json.c

Go to the documentation of this file.

/*-------------------------------------------------------------------------
 *
 * json.c
 *      JSON data type support.
 *
 * Portions Copyright (c) 1996-2013, PostgreSQL Global Development Group
 * Portions Copyright (c) 1994, Regents of the University of California
 *
 * IDENTIFICATION
 *    src/backend/utils/adt/json.c
 *
 *-------------------------------------------------------------------------
 */
#include "postgres.h"

#include "access/htup_details.h"
#include "access/transam.h"
#include "catalog/pg_cast.h"
#include "catalog/pg_type.h"
#include "executor/spi.h"
#include "lib/stringinfo.h"
#include "libpq/pqformat.h"
#include "mb/pg_wchar.h"
#include "parser/parse_coerce.h"
#include "utils/array.h"
#include "utils/builtins.h"
#include "utils/lsyscache.h"
#include "utils/json.h"
#include "utils/jsonapi.h"
#include "utils/typcache.h"
#include "utils/syscache.h"

/*
 * The context of the parser is maintained by the recursive descent
 * mechanism, but is passed explicitly to the error reporting routine
 * for better diagnostics.
 */
typedef enum                    /* contexts of JSON parser */
{
    JSON_PARSE_VALUE,           /* expecting a value */
    JSON_PARSE_STRING,          /* expecting a string (for a field name) */
    JSON_PARSE_ARRAY_START,     /* saw '[', expecting value or ']' */
    JSON_PARSE_ARRAY_NEXT,      /* saw array element, expecting ',' or ']' */
    JSON_PARSE_OBJECT_START,    /* saw '{', expecting label or '}' */
    JSON_PARSE_OBJECT_LABEL,    /* saw object label, expecting ':' */
    JSON_PARSE_OBJECT_NEXT,     /* saw object value, expecting ',' or '}' */
    JSON_PARSE_OBJECT_COMMA,    /* saw object ',', expecting next label */
    JSON_PARSE_END              /* saw the end of a document, expect nothing */
}   JsonParseContext;

static inline void json_lex(JsonLexContext *lex);
static inline void json_lex_string(JsonLexContext *lex);
static inline void json_lex_number(JsonLexContext *lex, char *s);
static inline void parse_scalar(JsonLexContext *lex, JsonSemAction sem);
static void parse_object_field(JsonLexContext *lex, JsonSemAction sem);
static void parse_object(JsonLexContext *lex, JsonSemAction sem);
static void parse_array_element(JsonLexContext *lex, JsonSemAction sem);
static void parse_array(JsonLexContext *lex, JsonSemAction sem);
static void report_parse_error(JsonParseContext ctx, JsonLexContext *lex);
static void report_invalid_token(JsonLexContext *lex);
static int  report_json_context(JsonLexContext *lex);
static char *extract_mb_char(char *s);
static void composite_to_json(Datum composite, StringInfo result,
                  bool use_line_feeds);
static void array_dim_to_json(StringInfo result, int dim, int ndims, int *dims,
                  Datum *vals, bool *nulls, int *valcount,
                  TYPCATEGORY tcategory, Oid typoutputfunc,
                  bool use_line_feeds);
static void array_to_json_internal(Datum array, StringInfo result,
                       bool use_line_feeds);

/* the null action object used for pure validation */
static jsonSemAction nullSemAction =
{
    NULL, NULL, NULL, NULL, NULL,
    NULL, NULL, NULL, NULL, NULL
};
static JsonSemAction NullSemAction = &nullSemAction;

/* Recursive Descent parser support routines */

/*
 * lex_peek
 *
 * what is the current look_ahead token?
*/
static inline JsonTokenType
lex_peek(JsonLexContext *lex)
{
    return lex->token_type;
}

/*
 * lex_accept
 *
 * accept the look_ahead token and move the lexer to the next token if the
 * look_ahead token matches the token parameter. In that case, and if required,
 * also hand back the de-escaped lexeme.
 *
 * returns true if the token matched, false otherwise.
 */
static inline bool
lex_accept(JsonLexContext *lex, JsonTokenType token, char **lexeme)
{
    if (lex->token_type == token)
    {
        if (lexeme != NULL)
        {
            if (lex->token_type == JSON_TOKEN_STRING)
            {
                if (lex->strval != NULL)
                    *lexeme = pstrdup(lex->strval->data);
            }
            else
            {
                int         len = (lex->token_terminator - lex->token_start);
                char       *tokstr = palloc(len + 1);

                memcpy(tokstr, lex->token_start, len);
                tokstr[len] = '\0';
                *lexeme = tokstr;
            }
        }
        json_lex(lex);
        return true;
    }
    return false;
}

/*
 * lex_accept
 *
 * move the lexer to the next token if the current look_ahead token matches
 * the parameter token. Otherwise, report an error.
 */
static inline void
lex_expect(JsonParseContext ctx, JsonLexContext *lex, JsonTokenType token)
{
    if (!lex_accept(lex, token, NULL))
        report_parse_error(ctx, lex);;
}

/*
 * All the defined  type categories are upper case , so use lower case here
 * so we avoid any possible clash.
 */
/* fake type category for JSON so we can distinguish it in datum_to_json */
#define TYPCATEGORY_JSON 'j'
/* fake category for types that have a cast to json */
#define TYPCATEGORY_JSON_CAST 'c'
/* letters appearing in numeric output that aren't valid in a JSON number */
#define NON_NUMERIC_LETTER "NnAaIiFfTtYy"
/* chars to consider as part of an alphanumeric token */
#define JSON_ALPHANUMERIC_CHAR(c)  \
    (((c) >= 'a' && (c) <= 'z') || \
     ((c) >= 'A' && (c) <= 'Z') || \
     ((c) >= '0' && (c) <= '9') || \
     (c) == '_' || \
     IS_HIGHBIT_SET(c))

/*
 * Input.
 */
Datum
json_in(PG_FUNCTION_ARGS)
{
    char       *json = PG_GETARG_CSTRING(0);
    text       *result = cstring_to_text(json);
    JsonLexContext *lex;

    /* validate it */
    lex = makeJsonLexContext(result, false);
    pg_parse_json(lex, NullSemAction);

    /* Internal representation is the same as text, for now */
    PG_RETURN_TEXT_P(result);
}

/*
 * Output.
 */
Datum
json_out(PG_FUNCTION_ARGS)
{
    /* we needn't detoast because text_to_cstring will handle that */
    Datum       txt = PG_GETARG_DATUM(0);

