diff options
Diffstat (limited to 'src/VBox/Runtime/common/misc/json.cpp')
| -rw-r--r-- | src/VBox/Runtime/common/misc/json.cpp | 1848 |
1 files changed, 1848 insertions, 0 deletions
diff --git a/src/VBox/Runtime/common/misc/json.cpp b/src/VBox/Runtime/common/misc/json.cpp new file mode 100644 index 00000000..4edaca9a --- /dev/null +++ b/src/VBox/Runtime/common/misc/json.cpp @@ -0,0 +1,1848 @@ +/* $Id: json.cpp $ */ +/** @file + * IPRT JSON parser API (JSON). + */ + +/* + * Copyright (C) 2016-2019 Oracle Corporation + * + * This file is part of VirtualBox Open Source Edition (OSE), as + * available from http://www.virtualbox.org. This file is free software; + * you can redistribute it and/or modify it under the terms of the GNU + * General Public License (GPL) as published by the Free Software + * Foundation, in version 2 as it comes in the "COPYING" file of the + * VirtualBox OSE distribution. VirtualBox OSE is distributed in the + * hope that it will be useful, but WITHOUT ANY WARRANTY of any kind. + * + * The contents of this file may alternatively be used under the terms + * of the Common Development and Distribution License Version 1.0 + * (CDDL) only, as it comes in the "COPYING.CDDL" file of the + * VirtualBox OSE distribution, in which case the provisions of the + * CDDL are applicable instead of those of the GPL. + * + * You may elect to license modified versions of this file under the + * terms and conditions of either the GPL or the CDDL or both. + */ + + +/********************************************************************************************************************************* +* Header Files * +*********************************************************************************************************************************/ +#include <iprt/json.h> +#include "internal/iprt.h" + +#include <iprt/asm.h> +#include <iprt/assert.h> +#include <iprt/ctype.h> +#include <iprt/err.h> +#include <iprt/mem.h> +#include <iprt/stream.h> +#include <iprt/string.h> +#include <iprt/utf16.h> + +#include <stdlib.h> /* strtod() */ +#include <errno.h> /* errno */ + + +/********************************************************************************************************************************* +* Structures and Typedefs * +*********************************************************************************************************************************/ + +/** + * JSON parser position information. + */ +typedef struct RTJSONPOS +{ + /** Line in the source. */ + size_t iLine; + /** Current start character .*/ + size_t iChStart; + /** Current end character. */ + size_t iChEnd; +} RTJSONPOS; +/** Pointer to a position. */ +typedef RTJSONPOS *PRTJSONPOS; + +/** + * JSON token class. + */ +typedef enum RTJSONTOKENCLASS +{ + /** Invalid. */ + RTJSONTOKENCLASS_INVALID = 0, + /** Array begin. */ + RTJSONTOKENCLASS_BEGIN_ARRAY, + /** Object begin. */ + RTJSONTOKENCLASS_BEGIN_OBJECT, + /** Array end. */ + RTJSONTOKENCLASS_END_ARRAY, + /** Object end. */ + RTJSONTOKENCLASS_END_OBJECT, + /** Separator for name/value pairs. */ + RTJSONTOKENCLASS_NAME_SEPARATOR, + /** Value separator. */ + RTJSONTOKENCLASS_VALUE_SEPARATOR, + /** String */ + RTJSONTOKENCLASS_STRING, + /** Integer number. */ + RTJSONTOKENCLASS_INTEGER, + /** Floating point number. */ + RTJSONTOKENCLASS_NUMBER, + /** null keyword. */ + RTJSONTOKENCLASS_NULL, + /** false keyword. */ + RTJSONTOKENCLASS_FALSE, + /** true keyword. */ + RTJSONTOKENCLASS_TRUE, + /** End of stream */ + RTJSONTOKENCLASS_EOS, + /** 32bit hack. */ + RTJSONTOKENCLASS_32BIT_HACK = 0x7fffffff +} RTJSONTOKENCLASS; +/** Pointer to a token class. */ +typedef RTJSONTOKENCLASS *PRTJSONTOKENCLASS; + +/** + * JSON token. + */ +typedef struct RTJSONTOKEN +{ + /** Token class. */ + RTJSONTOKENCLASS enmClass; + /** Token position in the source buffer. */ + RTJSONPOS Pos; + /** Data based on the token class. */ + union + { + /** String. */ + struct + { + /** Pointer to the start of the string. */ + char *pszStr; + } String; + /** Number. */ + struct + { + int64_t i64Num; + } Integer; + /** Floating point number. */ + double rdNum; + } Class; +} RTJSONTOKEN; +/** Pointer to a JSON token. */ +typedef RTJSONTOKEN *PRTJSONTOKEN; +/** Pointer to a const script token. */ +typedef const RTJSONTOKEN *PCRTJSONTOKEN; + +/** + * Tokenizer read input callback. + * + * @returns IPRT status code. + * @param pvUser Opaque user data for the callee. + * @param offInput Start offset from the start of the input stream to read from. + * @param pvBuf Where to store the read data. + * @param cbBuf How much to read. + * @param pcbRead Where to store the amount of data read on succcess. + */ +typedef DECLCALLBACK(int) FNRTJSONTOKENIZERREAD(void *pvUser, size_t offInput, void *pvBuf, size_t cbBuf, + size_t *pcbRead); +/** Pointer to a tokenizer read buffer callback. */ +typedef FNRTJSONTOKENIZERREAD *PFNRTJSONTOKENIZERREAD; + +/** + * Tokenizer state. + */ +typedef struct RTJSONTOKENIZER +{ + /** Read callback. */ + PFNRTJSONTOKENIZERREAD pfnRead; + /** Opaque user data. */ + void *pvUser; + /** Current offset into the input stream. */ + size_t offInput; + /** Number of valid bytes in the input buffer. */ + size_t cbBuf; + /** Current offset into the input buffer. */ + size_t offBuf; + /** Input cache buffer. */ + char achBuf[512]; + /** Current position into the input stream. */ + RTJSONPOS Pos; + /** Token 1. */ + RTJSONTOKEN Token1; + /** Token 2. */ + RTJSONTOKEN Token2; + /** Pointer to the current active token. */ + PRTJSONTOKEN pTokenCurr; + /** The next token in the input stream (used for peeking). */ + PRTJSONTOKEN pTokenNext; + /** The tokenizer error state. */ + int rcTok; + /** Where to return extended error information.