313 lines
9.4 KiB
C++
313 lines
9.4 KiB
C++
// core/json implementation — see json.h. Any behavioral change here changes
|
||
// every persisted-blob parser that shares this lexical layer at once.
|
||
|
||
#include "core/json/json.h"
|
||
|
||
#include <cerrno>
|
||
#include <climits>
|
||
#include <cstdio>
|
||
#include <cstdlib>
|
||
|
||
namespace reasampler::json {
|
||
|
||
// -- emit helpers -------------------------------------------------------
|
||
|
||
void writeEscaped(std::string& out, const std::string& s) {
|
||
out += '"';
|
||
for (char c : s) {
|
||
switch (c) {
|
||
case '"': out += "\\\""; break;
|
||
case '\\': out += "\\\\"; break;
|
||
case '\b': out += "\\b"; break;
|
||
case '\f': out += "\\f"; break;
|
||
case '\n': out += "\\n"; break;
|
||
case '\r': out += "\\r"; break;
|
||
case '\t': out += "\\t"; break;
|
||
default:
|
||
if (static_cast<unsigned char>(c) < 0x20) {
|
||
char buf[8];
|
||
std::snprintf(buf, sizeof(buf), "\\u%04x", static_cast<unsigned char>(c));
|
||
out += buf;
|
||
} else {
|
||
out += c;
|
||
}
|
||
}
|
||
}
|
||
out += '"';
|
||
}
|
||
|
||
std::string numToStr(double v) {
|
||
char buf[32];
|
||
std::snprintf(buf, sizeof(buf), "%.17g", v);
|
||
return buf;
|
||
}
|
||
|
||
std::string numToStr(std::int64_t v) {
|
||
char buf[32];
|
||
std::snprintf(buf, sizeof(buf), "%lld", static_cast<long long>(v));
|
||
return buf;
|
||
}
|
||
|
||
std::string numToStr(int v) {
|
||
char buf[16];
|
||
std::snprintf(buf, sizeof(buf), "%d", v);
|
||
return buf;
|
||
}
|
||
|
||
void writeStringArray(std::string& out, const std::vector<std::string>& v) {
|
||
out += '[';
|
||
for (std::size_t i = 0; i < v.size(); ++i) {
|
||
if (i) out += ',';
|
||
writeEscaped(out, v[i]);
|
||
}
|
||
out += ']';
|
||
}
|
||
|
||
void writeIntArray(std::string& out, const std::vector<int>& v) {
|
||
out += '[';
|
||
for (std::size_t i = 0; i < v.size(); ++i) {
|
||
if (i) out += ',';
|
||
out += numToStr(v[i]);
|
||
}
|
||
out += ']';
|
||
}
|
||
|
||
// -- Reader ---------------------------------------------------------------
|
||
|
||
void Reader::skipWs() {
|
||
while (!eof()) {
|
||
char c = s_[pos_];
|
||
if (c == ' ' || c == '\t' || c == '\n' || c == '\r') ++pos_;
|
||
else break;
|
||
}
|
||
}
|
||
|
||
bool Reader::consume(char c) {
|
||
skipWs();
|
||
if (eof() || s_[pos_] != c) return false;
|
||
++pos_;
|
||
return true;
|
||
}
|
||
|
||
// Parses a JSON string literal (with the escapes our writers emit, plus \uXXXX
|
||
// for control chars). Positioned before the opening quote (skips leading ws).