    PG_RETURN_CSTRING(TextDatumGetCString(txt));
}

/*
 * Binary send.
 */
Datum
json_send(PG_FUNCTION_ARGS)
{
    text       *t = PG_GETARG_TEXT_PP(0);
    StringInfoData buf;

    pq_begintypsend(&buf);
    pq_sendtext(&buf, VARDATA_ANY(t), VARSIZE_ANY_EXHDR(t));
    PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}

/*
 * Binary receive.
 */
Datum
json_recv(PG_FUNCTION_ARGS)
{
    StringInfo  buf = (StringInfo) PG_GETARG_POINTER(0);
    text       *result;
    char       *str;
    int         nbytes;
    JsonLexContext *lex;

    str = pq_getmsgtext(buf, buf->len - buf->cursor, &nbytes);

    result = palloc(nbytes + VARHDRSZ);
    SET_VARSIZE(result, nbytes + VARHDRSZ);
    memcpy(VARDATA(result), str, nbytes);

    /* Validate it. */
    lex = makeJsonLexContext(result, false);
    pg_parse_json(lex, NullSemAction);

    PG_RETURN_TEXT_P(result);
}

/*
 * makeJsonLexContext
 *
 * lex constructor, with or without StringInfo object
 * for de-escaped lexemes.
 *
 * Without is better as it makes the processing faster, so only make one
 * if really required.
 */
JsonLexContext *
makeJsonLexContext(text *json, bool need_escapes)
{
    JsonLexContext *lex = palloc0(sizeof(JsonLexContext));

    lex->input = lex->token_terminator = lex->line_start = VARDATA(json);
    lex->line_number = 1;
    lex->input_length = VARSIZE(json) - VARHDRSZ;
    if (need_escapes)
        lex->strval = makeStringInfo();
    return lex;
}

/*
 * pg_parse_json
 *
 * Publicly visible entry point for the JSON parser.
 *
 * lex is a lexing context, set up for the json to be processed by calling
 * makeJsonLexContext(). sem is a strucure of function pointers to semantic
 * action routines to be called at appropriate spots during parsing, and a
 * pointer to a state object to be passed to those routines.
 */
void
pg_parse_json(JsonLexContext *lex, JsonSemAction sem)
{
    JsonTokenType tok;

    /* get the initial token */
    json_lex(lex);

    tok = lex_peek(lex);

    /* parse by recursive descent */
    switch (tok)
    {
        case JSON_TOKEN_OBJECT_START:
            parse_object(lex, sem);
            break;
        case JSON_TOKEN_ARRAY_START:
            parse_array(lex, sem);
            break;
        default:
            parse_scalar(lex, sem);     /* json can be a bare scalar */
    }

    lex_expect(JSON_PARSE_END, lex, JSON_TOKEN_END);

}

/*
 *  Recursive Descent parse routines. There is one for each structural
 *  element in a json document:
 *    - scalar (string, number, true, false, null)
 *    - array  ( [ ] )
 *    - array element
 *    - object ( { } )
 *    - object field
 */
static inline void
parse_scalar(JsonLexContext *lex, JsonSemAction sem)
{
    char       *val = NULL;
    json_scalar_action sfunc = sem->scalar;
    char      **valaddr;
    JsonTokenType tok = lex_peek(lex);

    valaddr = sfunc == NULL ? NULL : &val;

    /* a scalar must be a string, a number, true, false, or null */
    switch (tok)
    {
        case JSON_TOKEN_TRUE:
            lex_accept(lex, JSON_TOKEN_TRUE, valaddr);
            break;
        case JSON_TOKEN_FALSE:
            lex_accept(lex, JSON_TOKEN_FALSE, valaddr);
            break;
        case JSON_TOKEN_NULL:
            lex_accept(lex, JSON_TOKEN_NULL, valaddr);
            break;
        case JSON_TOKEN_NUMBER:
            lex_accept(lex, JSON_TOKEN_NUMBER, valaddr);
            break;
        case JSON_TOKEN_STRING:
            lex_accept(lex, JSON_TOKEN_STRING, valaddr);
            break;
        default:
            report_parse_error(JSON_PARSE_VALUE, lex);
    }

    if (sfunc != NULL)
        (*sfunc) (sem->semstate, val, tok);
}

static void
parse_object_field(JsonLexContext *lex, JsonSemAction sem)
{
    /*
     * an object field is "fieldname" : value where value can be a scalar,
     * object or array
     */

    char       *fname = NULL;   /* keep compiler quiet */
    json_ofield_action ostart = sem->object_field_start;
    json_ofield_action oend = sem->object_field_end;
    bool        isnull;
    char      **fnameaddr = NULL;
    JsonTokenType tok;

    if (ostart != NULL || oend != NULL)
        fnameaddr = &fname;

    if (!lex_accept(lex, JSON_TOKEN_STRING, fnameaddr))
        report_parse_error(JSON_PARSE_STRING, lex);

    lex_expect(JSON_PARSE_OBJECT_LABEL, lex, JSON_TOKEN_COLON);

    tok = lex_peek(lex);
    isnull = tok == JSON_TOKEN_NULL;

    if (ostart != NULL)
        (*ostart) (sem->semstate, fname, isnull);

    switch (tok)
    {
        case JSON_TOKEN_OBJECT_START:
            parse_object(lex, sem);
            break;
        case JSON_TOKEN_ARRAY_START:
            parse_array(lex, sem);
            break;
        default:
            parse_scalar(lex, sem);
    }

    if (oend != NULL)
        (*oend) (sem->semstate, fname, isnull);

    if (fname != NULL)
        pfree(fname);
}

static void
parse_object(JsonLexContext *lex, JsonSemAction sem)
{
    /*
     * an object is a possibly empty sequence of object fields, separated by
     * commas and surrounde by curly braces.
     */
    json_struct_action ostart = sem->object_start;
    json_struct_action oend = sem->object_end;
    JsonTokenType tok;

    if (ostart != NULL)
        (*ostart) (sem->semstate);

    /*
     * Data inside an object at at a higher nesting level than the object
     * itself. Note that we increment this after we call the semantic routine
     * for the object start and restore it before we call the routine for the
     * object end.
     */
    lex->lex_level++;