*/ + PRTERRINFO pErrInfo; +} RTJSONTOKENIZER; +/** Pointer to a JSON tokenizer. */ +typedef RTJSONTOKENIZER *PRTJSONTOKENIZER; + +/** Pointer to the internal JSON value instance. */ +typedef struct RTJSONVALINT *PRTJSONVALINT; + +/** + * A JSON value. + */ +typedef struct RTJSONVALINT +{ + /** Type of the JSON value. */ + RTJSONVALTYPE enmType; + /** Reference count for this JSON value. */ + volatile uint32_t cRefs; + /** Type dependent data. */ + union + { + /** String type*/ + struct + { + /** Pointer to the string. */ + char *pszStr; + } String; + /** Number type. */ + struct + { + /** Signed 64-bit integer. */ + int64_t i64Num; + } Integer; + /** Floating point number . */ + double rdNum; + /** Array type. */ + struct + { + /** Number of elements in the array. */ + unsigned cItems; + /** Pointer to the array of items. */ + PRTJSONVALINT *papItems; + } Array; + /** Object type. */ + struct + { + /** Number of members. */ + unsigned cMembers; + /** Pointer to the array holding the member names. */ + char **papszNames; + /** Pointer to the array holding the values. */ + PRTJSONVALINT *papValues; + } Object; + } Type; +} RTJSONVALINT; + +/** + * A JSON iterator. + */ +typedef struct RTJSONITINT +{ + /** Referenced JSON value. */ + PRTJSONVALINT pJsonVal; + /** Current index. */ + unsigned idxCur; +} RTJSONITINT; +/** Pointer to the internal JSON iterator instance. */ +typedef RTJSONITINT *PRTJSONITINT; + +/** + * Passing arguments for the read callbacks. + */ +typedef struct RTJSONREADERARGS +{ + /** Buffer/File size */ + size_t cbData; + /** Data specific for one callback. */ + union + { + PRTSTREAM hStream; + const uint8_t *pbBuf; + } u; +} RTJSONREADERARGS; +/** Pointer to a readers argument. */ +typedef RTJSONREADERARGS *PRTJSONREADERARGS; + + +/********************************************************************************************************************************* +* Internal Functions * +*********************************************************************************************************************************/ +static int rtJsonParseValue(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN pToken, PRTJSONVALINT *ppJsonVal); + + +/** + * Fill the input buffer from the input stream. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state. + */ +static int rtJsonTokenizerRead(PRTJSONTOKENIZER pTokenizer) +{ + size_t cbRead = 0; + int rc = pTokenizer->pfnRead(pTokenizer->pvUser, pTokenizer->offInput, &pTokenizer->achBuf[0], + sizeof(pTokenizer->achBuf), &cbRead); + if (RT_SUCCESS(rc)) + { + pTokenizer->cbBuf = cbRead; + pTokenizer->offInput += cbRead; + pTokenizer->offBuf = 0; + /* Validate UTF-8 encoding. */ + rc = RTStrValidateEncodingEx(&pTokenizer->achBuf[0], cbRead, 0 /* fFlags */); + /* If we read less than requested we reached the end and fill the remainder with terminators. */ + if (cbRead < sizeof(pTokenizer->achBuf)) + memset(&pTokenizer->achBuf[cbRead], 0, sizeof(pTokenizer->achBuf) - cbRead); + } + + return rc; +} + +/** + * Skips the given amount of characters in the input stream. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state. + * @param cchSkip The amount of characters to skip. + */ +static int rtJsonTokenizerSkip(PRTJSONTOKENIZER pTokenizer, size_t cchSkip) +{ + int rc = VINF_SUCCESS; + + /* + * In case we reached the end of the stream don't even attempt to read new data. + * Safety precaution for possible bugs in the parser causing out of bounds reads + */ + if (pTokenizer->achBuf[pTokenizer->offBuf] == '\0') + return rc; + + while ( cchSkip > 0 + && pTokenizer->offBuf < pTokenizer->cbBuf + && RT_SUCCESS(rc)) + { + size_t cchThisSkip = RT_MIN(cchSkip, pTokenizer->cbBuf - pTokenizer->offBuf); + + pTokenizer->offBuf += cchThisSkip; + /* Read new data if required and we didn't reach the end yet. */ + if ( pTokenizer->offBuf == pTokenizer->cbBuf + && pTokenizer->cbBuf == sizeof(pTokenizer->achBuf)) + rc = rtJsonTokenizerRead(pTokenizer); + + cchSkip -= cchThisSkip; + } + + return rc; +} + + +/** + * Returns whether the tokenizer reached the end of the stream. + * + * @returns true if the tokenizer reached the end of stream marker + * false otherwise. + * @param pTokenizer The tokenizer state. + */ +DECLINLINE(bool) rtJsonTokenizerIsEos(PRTJSONTOKENIZER pTokenizer) +{ + return pTokenizer->achBuf[pTokenizer->offBuf] == '\0'; +} + +/** + * Skip one character in the input stream. + * + * @returns nothing. + * @param pTokenizer The tokenizer state. + */ +DECLINLINE(void) rtJsonTokenizerSkipCh(PRTJSONTOKENIZER pTokenizer) +{ + rtJsonTokenizerSkip(pTokenizer, 1); + pTokenizer->Pos.iChStart++; + pTokenizer->Pos.iChEnd++; +} + +/** + * Returns the next char in the input buffer without advancing it. + * + * @returns Next character in the input buffer. + * @param pTokenizer The tokenizer state. + */ +DECLINLINE(char) rtJsonTokenizerPeekCh(PRTJSONTOKENIZER pTokenizer) +{ + return !rtJsonTokenizerIsEos(pTokenizer) + ? pTokenizer->achBuf[pTokenizer->offBuf + 1] /** @todo Read out of bounds */ + : '\0'; +} + +/** + * Returns the next character in the input buffer advancing the internal + * position. + * + * @returns Next character in the stream. + * @param pTokenizer The tokenizer state. + */ +DECLINLINE(char) rtJsonTokenizerGetCh(PRTJSONTOKENIZER pTokenizer) +{ + char ch; + + if (!rtJsonTokenizerIsEos(pTokenizer)) + ch = pTokenizer->achBuf[pTokenizer->offBuf]; + else + ch = '\0'; + + return ch; +} + +/** + * Sets a new line for the tokenizer. + * + * @returns nothing. + * @param pTokenizer The tokenizer state. + * @param cSkip Amount of characters to skip making up the new line. + */ +DECLINLINE(void) rtJsonTokenizerNewLine(PRTJSONTOKENIZER pTokenizer, unsigned cSkip) +{ + rtJsonTokenizerSkip(pTokenizer, cSkip); + pTokenizer->Pos.iLine++; + pTokenizer->Pos.iChStart = 1; + pTokenizer->Pos.iChEnd = 1; +} + +/** + * Checks whether the current position in the input stream is a new line + * and skips it. + * + * @returns Flag whether there was a new line at the current position + * in the input buffer. + * @param pTokenizer The tokenizer state. + */ +DECLINLINE(bool) rtJsonTokenizerIsSkipNewLine(PRTJSONTOKENIZER pTokenizer) +{ + bool fNewline = true; + + if ( rtJsonTokenizerGetCh(pTokenizer) == '\r' + && rtJsonTokenizerPeekCh(pTokenizer) == '\n') + rtJsonTokenizerNewLine(pTokenizer, 2); + else if (rtJsonTokenizerGetCh(pTokenizer) == '\n') + rtJsonTokenizerNewLine(pTokenizer, 1); + else + fNewline = false; + + return fNewline; +} + +/** + * Skip all whitespace starting from the current input buffer position. + * Skips all present comments too. + * + * @returns nothing. + * @param pTokenizer The tokenizer state. + */ +DECLINLINE(void) rtJsonTokenizerSkipWhitespace(PRTJSONTOKENIZER pTokenizer) +{ + while (!rtJsonTokenizerIsEos(pTokenizer)) + { + while ( rtJsonTokenizerGetCh(pTokenizer) == ' ' + || rtJsonTokenizerGetCh(pTokenizer) == '\t') + rtJsonTokenizerSkipCh(pTokenizer); + + if ( !rtJsonTokenizerIsEos(pTokenizer) + && !rtJsonTokenizerIsSkipNewLine(pTokenizer)) + break; /* Skipped everything, next is some real content. */ + } +} + +/** + * Get an literal token from the tokenizer. