jsonapi.h File Reference

#include "lib/stringinfo.h"

Include dependency graph for jsonapi.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Data Structures
struct	JsonLexContext
struct	JsonSemAction

Macros
#define	JSONLEX_FREE_STRUCT   (1 << 0)
#define	JSONLEX_FREE_STRVAL   (1 << 1)

Typedefs
typedef enum JsonTokenType	JsonTokenType
typedef enum JsonParseErrorType	JsonParseErrorType
typedef struct JsonParserStack	JsonParserStack
typedef struct JsonIncrementalState	JsonIncrementalState
typedef struct JsonLexContext	JsonLexContext
typedef JsonParseErrorType(*	json_struct_action) (void *state)
typedef JsonParseErrorType(*	json_ofield_action) (void *state, char *fname, bool isnull)
typedef JsonParseErrorType(*	json_aelem_action) (void *state, bool isnull)
typedef JsonParseErrorType(*	json_scalar_action) (void *state, char *token, JsonTokenType tokentype)
typedef struct JsonSemAction	JsonSemAction

Enumerations
enum	JsonTokenType {   JSON_TOKEN_INVALID , JSON_TOKEN_STRING , JSON_TOKEN_NUMBER , JSON_TOKEN_OBJECT_START ,   JSON_TOKEN_OBJECT_END , JSON_TOKEN_ARRAY_START , JSON_TOKEN_ARRAY_END , JSON_TOKEN_COMMA ,   JSON_TOKEN_COLON , JSON_TOKEN_TRUE , JSON_TOKEN_FALSE , JSON_TOKEN_NULL ,   JSON_TOKEN_END }
enum	JsonParseErrorType {   JSON_SUCCESS , JSON_INCOMPLETE , JSON_INVALID_LEXER_TYPE , JSON_NESTING_TOO_DEEP ,   JSON_ESCAPING_INVALID , JSON_ESCAPING_REQUIRED , JSON_EXPECTED_ARRAY_FIRST , JSON_EXPECTED_ARRAY_NEXT ,   JSON_EXPECTED_COLON , JSON_EXPECTED_END , JSON_EXPECTED_JSON , JSON_EXPECTED_MORE ,   JSON_EXPECTED_OBJECT_FIRST , JSON_EXPECTED_OBJECT_NEXT , JSON_EXPECTED_STRING , JSON_INVALID_TOKEN ,   JSON_UNICODE_CODE_POINT_ZERO , JSON_UNICODE_ESCAPE_FORMAT , JSON_UNICODE_HIGH_ESCAPE , JSON_UNICODE_UNTRANSLATABLE ,   JSON_UNICODE_HIGH_SURROGATE , JSON_UNICODE_LOW_SURROGATE , JSON_SEM_ACTION_FAILED }

Functions
JsonParseErrorType	pg_parse_json (JsonLexContext *lex, JsonSemAction *sem)
JsonParseErrorType	pg_parse_json_incremental (JsonLexContext *lex, JsonSemAction *sem, char *json, int len, bool is_last)
JsonParseErrorType	json_count_array_elements (JsonLexContext *lex, int *elements)
JsonLexContext *	makeJsonLexContextCstringLen (JsonLexContext *lex, char *json, int len, int encoding, bool need_escapes)
JsonLexContext *	makeJsonLexContextIncremental (JsonLexContext *lex, int encoding, bool need_escapes)
void	freeJsonLexContext (JsonLexContext *lex)
JsonParseErrorType	json_lex (JsonLexContext *lex)
char *	json_errdetail (JsonParseErrorType error, JsonLexContext *lex)
bool	IsValidJsonNumber (const char *str, int len)

Variables
PGDLLIMPORT JsonSemAction	nullSemAction

Macro Definition Documentation

JSONLEX_FREE_STRUCT

#define JSONLEX_FREE_STRUCT   (1 << 0)

Definition at line 87 of file jsonapi.h.

JSONLEX_FREE_STRVAL

#define JSONLEX_FREE_STRVAL   (1 << 1)

Definition at line 88 of file jsonapi.h.