    /* we know this will succeeed, just clearing the token */
    lex_expect(JSON_PARSE_OBJECT_START, lex, JSON_TOKEN_OBJECT_START);

    tok = lex_peek(lex);
    switch (tok)
    {
        case JSON_TOKEN_STRING:
            parse_object_field(lex, sem);
            while (lex_accept(lex, JSON_TOKEN_COMMA, NULL))
                parse_object_field(lex, sem);
            break;
        case JSON_TOKEN_OBJECT_END:
            break;
        default:
            /* case of an invalid initial token inside the object */
            report_parse_error(JSON_PARSE_OBJECT_START, lex);
    }

    lex_expect(JSON_PARSE_OBJECT_NEXT, lex, JSON_TOKEN_OBJECT_END);

    lex->lex_level--;

    if (oend != NULL)
        (*oend) (sem->semstate);
}

static void
parse_array_element(JsonLexContext *lex, JsonSemAction sem)
{
    json_aelem_action astart = sem->array_element_start;
    json_aelem_action aend = sem->array_element_end;
    JsonTokenType tok = lex_peek(lex);

    bool        isnull;

    isnull = tok == JSON_TOKEN_NULL;

    if (astart != NULL)
        (*astart) (sem->semstate, isnull);

    /* an array element is any object, array or scalar */
    switch (tok)
    {
        case JSON_TOKEN_OBJECT_START:
            parse_object(lex, sem);
            break;
        case JSON_TOKEN_ARRAY_START:
            parse_array(lex, sem);
            break;
        default:
            parse_scalar(lex, sem);
    }

    if (aend != NULL)
        (*aend) (sem->semstate, isnull);
}

static void
parse_array(JsonLexContext *lex, JsonSemAction sem)
{
    /*
     * an array is a possibly empty sequence of array elements, separated by
     * commas and surrounded by square brackets.
     */
    json_struct_action astart = sem->array_start;
    json_struct_action aend = sem->array_end;

    if (astart != NULL)
        (*astart) (sem->semstate);

    /*
     * Data inside an array at at a higher nesting level than the array
     * itself. Note that we increment this after we call the semantic routine
     * for the array start and restore it before we call the routine for the
     * array end.
     */
    lex->lex_level++;

    lex_expect(JSON_PARSE_ARRAY_START, lex, JSON_TOKEN_ARRAY_START);
    if (lex_peek(lex) != JSON_TOKEN_ARRAY_END)
    {

        parse_array_element(lex, sem);

        while (lex_accept(lex, JSON_TOKEN_COMMA, NULL))
            parse_array_element(lex, sem);
    }

    lex_expect(JSON_PARSE_ARRAY_NEXT, lex, JSON_TOKEN_ARRAY_END);

    lex->lex_level--;

    if (aend != NULL)
        (*aend) (sem->semstate);
}

/*
 * Lex one token from the input stream.
 */
static inline void
json_lex(JsonLexContext *lex)
{
    char       *s;
    int         len;

    /* Skip leading whitespace. */
    s = lex->token_terminator;
    len = s - lex->input;
    while (len < lex->input_length &&
           (*s == ' ' || *s == '\t' || *s == '\n' || *s == '\r'))
    {
        if (*s == '\n')
            ++lex->line_number;
        ++s;
        ++len;
    }
    lex->token_start = s;

    /* Determine token type. */
    if (len >= lex->input_length)
    {
        lex->token_start = NULL;
        lex->prev_token_terminator = lex->token_terminator;
        lex->token_terminator = s;
        lex->token_type = JSON_TOKEN_END;
    }
    else
        switch (*s)
        {
                /* Single-character token, some kind of punctuation mark. */
            case '{':
                lex->prev_token_terminator = lex->token_terminator;
                lex->token_terminator = s + 1;
                lex->token_type = JSON_TOKEN_OBJECT_START;
                break;
            case '}':
                lex->prev_token_terminator = lex->token_terminator;
                lex->token_terminator = s + 1;
                lex->token_type = JSON_TOKEN_OBJECT_END;
                break;
            case '[':
                lex->prev_token_terminator = lex->token_terminator;
                lex->token_terminator = s + 1;
                lex->token_type = JSON_TOKEN_ARRAY_START;
                break;
            case ']':
                lex->prev_token_terminator = lex->token_terminator;
                lex->token_terminator = s + 1;
                lex->token_type = JSON_TOKEN_ARRAY_END;
                break;
            case ',':
                lex->prev_token_terminator = lex->token_terminator;
                lex->token_terminator = s + 1;
                lex->token_type = JSON_TOKEN_COMMA;
                break;
            case ':':
                lex->prev_token_terminator = lex->token_terminator;
                lex->token_terminator = s + 1;
                lex->token_type = JSON_TOKEN_COLON;
                break;
            case '"':
                /* string */
                json_lex_string(lex);
                lex->token_type = JSON_TOKEN_STRING;
                break;
            case '-':
                /* Negative number. */
                json_lex_number(lex, s + 1);
                lex->token_type = JSON_TOKEN_NUMBER;
                break;
            case '0':
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
                /* Positive number. */
                json_lex_number(lex, s);
                lex->token_type = JSON_TOKEN_NUMBER;
                break;
            default:
                {
                    char       *p;

                    /*
                     * We're not dealing with a string, number, legal
                     * punctuation mark, or end of string.  The only legal
                     * tokens we might find here are true, false, and null,
                     * but for error reporting purposes we scan until we see a
                     * non-alphanumeric character.  That way, we can report
                     * the whole word as an unexpected token, rather than just
                     * some unintuitive prefix thereof.
                     */
                    for (p = s; JSON_ALPHANUMERIC_CHAR(*p) && p - s < lex->input_length - len; p++)
                         /* skip */ ;

                    /*
                     * We got some sort of unexpected punctuation or an
                     * otherwise unexpected character, so just complain about
                     * that one character.
                     */
                    if (p == s)
                    {
                        lex->prev_token_terminator = lex->token_terminator;
                        lex->token_terminator = s + 1;
                        report_invalid_token(lex);
                    }

                    /*
                     * We've got a real alphanumeric token here.  If it
                     * happens to be true, false, or null, all is well.  If
                     * not, error out.
                     */
                    lex->prev_token_terminator = lex->token_terminator;
                    lex->token_terminator = p;
                    if (p - s == 4)
                    {
                        if (memcmp(s, "true", 4) == 0)
                            lex->token_type = JSON_TOKEN_TRUE;
                        else if (memcmp(s, "null", 4) == 0)
                            lex->token_type = JSON_TOKEN_NULL;
                        else
                            report_invalid_token(lex);
                    }
                    else if (p - s == 5 && memcmp(s, "false", 5) == 0)
                        lex->token_type = JSON_TOKEN_FALSE;
                    else
                        report_invalid_token(lex);

