From 272ae627c535b1ee11b82ba045c641945e34bdc5 Mon Sep 17 00:00:00 2001 From: drh <> Date: Fri, 28 Apr 2023 14:48:11 +0000 Subject: [PATCH] Add the json_error(X) function that returns the 1-based character offset to the first syntax error in JSON5 string X, or 0 if there are no errors. FossilOrigin-Name: 901ad995d5a722ca2672516205ff488e9acd703a828ca5fc43f11fca5f2af120 --- manifest | 12 +++--- manifest.uuid | 2 +- src/json.c | 115 +++++++++++++++++++++++++++++++++++++++++++++----- 3 files changed, 111 insertions(+), 18 deletions(-) diff --git a/manifest b/manifest index e9188432cd..77de439258 100644 --- a/manifest +++ b/manifest @@ -1,5 +1,5 @@ -C Fix\sindentation\sand\somit\strailing\swhitespace\sin\sthe\srandom\sJSON\sgenerator\nscript. -D 2023-04-28T13:25:35.641 +C Add\sthe\sjson_error(X)\sfunction\sthat\sreturns\sthe\s1-based\scharacter\soffset\sto\nthe\sfirst\ssyntax\serror\sin\sJSON5\sstring\sX,\sor\s0\sif\sthere\sare\sno\serrors. +D 2023-04-28T14:48:11.699 F .fossil-settings/empty-dirs dbb81e8fc0401ac46a1491ab34a7f2c7c0452f2f06b54ebb845d024ca8283ef1 F .fossil-settings/ignore-glob 35175cdfcf539b2318cb04a9901442804be81cd677d8b889fcc9149c21f239ea F LICENSE.md df5091916dbb40e6e9686186587125e1b2ff51f022cc334e886c19a0e9982724 @@ -593,7 +593,7 @@ F src/hash.h 3340ab6e1d13e725571d7cee6d3e3135f0779a7d8e76a9ce0a85971fa3953c51 F src/hwtime.h b638809e083b601b618df877b2e89cb87c2a47a01f4def10be4c4ebb54664ac7 F src/in-operator.md 10cd8f4bcd225a32518407c2fb2484089112fd71 F src/insert.c a8de1db43335fc4946370a7a7e47d89975ad678ddb15078a150e993ba2fb37d4 -F src/json.c 7d03a1185d7aee74f53a3f2381355ff4b71d955f287d29d30c0b0979afc96b67 +F src/json.c b532d42d310e570d7f620a692a26f21dc7306063b866088c4e43c647a17118b3 F src/legacy.c d7874bc885906868cd51e6c2156698f2754f02d9eee1bae2d687323c3ca8e5aa F src/loadext.c be5af440f3192c58681b5d43167dbca3ccbfce394d89faa22378a14264781136 F src/main.c 09bc5191f75dc48fc4dfddda143cb864c0c3dbc3297eb9a9c8e01fea58ff847d @@ -2066,8 +2066,8 @@ F vsixtest/vsixtest.tcl 6a9a6ab600c25a91a7acc6293828957a386a8a93 F vsixtest/vsixtest.vcxproj.data 2ed517e100c66dc455b492e1a33350c1b20fbcdc F vsixtest/vsixtest.vcxproj.filters 37e51ffedcdb064aad6ff33b6148725226cd608e F vsixtest/vsixtest_TemporaryKey.pfx e5b1b036facdb453873e7084e1cae9102ccc67a0 -P f56528d413d8e622f7c4f18b2f9f2e620bfb441c020461299b35a90072ee6c13 -R 5da141bb06598ed103d496559df7ae36 +P 629db09fceb7bf37561b52ccee06ebf4df261291e9a8ffcca82b243f6db5ff07 +R cda10fe16274848e02e4d5ea3713e4b9 U drh -Z 790f8e2b554134402e5889a19faba52a +Z 80f17384e6d82e3c8657040553f2124e # Remove this line to create a well-formed Fossil manifest. diff --git a/manifest.uuid b/manifest.uuid index d1f6aa531d..0b6e554ed2 100644 --- a/manifest.uuid +++ b/manifest.uuid @@ -1 +1 @@ -629db09fceb7bf37561b52ccee06ebf4df261291e9a8ffcca82b243f6db5ff07 \ No newline at end of file +901ad995d5a722ca2672516205ff488e9acd703a828ca5fc43f11fca5f2af120 \ No newline at end of file diff --git a/src/json.c b/src/json.c index 1152d9f916..d1a057b48e 100644 --- a/src/json.c +++ b/src/json.c @@ -1073,7 +1073,10 @@ json_parse_restart: iThis = jsonParseAddNode(pParse, JSON_OBJECT, 0, 0); if( iThis<0 ) return -1; for(j=i+1;;j++){ - if( ++pParse->iDepth > JSON_MAX_DEPTH ) return -1; + if( ++pParse->iDepth > JSON_MAX_DEPTH ){ + pParse->iErr = j; + return -1; + } x = jsonParseValue(pParse, j); if( x<=0 ){ if( x==(-2) ){ @@ -1090,6 +1093,7 @@ json_parse_restart: pParse->has5 = 1; x = k; }else{ + pParse->iErr = j; return -1; } } @@ -1109,13 +1113,19 @@ json_parse_restart: } } x = jsonParseValue(pParse, j); - if( x!