diff options
Diffstat (limited to 'src/univalue/lib')
-rw-r--r-- | src/univalue/lib/.gitignore | 2 | ||||
-rw-r--r-- | src/univalue/lib/univalue.cpp | 242 | ||||
-rw-r--r-- | src/univalue/lib/univalue_escapes.h | 262 | ||||
-rw-r--r-- | src/univalue/lib/univalue_get.cpp | 147 | ||||
-rw-r--r-- | src/univalue/lib/univalue_read.cpp | 449 | ||||
-rw-r--r-- | src/univalue/lib/univalue_utffilter.h | 119 | ||||
-rw-r--r-- | src/univalue/lib/univalue_write.cpp | 113 |
7 files changed, 1334 insertions, 0 deletions
diff --git a/src/univalue/lib/.gitignore b/src/univalue/lib/.gitignore new file mode 100644 index 0000000000..ee7fc2851c --- /dev/null +++ b/src/univalue/lib/.gitignore @@ -0,0 +1,2 @@ +gen +.libs diff --git a/src/univalue/lib/univalue.cpp b/src/univalue/lib/univalue.cpp new file mode 100644 index 0000000000..4c9c15d63e --- /dev/null +++ b/src/univalue/lib/univalue.cpp @@ -0,0 +1,242 @@ +// Copyright 2014 BitPay Inc. +// Copyright 2015 Bitcoin Core Developers +// Distributed under the MIT software license, see the accompanying +// file COPYING or http://www.opensource.org/licenses/mit-license.php. + +#include <stdint.h> +#include <iomanip> +#include <sstream> +#include <stdlib.h> + +#include "univalue.h" + +const UniValue NullUniValue; + +void UniValue::clear() +{ + typ = VNULL; + val.clear(); + keys.clear(); + values.clear(); +} + +bool UniValue::setNull() +{ + clear(); + return true; +} + +bool UniValue::setBool(bool val_) +{ + clear(); + typ = VBOOL; + if (val_) + val = "1"; + return true; +} + +static bool validNumStr(const std::string& s) +{ + std::string tokenVal; + unsigned int consumed; + enum jtokentype tt = getJsonToken(tokenVal, consumed, s.data(), s.data() + s.size()); + return (tt == JTOK_NUMBER); +} + +bool UniValue::setNumStr(const std::string& val_) +{ + if (!validNumStr(val_)) + return false; + + clear(); + typ = VNUM; + val = val_; + return true; +} + +bool UniValue::setInt(uint64_t val_) +{ + std::ostringstream oss; + + oss << val_; + + return setNumStr(oss.str()); +} + +bool UniValue::setInt(int64_t val_) +{ + std::ostringstream oss; + + oss << val_; + + return setNumStr(oss.str()); +} + +bool UniValue::setFloat(double val_) +{ + std::ostringstream oss; + + oss << std::setprecision(16) << val_; + + bool ret = setNumStr(oss.str()); + typ = VNUM; + return ret; +} + +bool UniValue::setStr(const std::string& val_) +{ + clear(); + typ = VSTR; + val = val_; + return true; +} + +bool UniValue::setArray() +{ + clear(); + typ = VARR; + return true; +} + +bool UniValue::setObject() +{ + clear(); + typ = VOBJ; + return true; +} + +bool UniValue::push_back(const UniValue& val_) +{ + if (typ != VARR) + return false; + + values.push_back(val_); + return true; +} + +bool UniValue::push_backV(const std::vector<UniValue>& vec) +{ + if (typ != VARR) + return false; + + values.insert(values.end(), vec.begin(), vec.end()); + + return true; +} + +void UniValue::__pushKV(const std::string& key, const UniValue& val_) +{ + keys.push_back(key); + values.push_back(val_); +} + +bool UniValue::pushKV(const std::string& key, const UniValue& val_) +{ + if (typ != VOBJ) + return false; + + size_t idx; + if (findKey(key, idx)) + values[idx] = val_; + else + __pushKV(key, val_); + return true; +} + +bool UniValue::pushKVs(const UniValue& obj) +{ + if (typ != VOBJ || obj.typ != VOBJ) + return false; + + for (size_t i = 0; i < obj.keys.size(); i++) + __pushKV(obj.keys[i], obj.values.at(i)); + + return true; +} + +void UniValue::getObjMap(std::map<std::string,UniValue>& kv) const +{ + if (typ != VOBJ) + return; + + kv.clear(); + for (size_t i = 0; i < keys.size(); i++) + kv[keys[i]] = values[i]; +} + +bool UniValue::findKey(const std::string& key, size_t& retIdx) const +{ + for (size_t