Improved robustness in the decoding of JSON5 text escape sequences found

in malformed JSONB.

FossilOrigin-Name: 35e0108af2bdd830375c31c525f8ed0e8df64959d89649a88402dc1a5c376612

diff --git a/manifest b/manifest
index 8b71972a229120da568606e92f46297e79c6df7f..1d877db3b81dd611759f8312918732214fe6748b 100644 (file)
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C Fix\san\soff-by-one\serror\sin\sthe\schanges\sfrom\sthe\sprevious\scheck-in\son\nthe\sjsonb\sbranch.
-D 2023-10-10T18:55:29.425
+C Improved\srobustness\sin\sthe\sdecoding\sof\sJSON5\stext\sescape\ssequences\sfound\nin\smalformed\sJSONB.
+D 2023-10-10T23:02:31.182
 F .fossil-settings/empty-dirs dbb81e8fc0401ac46a1491ab34a7f2c7c0452f2f06b54ebb845d024ca8283ef1
 F .fossil-settings/ignore-glob 35175cdfcf539b2318cb04a9901442804be81cd677d8b889fcc9149c21f239ea
 F LICENSE.md df5091916dbb40e6e9686186587125e1b2ff51f022cc334e886c19a0e9982724
@@ -674,7 +674,7 @@ 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 3a0d5a277dfe592998f803b166f2bcd761259643446c4ef3b58f662f4e6e121a
+F src/json.c d86e0d27c8d99c4a9e9bbe44401224270d667118742df371a3928e0406243647
 F src/legacy.c d7874bc885906868cd51e6c2156698f2754f02d9eee1bae2d687323c3ca8e5aa
 F src/loadext.c 98cfba10989b3da6f1807ad42444017742db7f100a54f1032af7a8b1295912c0
 F src/main.c 618aeb399e993cf561864f4b0cf6a331ee4f355cf663635f8d9da3193a46aa40
@@ -2129,8 +2129,8 @@ F vsixtest/vsixtest.tcl 6a9a6ab600c25a91a7acc6293828957a386a8a93
 F vsixtest/vsixtest.vcxproj.data 2ed517e100c66dc455b492e1a33350c1b20fbcdc
 F vsixtest/vsixtest.vcxproj.filters 37e51ffedcdb064aad6ff33b6148725226cd608e
 F vsixtest/vsixtest_TemporaryKey.pfx e5b1b036facdb453873e7084e1cae9102ccc67a0
-P 0caa320d9099adbaf98e3719003dbdc4d158abcb3d8a1af20fbcd4c08c970f4a
-R 2be00e287cc1d614a9c2f8c8c8493433
+P 216191b113da43516d31301fb133173add66358d503677ab997bc82cc88bfea4
+R 3e258cb31e76158544ba210b63aec864
 U drh
-Z f13f9b96095bfab368044d04f5c134b6
+Z 26ce59215821be9c842a531e0d11f121
 # Remove this line to create a well-formed Fossil manifest.

diff --git a/manifest.uuid b/manifest.uuid
index a728066e523076b198b07c3fd7f6b66d288c066f..37779a76ba3e6128486327aafaed9db595b1504f 100644 (file)
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-216191b113da43516d31301fb133173add66358d503677ab997bc82cc88bfea4
\ No newline at end of file
+35e0108af2bdd830375c31c525f8ed0e8df64959d89649a88402dc1a5c376612
\ No newline at end of file