=(-5) ) return -1; + if( x!=(-5) ){ + pParse->iErr = j; + return -1; + } j = pParse->iErr+1; } parse_object_value: x = jsonParseValue(pParse, j); pParse->iDepth--; - if( x<=0 ) return -1; + if( x<=0 ){ + pParse->iErr = j; + return -1; + } j = x; if( z[j]==',' ){ continue; @@ -1140,6 +1150,7 @@ json_parse_restart: break; } } + pParse->iErr = j; return -1; } pParse->aNode[iThis].n = pParse->nNode - (u32)iThis - 1; @@ -1151,7 +1162,10 @@ json_parse_restart: if( iThis<0 ) return -1; memset(&pParse->aNode[iThis].u, 0, sizeof(pParse->aNode[iThis].u)); for(j=i+1;;j++){ - if( ++pParse->iDepth > JSON_MAX_DEPTH ) return -1; + if( ++pParse->iDepth > JSON_MAX_DEPTH ){ + pParse->iErr = j; + return -1; + } x = jsonParseValue(pParse, j); pParse->iDepth--; if( x<=0 ){ @@ -1160,6 +1174,7 @@ json_parse_restart: if( pParse->nNode!=(u32)iThis+1 ) pParse->has5 = 1; break; } + pParse->iErr = j; return -1; } j = x; @@ -1186,6 +1201,7 @@ json_parse_restart: break; } } + pParse->iErr = j; return -1; } pParse->aNode[iThis].n = pParse->nNode - (u32)iThis - 1; @@ -1207,6 +1223,7 @@ json_parse_restart: c = z[j]; if( (c & ~0x1f)==0 ){ /* Control characters are not allowed in strings */ + pParse->iErr = j; return -1; } if( c=='\\' ){ @@ -1227,6 +1244,7 @@ json_parse_restart: jnFlags |= (JNODE_ESCAPE|JNODE_JSON5); pParse->has5 = 1; }else{ + pParse->iErr = j; return -1; } }else if( c==cDelim ){ @@ -1242,6 +1260,7 @@ json_parse_restart: jsonParseAddNode(pParse, JSON_NULL, 0, 0); return i+4; } + pParse->iErr = i; return -1; } case 't': { @@ -1249,6 +1268,7 @@ json_parse_restart: jsonParseAddNode(pParse, JSON_TRUE, 0, 0); return i+4; } + pParse->iErr = i; return -1; } case 'f': { @@ -1256,6 +1276,7 @@ json_parse_restart: jsonParseAddNode(pParse, JSON_FALSE, 0, 0); return i+5; } + pParse->iErr = i; return -1; } case '+': { @@ -1271,6 +1292,7 @@ json_parse_restart: seenDP = JSON_REAL; goto parse_number_2; } + pParse->iErr = i; return -1; case '-': case '0': @@ -1338,6 +1360,7 @@ json_parse_restart: jnFlags |= JNODE_JSON5; goto parse_number_2; } + pParse->iErr = i; return -1; } if( z[i+1]=='0' ){ @@ -1358,7 +1381,10 @@ json_parse_restart: c = z[j]; if( sqlite3Isdigit(c) ) continue; if( c=='.' ){ - if( seenDP==JSON_REAL ) return -1; + if( seenDP==JSON_REAL ){ + pParse->iErr = j; + return -1; + } seenDP = JSON_REAL; continue; } @@ -1368,10 +1394,14 @@ json_parse_restart: pParse->has5 = 1; jnFlags |= JNODE_JSON5; }else{ + pParse->iErr = j; return -1; } } - if( seenE ) return -1; + if( seenE ){ + pParse->iErr = j; + return -1; + } seenDP = JSON_REAL; seenE = 1; c = z[j+1]; @@ -1379,7 +1409,10 @@ json_parse_restart: j++; c = z[j+1]; } - if( c<'0' || c>'9' ) return -1; + if( c<'0' || c>'9' ){ + pParse->iErr = j; + return -1; + } continue; } break; @@ -1389,6 +1422,7 @@ json_parse_restart: pParse->has5 = 1; jnFlags |= JNODE_JSON5; }else{ + pParse->iErr = j; return -1; } } @@ -1402,6 +1436,7 @@ json_parse_restart: pParse->has5 = 1; return i+3; } + pParse->iErr = i; return -1; } case 'I': { @@ -1410,6 +1445,7 @@ json_parse_restart: pParse->has5 = 1; return i+8; } + pParse->iErr = i; return -1; } case '}': { @@ -1453,6 +1489,7 @@ json_parse_restart: pParse->has5 = 1; goto json_parse_restart; } + pParse->iErr = i; return -1; } default: { @@ -1471,6 +1508,7 @@ json_parse_restart: return i + nn; } #endif + pParse->iErr = i; return -1; /* Syntax error */ } } /* End switch(z[i]) */ @@ -1575,6 +1613,15 @@ static int jsonParseFindParents(JsonParse *pParse){ ** is no longer valid, parse the JSON again and return the new parse, ** and also register the new parse so that it will be available for ** future sqlite3_get_auxdata() calls. +** +** If an error occurs and pErrCtx!=0 then report the error on pErrCtx +** and return NULL. +** +** If an error occurs and pErrCtx==0 then return the Parse object with +** JsonParse.nErr non-zero. If the caller invokes this routine with +** pErrCtx==0 and it gets back a JsonParse with nErr!=0, then the caller +** is responsible for invoking jsonParseFree() on the returned value. +** But the caller may invoke jsonParseFree() *only* if pParse->nErr!=0. */ static JsonParse *jsonParseCached( sqlite3_context *pCtx, @@ -1624,6 +1671,10 @@ static JsonParse *jsonParseCached( p->zJson = (char*)&p[1]; memcpy((char*)p->zJson, zJson, nJson+1); if( jsonParse(p, pErrCtx, p->zJson) ){ + if( pErrCtx==0 ){ + p->nErr = 1; + return p; + } sqlite3_free(p); return 0; } @@ -2501,8 +2552,8 @@ static void jsonTypeFunc( /* ** json_valid(JSON) ** -** Return 1 if JSON is a well-formed JSON string according to RFC-7159. -** Return 0 otherwise. +** Return 1 if JSON is a well-formed canonical JSON string according +** to RFC-7159. Return 0 otherwise. */ static void jsonValidFunc( sqlite3_context *ctx, @@ -2512,8 +2563,17 @@ static void jsonValidFunc( JsonParse *p; /* The parse */ UNUSED_PARAMETER(argc); p = jsonParseCached(ctx, argv, 0); - sqlite3_result_int(ctx, p!=0 && p->has5==0); + sqlite3_result_int(ctx, p!=0 && p->nErr==0 && p->has5==0); + if( p!=0 && p->nErr ) jsonParseFree(p); } + + +/* +** json_valid5(JSON) +** +** Return 1 if JSON is a well-formed JSON5 string. +** Return 0 otherwise. +*/ static void jsonValid5Func( sqlite3_context *ctx, int argc, @@ -2522,7 +2582,39 @@ static void jsonValid5Func( JsonParse *p; /* The parse */ UNUSED_PARAMETER(argc); p = jsonParseCached(ctx, argv, 0); - sqlite3_result_int(ctx, p!=0); + sqlite3_result_int(ctx, p!=0 && p->nErr==0); + if( p!=0 && p->nErr ) jsonParseFree(p); +} + + + +/* +** json_error(JSON) +** +** If JSON is not a well-formed JSON5 string, then return the 1-based +** character offset to the location of the first error in that string. +** Return 0 otherwise. +*/ +static void jsonErrorFunc( + sqlite3_context *ctx, + int argc, + sqlite3_value **argv +){ + JsonParse *p; /* The parse */ + UNUSED_PARAMETER(argc); + p = jsonParseCached(ctx, argv, 0); + if( p==0 || p->oom || p->nErr==0 ){ + sqlite3_result_int(ctx, 0); + }else{ + int n = 0; + int i; + const char *z = p->zJson; + for(i=0; iiErr && z[i]; i++){ + if( (z[i]&0xc0)!=80 ) n++; + } + sqlite3_result_int(ctx, n); + jsonParseFree(p); + } } @@ -3235,6 +3327,7 @@ void sqlite3RegisterJsonFunctions(void){ JFUNCTION(json_array, -1, 0, jsonArrayFunc), JFUNCTION(json_array_length, 1, 0, jsonArrayLengthFunc), JFUNCTION(json_array_length, 2, 0, jsonArrayLengthFunc), + JFUNCTION(json_error, 1, 0, jsonErrorFunc), JFUNCTION(json_extract, -1, 0, jsonExtractFunc), JFUNCTION(->, 2, JSON_JSON, jsonExtractFunc), JFUNCTION(->>, 2, JSON_SQL, jsonExtractFunc), -- 2.47.2