i = 0; i < keys.size(); i++) { + if (keys[i] == key) { + retIdx = i; + return true; + } + } + + return false; +} + +bool UniValue::checkObject(const std::map<std::string,UniValue::VType>& t) const +{ + if (typ != VOBJ) + return false; + + for (std::map<std::string,UniValue::VType>::const_iterator it = t.begin(); + it != t.end(); ++it) { + size_t idx = 0; + if (!findKey(it->first, idx)) + return false; + + if (values.at(idx).getType() != it->second) + return false; + } + + return true; +} + +const UniValue& UniValue::operator[](const std::string& key) const +{ + if (typ != VOBJ) + return NullUniValue; + + size_t index = 0; + if (!findKey(key, index)) + return NullUniValue; + + return values.at(index); +} + +const UniValue& UniValue::operator[](size_t index) const +{ + if (typ != VOBJ && typ != VARR) + return NullUniValue; + if (index >= values.size()) + return NullUniValue; + + return values.at(index); +} + +const char *uvTypeName(UniValue::VType t) +{ + switch (t) { + case UniValue::VNULL: return "null"; + case UniValue::VBOOL: return "bool"; + case UniValue::VOBJ: return "object"; + case UniValue::VARR: return "array"; + case UniValue::VSTR: return "string"; + case UniValue::VNUM: return "number"; + } + + // not reached + return NULL; +} + +const UniValue& find_value(const UniValue& obj, const std::string& name) +{ + for (unsigned int i = 0; i < obj.keys.size(); i++) + if (obj.keys[i] == name) + return obj.values.at(i); + + return NullUniValue; +} + diff --git a/src/univalue/lib/univalue_escapes.h b/src/univalue/lib/univalue_escapes.h new file mode 100644 index 0000000000..74596aab6d --- /dev/null +++ b/src/univalue/lib/univalue_escapes.h @@ -0,0 +1,262 @@ +// Automatically generated file. Do not modify. +#ifndef BITCOIN_UNIVALUE_UNIVALUE_ESCAPES_H +#define BITCOIN_UNIVALUE_UNIVALUE_ESCAPES_H +static const char *escapes[256] = { + "\\u0000", + "\\u0001", + "\\u0002", + "\\u0003", + "\\u0004", + "\\u0005", + "\\u0006", + "\\u0007", + "\\b", + "\\t", + "\\n", + "\\u000b", + "\\f", + "\\r", + "\\u000e", + "\\u000f", + "\\u0010", + "\\u0011", + "\\u0012", + "\\u0013", + "\\u0014", + "\\u0015", + "\\u0016", + "\\u0017", + "\\u0018", + "\\u0019", + "\\u001a", + "\\u001b", + "\\u001c", + "\\u001d", + "\\u001e", + "\\u001f", + NULL, + NULL, + "\\\"", + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + "\\\\", + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + "\\u007f", + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, + NULL, +}; +#endif // BITCOIN_UNIVALUE_UNIVALUE_ESCAPES_H diff --git a/src/univalue/lib/univalue_get.cpp b/src/univalue/lib/univalue_get.cpp new file mode 100644 index 0000000000..eabcf2dad1 --- /dev/null +++ b/src/univalue/lib/univalue_get.cpp @@ -0,0 +1,147 @@ +// Copyright 2014 BitPay Inc. +// Copyright 2015 Bitcoin Core Developers +// Distributed under the MIT software license, see the accompanying +// file COPYING or http://www.opensource.org/licenses/mit-license.php. + +#include <stdint.h> +#include <errno.h> +#include <string.h> +#include <stdlib.h> +#include <stdexcept> +#include <vector> +#include <limits> +#include <string> + +#include "univalue.h" + +namespace +{ +static bool ParsePrechecks(const std::string& str) +{ + if (str.empty()) // No empty string allowed + return false; + if (str.size() >= 1 && (json_isspace(str[0]) || json_isspace(str[str.size()-1]))) // No padding allowed + return false; + if (str.size() != strlen(str.c_str())) // No embedded NUL characters allowed + return false; + return true; +} + +bool ParseInt32(const std::string& str, int32_t *out) +{ + if (!ParsePrechecks(str)) + return false; + char *endp = NULL; + errno = 0; // strtol will not set errno if valid + long int n = strtol(str.c_str(), &endp, 10); + if(out) *out = (int32_t)n; + // Note that strtol returns a *long int*, so even if strtol doesn't report a over/underflow + // we still have to check that the returned value is within the range of an *int32_t*. On 64-bit + // platforms the size of these types may be different. + return endp && *endp == 0 && !errno && + n >= std::numeric_limits<int32_t>::min() && + n <= std::numeric_limits<int32_t>::max(); +} + +bool ParseInt64(const std::string& str, int64_t *out) +{ + if (!ParsePrechecks(str)) + return false; + char *endp = NULL; + errno = 0; // strtoll will not set errno if valid + long long int n = strtoll(str.c_str(), &endp, 10); + if(out) *out = (int64_t)n; + // Note that strtoll returns a *long long int*, so even if strtol doesn't report a over/underflow + // we still have to check that the returned value is within the range of an *int64_t*. + return endp && *endp == 0 && !errno && + n >= std::numeric_limits<int64_t>::min() && + n <= std::numeric_limits<int64_t>::max(); +} + +bool ParseDouble(const std::string& str, double *out) +{ + if (!ParsePrechecks(str)) + return false; + if (str.size() >= 2 && str[0] == '0' && str[1] == 'x') // No hexadecimal floats allowed + return false; + std::istringstream text(str); + text.imbue(std::locale::classic()); + double result; + text >> result; + if(out) *out = result; + return text.eof() && !text.fail(); +} +} + +const std::vector<std::string>& UniValue::getKeys() const +{ + if (typ != VOBJ) + throw std::runtime_error("JSON value is not an object as expected"); + return keys; +} + +const std::vector<UniValue>& UniValue::getValues() const +{ + if (typ != VOBJ && typ != VARR) + throw std::runtime_error("JSON value is not an object or array as expected"); + return values; +} + +bool UniValue::get_bool() const +{ + if (typ != VBOOL) + throw std::runtime_error("JSON value is not a boolean as expected"); + return getBool(); +} + +const std::string& UniValue::get_str() const +{ + if (typ != VSTR) + throw std::runtime_error("JSON value is not a string as expected"); + return getValStr(); +} + +int UniValue::get_int() const +{ + if (typ != VNUM) + throw std::runtime_error("JSON value is not an integer as expected"); + int32_t retval; + if (!ParseInt32(getValStr(), &retval)) + throw std::runtime_error("JSON integer out of range"); + return retval; +} + +int64_t UniValue::get_int64() const +{ + if (typ != VNUM) + throw std::runtime_error("JSON value is not an integer as expected"); + int64_t retval; + if (!ParseInt64(getValStr(), &retval)) + throw std::runtime_error("JSON integer out of range"); + return retval; +} + +double UniValue::get_real() const +{ + if (typ != VNUM) + throw std::runtime_error("JSON value is not a number as expected"); + double retval; + if (!ParseDouble(getValStr(), &retval)) + throw std::runtime_error("JSON double out of range"); + return retval; +} + +const UniValue& UniValue::get_obj() const +{ + if (typ != VOBJ) + throw std::runtime_error("JSON value is not an object as expected"); + return *this; +} + +const UniValue& UniValue::get_array() const +{ + if (typ != VARR) + throw std::runtime_error("JSON value is not an array as expected"); + return *this; +} + diff --git a/src/univalue/lib/univalue_read.cpp b/src/univalue/lib/univalue_read.cpp new file mode 100644 index 0000000000..14834db24d --- /dev/null +++ b/src/univalue/lib/univalue_read.cpp @@ -0,0 +1,449 @@ +// Copyright 2014 BitPay Inc. +// Distributed under the MIT software license, see the accompanying +// file COPYING or http://www.opensource.org/licenses/mit-license.php. + +#include <string.h> +#include <vector> +#include <stdio.h> +#include "univalue.h" +#include "univalue_utffilter.h" + +static bool json_isdigit(int ch) +{ + return ((ch >= '0') && (ch <= '9')); +} + +// convert hexadecimal string to unsigned integer +static const char *hatoui(const char *first, const char *last, + unsigned int& out) +{ + unsigned int result = 0; + for (; first != last; ++first) + { + int digit; + if (json_isdigit(*first)) + digit = *first - '0'; + + else if (*first >= 'a' && *first <= 'f') + digit = *first - 'a' + 10; + + else if (*first >= 'A' && *first <= 'F') + digit = *first - 'A' + 10; + + else + break; + + result = 16 * result + digit; + } + out = result; + + return first; +} + +enum jtokentype getJsonToken(std::string& tokenVal, unsigned int& consumed, + const char *raw, const char *end) +{ + tokenVal.clear(); + consumed = 0; + + const char *rawStart = raw; + + while (raw < end && (json_isspace(*raw))) // skip whitespace + raw++; + + if (raw >= end) + return JTOK_NONE; + + switch (*raw) { + + case '{': + raw++; + consumed = (raw - rawStart); + return JTOK_OBJ_OPEN; + case '}': + raw++; + consumed = (raw - rawStart); + return JTOK_OBJ_CLOSE; + case '[': + raw++; + consumed = (raw - rawStart); + return JTOK_ARR_OPEN; + case ']': + raw++; + consumed = (raw - rawStart); + return JTOK_ARR_CLOSE; + + case ':': + raw++; + consumed = (raw - rawStart); + return JTOK_COLON; + case ',': + raw++; + consumed = (raw - rawStart); + return JTOK_COMMA; + + case 'n': + case 't': + case 'f': + if (!strncmp(raw, "null", 4)) { + raw += 4; + consumed = (raw - rawStart); + return JTOK_KW_NULL; + } else if (!strncmp(raw, "true", 4)) { + raw += 4; + consumed = (raw - rawStart); + return JTOK_KW_TRUE; + } else if (!strncmp(raw, "false", 5)) { + raw += 5; + consumed = (raw - rawStart); + return JTOK_KW_FALSE; + } else + return JTOK_ERR; + + case '-': + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': { + // part 1: int + std::string numStr; + + const char *first = raw; + + const char *firstDigit = first; + if (!json_isdigit(*firstDigit)) + firstDigit++; + if ((*firstDigit == '0') && json_isdigit(firstDigit[1])) + return JTOK_ERR; + + numStr += *raw; // copy first char + raw++; + + if ((*first == '-') && (raw < end) && (!json_isdigit(*raw))) + return JTOK_ERR; + + while (raw < end && json_isdigit(*raw)) { // copy digits + numStr += *raw; + raw++; + } + + // part 2: frac + if (raw < end && *raw == '.') { + numStr += *raw; // copy . + raw++; + + if (raw >= end || !json_isdigit(*raw)) + return JTOK_ERR; + while (raw < end && json_isdigit(*raw)) { // copy digits + numStr += *raw; + raw++; + } + } + + // part 3: exp + if (raw < end && (*raw == 'e' || *raw == 'E')) { + numStr += *raw; // copy E + raw++; + + if (raw < end && (*raw == '-' || *raw == '+')) { // copy +/- + numStr += *raw; + raw++; + } + + if (raw >= end || !json_isdigit(*raw)) + return JTOK_ERR; + while (raw < end && json_isdigit(*raw)) { // copy digits + numStr += *raw; + raw++; + } + } + + tokenVal = numStr; + consumed = (raw - rawStart); + return JTOK_NUMBER; + } + + case '"': { + raw++; // skip " + + std::string valStr; + JSONUTF8StringFilter writer(valStr); + + while (true) { + if (raw >= end || (unsigned char)*raw < 0x20) + return JTOK_ERR; + + else if (*raw == '\\') { + raw++; // skip backslash + + if (raw >= end) + return JTOK_ERR; + + switch (*raw) { + case '"': writer.push_back('\"'); break; + case '\\': writer.push_back('\\'); break; + case '/': writer.push_back('/'); break; + case 'b': writer.push_back('\b'); break; + case 'f': writer.push_back('\f'); break; + case 'n': writer.push_back('\n'); break; + case 'r': writer.push_back('\r'); break; + case 't': writer.push_back('\t'); break; + + case 'u': { + unsigned int codepoint; + if (raw + 1 + 4 >= end || + hatoui(raw + 1, raw + 1 + 4, codepoint) != + raw + 1 + 4) + return JTOK_ERR; + writer.push_back_u(codepoint); + raw += 4; + break; + } + default: + return JTOK_ERR; + + } + + raw++; // skip esc'd char + } + + else if (*raw == '"') { + raw++; // skip " + break; // stop scanning + } + + else { + writer.push_back(*raw); + raw++; + } + } + + if (!writer.finalize()) + return JTOK_ERR; + tokenVal = valStr; + consumed = (raw - rawStart); + return JTOK_STRING; + } + + default: + return JTOK_ERR; + } +} + +enum expect_bits { + EXP_OBJ_NAME = (1U << 0), + EXP_COLON = (1U << 1), + EXP_ARR_VALUE = (1U << 2), + EXP_VALUE = (1U << 3), + EXP_NOT_VALUE = (1U << 4), +}; + +#define expect(bit) (expectMask & (EXP_##bit)) +#define setExpect(bit) (expectMask |= EXP_##bit) +#define clearExpect(bit) (expectMask &= ~EXP_##bit) + +bool UniValue::read(const char *raw, size_t size) +{ + clear(); + + uint32_t expectMask = 0; + std::vector<UniValue*> stack; + + std::string tokenVal; + unsigned int consumed; + enum jtokentype tok = JTOK_NONE; + enum jtokentype last_tok = JTOK_NONE; + const char* end = raw + size; + do { + last_tok = tok; + + tok = getJsonToken(tokenVal, consumed, raw, end); + if (tok == JTOK_NONE || tok == JTOK_ERR) + return false; + raw += consumed; + + bool isValueOpen = jsonTokenIsValue(tok) || + tok == JTOK_OBJ_OPEN || tok == JTOK_ARR_OPEN; + + if (expect(VALUE)) { + if (!isValueOpen) + return false; + clearExpect(VALUE); + + } else if (expect(ARR_VALUE)) { + bool isArrValue = isValueOpen || (tok == JTOK_ARR_CLOSE); + if (!isArrValue) + return false; + + clearExpect(ARR_VALUE); + + } else if (expect(OBJ_NAME)) { + bool isObjName = (tok == JTOK_OBJ_CLOSE || tok == JTOK_STRING); + if (!isObjName) + return false; + + } else if (expect(COLON)) { + if (tok != JTOK_COLON) + return false; + clearExpect(COLON); + + } else if (!expect(COLON) && (tok == JTOK_COLON)) { + return false; + } + + if (expect(NOT_VALUE)) { + if (isValueOpen) + return false; + clearExpect(NOT_VALUE); + } + + switch (tok) { + + case JTOK_OBJ_OPEN: + case JTOK_ARR_OPEN: { + VType utyp = (tok == JTOK_OBJ_OPEN ? VOBJ : VARR); + if (!stack.size()) { + if (utyp == VOBJ) + setObject(); + else + setArray(); + stack.push_back(this); + } else { + UniValue tmpVal(utyp); + UniValue *top = stack.back(); + top->values.push_back(tmpVal); + + UniValue *newTop = &(top->values.back()); + stack.push_back(newTop); + } + + if (utyp == VOBJ) + setExpect(OBJ_NAME); + else + setExpect(ARR_VALUE); + break; + } + + case JTOK_OBJ_CLOSE: + case JTOK_ARR_CLOSE: { + if (!stack.size() || (last_tok == JTOK_COMMA)) + return false; + + VType utyp = (tok == JTOK_OBJ_CLOSE ? VOBJ : VARR); + UniValue *top = stack.back(); + if (utyp != top->getType()) + return false; + + stack.pop_back(); + clearExpect(OBJ_NAME); + setExpect(NOT_VALUE); + break; + } + + case JTOK_COLON: { + if (!stack.size()) + return false; + + UniValue *top = stack.back(); + if (top->getType() != VOBJ) + return false; + + setExpect(VALUE); + break; + } + + case JTOK_COMMA: { + if (!stack.size() || + (last_tok == JTOK_COMMA) || (last_tok == JTOK_ARR_OPEN)) + return false; + + UniValue *top = stack.back(); + if (top->getType() == VOBJ) + setExpect(OBJ_NAME); + else + setExpect(ARR_VALUE); + break; + } + + case JTOK_KW_NULL: + case JTOK_KW_TRUE: + case JTOK_KW_FALSE: { + UniValue tmpVal; + switch (tok) { + case JTOK_KW_NULL: + // do nothing more + break; + case JTOK_KW_TRUE: + tmpVal.setBool(true); + break; + case JTOK_KW_FALSE: + tmpVal.setBool(false); + break; + default: /* impossible */ break; + } + + if (!stack.size()) { + *this = tmpVal; + break; + } + + UniValue *top = stack.back(); + top->values.push_back(tmpVal); + + setExpect(NOT_VALUE); + break; + } + + case JTOK_NUMBER: { + UniValue tmpVal(VNUM, tokenVal); + if (!stack.size()) { + *this = tmpVal; + break; + } + + UniValue *top = stack.back(); + top->values.push_back(tmpVal); + + setExpect(NOT_VALUE); + break; + } + + case JTOK_STRING: { + if (expect(OBJ_NAME)) { + UniValue *top = stack.back(); + top->keys.push_back(tokenVal); + clearExpect(OBJ_NAME); + setExpect(COLON); + } else { + UniValue tmpVal(VSTR, tokenVal); + if (!stack.size()) { + *this = tmpVal; + break; + } + UniValue *top = stack.back(); + top->values.push_back(tmpVal); + } + + setExpect(NOT_VALUE); + break; + } + + default: + return false; + } + } while (!stack.empty ()); + + /* Check that nothing follows the initial construct (parsed above). */ + tok = getJsonToken(tokenVal, consumed, raw, end); + if (tok != JTOK_NONE) + return false; + + return true; +} + diff --git a/src/univalue/lib/univalue_utffilter.h b/src/univalue/lib/univalue_utffilter.h new file mode 100644 index 0000000000..20d4043009 --- /dev/null +++ b/src/univalue/lib/univalue_utffilter.h @@ -0,0 +1,119 @@ +// Copyright 2016 Wladimir J. van der Laan +// Distributed under the MIT software license, see the accompanying +// file COPYING or http://www.opensource.org/licenses/mit-license.php. +#ifndef UNIVALUE_UTFFILTER_H +#define UNIVALUE_UTFFILTER_H + +#include <string> + +/** + * Filter that generates and validates UTF-8, as well as collates UTF-16 + * surrogate pairs as specified in RFC4627. + */ +class JSONUTF8StringFilter +{ +public: + explicit JSONUTF8StringFilter(std::string &s): + str(s), is_valid(true), codepoint(0), state(0), surpair(0) + { + } + // Write single 8-bit char (may be part of UTF-8 sequence) + void push_back(unsigned char ch) + { + if (state == 0) { + if (ch < 0x80) // 7-bit ASCII, fast direct pass-through + str.push_back(ch); + else if (ch < 0xc0) // Mid-sequence character, invalid in this state + is_valid = false; + else if (ch < 0xe0) { // Start of 2-byte sequence + codepoint = (ch & 0x1f) << 6; + state = 6; + } else if (ch < 0xf0) { // Start of 3-byte sequence + codepoint = (ch & 0x0f) << 12; + state = 12; + } else if (ch < 0xf8) { // Start of 4-byte sequence + codepoint = (ch & 0x07) << 18; + state = 18; + } else // Reserved, invalid + is_valid = false; + } else { + if ((ch & 0xc0) != 0x80) // Not a continuation, invalid + is_valid = false; + state -= 6; + codepoint |= (ch & 0x3f) << state; + if (state == 0) + push_back_u(codepoint); + } + } + // Write codepoint directly, possibly collating surrogate pairs + void push_back_u(unsigned int codepoint_) + { + if (state) // Only accept full codepoints in open state + is_valid = false; + if (codepoint_ >= 0xD800 && codepoint_ < 0xDC00) { // First half of surrogate pair + if (surpair) // Two subsequent surrogate pair openers - fail + is_valid = false; + else + surpair = codepoint_; + } else if (codepoint_ >= 0xDC00 && codepoint_ < 0xE000) { // Second half of surrogate pair + if (surpair) { // Open surrogate pair, expect second half + // Compute code point from UTF-16 surrogate pair + append_codepoint(0x10000 | ((surpair - 0xD800)<<10) | (codepoint_ - 0xDC00)); + surpair = 0; + } else // Second half doesn't follow a first half - fail + is_valid = false; + } else { + if (surpair) // First half of surrogate pair not followed by second - fail + is_valid = false; + else + append_codepoint(codepoint_); + } + } + // Check that we're in a state where the string can be ended + // No open sequences, no open surrogate pairs, etc + bool finalize() + { + if (state || surpair) + is_valid = false; + return is_valid; + } +private: + std::string &str; + bool is_valid; + // Current UTF-8 decoding state + unsigned int codepoint; + int state; // Top bit to be filled in for next UTF-8 byte, or 0 + + // Keep track of the following state to handle the following section of + // RFC4627: + // + // To escape an extended character that is not in the Basic