                }
        }                       /* end of switch */
}

/*
 * The next token in the input stream is known to be a string; lex it.
 */
static inline void
json_lex_string(JsonLexContext *lex)
{
    char       *s;
    int         len;

    if (lex->strval != NULL)
        resetStringInfo(lex->strval);

    len = lex->token_start - lex->input;
    len++;
    for (s = lex->token_start + 1; *s != '"'; s++, len++)
    {
        /* Premature end of the string. */
        if (len >= lex->input_length)
        {
            lex->token_terminator = s;
            report_invalid_token(lex);
        }
        else if ((unsigned char) *s < 32)
        {
            /* Per RFC4627, these characters MUST be escaped. */
            /* Since *s isn't printable, exclude it from the context string */
            lex->token_terminator = s;
            ereport(ERROR,
                    (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                     errmsg("invalid input syntax for type json"),
                     errdetail("Character with value 0x%02x must be escaped.",
                               (unsigned char) *s),
                     report_json_context(lex)));
        }
        else if (*s == '\\')
        {
            /* OK, we have an escape character. */
            s++;
            len++;
            if (len >= lex->input_length)
            {
                lex->token_terminator = s;
                report_invalid_token(lex);
            }
            else if (*s == 'u')
            {
                int         i;
                int         ch = 0;

                for (i = 1; i <= 4; i++)
                {
                    s++;
                    len++;
                    if (len >= lex->input_length)
                    {
                        lex->token_terminator = s;
                        report_invalid_token(lex);
                    }
                    else if (*s >= '0' && *s <= '9')
                        ch = (ch * 16) + (*s - '0');
                    else if (*s >= 'a' && *s <= 'f')
                        ch = (ch * 16) + (*s - 'a') + 10;
                    else if (*s >= 'A' && *s <= 'F')
                        ch = (ch * 16) + (*s - 'A') + 10;
                    else
                    {
                        lex->token_terminator = s + pg_mblen(s);
                        ereport(ERROR,
                                (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                                 errmsg("invalid input syntax for type json"),
                                 errdetail("\"\\u\" must be followed by four hexadecimal digits."),
                                 report_json_context(lex)));
                    }
                }
                if (lex->strval != NULL)
                {
                    char        utf8str[5];
                    int         utf8len;
                    char       *converted;

                    unicode_to_utf8(ch, (unsigned char *) utf8str);
                    utf8len = pg_utf_mblen((unsigned char *) utf8str);
                    utf8str[utf8len] = '\0';
                    converted = pg_any_to_server(utf8str, utf8len, PG_UTF8);
                    appendStringInfoString(lex->strval, converted);
                    if (converted != utf8str)
                        pfree(converted);

                }
            }
            else if (lex->strval != NULL)
            {
                switch (*s)
                {
                    case '"':
                    case '\\':
                    case '/':
                        appendStringInfoChar(lex->strval, *s);
                        break;
                    case 'b':
                        appendStringInfoChar(lex->strval, '\b');
                        break;
                    case 'f':
                        appendStringInfoChar(lex->strval, '\f');
                        break;
                    case 'n':
                        appendStringInfoChar(lex->strval, '\n');
                        break;
                    case 'r':
                        appendStringInfoChar(lex->strval, '\r');
                        break;
                    case 't':
                        appendStringInfoChar(lex->strval, '\t');
                        break;
                    default:
                        /* Not a valid string escape, so error out. */
                        lex->token_terminator = s + pg_mblen(s);
                        ereport(ERROR,
                                (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                                 errmsg("invalid input syntax for type json"),
                            errdetail("Escape sequence \"\\%s\" is invalid.",
                                      extract_mb_char(s)),
                                 report_json_context(lex)));
                }
            }
            else if (strchr("\"\\/bfnrt", *s) == NULL)
            {
                /*
                 * Simpler processing if we're not bothered about de-escaping
                 *
                 * It's very tempting to remove the strchr() call here and
                 * replace it with a switch statement, but testing so far has
                 * shown it's not a performance win.
                 */
                lex->token_terminator = s + pg_mblen(s);
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                         errdetail("Escape sequence \"\\%s\" is invalid.",
                                   extract_mb_char(s)),
                         report_json_context(lex)));
            }

        }
        else if (lex->strval != NULL)
        {
            appendStringInfoChar(lex->strval, *s);
        }

    }

    /* Hooray, we found the end of the string! */
    lex->prev_token_terminator = lex->token_terminator;
    lex->token_terminator = s + 1;
}

/*-------------------------------------------------------------------------
 * The next token in the input stream is known to be a number; lex it.
 *
 * In JSON, a number consists of four parts:
 *
 * (1) An optional minus sign ('-').
 *
 * (2) Either a single '0', or a string of one or more digits that does not
 *     begin with a '0'.
 *
 * (3) An optional decimal part, consisting of a period ('.') followed by
 *     one or more digits.  (Note: While this part can be omitted
 *     completely, it's not OK to have only the decimal point without
 *     any digits afterwards.)
 *
 * (4) An optional exponent part, consisting of 'e' or 'E', optionally
 *     followed by '+' or '-', followed by one or more digits.  (Note:
 *     As with the decimal part, if 'e' or 'E' is present, it must be
 *     followed by at least one digit.)
 *
 * The 's' argument to this function points to the ostensible beginning
 * of part 2 - i.e. the character after any optional minus sign, and the
 * first character of the string if there is none.
 *
 *-------------------------------------------------------------------------
 */
static inline void
json_lex_number(JsonLexContext *lex, char *s)
{
    bool        error = false;
    char       *p;
    int         len;

    len = s - lex->input;
    /* Part (1): leading sign indicator. */
    /* Caller already did this for us; so do nothing. */

    /* Part (2): parse main digit string. */
    if (*s == '0')
    {
        s++;
        len++;
    }
    else if (*s >= '1' && *s <= '9')
    {
        do
        {
            s++;
            len++;
        } while (*s >= '0' && *s <= '9' && len < lex->input_length);
    }
    else
        error = true;