|
||
bool Reader::parseString(std::string& out) {
|
||
skipWs();
|
||
if (eof() || s_[pos_] != '"') return false;
|
||
++pos_;
|
||
out.clear();
|
||
while (!eof()) {
|
||
char c = s_[pos_++];
|
||
if (c == '"') return true;
|
||
if (c == '\\') {
|
||
if (eof()) return false;
|
||
char e = s_[pos_++];
|
||
switch (e) {
|
||
case '"': out += '"'; break;
|
||
case '\\': out += '\\'; break;
|
||
case '/': out += '/'; break;
|
||
case 'b': out += '\b'; break;
|
||
case 'f': out += '\f'; break;
|
||
case 'n': out += '\n'; break;
|
||
case 'r': out += '\r'; break;
|
||
case 't': out += '\t'; break;
|
||
case 'u': {
|
||
auto readHex4 = [&](unsigned int& cp) -> bool {
|
||
if (pos_ + 4 > s_.size()) return false;
|
||
cp = 0;
|
||
for (int i = 0; i < 4; ++i) {
|
||
char h = s_[pos_++];
|
||
cp <<= 4;
|
||
if (h >= '0' && h <= '9') cp |= static_cast<unsigned>(h - '0');
|
||
else if (h >= 'a' && h <= 'f') cp |= static_cast<unsigned>(h - 'a' + 10);
|
||
else if (h >= 'A' && h <= 'F') cp |= static_cast<unsigned>(h - 'A' + 10);
|
||
else return false;
|
||
}
|
||
return true;
|
||
};
|
||
|
||
unsigned int hi = 0;
|
||
if (!readHex4(hi)) return false;
|
||
|
||
unsigned int codePoint = hi;
|
||
if (hi >= 0xD800 && hi <= 0xDBFF) {
|
||
// High surrogate — must be followed by \uDC00–\uDFFF.
|
||
if (pos_ + 6 > s_.size()) return false;
|
||
if (s_[pos_] != '\\' || s_[pos_ + 1] != 'u') return false;
|
||
pos_ += 2;
|
||
unsigned int lo = 0;
|
||
if (!readHex4(lo)) return false;
|
||
if (lo < 0xDC00 || lo > 0xDFFF) return false; // unpaired high surrogate
|
||
codePoint = 0x10000 + ((hi - 0xD800) << 10) + (lo - 0xDC00);
|
||
} else if (hi >= 0xDC00 && hi <= 0xDFFF) {
|
||
return false; // unpaired low surrogate — malformed
|
||
}
|
||
|
||
if (codePoint <= 0x7F) {
|
||
out += static_cast<char>(codePoint);
|
||
} else if (codePoint <= 0x7FF) {
|
||
out += static_cast<char>(0xC0 | (codePoint >> 6));
|
||
out += static_cast<char>(0x80 | (codePoint & 0x3F));
|
||
} else if (codePoint <= 0xFFFF) {
|
||
out += static_cast<char>(0xE0 | (codePoint >> 12));
|
||
out += static_cast<char>(0x80 | ((codePoint >> 6) & 0x3F));
|
||
out += static_cast<char>(0x80 | (codePoint & 0x3F));
|
||
} else {
|
||
out += static_cast<char>(0xF0 | (codePoint >> 18));
|
||
out += static_cast<char>(0x80 | ((codePoint >> 12) & 0x3F));
|
||
out += static_cast<char>(0x80 | ((codePoint >> 6) & 0x3F));
|
||
out += static_cast<char>(0x80 | (codePoint & 0x3F));
|
||
}
|
||
break;
|
||
}
|
||
default: return false;
|
||
}
|
||
} else {
|
||
out += c;
|
||
}
|
||
}
|
||
return false; // unterminated string
|
||
}
|
||
|
||
bool Reader::parseRawScalar(std::string& out) {
|
||
skipWs();
|
||
std::size_t start = pos_;
|
||
while (!eof()) {
|
||
char c = s_[pos_];
|
||
if (c == ',' || c == '}' || c == ']' || c == ' ' || c == '\t' ||
|
||
c == '\n' || c == '\r')
|
||
break;
|
||
++pos_;
|
||
}
|
||
if (pos_ == start) return false;
|
||
out.assign(s_, start, pos_ - start);
|
||
return true;
|
||
}
|
||
|
||
bool Reader::parseDouble(double& out) {
|
||
std::string tok;
|
||
if (!parseRawScalar(tok)) return false;
|
||
const char* b = tok.c_str();
|
||
char* end = nullptr;
|
||
errno = 0;
|
||
double v = std::strtod(b, &end);
|
||
if (end != b + tok.size()) return false;
|