Multilingual + // Plane, the character is represented as a twelve-character sequence, + // encoding the UTF-16 surrogate pair. So, for example, a string + // containing only the G clef character (U+1D11E) may be represented as + // "\uD834\uDD1E". + // + // Two subsequent \u.... may have to be replaced with one actual codepoint. + unsigned int surpair; // First half of open UTF-16 surrogate pair, or 0 + + void append_codepoint(unsigned int codepoint_) + { + if (codepoint_ <= 0x7f) + str.push_back((char)codepoint_); + else if (codepoint_ <= 0x7FF) { + str.push_back((char)(0xC0 | (codepoint_ >> 6))); + str.push_back((char)(0x80 | (codepoint_ & 0x3F))); + } else if (codepoint_ <= 0xFFFF) { + str.push_back((char)(0xE0 | (codepoint_ >> 12))); + str.push_back((char)(0x80 | ((codepoint_ >> 6) & 0x3F))); + str.push_back((char)(0x80 | (codepoint_ & 0x3F))); + } else if (codepoint_ <= 0x1FFFFF) { + str.push_back((char)(0xF0 | (codepoint_ >> 18))); + str.push_back((char)(0x80 | ((codepoint_ >> 12) & 0x3F))); + str.push_back((char)(0x80 | ((codepoint_ >> 6) & 0x3F))); + str.push_back((char)(0x80 | (codepoint_ & 0x3F))); + } + } +}; + +#endif diff --git a/src/univalue/lib/univalue_write.cpp b/src/univalue/lib/univalue_write.cpp new file mode 100644 index 0000000000..827eb9b271 --- /dev/null +++ b/src/univalue/lib/univalue_write.cpp @@ -0,0 +1,113 @@ +// Copyright 2014 BitPay Inc. +// Distributed under the MIT software license, see the accompanying +// file COPYING or http://www.opensource.org/licenses/mit-license.php. + +#include <iomanip> +#include <sstream> +#include <stdio.h> +#include "univalue.h" +#include "univalue_escapes.h" + +static std::string json_escape(const std::string& inS) +{ + std::string outS; + outS.reserve(inS.size() * 2); + + for (unsigned int i = 0; i < inS.size(); i++) { + unsigned char ch = inS[i]; + const char *escStr = escapes[ch]; + + if (escStr) + outS += escStr; + else + outS += ch; + } + + return outS; +} + +std::string UniValue::write(unsigned int prettyIndent, + unsigned int indentLevel) const +{ + std::string s; + s.reserve(1024); + + unsigned int modIndent = indentLevel; + if (modIndent == 0) + modIndent = 1; + + switch (typ) { + case VNULL: + s += "null"; + break; + case VOBJ: + writeObject(prettyIndent, modIndent, s); + break; + case VARR: + writeArray(prettyIndent, modIndent, s); + break; + case VSTR: + s += "\"" + json_escape(val) + "\""; + break; + case VNUM: + s += val; + break; + case VBOOL: + s += (val == "1" ? "true" : "false"); + break; + } + + return s; +} + +static void indentStr(unsigned int prettyIndent, unsigned int indentLevel, std::string& s) +{ + s.append(prettyIndent * indentLevel, ' '); +} + +void UniValue::writeArray(unsigned int prettyIndent, unsigned int indentLevel, std::string& s) const +{ + s += "["; + if (prettyIndent) + s += "\n"; + + for (unsigned int i = 0; i < values.size(); i++) { + if (prettyIndent) + indentStr(prettyIndent, indentLevel, s); + s += values[i].write(prettyIndent, indentLevel + 1); + if (i != (values.size() - 1)) { + s += ","; + } + if (prettyIndent) + s += "\n"; + } + + if (prettyIndent) + indentStr(prettyIndent, indentLevel - 1, s); + s += "]"; +} + +void UniValue::writeObject(unsigned int prettyIndent, unsigned int indentLevel, std::string& s) const +{ + s += "{"; + if (prettyIndent) + s += "\n"; + + for (unsigned int i = 0; i < keys.size(); i++) { + if (prettyIndent) + indentStr(prettyIndent, indentLevel, s); + s += "\"" + json_escape(keys[i]) + "\":"; + if (prettyIndent) + s += " "; + s += values.at(i).write(prettyIndent, indentLevel + 1); + if (i != (values.size() - 1)) + s += ","; + if (prettyIndent) + s += "\n"; + } + + if (prettyIndent) + indentStr(prettyIndent, indentLevel - 1, s); + s += "}"; +} + |