diff --git a/src/json.c b/src/json.c
index b9f13f12f120be2b8dfaa2f687f6a342bdd2b53b..8f263556a7c400a719421d4812fc5e832c24484d 100644 (file)
--- a/src/json.c
+++ b/src/json.c
@@ -577,9 +577,13 @@ static void jsonAppendNormalizedString(JsonString *p, const char *zIn, u32 N){
     for(i=0; i<N && zIn[i]!='\\'; i++){}
     if( i>0 ){
       jsonAppendRawNZ(p, zIn, i);
+      if( i>=N ) break;
       zIn += i;
       N -= i;
-      if( N==0 ) break;
+    }
+    if( N<2 ){
+      p->eErr |= JSTRING_MALFORMED;
+      break;
     }
     assert( zIn[0]=='\\' );
     switch( (u8)zIn[1] ){
@@ -590,6 +594,11 @@ static void jsonAppendNormalizedString(JsonString *p, const char *zIn, u32 N){
         jsonAppendRawNZ(p, "\\u0009", 6);
         break;
       case 'x':
+        if( N<4 ){
+          N = 2;
+          p->eErr |= JSTRING_MALFORMED;
+          break;
+        }
         jsonAppendRawNZ(p, "\\u00", 4);
         jsonAppendRawNZ(p, &zIn[2], 2);
         zIn += 2;
@@ -599,14 +608,22 @@ static void jsonAppendNormalizedString(JsonString *p, const char *zIn, u32 N){
         jsonAppendRawNZ(p, "\\u0000", 6);
         break;
       case '\r':
-        if( zIn[2]=='\n' ){
+        if( N>2 && zIn[2]=='\n' ){
           zIn++;
           N--;
         }
         break;
       case '\n':
         break;
-      case 0xe2:
+      case 0xe2:  /* \ followed by U+2028 or U+2029 line terminator ignored */
+        if( N<4
+         || 0x80!=(u8)zIn[2]
+         || (0xa8!=(u8)zIn[3] && 0xa9!=(u8)zIn[3])
+        ){
+          N = 2;
+          p->eErr |= JSTRING_MALFORMED;
+          break;
+        }
         assert( N>=4 );
         assert( 0x80==(u8)zIn[2] );
         assert( 0xa8==(u8)zIn[3] || 0xa9==(u8)zIn[3] );
@@ -617,6 +634,7 @@ static void jsonAppendNormalizedString(JsonString *p, const char *zIn, u32 N){
         jsonAppendRawNZ(p, zIn, 2);
         break;
     }
+    assert( N>=2 );
     zIn += 2;
     N -= 2;
   }
@@ -3353,13 +3371,16 @@ static u32 jsonXlateBlobToText(
         for(k=0; k<sz2 && zIn[k]!='\\'; k++){}
         if( k>0 ){
           jsonAppendRawNZ(pOut, zIn, k);
-          if( sz2<=k ){
-            if( sz2<k ) pOut->eErr |= JSTRING_MALFORMED;
+          if( k>=sz2 ){
             break;
           }
           zIn += k;
           sz2 -= k;
         }
+        if( sz2<2 ){
+          if( sz2>0 ) pOut->eErr |= JSTRING_MALFORMED;
+          if( sz2==0 ) break;
+        }
         assert( zIn[0]=='\\' );
         switch( (u8)zIn[1] ){
           case '\'':
@@ -3369,21 +3390,21 @@ static u32 jsonXlateBlobToText(
             jsonAppendRawNZ(pOut, "\\u0009", 6);
             break;
           case 'x':
-            jsonAppendRawNZ(pOut, "\\u00", 4);
-            jsonAppendRawNZ(pOut, &zIn[2], 2);
             if( sz2<2 ){
               pOut->eErr |= JSTRING_MALFORMED;
               sz2 = 0;
-            }else{
-              zIn += 2;
-              sz2 -= 2;
+              break;
             }
+            jsonAppendRawNZ(pOut, "\\u00", 4);
+            jsonAppendRawNZ(pOut, &zIn[2], 2);
+            zIn += 2;
+            sz2 -= 2;
             break;
           case '0':
             jsonAppendRawNZ(pOut, "\\u0000", 6);
             break;
           case '\r':
-            if( zIn[2]=='\n' ){
+            if( sz2>2 && zIn[2]=='\n' ){
               zIn++;
               sz2--;
             }
@@ -3391,6 +3412,9 @@ static u32 jsonXlateBlobToText(
           case '\n':
             break;
           case 0xe2:
+            /* '\' followed by either U+2028 or U+2029 is ignored as
+            ** whitespace.  Not that in UTF8, U+2028 is 0xe2 0x80 0x29.
+            ** U+2029 is the same except for the last byte */
             if( sz2<4
              || 0x80!=(u8)zIn[2]
              || (0xa8!=(u8)zIn[3] && 0xa9!=(u8)zIn[3])
@@ -3407,6 +3431,7 @@ static u32 jsonXlateBlobToText(
             break;
         }
         if( sz2<2 ){
+          sz2 = 0;
           pOut->eErr |= JSTRING_MALFORMED;
           break;
         }