    /* Part (3): parse optional decimal portion. */
    if (len < lex->input_length && *s == '.')
    {
        s++;
        len++;
        if (len == lex->input_length || *s < '0' || *s > '9')
            error = true;
        else
        {
            do
            {
                s++;
                len++;
            } while (*s >= '0' && *s <= '9' && len < lex->input_length);
        }
    }

    /* Part (4): parse optional exponent. */
    if (len < lex->input_length && (*s == 'e' || *s == 'E'))
    {
        s++;
        len++;
        if (len < lex->input_length && (*s == '+' || *s == '-'))
        {
            s++;
            len++;
        }
        if (len == lex->input_length || *s < '0' || *s > '9')
            error = true;
        else
        {
            do
            {
                s++;
                len++;
            } while (len < lex->input_length && *s >= '0' && *s <= '9');
        }
    }

    /*
     * Check for trailing garbage.  As in json_lex(), any alphanumeric stuff
     * here should be considered part of the token for error-reporting
     * purposes.
     */
    for (p = s; JSON_ALPHANUMERIC_CHAR(*p) && len < lex->input_length; p++, len++)
        error = true;
    lex->prev_token_terminator = lex->token_terminator;
    lex->token_terminator = p;
    if (error)
        report_invalid_token(lex);
}

/*
 * Report a parse error.
 *
 * lex->token_start and lex->token_terminator must identify the current token.
 */
static void
report_parse_error(JsonParseContext ctx, JsonLexContext *lex)
{
    char       *token;
    int         toklen;

    /* Handle case where the input ended prematurely. */
    if (lex->token_start == NULL || lex->token_type == JSON_TOKEN_END)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                 errmsg("invalid input syntax for type json"),
                 errdetail("The input string ended unexpectedly."),
                 report_json_context(lex)));

    /* Separate out the current token. */
    toklen = lex->token_terminator - lex->token_start;
    token = palloc(toklen + 1);
    memcpy(token, lex->token_start, toklen);
    token[toklen] = '\0';

    /* Complain, with the appropriate detail message. */
    if (ctx == JSON_PARSE_END)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                 errmsg("invalid input syntax for type json"),
                 errdetail("Expected end of input, but found \"%s\".",
                           token),
                 report_json_context(lex)));
    else
    {
        switch (ctx)
        {
            case JSON_PARSE_VALUE:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                         errdetail("Expected JSON value, but found \"%s\".",
                                   token),
                         report_json_context(lex)));
                break;
            case JSON_PARSE_STRING:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                         errdetail("Expected string, but found \"%s\".",
                                   token),
                         report_json_context(lex)));
                break;
            case JSON_PARSE_ARRAY_START:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                         errdetail("Expected array element or \"]\", but found \"%s\".",
                                   token),
                         report_json_context(lex)));
                break;
            case JSON_PARSE_ARRAY_NEXT:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                      errdetail("Expected \",\" or \"]\", but found \"%s\".",
                                token),
                         report_json_context(lex)));
                break;
            case JSON_PARSE_OBJECT_START:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                     errdetail("Expected string or \"}\", but found \"%s\".",
                               token),
                         report_json_context(lex)));
                break;
            case JSON_PARSE_OBJECT_LABEL:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                         errdetail("Expected \":\", but found \"%s\".",
                                   token),
                         report_json_context(lex)));
                break;
            case JSON_PARSE_OBJECT_NEXT:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                      errdetail("Expected \",\" or \"}\", but found \"%s\".",
                                token),
                         report_json_context(lex)));
                break;
            case JSON_PARSE_OBJECT_COMMA:
                ereport(ERROR,
                        (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
                         errmsg("invalid input syntax for type json"),
                         errdetail("Expected string, but found \"%s\".",
                                   token),
                         report_json_context(lex)));
                break;
            default:
                elog(ERROR, "unexpected json parse state: %d", ctx);
        }
    }
}

/*
 * Report an invalid input token.
 *
 * lex->token_start and lex->token_terminator must identify the token.
 */
static void
report_invalid_token(JsonLexContext *lex)
{
    char       *token;
    int         toklen;

    /* Separate out the offending token. */
    toklen = lex->token_terminator - lex->token_start;
    token = palloc(toklen + 1);
    memcpy(token, lex->token_start, toklen);
    token[toklen] = '\0';

    ereport(ERROR,
            (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
             errmsg("invalid input syntax for type json"),
             errdetail("Token \"%s\" is invalid.", token),
             report_json_context(lex)));
}

/*
 * Report a CONTEXT line for bogus JSON input.
 *
 * lex->token_terminator must be set to identify the spot where we detected
 * the error.  Note that lex->token_start might be NULL, in case we recognized
 * error at EOF.
 *
 * The return value isn't meaningful, but we make it non-void so that this
 * can be invoked inside ereport().
 */
static int
report_json_context(JsonLexContext *lex)
{
    const char *context_start;
    const char *context_end;
    const char *line_start;
    int         line_number;
    char       *ctxt;
    int         ctxtlen;
    const char *prefix;
    const char *suffix;

    /* Choose boundaries for the part of the input we will display */
    context_start = lex->input;
    context_end = lex->token_terminator;
    line_start = context_start;
    line_number = 1;
    for (;;)
    {
        /* Always advance over newlines (context_end test is just paranoia) */
        if (*context_start == '\n' && context_start < context_end)
        {
            context_start++;
            line_start = context_start;
            line_number++;
            continue;
        }
        /* Otherwise, done as soon as we are close enough to context_end */
        if (context_end - context_start < 50)
            break;
        /* Advance to next multibyte character */
        if (IS_HIGHBIT_SET(*context_start))
            context_start += pg_mblen(context_start);
        else
            context_start++;
    }

    /*
     * We add "..." to indicate that the excerpt doesn't start at the
     * beginning of the line ... but if we're within 3 characters of the
     * beginning of the line, we might as well just show the whole line.
     */
    if (context_start - line_start <= 3)
        context_start = line_start;

    /* Get a null-terminated copy of the data to present */
    ctxtlen = context_end - context_start;
    ctxt = palloc(ctxtlen + 1);
    memcpy(ctxt, context_start, ctxtlen);
    ctxt[ctxtlen] = '\0';

