction\snow\sreports\san\sapproximate\sbyte\soffset\nto\sthe\sproblem\sin\sa\sJSONB\sif\sthere\sis\sa\sproblem.
-D 2023-12-11T21:00:55.202
+C Validity\schecking\sof\stext\snodes\sin\sJSONB.
+D 2023-12-12T02:31:12.655
 F .fossil-settings/empty-dirs dbb81e8fc0401ac46a1491ab34a7f2c7c0452f2f06b54ebb845d024ca8283ef1
 F .fossil-settings/ignore-glob 35175cdfcf539b2318cb04a9901442804be81cd677d8b889fcc9149c21f239ea
 F LICENSE.md df5091916dbb40e6e9686186587125e1b2ff51f022cc334e886c19a0e9982724
@@ -696,7 +696,7 @@ 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 387dd9d7d63bac1e0068d66e8b40d86e1e2f0edc6930da28c67584c7a5a32872
+F src/json.c 34f829db190ae485926855da651d349eedb35fa56010266e14c0b094f536c9f5
 F src/legacy.c d7874bc885906868cd51e6c2156698f2754f02d9eee1bae2d687323c3ca8e5aa
 F src/loadext.c 7432c944ff197046d67a1207790a1b13eec4548c85a9457eb0896bb3641dfb36
 F src/main.c ce714ee501122c76eb2e69b292bebe443aba611fc3b88e6786eb910285515fe4
@@ -2153,8 +2153,8 @@ F vsixtest/vsixtest.tcl 6a9a6ab600c25a91a7acc6293828957a386a8a93
 F vsixtest/vsixtest.vcxproj.data 2ed517e100c66dc455b492e1a33350c1b20fbcdc
 F vsixtest/vsixtest.vcxproj.filters 37e51ffedcdb064aad6ff33b6148725226cd608e
 F vsixtest/vsixtest_TemporaryKey.pfx e5b1b036facdb453873e7084e1cae9102ccc67a0
-P c3d60cf7028a333b825d5b89516945a73e0c158ac81d8bcc117d21bfd98602c8
-R eef20549bb0818034c7a80cb1490f371
+P 80d5d94dff6a2d2557039be3d7d47c1a6003c4b98defe0bd411acfeb963ad5dd
+R 8ac47232b4f02631d44b831db0790bc1
 U drh
-Z 99dd825f602cc4286df6f3fcbe1570d6
+Z 8178cefe916c4486a65df71a751de097
 # Remove this line to create a well-formed Fossil manifest.

diff --git a/manifest.uuid b/manifest.uuid
index 07345f01928f4af2580cf70599d59b6b33cdaa81..208d3cd26d3a240f104fc7c68ff9d490d556fcb7 100644 (file)
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-80d5d94dff6a2d2557039be3d7d47c1a6003c4b98defe0bd411acfeb963ad5dd
\ No newline at end of file
+fa5160687c2f970d407e8af73c246f7cd806bb4ce35f29a79ac534a8646a6c8e
\ No newline at end of file

diff --git a/src/json.c b/src/json.c
index 31367fe5446daf7959f4ef38a59048fa2f8a4683..17ebc769cc30952163ec1a32ad14c96954ae3907 100644 (file)
--- a/src/json.c
+++ b/src/json.c
@@ -176,7 +176,9 @@ static const char jsonSpaces[] = "\011\012\015\040";
 
 /*
 ** Characters that are special to JSON.  Control characters,
-** '"' and '\\'.
+** '"' and '\\' and '\''.  Actually, '\'' is not special to
+** canonical JSON, but it is special in JSON-5, so we include
+** it in the set of special characters.
 */
 static const char jsonIsOk[256] = {
   0, 0, 0, 0, 0, 0, 0, 0,  0, 0, 0, 0, 0, 0, 0, 0,
@@ -337,6 +339,7 @@ static void jsonReturnParse(sqlite3_context*,JsonParse*);
 static JsonParse *jsonParseFuncArg(sqlite3_context*,sqlite3_value*,u32);
 static void jsonParseFree(JsonParse*);
 static u32 jsonbPayloadSize(const JsonParse*, u32, u32*);
+static u32 jsonUnescapeOneChar(const char*, u32, u32*);
 
 /**************************************************************************
 ** Utility routines for dealing with JsonCache objects
@@ -1346,10 +1349,42 @@ static u32 jsonbValidityCheck(
       return 0;
     }
     case JSONB_TEXT: {
+      j = i+n;
+      k = j+sz;
+      while( j<k ){
+        if( !jsonIsOk[z[j]] && z[j]!='\'' ) return j+1;
+        j++;
+      }
       return 0;
     }
     case JSONB_TEXTJ:
     case JSONB_TEXT5: {
+      j = i+n;
+      k = j+sz;
+      while( j<k ){
+        if( !jsonIsOk[z[j]] && z[j]!='\'' ){
+          if( z[j]=='"' ){
+            if( x==JSONB_TEXTJ ) return j+1;
+            j++;
+          }else if( z[j]!='\\' || j+1>=k ){
+            return j+1;
+          }else if( strchr("\"\\/bfnrt",z[j+1])!=0 ){
+            j++;
+          }else if( z[j+1]=='u' ){
+            if( j+5>=k ) return j+1;
+            if( !jsonIs4Hex((const char*)&z[j+2]) ) return j+1;
+            j++;
+          }else if( x!=JSONB_TEXT5 ){
+            return j+1;
+          }else{
+            u32 c = 0;
+            u32 szC = jsonUnescapeOneChar((const char*)&z[j], k-j, &c);
+            if( c==0xfffd ) return j+1;
+            j += szC - 1;
+          }
+        }
+        j++;
+      }
       return 0;
     }
     case JSONB_TEXTRAW: {
@@ -1625,6 +1660,8 @@ json_parse_restart:
         /* Control characters are not allowed in strings */
         pParse->iErr = j;
         return -1;
+      }else if( c=='"' ){
+        opcode = JSONB_TEXT5;
       }
       j++;
     }
@@ -2084,13 +2121,13 @@ static u32 jsonXlateBlobToText(
       }
       break;
     }
+    case JSONB_TEXT:
     case JSONB_TEXTJ: {
       jsonAppendChar(pOut, '"');
       jsonAppendRaw(pOut, (const char*)&pParse->aBlob[i+n], sz);
       jsonAppendChar(pOut, '"');
       break;
     }
-    case JSONB_TEXT:
     case JSONB_TEXT5: {
       const char *zIn;
       u32 k;