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state. + * @param pToken The uninitialized token. + */ +static int rtJsonTokenizerGetLiteral(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN pToken) +{ + int rc = VINF_SUCCESS; + char ch = rtJsonTokenizerGetCh(pTokenizer); + size_t cchLiteral = 0; + char szLiteral[6]; /* false + 0 terminator as the lingest possible literal. */ + RT_ZERO(szLiteral); + + pToken->Pos = pTokenizer->Pos; + + Assert(RT_C_IS_ALPHA(ch)); + + while ( RT_C_IS_ALPHA(ch) + && cchLiteral < RT_ELEMENTS(szLiteral) - 1) + { + szLiteral[cchLiteral] = ch; + cchLiteral++; + rtJsonTokenizerSkipCh(pTokenizer); + ch = rtJsonTokenizerGetCh(pTokenizer); + } + + if (!RTStrNCmp(&szLiteral[0], "false", RT_ELEMENTS(szLiteral))) + pToken->enmClass = RTJSONTOKENCLASS_FALSE; + else if (!RTStrNCmp(&szLiteral[0], "true", RT_ELEMENTS(szLiteral))) + pToken->enmClass = RTJSONTOKENCLASS_TRUE; + else if (!RTStrNCmp(&szLiteral[0], "null", RT_ELEMENTS(szLiteral))) + pToken->enmClass = RTJSONTOKENCLASS_NULL; + else + { + pToken->enmClass = RTJSONTOKENCLASS_INVALID; + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "malformed literal '%.6s' (line %zu col %zu)", + &szLiteral[0], pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + + pToken->Pos.iChEnd += cchLiteral; + return rc; +} + +/** + * Get a numerical constant from the tokenizer. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state. + * @param pToken The uninitialized token. + */ +static int rtJsonTokenizerGetNumber(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN pToken) +{ + size_t cchNum = 0; + char szTmp[128]; /* Everything larger is not possible to display in signed 64bit. */ + + pToken->enmClass = RTJSONTOKENCLASS_INTEGER; + + char ch = rtJsonTokenizerGetCh(pTokenizer); + if (ch == '-') + { + szTmp[cchNum++] = '-'; + rtJsonTokenizerSkipCh(pTokenizer); + ch = rtJsonTokenizerGetCh(pTokenizer); + } + + while ( RT_C_IS_DIGIT(ch) + && cchNum < sizeof(szTmp) - 1) + { + szTmp[cchNum] = ch; + cchNum++; + rtJsonTokenizerSkipCh(pTokenizer); + ch = rtJsonTokenizerGetCh(pTokenizer); + } + + int rc = VINF_SUCCESS; + if (RT_C_IS_DIGIT(ch) && cchNum >= sizeof(szTmp) - 1) + rc = VERR_NUMBER_TOO_BIG; + else if (ch != '.') + { + szTmp[cchNum] = '\0'; + rc = RTStrToInt64Ex(&szTmp[0], NULL, 10, &pToken->Class.Integer.i64Num); + Assert(RT_SUCCESS(rc) || rc == VWRN_NUMBER_TOO_BIG); + if (rc == VWRN_NUMBER_TOO_BIG) + rc = VERR_NUMBER_TOO_BIG; + } + else + { + /* + * A floating point value. + */ + pToken->enmClass = RTJSONTOKENCLASS_NUMBER; + rtJsonTokenizerSkipCh(pTokenizer); + szTmp[cchNum++] = '.'; + + ch = rtJsonTokenizerGetCh(pTokenizer); + while ( RT_C_IS_DIGIT(ch) + && cchNum < sizeof(szTmp) - 1) + { + szTmp[cchNum++] = ch; + rtJsonTokenizerSkipCh(pTokenizer); + ch = rtJsonTokenizerGetCh(pTokenizer); + } + if ( (ch == 'e' || ch == 'E') + && cchNum < sizeof(szTmp) - 2) + { + szTmp[cchNum++] = 'e'; + rtJsonTokenizerSkipCh(pTokenizer); + ch = rtJsonTokenizerGetCh(pTokenizer); + if (ch == '+' || ch == '-') + { + szTmp[cchNum++] = ch; + rtJsonTokenizerSkipCh(pTokenizer); + ch = rtJsonTokenizerGetCh(pTokenizer); + } + while ( RT_C_IS_DIGIT(ch) + && cchNum < sizeof(szTmp) - 1) + { + szTmp[cchNum++] = ch; + rtJsonTokenizerSkipCh(pTokenizer); + ch = rtJsonTokenizerGetCh(pTokenizer); + } + } + if (cchNum < sizeof(szTmp) - 1) + { + szTmp[cchNum] = '\0'; + + /** @todo Not sure if strtod does the 100% right thing here... */ + errno = 0; + char *pszNext = NULL; + pToken->Class.rdNum = strtod(szTmp, &pszNext); + if (errno == 0) + { + rc = VINF_SUCCESS; + Assert(!pszNext || *pszNext == '\0'); + } + else + rc = RTErrConvertFromErrno(errno); + } + } + + return rc; +} + +/** + * Parses a string constant. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state. + * @param pToken The uninitialized token. + */ +static int rtJsonTokenizerGetString(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN pToken) +{ + size_t cchStrMax = 64; + char *pszDecoded = (char *)RTStrAlloc(cchStrMax); + AssertReturn(pszDecoded, VERR_NO_STR_MEMORY); + + Assert(rtJsonTokenizerGetCh(pTokenizer) == '\"'); + rtJsonTokenizerSkipCh(pTokenizer); /* Skip " */ + + pToken->enmClass = RTJSONTOKENCLASS_STRING; + pToken->Pos = pTokenizer->Pos; + + size_t cchStr = 0; + char ch = rtJsonTokenizerGetCh(pTokenizer); + while ( ch != '\"' + && ch != '\0') + { + if (ch != '\\') + { + pszDecoded[cchStr++] = ch; + rtJsonTokenizerSkipCh(pTokenizer); + } + else + { + /* Escape sequence, check the next character */ + rtJsonTokenizerSkipCh(pTokenizer); + char chNext = rtJsonTokenizerGetCh(pTokenizer); + switch (chNext) + { + case '\"': + pszDecoded[cchStr++] = '\"'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case '\\': + pszDecoded[cchStr++] = '\\'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case '/': + pszDecoded[cchStr++] = '/'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case 'b': + pszDecoded[cchStr++] = '\b'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case 'n': + pszDecoded[cchStr++] = '\n'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case 'f': + pszDecoded[cchStr++] = '\f'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case 'r': + pszDecoded[cchStr++] = '\r'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case 't': + pszDecoded[cchStr++] = '\t'; + rtJsonTokenizerSkipCh(pTokenizer); + break; + case 'u': + { + /* \uXXXX */ + int rc = VERR_JSON_INVALID_UTF16_ESCAPE_SEQUENCE; + rtJsonTokenizerSkipCh(pTokenizer); + char chX1 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX1)) + { + rtJsonTokenizerSkipCh(pTokenizer); + char chX2 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX2)) + { + rtJsonTokenizerSkipCh(pTokenizer); + char chX3 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX3)) + { + rtJsonTokenizerSkipCh(pTokenizer); + char chX4 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX4)) + { + rtJsonTokenizerSkipCh(pTokenizer); + + RTUNICP uc = ((RTUTF16)(chX1 <= '9' ? chX1 - '0' : (chX1 & 7) + 9) << 12) + | ((RTUTF16)(chX2 <= '9' ? chX2 - '0' : (chX2 & 7) + 9) << 8) + | ((RTUTF16)(chX3 <= '9' ? chX3 - '0' : (chX3 & 7) + 9) << 4) + | ((RTUTF16)(chX4 <= '9' ? chX4 - '0' : (chX4 & 7) + 9)); + if ( !RTUtf16IsHighSurrogate((RTUTF16)uc) + && !RTUtf16IsLowSurrogate((RTUTF16)uc)) + rc = VINF_SUCCESS; + else if (RTUtf16IsHighSurrogate((RTUTF16)uc)) + { + /* The