    /*
     * Show the context, prefixing "..." if not starting at start of line, and
     * suffixing "..." if not ending at end of line.
     */
    prefix = (context_start > line_start) ? "..." : "";
    suffix = (lex->token_type != JSON_TOKEN_END && context_end - lex->input < lex->input_length && *context_end != '\n' && *context_end != '\r') ? "..." : "";

    return errcontext("JSON data, line %d: %s%s%s",
                      line_number, prefix, ctxt, suffix);
}

/*
 * Extract a single, possibly multi-byte char from the input string.
 */
static char *
extract_mb_char(char *s)
{
    char       *res;
    int         len;

    len = pg_mblen(s);
    res = palloc(len + 1);
    memcpy(res, s, len);
    res[len] = '\0';

    return res;
}

/*
 * Turn a scalar Datum into JSON, appending the string to "result".
 *
 * Hand off a non-scalar datum to composite_to_json or array_to_json_internal
 * as appropriate.
 */
static void
datum_to_json(Datum val, bool is_null, StringInfo result,
              TYPCATEGORY tcategory, Oid typoutputfunc)
{
    char       *outputstr;
    text       *jsontext;

    if (is_null)
    {
        appendStringInfoString(result, "null");
        return;
    }

    switch (tcategory)
    {
        case TYPCATEGORY_ARRAY:
            array_to_json_internal(val, result, false);
            break;
        case TYPCATEGORY_COMPOSITE:
            composite_to_json(val, result, false);
            break;
        case TYPCATEGORY_BOOLEAN:
            if (DatumGetBool(val))
                appendStringInfoString(result, "true");
            else
                appendStringInfoString(result, "false");
            break;
        case TYPCATEGORY_NUMERIC:
            outputstr = OidOutputFunctionCall(typoutputfunc, val);

            /*
             * Don't call escape_json here if it's a valid JSON number.
             * Numeric output should usually be a valid JSON number and JSON
             * numbers shouldn't be quoted. Quote cases like "Nan" and
             * "Infinity", however.
             */
            if (strpbrk(outputstr, NON_NUMERIC_LETTER) == NULL)
                appendStringInfoString(result, outputstr);
            else
                escape_json(result, outputstr);
            pfree(outputstr);
            break;
        case TYPCATEGORY_JSON:
            /* JSON will already be escaped */
            outputstr = OidOutputFunctionCall(typoutputfunc, val);
            appendStringInfoString(result, outputstr);
            pfree(outputstr);
            break;
        case TYPCATEGORY_JSON_CAST:
            jsontext = DatumGetTextP(OidFunctionCall1(typoutputfunc, val));
            outputstr = text_to_cstring(jsontext);
            appendStringInfoString(result, outputstr);
            pfree(outputstr);
            pfree(jsontext);
            break;
        default:
            outputstr = OidOutputFunctionCall(typoutputfunc, val);
            escape_json(result, outputstr);
            pfree(outputstr);
            break;
    }
}

/*
 * Process a single dimension of an array.
 * If it's the innermost dimension, output the values, otherwise call
 * ourselves recursively to process the next dimension.
 */
static void
array_dim_to_json(StringInfo result, int dim, int ndims, int *dims, Datum *vals,
                  bool *nulls, int *valcount, TYPCATEGORY tcategory,
                  Oid typoutputfunc, bool use_line_feeds)
{
    int         i;
    const char *sep;

    Assert(dim < ndims);

    sep = use_line_feeds ? ",\n " : ",";

    appendStringInfoChar(result, '[');

    for (i = 1; i <= dims[dim]; i++)
    {
        if (i > 1)
            appendStringInfoString(result, sep);

        if (dim + 1 == ndims)
        {
            datum_to_json(vals[*valcount], nulls[*valcount], result, tcategory,
                          typoutputfunc);
            (*valcount)++;
        }
        else
        {
            /*
             * Do we want line feeds on inner dimensions of arrays? For now
             * we'll say no.
             */
            array_dim_to_json(result, dim + 1, ndims, dims, vals, nulls,
                              valcount, tcategory, typoutputfunc, false);
        }
    }

    appendStringInfoChar(result, ']');
}

/*
 * Turn an array into JSON.
 */
static void
array_to_json_internal(Datum array, StringInfo result, bool use_line_feeds)
{
    ArrayType  *v = DatumGetArrayTypeP(array);
    Oid         element_type = ARR_ELEMTYPE(v);
    int        *dim;
    int         ndim;
    int         nitems;
    int         count = 0;
    Datum      *elements;
    bool       *nulls;
    int16       typlen;
    bool        typbyval;
    char        typalign,
                typdelim;
    Oid         typioparam;
    Oid         typoutputfunc;
    TYPCATEGORY tcategory;
    Oid         castfunc = InvalidOid;

    ndim = ARR_NDIM(v);
    dim = ARR_DIMS(v);
    nitems = ArrayGetNItems(ndim, dim);

    if (nitems <= 0)
    {
        appendStringInfoString(result, "[]");
        return;
    }

    get_type_io_data(element_type, IOFunc_output,
                     &typlen, &typbyval, &typalign,
                     &typdelim, &typioparam, &typoutputfunc);

    if (element_type > FirstNormalObjectId)
    {
        HeapTuple   tuple;
        Form_pg_cast castForm;

        tuple = SearchSysCache2(CASTSOURCETARGET,
                                ObjectIdGetDatum(element_type),
                                ObjectIdGetDatum(JSONOID));
        if (HeapTupleIsValid(tuple))
        {
            castForm = (Form_pg_cast) GETSTRUCT(tuple);

            if (castForm->castmethod == COERCION_METHOD_FUNCTION)
                castfunc = typoutputfunc = castForm->castfunc;

            ReleaseSysCache(tuple);
        }
    }

    deconstruct_array(v, element_type, typlen, typbyval,
                      typalign, &elements, &nulls,
                      &nitems);

    if (castfunc != InvalidOid)
        tcategory = TYPCATEGORY_JSON_CAST;
    else if (element_type == RECORDOID)
        tcategory = TYPCATEGORY_COMPOSITE;
    else if (element_type == JSONOID)
        tcategory = TYPCATEGORY_JSON;
    else
        tcategory = TypeCategory(element_type);

    array_dim_to_json(result, 0, ndim, dim, elements, nulls, &count, tcategory,
                      typoutputfunc, use_line_feeds);

    pfree(elements);
    pfree(nulls);
}

/*
 * Turn a composite / record into JSON.
 */
static void
composite_to_json(Datum composite, StringInfo result, bool use_line_feeds)
{
    HeapTupleHeader td;
    Oid         tupType;
    int32       tupTypmod;
    TupleDesc   tupdesc;
    HeapTupleData tmptup,
               *tuple;
    int         i;
    bool        needsep = false;
    const char *sep;

    sep = use_line_feeds ? ",\n " : ",";

    td = DatumGetHeapTupleHeader(composite);

