soa/types/json_parsing.cc

/* json_parsing.cc
   Jeremy Barnes, 8 March 2013
   Copyright (c) 2013 Datacratic Inc.  All rights reserved.

*/

#include "json_parsing.h"
#include "string.h"

using namespace std;
using namespace ML;

namespace Datacratic {

Utf8String
StreamingJsonParsingContext::
expectStringUtf8()
{
    skipJsonWhitespace((*context));
    context->expect_literal('"');

    char internalBuffer[4096];

    char * buffer = internalBuffer;
    size_t bufferSize = 4096;
    size_t pos = 0;

    // Keep expanding until it fits
    while (!context->match_literal('"')) {
        // We need up to 4 characters to add a new UTF-8 code point
        if (pos >= bufferSize - 4) {
            size_t newBufferSize = bufferSize * 8;
            char * newBuffer = new char[newBufferSize];
            std::copy(buffer, buffer + bufferSize, newBuffer);
            if (buffer != internalBuffer)
                delete[] buffer;
            buffer = newBuffer;
            bufferSize = newBufferSize;
        }

        int c = *(*context);
        
        //cerr << "c = " << c << " " << (char)c << endl;

        if (c < 0 || c > 127) {
            // Unicode
            c = utf8::unchecked::next(*context);

            char * p1 = buffer + pos;
            char * p2 = p1;
            pos += utf8::append(c, p2) - p1;

            continue;
        }
        ++(*context);

        if (c == '\\') {
            c = *(*context)++;
            switch (c) {
            case 't': c = '\t';  break;
            case 'n': c = '\n';  break;
            case 'r': c = '\r';  break;
            case 'f': c = '\f';  break;
            case '/': c = '/';   break;
            case '\\':c = '\\';  break;
            case '"': c = '"';   break;
            case 'u': {
                int code = 0;
                for (unsigned i = 0;  i < 4;  ++i) {
                    int c = *(*context)++;
                    int digit;
                    if (c >= '0' && c <= '9')
                        digit = c - '0';
                    else if (c >= 'a' && c <= 'f')
                        digit = c - 'a';
                    else if (c >= 'A' && c <= 'F')
                        digit = c - 'A';
                    else context->exception("invalid hexadecimal in code");

                    code = (code << 4) | digit;
                }
                //cerr << "code = " << code << endl;
                c = code;
                break;
            }
            default:
                context->exception("invalid escaped char");
            }
        }

        if (c < ' ' || c >= 127) {
            char * p1 = buffer + pos;
            char * p2 = p1;
            pos += utf8::append(c, p2) - p1;
        }
        else buffer[pos++] = c;
    }

    Utf8String result(string(buffer, buffer + pos));
    if (buffer != internalBuffer)
        delete[] buffer;
    
    return result;
}


}  // namespace Datacratic