must be a low surrogate pair following the high one: */ + rc = VINF_SUCCESS; + ch = rtJsonTokenizerGetCh(pTokenizer); + if (ch == '\\') + rtJsonTokenizerSkipCh(pTokenizer); + else + rc = VERR_JSON_MISSING_SURROGATE_PAIR; + ch = rtJsonTokenizerGetCh(pTokenizer); + if (ch == 'u') + rtJsonTokenizerSkipCh(pTokenizer); + else + rc = VERR_JSON_MISSING_SURROGATE_PAIR; + chX1 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX1)) + rtJsonTokenizerSkipCh(pTokenizer); + else if (RT_SUCCESS_NP(rc)) + rc = VERR_JSON_INVALID_UTF16_ESCAPE_SEQUENCE; + chX2 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX2)) + rtJsonTokenizerSkipCh(pTokenizer); + else if (RT_SUCCESS_NP(rc)) + rc = VERR_JSON_INVALID_UTF16_ESCAPE_SEQUENCE; + chX3 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX3)) + rtJsonTokenizerSkipCh(pTokenizer); + else if (RT_SUCCESS_NP(rc)) + rc = VERR_JSON_INVALID_UTF16_ESCAPE_SEQUENCE; + chX4 = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_XDIGIT(chX4)) + rtJsonTokenizerSkipCh(pTokenizer); + else if (RT_SUCCESS_NP(rc)) + rc = VERR_JSON_INVALID_UTF16_ESCAPE_SEQUENCE; + if (RT_SUCCESS(rc)) + { + RTUTF16 wc2 = ((RTUTF16)(chX1 <= '9' ? chX1 - '0' : (chX1 & 7) + 9) << 12) + | ((RTUTF16)(chX2 <= '9' ? chX2 - '0' : (chX2 & 7) + 9) << 8) + | ((RTUTF16)(chX3 <= '9' ? chX3 - '0' : (chX3 & 7) + 9) << 4) + | ((RTUTF16)(chX4 <= '9' ? chX4 - '0' : (chX4 & 7) + 9)); + if (RTUtf16IsLowSurrogate(wc2)) + uc = 0x10000 + (((uc & 0x3ff) << 10) | (wc2 & 0x3ff)); + else + rc = VERR_JSON_BAD_SURROGATE_PAIR_SEQUENCE; + } + } + else + rc = VERR_JSON_BAD_SURROGATE_PAIR_SEQUENCE; + if (RT_SUCCESS(rc)) + { + if ( uc != 0 + && uc != 0xfffe + && uc != 0xffff) + { + Assert(cchStr + RTStrCpSize(uc) < cchStrMax); + char *pszNext = RTStrPutCp(&pszDecoded[cchStr], uc); + Assert((size_t)(pszNext - &pszDecoded[cchStr]) == RTStrCpSize(uc)); + cchStr += pszNext - &pszDecoded[cchStr]; + break; + } + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_INVALID_CODEPOINT, + "Invalid \\u code point: %#x (line %zu col %zu)", + uc, pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + } + } + } + } + RTStrFree(pszDecoded); + if (rc == VERR_JSON_INVALID_UTF16_ESCAPE_SEQUENCE) + rc = RTErrInfoSetF(pTokenizer->pErrInfo, rc, "Invalid \\u escape sequence (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + else if (rc == VERR_JSON_MISSING_SURROGATE_PAIR) + rc = RTErrInfoSetF(pTokenizer->pErrInfo, rc, "Missing UTF-16 surrogate pair (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + else if (rc == VERR_JSON_BAD_SURROGATE_PAIR_SEQUENCE) + rc = RTErrInfoSetF(pTokenizer->pErrInfo, rc, "Invalid UTF-16 surrogate pair (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + return rc; + } + + default: + RTStrFree(pszDecoded); + return RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "bad escape sequence (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + } + + + if (cchStr < cchStrMax - 4) + { /* likely */ } + else + { + /* Increase string space. */ + size_t cchStrMaxNew = cchStrMax < _4K ? cchStrMax * 2 : cchStrMax + _4K; + int rc = RTStrRealloc(&pszDecoded, cchStrMaxNew); + if (RT_SUCCESS(rc)) + cchStrMax = cchStrMaxNew; + else + { + RTStrFree(pszDecoded); + return rc; + } + } + ch = rtJsonTokenizerGetCh(pTokenizer); + } + + if (ch == '\"') + rtJsonTokenizerSkipCh(pTokenizer); /* Skip closing " */ + + Assert(cchStr < cchStrMax); + pszDecoded[cchStr] = '\0'; + if (cchStrMax - cchStr >= cchStrMax / 2) + RTStrRealloc(&pszDecoded, cchStr + 1); + pToken->Class.String.pszStr = pszDecoded; + + pToken->Pos.iChEnd = pTokenizer->Pos.iChEnd; + return VINF_SUCCESS; +} + +/** + * Get the end of stream token. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state. + * @param pToken The uninitialized token. + */ +static int rtJsonTokenizerGetEos(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN pToken) +{ + Assert(rtJsonTokenizerGetCh(pTokenizer) == '\0'); + + pToken->enmClass = RTJSONTOKENCLASS_EOS; + pToken->Pos = pTokenizer->Pos; + return VINF_SUCCESS; +} + +/** + * Read the next token from the tokenizer stream. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer to read from. + * @param pToken Uninitialized token to fill the token data into. + */ +static int rtJsonTokenizerReadNextToken(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN pToken) +{ + int rc = VINF_SUCCESS; + + /* Skip all eventually existing whitespace and newlines first. */ + rtJsonTokenizerSkipWhitespace(pTokenizer); + + char ch = rtJsonTokenizerGetCh(pTokenizer); + if (RT_C_IS_ALPHA(ch)) + rc = rtJsonTokenizerGetLiteral(pTokenizer, pToken); + else if (RT_C_IS_DIGIT(ch) || ch == '-') + rc = rtJsonTokenizerGetNumber(pTokenizer, pToken); + else if (ch == '\"') + rc = rtJsonTokenizerGetString(pTokenizer, pToken); + else if (ch == '\0') + rc = rtJsonTokenizerGetEos(pTokenizer, pToken); + else if (ch == '{') + { + pToken->enmClass = RTJSONTOKENCLASS_BEGIN_OBJECT; + rtJsonTokenizerSkipCh(pTokenizer); + } + else if (ch == '}') + { + pToken->enmClass = RTJSONTOKENCLASS_END_OBJECT; + rtJsonTokenizerSkipCh(pTokenizer); + } + else if (ch == '[') + { + pToken->enmClass = RTJSONTOKENCLASS_BEGIN_ARRAY; + rtJsonTokenizerSkipCh(pTokenizer); + } + else if (ch == ']') + { + pToken->enmClass = RTJSONTOKENCLASS_END_ARRAY; + rtJsonTokenizerSkipCh(pTokenizer); + } + else if (ch == ':') + { + pToken->enmClass = RTJSONTOKENCLASS_NAME_SEPARATOR; + rtJsonTokenizerSkipCh(pTokenizer); + } + else if (ch == ',') + { + pToken->enmClass = RTJSONTOKENCLASS_VALUE_SEPARATOR; + rtJsonTokenizerSkipCh(pTokenizer); + } + else + { + pToken->enmClass = RTJSONTOKENCLASS_INVALID; + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "bad token '%c' (line %zu col %zu)", + ch, pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + + if (RT_FAILURE(rc)) + pTokenizer->rcTok = rc; + + return rc; +} + +/** + * Create a new tokenizer. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state to initialize. + * @param pfnRead Read callback for the input stream. + * @param pvUser Opaque user data to pass to the callback. + * @param pErrInfo Where to return extended error info. + */ +static int rtJsonTokenizerInit(PRTJSONTOKENIZER pTokenizer, PFNRTJSONTOKENIZERREAD pfnRead, void *pvUser, PRTERRINFO pErrInfo) +{ + pTokenizer->pfnRead = pfnRead; + pTokenizer->pvUser = pvUser; + pTokenizer->offInput = 0; + pTokenizer->cbBuf = 0; + pTokenizer->offBuf = 0; + pTokenizer->Pos.iLine = 1; + pTokenizer->Pos.iChStart = 1; + pTokenizer->Pos.iChEnd = 1; + pTokenizer->pTokenCurr = &pTokenizer->Token1; + pTokenizer->pTokenNext = &pTokenizer->Token2; + pTokenizer->rcTok = VINF_SUCCESS; + pTokenizer->pErrInfo = pErrInfo; + + RT_ZERO(pTokenizer->achBuf); + + /* Fill the input buffer. */ + int rc = rtJsonTokenizerRead(pTokenizer); + + /* Fill the tokenizer with two first tokens. */ + if (RT_SUCCESS(rc)) + rc = rtJsonTokenizerReadNextToken(pTokenizer, pTokenizer->pTokenCurr); + if (RT_SUCCESS(rc)) + rc = rtJsonTokenizerReadNextToken(pTokenizer, pTokenizer->pTokenNext); + + return rc; +} + +/** + * Destroys a given tokenizer state. + * + * @returns nothing. + * @param pTokenizer The tokenizer to destroy. + */ +static void rtJsonTokenizerDestroy(PRTJSONTOKENIZER pTokenizer) +{ + RT_NOREF_PV(pTokenizer); +} + +/** + * Get the current token in the input stream. + * + * @returns Pointer to the next token in the stream. + * @param pTokenizer The tokenizer state. + * @param ppToken Where to store the pointer to the current token on success. + */ +DECLINLINE(int) rtJsonTokenizerGetToken(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN *ppToken) +{ + if (RT_SUCCESS(pTokenizer->rcTok)) + { + *ppToken = pTokenizer->pTokenCurr; + return VINF_SUCCESS; + } + + return pTokenizer->rcTok; +} + +/** + * Consume the current token advancing to the next in the stream. + * + * @returns nothing. + * @param pTokenizer The tokenizer state. + */ +static void rtJsonTokenizerConsume(PRTJSONTOKENIZER pTokenizer) +{ + PRTJSONTOKEN pTokenTmp = pTokenizer->pTokenCurr; + + /* Switch next token to current token and read in the next token. */ + pTokenizer->pTokenCurr = pTokenizer->pTokenNext; + pTokenizer->pTokenNext = pTokenTmp; + rtJsonTokenizerReadNextToken(pTokenizer, pTokenizer->pTokenNext); +} + +/** + * Consumes the current token if it matches the given class returning an indicator. + * + * @returns true if the class matched and the token was consumed. + * @retval false otherwise. + * @param pTokenizer The tokenizer state. + * @param enmClass The token class to match against. + */ +static bool rtJsonTokenizerConsumeIfMatched(PRTJSONTOKENIZER pTokenizer, RTJSONTOKENCLASS enmClass) +{ + PRTJSONTOKEN pToken = NULL; + int rc = rtJsonTokenizerGetToken(pTokenizer, &pToken); + if (RT_SUCCESS(rc)) + { + if (pToken->enmClass == enmClass) + { + rtJsonTokenizerConsume(pTokenizer); + return true; + } + } + + return false; +} + +/** + * Destroys a given JSON value releasing the reference to all child values. + * + * @returns nothing. + * @param pThis The JSON value to destroy. + */ +static void rtJsonValDestroy(PRTJSONVALINT pThis) +{ + switch (pThis->enmType) + { + case RTJSONVALTYPE_OBJECT: + for (unsigned i = 0; i < pThis->Type.Object.cMembers; i++) + { + RTStrFree(pThis->Type.Object.papszNames[i]); + RTJsonValueRelease(pThis->Type.Object.papValues[i]); + } + RTMemFree(pThis->Type.Object.papszNames); + RTMemFree(pThis->Type.Object.papValues); + break; + case RTJSONVALTYPE_ARRAY: + for (unsigned i = 0; i < pThis->Type.Array.cItems; i++) + RTJsonValueRelease(pThis->Type.Array.papItems[i]); + RTMemFree(pThis->Type.Array.papItems); + break; + case RTJSONVALTYPE_STRING: + RTStrFree(pThis->Type.String.pszStr); + break; + case RTJSONVALTYPE_INTEGER: + case RTJSONVALTYPE_NUMBER: + case RTJSONVALTYPE_NULL: + case RTJSONVALTYPE_TRUE: + case RTJSONVALTYPE_FALSE: + /* nothing to do. */ + break; + default: + AssertMsgFailed(("Invalid JSON value type: %u\n", pThis->enmType)); + } + RTMemFree(pThis); +} + +/** + * Creates a new JSON value with the given type. + * + * @returns Pointer to JSON value on success, NULL if out of memory. + * @param enmType The JSON value type. + */ +static PRTJSONVALINT rtJsonValueCreate(RTJSONVALTYPE enmType) +{ + PRTJSONVALINT pThis = (PRTJSONVALINT)RTMemAllocZ(sizeof(RTJSONVALINT)); + if (RT_LIKELY(pThis)) + { + pThis->enmType = enmType; + pThis->cRefs = 1; + } + + return pThis; +} + +/** + * Parses an JSON array. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer to use. + * @param pJsonVal The JSON array value to fill in. + */ +static int rtJsonParseArray(PRTJSONTOKENIZER pTokenizer, PRTJSONVALINT pJsonVal) +{ + int rc = VINF_SUCCESS; + PRTJSONTOKEN pToken = NULL; + uint32_t cItems = 0; + uint32_t cItemsMax = 0; + PRTJSONVALINT *papItems = NULL; + + rc = rtJsonTokenizerGetToken(pTokenizer, &pToken); + while ( RT_SUCCESS(rc) + && pToken->enmClass != RTJSONTOKENCLASS_END_ARRAY + && pToken->enmClass != RTJSONTOKENCLASS_EOS) + { + PRTJSONVALINT pVal = NULL; + rc = rtJsonParseValue(pTokenizer, pToken, &pVal); + if (RT_SUCCESS(rc)) + { + if (cItems == cItemsMax) + { + cItemsMax += 10; + PRTJSONVALINT *papItemsNew = (PRTJSONVALINT *)RTMemRealloc(papItems, cItemsMax * sizeof(PRTJSONVALINT)); + if (RT_UNLIKELY(!papItemsNew)) + { + rc = VERR_NO_MEMORY; + break; + } + papItems = papItemsNew; + } + + Assert(cItems < cItemsMax); + papItems[cItems] = pVal; + cItems++; + } + + /* Skip value separator and continue with next token. */ + bool fSkippedSep = rtJsonTokenizerConsumeIfMatched(pTokenizer, RTJSONTOKENCLASS_VALUE_SEPARATOR); + rc = rtJsonTokenizerGetToken(pTokenizer, &pToken); + + if ( RT_SUCCESS(rc) + && !fSkippedSep + && pToken->enmClass != RTJSONTOKENCLASS_END_ARRAY) + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "expected end of array (#1) (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + + if (RT_SUCCESS(rc)) + { + if (pToken->enmClass == RTJSONTOKENCLASS_END_ARRAY) + { + rtJsonTokenizerConsume(pTokenizer); + pJsonVal->Type.Array.cItems = cItems; + pJsonVal->Type.Array.papItems = papItems; + } + else + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "expected end of array (#2) (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + + if (RT_FAILURE(rc)) + { + for (uint32_t i = 0; i < cItems; i++) + RTJsonValueRelease(papItems[i]); + RTMemFree(papItems); + } + + return rc; +} + +/** + * Parses an JSON object. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer to use. + * @param pJsonVal The JSON object value to fill in. + */ +static int rtJsonParseObject(PRTJSONTOKENIZER pTokenizer, PRTJSONVALINT pJsonVal) +{ + int rc = VINF_SUCCESS; + PRTJSONTOKEN pToken = NULL; + uint32_t cMembers = 0; + uint32_t cMembersMax = 0; + PRTJSONVALINT *papValues = NULL; + char **papszNames = NULL; + + rc = rtJsonTokenizerGetToken(pTokenizer, &pToken); + while ( RT_SUCCESS(rc) + && pToken->enmClass == RTJSONTOKENCLASS_STRING) + { + char *pszName = pToken->Class.String.pszStr; /* We can consume this string as it was allocated. */ + + rtJsonTokenizerConsume(pTokenizer); + if (rtJsonTokenizerConsumeIfMatched(pTokenizer, RTJSONTOKENCLASS_NAME_SEPARATOR)) + { + PRTJSONVALINT pVal = NULL; + rc = rtJsonTokenizerGetToken(pTokenizer, &pToken); + if (RT_SUCCESS(rc)) + rc = rtJsonParseValue(pTokenizer, pToken, &pVal); + if (RT_SUCCESS(rc)) + { + if (cMembers == cMembersMax) + { + cMembersMax += 10; + PRTJSONVALINT *papValuesNew = (PRTJSONVALINT *)RTMemRealloc(papValues, cMembersMax * sizeof(PRTJSONVALINT)); + char **papszNamesNew = (char **)RTMemRealloc(papszNames, cMembersMax * sizeof(char *)); + if (RT_UNLIKELY(!papValuesNew || !papszNamesNew)) + { + if (papValuesNew) + RTMemFree(papValuesNew); + if (papszNamesNew) + RTMemFree(papszNamesNew); + rc = VERR_NO_MEMORY; + break; + } + + papValues = papValuesNew; + papszNames = papszNamesNew; + } + + Assert(cMembers < cMembersMax); + papszNames[cMembers] = pszName; + papValues[cMembers] = pVal; + cMembers++; + + /* Skip value separator and continue with next token. */ + bool fSkippedSep = rtJsonTokenizerConsumeIfMatched(pTokenizer, RTJSONTOKENCLASS_VALUE_SEPARATOR); + rc = rtJsonTokenizerGetToken(pTokenizer, &pToken); + + if ( RT_SUCCESS(rc) + && !fSkippedSep + && pToken->enmClass != RTJSONTOKENCLASS_END_OBJECT) + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "expected end of object (#1) (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + } + else + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "expected name separator (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + + if (RT_SUCCESS(rc)) + { + if (pToken->enmClass == RTJSONTOKENCLASS_END_OBJECT) + { + rtJsonTokenizerConsume(pTokenizer); + pJsonVal->Type.Object.cMembers = cMembers; + pJsonVal->Type.Object.papValues = papValues; + pJsonVal->Type.Object.papszNames = papszNames; + } + else + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "expected end of object (#2) (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + } + + if (RT_FAILURE(rc)) + { + for (uint32_t i = 0; i < cMembers; i++) + { + RTJsonValueRelease(papValues[i]); + RTStrFree(papszNames[i]); + } + RTMemFree(papValues); + RTMemFree(papszNames); + } + + return rc; +} + +/** + * Parses a single JSON value and returns it on success. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer to use. + * @param pToken The token to parse. + * @param ppJsonVal Where to store the pointer to the JSON value on success. + */ +static int rtJsonParseValue(PRTJSONTOKENIZER pTokenizer, PRTJSONTOKEN pToken, PRTJSONVALINT *ppJsonVal) +{ + int rc = VINF_SUCCESS; + PRTJSONVALINT pVal = NULL; + + switch (pToken->enmClass) + { + case RTJSONTOKENCLASS_BEGIN_ARRAY: + rtJsonTokenizerConsume(pTokenizer); + pVal = rtJsonValueCreate(RTJSONVALTYPE_ARRAY); + if (RT_LIKELY(pVal)) + rc = rtJsonParseArray(pTokenizer, pVal); + break; + case RTJSONTOKENCLASS_BEGIN_OBJECT: + rtJsonTokenizerConsume(pTokenizer); + pVal = rtJsonValueCreate(RTJSONVALTYPE_OBJECT); + if (RT_LIKELY(pVal)) + rc = rtJsonParseObject(pTokenizer, pVal); + break; + case RTJSONTOKENCLASS_STRING: + pVal = rtJsonValueCreate(RTJSONVALTYPE_STRING); + if (RT_LIKELY(pVal)) + pVal->Type.String.pszStr = pToken->Class.String.pszStr; + rtJsonTokenizerConsume(pTokenizer); + break; + case RTJSONTOKENCLASS_INTEGER: + pVal = rtJsonValueCreate(RTJSONVALTYPE_INTEGER); + if (RT_LIKELY(pVal)) + pVal->Type.Integer.i64Num = pToken->Class.Integer.i64Num; + rtJsonTokenizerConsume(pTokenizer); + break; + case RTJSONTOKENCLASS_NUMBER: + pVal = rtJsonValueCreate(RTJSONVALTYPE_NUMBER); + if (RT_LIKELY(pVal)) + pVal->Type.rdNum = pToken->Class.rdNum; + rtJsonTokenizerConsume(pTokenizer); + break; + case RTJSONTOKENCLASS_NULL: + rtJsonTokenizerConsume(pTokenizer); + pVal = rtJsonValueCreate(RTJSONVALTYPE_NULL); + break; + case RTJSONTOKENCLASS_FALSE: + rtJsonTokenizerConsume(pTokenizer); + pVal = rtJsonValueCreate(RTJSONVALTYPE_FALSE); + break; + case RTJSONTOKENCLASS_TRUE: + rtJsonTokenizerConsume(pTokenizer); + pVal = rtJsonValueCreate(RTJSONVALTYPE_TRUE); + break; + + case RTJSONTOKENCLASS_INVALID: + Assert(!pTokenizer->pErrInfo || RTErrInfoIsSet(pTokenizer->pErrInfo)); + rc = VERR_JSON_MALFORMED; + break; + case RTJSONTOKENCLASS_END_ARRAY: + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "unexpected '}' (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + break; + case RTJSONTOKENCLASS_END_OBJECT: + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "unexpected ']' (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + break; + case RTJSONTOKENCLASS_NAME_SEPARATOR: + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "unexpected ':' (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + break; + case RTJSONTOKENCLASS_VALUE_SEPARATOR: + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "unexpected ',' (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + break; + case RTJSONTOKENCLASS_EOS: + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "expected end of object (#1) (line %zu col %zu)", + pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + break; + default: + rc = RTErrInfoSetF(pTokenizer->pErrInfo, VERR_JSON_MALFORMED, "Unexpected token class %d (line %zu col %zu)", + pToken->enmClass, pTokenizer->Pos.iLine, pTokenizer->Pos.iChStart); + break; + } + + if (RT_SUCCESS(rc)) + { + if (pVal) + *ppJsonVal = pVal; + else + rc = VERR_NO_MEMORY; + } + else if (pVal) + rtJsonValDestroy(pVal); + + return rc; +} + +/** + * Entry point to parse a JSON document. + * + * @returns IPRT status code. + * @param pTokenizer The tokenizer state. + * @param ppJsonVal Where to store the root JSON value on success. + */ +static int rtJsonParse(PRTJSONTOKENIZER pTokenizer, PRTJSONVALINT *ppJsonVal) +{ + PRTJSONTOKEN pToken = NULL; + int rc = rtJsonTokenizerGetToken(pTokenizer, &pToken); + if (RT_SUCCESS(rc)) + rc = rtJsonParseValue(pTokenizer, pToken, ppJsonVal); + + return rc; +} + +/** + * Read callback for RTJsonParseFromBuf(). + */ +static DECLCALLBACK(int) rtJsonTokenizerParseFromBuf(void *pvUser, size_t offInput, + void *pvBuf, size_t cbBuf, + size_t *pcbRead) +{ + PRTJSONREADERARGS pArgs = (PRTJSONREADERARGS)pvUser; + size_t cbLeft = offInput < pArgs->cbData ? pArgs->cbData - offInput : 0; + + if (cbLeft) + memcpy(pvBuf, &pArgs->u.pbBuf[offInput], RT_MIN(cbLeft, cbBuf)); + + *pcbRead = RT_MIN(cbLeft, cbBuf); + + return VINF_SUCCESS; +} + +/** + * Read callback for RTJsonParseFromString(). + */ +static DECLCALLBACK(int) rtJsonTokenizerParseFromString(void *pvUser, size_t offInput, + void *pvBuf, size_t cbBuf, + size_t *pcbRead) +{ + const char *pszStr = (const char *)pvUser; + size_t cchStr = strlen(pszStr) + 1; /* Include zero terminator. */ + size_t cbLeft = offInput < cchStr ? cchStr - offInput : 0; + + if (cbLeft) + memcpy(pvBuf, &pszStr[offInput], RT_MIN(cbLeft, cbBuf)); + + *pcbRead = RT_MIN(cbLeft, cbBuf); + + return VINF_SUCCESS; +} + +/** + * Read callback for RTJsonParseFromFile(). + */ +static DECLCALLBACK(int) rtJsonTokenizerParseFromFile(void *pvUser, size_t offInput, + void *pvBuf, size_t cbBuf, + size_t *pcbRead) +{ + PRTJSONREADERARGS pArgs = (PRTJSONREADERARGS)pvUser; + + RT_NOREF_PV(offInput); + + size_t cbRead = 0; + int rc = RTStrmReadEx(pArgs->u.hStream, pvBuf, cbBuf, &cbRead); + if (RT_SUCCESS(rc)) + *pcbRead = cbRead; + + return rc; +} + +RTDECL(int) RTJsonParseFromBuf(PRTJSONVAL phJsonVal, const uint8_t *pbBuf, size_t cbBuf, PRTERRINFO pErrInfo) +{ + AssertPtrReturn(phJsonVal, VERR_INVALID_POINTER); + AssertPtrReturn(pbBuf, VERR_INVALID_POINTER); + AssertReturn(cbBuf > 0, VERR_INVALID_PARAMETER); + + RTJSONTOKENIZER Tokenizer; + RTJSONREADERARGS Args; + Args.cbData = cbBuf; + Args.u.pbBuf = pbBuf; + + int rc = rtJsonTokenizerInit(&Tokenizer, rtJsonTokenizerParseFromBuf, &Args, pErrInfo); + if (RT_SUCCESS(rc)) + { + rc = rtJsonParse(&Tokenizer, phJsonVal); + rtJsonTokenizerDestroy(&Tokenizer); + } + + return rc; +} + +RTDECL(int) RTJsonParseFromString(PRTJSONVAL phJsonVal, const char *pszStr, PRTERRINFO pErrInfo) +{ + AssertPtrReturn(phJsonVal, VERR_INVALID_POINTER); + AssertPtrReturn(pszStr, VERR_INVALID_POINTER); + + /** @todo r=bird: The rtJsonTokenizerParseFromString function does + * strlen() on the whole pszStr for each read. For larger strings ( + * longer than sizeof(Tokenizer.achBuf)) it would be good to join + * forces with RTJsonParseFromBuf. */ + RTJSONTOKENIZER Tokenizer; + int rc = rtJsonTokenizerInit(&Tokenizer, rtJsonTokenizerParseFromString, (void *)pszStr, pErrInfo); + if (RT_SUCCESS(rc)) + { + rc = rtJsonParse(&Tokenizer, phJsonVal); + rtJsonTokenizerDestroy(&Tokenizer); + } + + return rc; +} + +RTDECL(int) RTJsonParseFromFile(PRTJSONVAL phJsonVal, const char *pszFilename, PRTERRINFO pErrInfo) +{ + AssertPtrReturn(phJsonVal, VERR_INVALID_POINTER); + AssertPtrReturn(pszFilename, VERR_INVALID_POINTER); + + int rc = VINF_SUCCESS; + RTJSONREADERARGS Args; + + Args.cbData = 0; + rc = RTStrmOpen(pszFilename, "r", &Args.u.hStream); + if (RT_SUCCESS(rc)) + { + RTJSONTOKENIZER Tokenizer; + + rc = rtJsonTokenizerInit(&Tokenizer, rtJsonTokenizerParseFromFile, &Args, pErrInfo); + if (RT_SUCCESS(rc)) + { + rc = rtJsonParse(&Tokenizer, phJsonVal); + rtJsonTokenizerDestroy(&Tokenizer); + } + RTStrmClose(Args.u.hStream); + } + + return rc; +} + +RTDECL(uint32_t) RTJsonValueRetain(RTJSONVAL hJsonVal) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, UINT32_MAX); + + uint32_t cRefs = ASMAtomicIncU32(&pThis->cRefs); + AssertMsg(cRefs > 1 && cRefs < _1M, ("%#x %p\n", cRefs, pThis)); + return cRefs; +} + +RTDECL(uint32_t) RTJsonValueRelease(RTJSONVAL hJsonVal) +{ + PRTJSONVALINT pThis = hJsonVal; + if (pThis == NIL_RTJSONVAL) + return 0; + AssertPtrReturn(pThis, UINT32_MAX); + + uint32_t cRefs = ASMAtomicDecU32(&pThis->cRefs); + AssertMsg(cRefs < _1M, ("%#x %p\n", cRefs, pThis)); + if (cRefs == 0) + rtJsonValDestroy(pThis); + return cRefs; +} + +RTDECL(RTJSONVALTYPE) RTJsonValueGetType(RTJSONVAL hJsonVal) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, RTJSONVALTYPE_INVALID); + + if (pThis == NIL_RTJSONVAL) + return RTJSONVALTYPE_INVALID; + + return pThis->enmType; +} + + +RTDECL(const char *) RTJsonValueTypeName(RTJSONVALTYPE enmType) +{ + switch (enmType) + { + case RTJSONVALTYPE_INVALID: return "invalid"; + case RTJSONVALTYPE_OBJECT: return "object"; + case RTJSONVALTYPE_ARRAY: return "array"; + case RTJSONVALTYPE_STRING: return "string"; + case RTJSONVALTYPE_INTEGER: return "integer"; + case RTJSONVALTYPE_NUMBER: return "number"; + case RTJSONVALTYPE_NULL: return "null"; + case RTJSONVALTYPE_TRUE: return "true"; + case RTJSONVALTYPE_FALSE: return "false"; + default: return "???"; + } +} + + +#define RTJSON_ASSERT_VALID_HANDLE_AND_TYPE_RETURN(pJson, enmExpectedType, ret) do { \ + AssertPtrReturn((pJson), (ret)); \ + AssertReturn((pJson) != NIL_RTJSONVAL, (ret)); \ + AssertReturn((pJson)->enmType == (enmExpectedType), (ret)); \ + } while (0) + +#define RTJSON_TYPECHECK_RETURN(pJson, enmExpectedType) do {\ + if ((pJson)->enmType == (enmExpectedType)) { /*likely*/ } \ + else { /*AssertFailed();*/ return VERR_JSON_VALUE_INVALID_TYPE; } \ + } while (0) + + +#define RTJSON_TYPECHECK_CONTAINER_RETURN(pJson) do { \ + if ( (pJson)->enmType == RTJSONVALTYPE_ARRAY \ + || (pJson)->enmType == RTJSONVALTYPE_OBJECT) \ + { /* likely */ } \ + else { /*AssertFailed();*/ return VERR_JSON_VALUE_INVALID_TYPE;} \ + } while (0) + + +RTDECL(const char *) RTJsonValueGetString(RTJSONVAL hJsonVal) +{ + PRTJSONVALINT pThis = hJsonVal; + RTJSON_ASSERT_VALID_HANDLE_AND_TYPE_RETURN(pThis, RTJSONVALTYPE_STRING, NULL); + + return pThis->Type.String.pszStr; +} + + +RTDECL(int) RTJsonValueQueryString(RTJSONVAL hJsonVal, const char **ppszStr) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(ppszStr, VERR_INVALID_POINTER); + + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_STRING); + *ppszStr = pThis->Type.String.pszStr; + return VINF_SUCCESS; +} + +RTDECL(int) RTJsonValueQueryInteger(RTJSONVAL hJsonVal, int64_t *pi64Num) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(pi64Num, VERR_INVALID_POINTER); + + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_INTEGER); + *pi64Num = pThis->Type.Integer.i64Num; + return VINF_SUCCESS; +} + +RTDECL(int) RTJsonValueQueryNumber(RTJSONVAL hJsonVal, double *prdNum) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(prdNum, VERR_INVALID_POINTER); + + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_NUMBER); + *prdNum = pThis->Type.rdNum; + return VINF_SUCCESS; +} + +RTDECL(int) RTJsonValueQueryByName(RTJSONVAL hJsonVal, const char *pszName, PRTJSONVAL phJsonVal) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(pszName, VERR_INVALID_POINTER); + AssertPtrReturn(phJsonVal, VERR_INVALID_POINTER); + + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_OBJECT); + + int