    /* Extract rowtype info and find a tupdesc */
    tupType = HeapTupleHeaderGetTypeId(td);
    tupTypmod = HeapTupleHeaderGetTypMod(td);
    tupdesc = lookup_rowtype_tupdesc(tupType, tupTypmod);

    /* Build a temporary HeapTuple control structure */
    tmptup.t_len = HeapTupleHeaderGetDatumLength(td);
    tmptup.t_data = td;
    tuple = &tmptup;

    appendStringInfoChar(result, '{');

    for (i = 0; i < tupdesc->natts; i++)
    {
        Datum       val,
                    origval;
        bool        isnull;
        char       *attname;
        TYPCATEGORY tcategory;
        Oid         typoutput;
        bool        typisvarlena;
        Oid         castfunc = InvalidOid;

        if (tupdesc->attrs[i]->attisdropped)
            continue;

        if (needsep)
            appendStringInfoString(result, sep);
        needsep = true;

        attname = NameStr(tupdesc->attrs[i]->attname);
        escape_json(result, attname);
        appendStringInfoChar(result, ':');

        origval = heap_getattr(tuple, i + 1, tupdesc, &isnull);

        getTypeOutputInfo(tupdesc->attrs[i]->atttypid,
                          &typoutput, &typisvarlena);

        if (tupdesc->attrs[i]->atttypid > FirstNormalObjectId)
        {
            HeapTuple   cast_tuple;
            Form_pg_cast castForm;

            cast_tuple = SearchSysCache2(CASTSOURCETARGET,
                               ObjectIdGetDatum(tupdesc->attrs[i]->atttypid),
                                         ObjectIdGetDatum(JSONOID));
            if (HeapTupleIsValid(cast_tuple))
            {
                castForm = (Form_pg_cast) GETSTRUCT(cast_tuple);

                if (castForm->castmethod == COERCION_METHOD_FUNCTION)
                    castfunc = typoutput = castForm->castfunc;

                ReleaseSysCache(cast_tuple);
            }
        }

        if (castfunc != InvalidOid)
            tcategory = TYPCATEGORY_JSON_CAST;
        else if (tupdesc->attrs[i]->atttypid == RECORDARRAYOID)
            tcategory = TYPCATEGORY_ARRAY;
        else if (tupdesc->attrs[i]->atttypid == RECORDOID)
            tcategory = TYPCATEGORY_COMPOSITE;
        else if (tupdesc->attrs[i]->atttypid == JSONOID)
            tcategory = TYPCATEGORY_JSON;
        else
            tcategory = TypeCategory(tupdesc->attrs[i]->atttypid);

        /*
         * If we have a toasted datum, forcibly detoast it here to avoid
         * memory leakage inside the type's output routine.
         */
        if (typisvarlena && !isnull)
            val = PointerGetDatum(PG_DETOAST_DATUM(origval));
        else
            val = origval;

        datum_to_json(val, isnull, result, tcategory, typoutput);

        /* Clean up detoasted copy, if any */
        if (val != origval)
            pfree(DatumGetPointer(val));
    }

    appendStringInfoChar(result, '}');
    ReleaseTupleDesc(tupdesc);
}

/*
 * SQL function array_to_json(row)
 */
extern Datum
array_to_json(PG_FUNCTION_ARGS)
{
    Datum       array = PG_GETARG_DATUM(0);
    StringInfo  result;

    result = makeStringInfo();

    array_to_json_internal(array, result, false);

    PG_RETURN_TEXT_P(cstring_to_text(result->data));
}

/*
 * SQL function array_to_json(row, prettybool)
 */
extern Datum
array_to_json_pretty(PG_FUNCTION_ARGS)
{
    Datum       array = PG_GETARG_DATUM(0);
    bool        use_line_feeds = PG_GETARG_BOOL(1);
    StringInfo  result;

    result = makeStringInfo();

    array_to_json_internal(array, result, use_line_feeds);

    PG_RETURN_TEXT_P(cstring_to_text(result->data));
}

/*
 * SQL function row_to_json(row)
 */
extern Datum
row_to_json(PG_FUNCTION_ARGS)
{
    Datum       array = PG_GETARG_DATUM(0);
    StringInfo  result;

    result = makeStringInfo();

    composite_to_json(array, result, false);

    PG_RETURN_TEXT_P(cstring_to_text(result->data));
}

/*
 * SQL function row_to_json(row, prettybool)
 */
extern Datum
row_to_json_pretty(PG_FUNCTION_ARGS)
{
    Datum       array = PG_GETARG_DATUM(0);
    bool        use_line_feeds = PG_GETARG_BOOL(1);
    StringInfo  result;

    result = makeStringInfo();

    composite_to_json(array, result, use_line_feeds);

    PG_RETURN_TEXT_P(cstring_to_text(result->data));
}

/*
 * SQL function to_json(anyvalue)
 */
Datum
to_json(PG_FUNCTION_ARGS)
{
    Oid         val_type = get_fn_expr_argtype(fcinfo->flinfo, 0);
    StringInfo  result;
    Datum       orig_val,
                val;
    TYPCATEGORY tcategory;
    Oid         typoutput;
    bool        typisvarlena;
    Oid         castfunc = InvalidOid;

    if (val_type == InvalidOid)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
                 errmsg("could not determine input data type")));


    result = makeStringInfo();

    orig_val = PG_ARGISNULL(0) ? (Datum) 0 : PG_GETARG_DATUM(0);

    getTypeOutputInfo(val_type, &typoutput, &typisvarlena);

    if (val_type > FirstNormalObjectId)
    {
        HeapTuple   tuple;
        Form_pg_cast castForm;

        tuple = SearchSysCache2(CASTSOURCETARGET,
                                ObjectIdGetDatum(val_type),
                                ObjectIdGetDatum(JSONOID));
        if (HeapTupleIsValid(tuple))
        {
            castForm = (Form_pg_cast) GETSTRUCT(tuple);

            if (castForm->castmethod == COERCION_METHOD_FUNCTION)
                castfunc = typoutput = castForm->castfunc;