||
if (errno == ERANGE) return false; // overflow / underflow -> malformed
|
||
out = v;
|
||
return true;
|
||
}
|
||
|
||
bool Reader::parseInt64(std::int64_t& out) {
|
||
std::string tok;
|
||
if (!parseRawScalar(tok)) return false;
|
||
const char* b = tok.c_str();
|
||
char* end = nullptr;
|
||
errno = 0;
|
||
long long v = std::strtoll(b, &end, 10);
|
||
if (end != b + tok.size()) return false;
|
||
if (errno == ERANGE) return false; // overflow -> malformed
|
||
out = static_cast<std::int64_t>(v);
|
||
return true;
|
||
}
|
||
|
||
bool Reader::parseInt(int& out) {
|
||
std::int64_t v = 0;
|
||
if (!parseInt64(v)) return false;
|
||
if (v < INT_MIN || v > INT_MAX) return false;
|
||
out = static_cast<int>(v);
|
||
return true;
|
||
}
|
||
|
||
bool Reader::parseBool(bool& out) {
|
||
std::string tok;
|
||
if (!parseRawScalar(tok)) return false;
|
||
if (tok == "true") { out = true; return true; }
|
||
if (tok == "false") { out = false; return true; }
|
||
return false;
|
||
}
|
||
|
||
bool Reader::expectNullOr(bool& wasNull) {
|
||
skipWs();
|
||
if (eof()) return false;
|
||
if (s_.compare(pos_, 4, "null") == 0) {
|
||
pos_ += 4;
|
||
wasNull = true;
|
||
} else {
|
||
wasNull = false;
|
||
}
|
||
return true;
|
||
}
|
||
|
||
bool Reader::parseKey(std::string& key) {
|
||
if (!parseString(key)) return false;
|
||
return consume(':');
|
||
}
|
||
|
||
bool Reader::parseStringArray(std::vector<std::string>& out) {
|
||
if (!consume('[')) return false;
|
||
skipWs();
|
||
if (consume(']')) return true; // empty array
|
||
do {
|
||
std::string s;
|
||
if (!parseString(s)) return false;
|
||
out.push_back(std::move(s));
|
||
} while (consume(','));
|
||
return consume(']');
|
||
}
|
||
|
||
bool Reader::parseIntArray(std::vector<int>& out) {
|
||
if (!consume('[')) return false;
|
||
skipWs();
|
||
if (consume(']')) return true;
|
||
do {
|
||
int v = 0;
|
||
if (!parseInt(v)) return false;
|
||
out.push_back(v);
|
||
} while (consume(','));
|
||
return consume(']');
|
||
}
|
||
|
||
bool Reader::skipValue() {
|
||
std::string raw;
|
||
return captureValue(raw);
|
||
}
|
||
|
||
// Records the raw source span of one JSON value starting at the current position
|
||
// (after whitespace). Handles nested objects/arrays with string-aware brace
|
||
// matching (braces inside strings ignored).
|
||
bool Reader::captureValue(std::string& raw) {
|
||
skipWs();
|
||
if (eof()) return false;
|
||
std::size_t start = pos_;
|
||
char c = s_[pos_];
|
||
if (c == '"') {
|
||
std::string tmp;
|
||
if (!parseString(tmp)) return false;
|
||
raw.assign(s_, start, pos_ - start);
|
||
return true;
|
||
}
|
||
if (c == '{' || c == '[') {
|
||
char open = c, close = (c == '{') ? '}' : ']';
|
||
++pos_;
|
||
int depth = 1;
|
||
while (!eof() && depth > 0) {
|
||
char d = s_[pos_];
|
||
if (d == '"') {
|
||
std::string tmp;
|
||
if (!parseString(tmp)) return false; // advances past the string
|
||
continue;
|
||
}
|
||
if (d == open) ++depth;
|
||
else if (d == close) --depth;
|
||
++pos_;
|
||
}
|
||
if (depth != 0) return false;
|
||
raw.assign(s_, start, pos_ - start);
|
||
return true;
|
||
}
|
||
// bare scalar (number / true / false / null)
|
||
return parseRawScalar(raw);
|
||
}
|
||
|
||
} // namespace reasampler::json
|