rc = VERR_NOT_FOUND; + for (unsigned i = 0; i < pThis->Type.Object.cMembers; i++) + { + if (!RTStrCmp(pThis->Type.Object.papszNames[i], pszName)) + { + RTJsonValueRetain(pThis->Type.Object.papValues[i]); + *phJsonVal = pThis->Type.Object.papValues[i]; + rc = VINF_SUCCESS; + break; + } + } + + return rc; +} + +RTDECL(int) RTJsonValueQueryIntegerByName(RTJSONVAL hJsonVal, const char *pszName, int64_t *pi64Num) +{ + RTJSONVAL hJsonValNum = NIL_RTJSONVAL; + int rc = RTJsonValueQueryByName(hJsonVal, pszName, &hJsonValNum); + if (RT_SUCCESS(rc)) + { + rc = RTJsonValueQueryInteger(hJsonValNum, pi64Num); + RTJsonValueRelease(hJsonValNum); + } + + return rc; +} + +RTDECL(int) RTJsonValueQueryNumberByName(RTJSONVAL hJsonVal, const char *pszName, double *prdNum) +{ + RTJSONVAL hJsonValNum = NIL_RTJSONVAL; + int rc = RTJsonValueQueryByName(hJsonVal, pszName, &hJsonValNum); + if (RT_SUCCESS(rc)) + { + rc = RTJsonValueQueryNumber(hJsonValNum, prdNum); + RTJsonValueRelease(hJsonValNum); + } + + return rc; +} + +RTDECL(int) RTJsonValueQueryStringByName(RTJSONVAL hJsonVal, const char *pszName, char **ppszStr) +{ + RTJSONVAL hJsonValStr = NIL_RTJSONVAL; + int rc = RTJsonValueQueryByName(hJsonVal, pszName, &hJsonValStr); + if (RT_SUCCESS(rc)) + { + const char *pszStr = NULL; + rc = RTJsonValueQueryString(hJsonValStr, &pszStr); + if (RT_SUCCESS(rc)) + { + *ppszStr = RTStrDup(pszStr); + if (!*ppszStr) + rc = VERR_NO_STR_MEMORY; + } + RTJsonValueRelease(hJsonValStr); + } + + return rc; +} + +RTDECL(int) RTJsonValueQueryBooleanByName(RTJSONVAL hJsonVal, const char *pszName, bool *pfBoolean) +{ + AssertPtrReturn(pfBoolean, VERR_INVALID_POINTER); + + RTJSONVAL hJsonValBool = NIL_RTJSONVAL; + int rc = RTJsonValueQueryByName(hJsonVal, pszName, &hJsonValBool); + if (RT_SUCCESS(rc)) + { + RTJSONVALTYPE enmType = RTJsonValueGetType(hJsonValBool); + if (enmType == RTJSONVALTYPE_TRUE) + *pfBoolean = true; + else if (enmType == RTJSONVALTYPE_FALSE) + *pfBoolean = false; + else + rc = VERR_JSON_VALUE_INVALID_TYPE; + RTJsonValueRelease(hJsonValBool); + } + + return rc; +} + +RTDECL(unsigned) RTJsonValueGetArraySize(RTJSONVAL hJsonVal) +{ + PRTJSONVALINT pThis = hJsonVal; + RTJSON_ASSERT_VALID_HANDLE_AND_TYPE_RETURN(pThis, RTJSONVALTYPE_ARRAY, 0); + + return pThis->Type.Array.cItems; +} + +RTDECL(int) RTJsonValueQueryArraySize(RTJSONVAL hJsonVal, unsigned *pcItems) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(pcItems, VERR_INVALID_POINTER); + + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_ARRAY); + *pcItems = pThis->Type.Array.cItems; + return VINF_SUCCESS; +} + +RTDECL(int) RTJsonValueQueryByIndex(RTJSONVAL hJsonVal, unsigned idx, PRTJSONVAL phJsonVal) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(phJsonVal, VERR_INVALID_POINTER); + + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_ARRAY); + if (RT_UNLIKELY(idx >= pThis->Type.Array.cItems)) + return VERR_OUT_OF_RANGE; + + RTJsonValueRetain(pThis->Type.Array.papItems[idx]); + *phJsonVal = pThis->Type.Array.papItems[idx]; + return VINF_SUCCESS; +} + +static int rtJsonIteratorBeginWorker(PRTJSONVALINT pThis, PRTJSONIT phJsonIt) +{ + PRTJSONITINT pIt = (PRTJSONITINT)RTMemTmpAllocZ(sizeof(RTJSONITINT)); + if (pIt) + { + RTJsonValueRetain(pThis); + pIt->pJsonVal = pThis; + pIt->idxCur = 0; + + *phJsonIt = pIt; + return VINF_SUCCESS; + } + return VERR_NO_MEMORY; +} + +RTDECL(int) RTJsonIteratorBegin(RTJSONVAL hJsonVal, PRTJSONIT phJsonIt) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(phJsonIt, VERR_INVALID_POINTER); + RTJSON_TYPECHECK_CONTAINER_RETURN(pThis); + + return rtJsonIteratorBeginWorker(pThis, phJsonIt); +} + +RTDECL(int) RTJsonIteratorBeginArray(RTJSONVAL hJsonVal, PRTJSONIT phJsonIt) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(phJsonIt, VERR_INVALID_POINTER); + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_ARRAY); + + if (pThis->Type.Array.cItems > 0) + return rtJsonIteratorBeginWorker(pThis, phJsonIt); + return VERR_JSON_IS_EMPTY; +} + +RTDECL(int) RTJsonIteratorBeginObject(RTJSONVAL hJsonVal, PRTJSONIT phJsonIt) +{ + PRTJSONVALINT pThis = hJsonVal; + AssertPtrReturn(pThis, VERR_INVALID_HANDLE); + AssertPtrReturn(phJsonIt, VERR_INVALID_POINTER); + RTJSON_TYPECHECK_RETURN(pThis, RTJSONVALTYPE_OBJECT); + + if (pThis->Type.Object.cMembers > 0) + return rtJsonIteratorBeginWorker(pThis, phJsonIt); + return VERR_JSON_IS_EMPTY; +} + +RTDECL(int) RTJsonIteratorQueryValue(RTJSONIT hJsonIt, PRTJSONVAL phJsonVal, const char **ppszName) +{ + PRTJSONITINT pIt = hJsonIt; + AssertPtrReturn(pIt, VERR_INVALID_HANDLE); + AssertReturn(pIt != NIL_RTJSONIT, VERR_INVALID_HANDLE); + AssertPtrReturn(phJsonVal, VERR_INVALID_POINTER); + + int rc = VINF_SUCCESS; + PRTJSONVALINT pThis = pIt->pJsonVal; + if (pThis->enmType == RTJSONVALTYPE_ARRAY) + { + if (pIt->idxCur < pThis->Type.Array.cItems) + { + if (ppszName) + *ppszName = NULL; + + RTJsonValueRetain(pThis->Type.Array.papItems[pIt->idxCur]); + *phJsonVal = pThis->Type.Array.papItems[pIt->idxCur]; + } + else + rc = VERR_JSON_ITERATOR_END; + } + else + { + Assert(pThis->enmType == RTJSONVALTYPE_OBJECT); + + if (pIt->idxCur < pThis->Type.Object.cMembers) + { + if (ppszName) + *ppszName = pThis->Type.Object.papszNames[pIt->idxCur]; + + RTJsonValueRetain(pThis->Type.Object.papValues[pIt->idxCur]); + *phJsonVal = pThis->Type.Object.papValues[pIt->idxCur]; + } + else + rc = VERR_JSON_ITERATOR_END; + } + + return rc; +} + +RTDECL(int) RTJsonIteratorNext(RTJSONIT hJsonIt) +{ + PRTJSONITINT pIt = hJsonIt; + AssertPtrReturn(pIt, VERR_INVALID_HANDLE); + AssertReturn(pIt != NIL_RTJSONIT, VERR_INVALID_HANDLE); + + int rc = VINF_SUCCESS; + PRTJSONVALINT pThis = pIt->pJsonVal; + if (pThis->enmType == RTJSONVALTYPE_ARRAY) + { + if (pIt->idxCur < pThis->Type.Array.cItems) + pIt->idxCur++; + + if (pIt->idxCur == pThis->Type.Object.cMembers) + rc = VERR_JSON_ITERATOR_END; + } + else + { + Assert(pThis->enmType == RTJSONVALTYPE_OBJECT); + + if (pIt->idxCur < pThis->Type.Object.cMembers) + pIt->idxCur++; + + if (pIt->idxCur == pThis->Type.Object.cMembers) + rc = VERR_JSON_ITERATOR_END; + } + + return rc; +} + +RTDECL(void) RTJsonIteratorFree(RTJSONIT hJsonIt) +{ + PRTJSONITINT pThis = hJsonIt; + AssertPtrReturnVoid(pThis); + + if (pThis == NIL_RTJSONIT) + return; + + RTJsonValueRelease(pThis->pJsonVal); + RTMemTmpFree(pThis); +} |