            ReleaseSysCache(tuple);
        }
    }

    if (castfunc != InvalidOid)
        tcategory = TYPCATEGORY_JSON_CAST;
    else if (val_type == RECORDARRAYOID)
        tcategory = TYPCATEGORY_ARRAY;
    else if (val_type == RECORDOID)
        tcategory = TYPCATEGORY_COMPOSITE;
    else if (val_type == JSONOID)
        tcategory = TYPCATEGORY_JSON;
    else
        tcategory = TypeCategory(val_type);

    /*
     * If we have a toasted datum, forcibly detoast it here to avoid memory
     * leakage inside the type's output routine.
     */
    if (typisvarlena && orig_val != (Datum) 0)
        val = PointerGetDatum(PG_DETOAST_DATUM(orig_val));
    else
        val = orig_val;

    datum_to_json(val, false, result, tcategory, typoutput);

    /* Clean up detoasted copy, if any */
    if (val != orig_val)
        pfree(DatumGetPointer(val));

    PG_RETURN_TEXT_P(cstring_to_text(result->data));
}

/*
 * json_agg transition function
 */
Datum
json_agg_transfn(PG_FUNCTION_ARGS)
{
    Oid         val_type = get_fn_expr_argtype(fcinfo->flinfo, 1);
    MemoryContext aggcontext,
                oldcontext;
    StringInfo  state;
    Datum       orig_val,
                val;
    TYPCATEGORY tcategory;
    Oid         typoutput;
    bool        typisvarlena;
    Oid         castfunc = InvalidOid;

    if (val_type == InvalidOid)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
                 errmsg("could not determine input data type")));

    if (!AggCheckCallContext(fcinfo, &aggcontext))
    {
        /* cannot be called directly because of internal-type argument */
        elog(ERROR, "json_agg_transfn called in non-aggregate context");
    }

    if (PG_ARGISNULL(0))
    {
        /*
         * Make this StringInfo in a context where it will persist for the
         * duration off the aggregate call. It's only needed for this initial
         * piece, as the StringInfo routines make sure they use the right
         * context to enlarge the object if necessary.
         */
        oldcontext = MemoryContextSwitchTo(aggcontext);
        state = makeStringInfo();
        MemoryContextSwitchTo(oldcontext);

        appendStringInfoChar(state, '[');
    }
    else
    {
        state = (StringInfo) PG_GETARG_POINTER(0);
        appendStringInfoString(state, ", ");
    }

    /* fast path for NULLs */
    if (PG_ARGISNULL(1))
    {
        orig_val = (Datum) 0;
        datum_to_json(orig_val, true, state, 0, InvalidOid);
        PG_RETURN_POINTER(state);
    }


    orig_val = PG_GETARG_DATUM(1);

    getTypeOutputInfo(val_type, &typoutput, &typisvarlena);

    if (val_type > FirstNormalObjectId)
    {
        HeapTuple   tuple;
        Form_pg_cast castForm;

        tuple = SearchSysCache2(CASTSOURCETARGET,
                                ObjectIdGetDatum(val_type),
                                ObjectIdGetDatum(JSONOID));
        if (HeapTupleIsValid(tuple))
        {
            castForm = (Form_pg_cast) GETSTRUCT(tuple);

            if (castForm->castmethod == COERCION_METHOD_FUNCTION)
                castfunc = typoutput = castForm->castfunc;

            ReleaseSysCache(tuple);
        }
    }

    if (castfunc != InvalidOid)
        tcategory = TYPCATEGORY_JSON_CAST;
    else if (val_type == RECORDARRAYOID)
        tcategory = TYPCATEGORY_ARRAY;
    else if (val_type == RECORDOID)
        tcategory = TYPCATEGORY_COMPOSITE;
    else if (val_type == JSONOID)
        tcategory = TYPCATEGORY_JSON;
    else
        tcategory = TypeCategory(val_type);

    /*
     * If we have a toasted datum, forcibly detoast it here to avoid memory
     * leakage inside the type's output routine.
     */
    if (typisvarlena)
        val = PointerGetDatum(PG_DETOAST_DATUM(orig_val));
    else
        val = orig_val;

    if (!PG_ARGISNULL(0) &&
      (tcategory == TYPCATEGORY_ARRAY || tcategory == TYPCATEGORY_COMPOSITE))
    {
        appendStringInfoString(state, "\n ");
    }

    datum_to_json(val, false, state, tcategory, typoutput);

    /* Clean up detoasted copy, if any */
    if (val != orig_val)
        pfree(DatumGetPointer(val));

    /*
     * The transition type for array_agg() is declared to be "internal", which
     * is a pass-by-value type the same size as a pointer.  So we can safely
     * pass the ArrayBuildState pointer through nodeAgg.c's machinations.
     */
    PG_RETURN_POINTER(state);
}

/*
 * json_agg final function
 */
Datum
json_agg_finalfn(PG_FUNCTION_ARGS)
{
    StringInfo  state;

    /* cannot be called directly because of internal-type argument */
    Assert(AggCheckCallContext(fcinfo, NULL));

    state = PG_ARGISNULL(0) ? NULL : (StringInfo) PG_GETARG_POINTER(0);

    if (state == NULL)
        PG_RETURN_NULL();

    appendStringInfoChar(state, ']');

    PG_RETURN_TEXT_P(cstring_to_text(state->data));
}

/*
 * Produce a JSON string literal, properly escaping characters in the text.
 */
void
escape_json(StringInfo buf, const char *str)
{
    const char *p;

    appendStringInfoCharMacro(buf, '\"');
    for (p = str; *p; p++)
    {
        switch (*p)
        {
            case '\b':
                appendStringInfoString(buf, "\\b");
                break;
            case '\f':
                appendStringInfoString(buf, "\\f");
                break;
            case '\n':
                appendStringInfoString(buf, "\\n");
                break;
            case '\r':
                appendStringInfoString(buf, "\\r");
                break;
            case '\t':
                appendStringInfoString(buf, "\\t");
                break;
            case '"':
                appendStringInfoString(buf, "\\\"");
                break;
            case '\\':
                appendStringInfoString(buf, "\\\\");
                break;
            default:
                if ((unsigned char) *p < ' ')
                    appendStringInfo(buf, "\\u%04x", (int) *p);
                else
                    appendStringInfoCharMacro(buf, *p);
                break;
        }
    }
    appendStringInfoCharMacro(buf, '\"');
}