Typedef Documentation

json_aelem_action

typedef JsonParseErrorType(* json_aelem_action) (void *state, bool isnull)

Definition at line 111 of file jsonapi.h.

json_ofield_action

typedef JsonParseErrorType(* json_ofield_action) (void *state, char *fname, bool isnull)

Definition at line 110 of file jsonapi.h.

json_scalar_action

typedef JsonParseErrorType(* json_scalar_action) (void *state, char *token, JsonTokenType tokentype)

Definition at line 112 of file jsonapi.h.

json_struct_action

typedef JsonParseErrorType(* json_struct_action) (void *state)

Definition at line 109 of file jsonapi.h.

JsonIncrementalState

typedef struct JsonIncrementalState JsonIncrementalState

Definition at line 1 of file jsonapi.h.

JsonLexContext

typedef struct JsonLexContext JsonLexContext

JsonParseErrorType

typedef enum JsonParseErrorType JsonParseErrorType

JsonParserStack

typedef struct JsonParserStack JsonParserStack

Definition at line 1 of file jsonapi.h.

JsonSemAction

typedef struct JsonSemAction JsonSemAction

JsonTokenType

typedef enum JsonTokenType JsonTokenType

Enumeration Type Documentation

JsonParseErrorType

enum JsonParseErrorType

Enumerator
JSON_SUCCESS
JSON_INCOMPLETE
JSON_INVALID_LEXER_TYPE
JSON_NESTING_TOO_DEEP
JSON_ESCAPING_INVALID
JSON_ESCAPING_REQUIRED
JSON_EXPECTED_ARRAY_FIRST
JSON_EXPECTED_ARRAY_NEXT
JSON_EXPECTED_COLON
JSON_EXPECTED_END
JSON_EXPECTED_JSON
JSON_EXPECTED_MORE
JSON_EXPECTED_OBJECT_FIRST
JSON_EXPECTED_OBJECT_NEXT
JSON_EXPECTED_STRING
JSON_INVALID_TOKEN
JSON_UNICODE_CODE_POINT_ZERO
JSON_UNICODE_ESCAPE_FORMAT
JSON_UNICODE_HIGH_ESCAPE
JSON_UNICODE_UNTRANSLATABLE
JSON_UNICODE_HIGH_SURROGATE
JSON_UNICODE_LOW_SURROGATE
JSON_SEM_ACTION_FAILED

Definition at line 36 of file jsonapi.h.

{
    JSON_SUCCESS,
    JSON_INCOMPLETE,
    JSON_INVALID_LEXER_TYPE,
    JSON_NESTING_TOO_DEEP,
    JSON_ESCAPING_INVALID,
    JSON_ESCAPING_REQUIRED,
    JSON_EXPECTED_ARRAY_FIRST,
    JSON_EXPECTED_ARRAY_NEXT,
    JSON_EXPECTED_COLON,
    JSON_EXPECTED_END,
    JSON_EXPECTED_JSON,
    JSON_EXPECTED_MORE,
    JSON_EXPECTED_OBJECT_FIRST,
    JSON_EXPECTED_OBJECT_NEXT,
    JSON_EXPECTED_STRING,
    JSON_INVALID_TOKEN,
    JSON_UNICODE_CODE_POINT_ZERO,
    JSON_UNICODE_ESCAPE_FORMAT,
    JSON_UNICODE_HIGH_ESCAPE,
    JSON_UNICODE_UNTRANSLATABLE,
    JSON_UNICODE_HIGH_SURROGATE,
    JSON_UNICODE_LOW_SURROGATE,
    JSON_SEM_ACTION_FAILED,     /* error should already be reported */
} JsonParseErrorType;

JsonTokenType

enum JsonTokenType

Enumerator
JSON_TOKEN_INVALID
JSON_TOKEN_STRING
JSON_TOKEN_NUMBER
JSON_TOKEN_OBJECT_START
JSON_TOKEN_OBJECT_END
JSON_TOKEN_ARRAY_START
JSON_TOKEN_ARRAY_END
JSON_TOKEN_COMMA
JSON_TOKEN_COLON
JSON_TOKEN_TRUE
JSON_TOKEN_FALSE
JSON_TOKEN_NULL
JSON_TOKEN_END

Definition at line 19 of file jsonapi.h.

{
    JSON_TOKEN_INVALID,
    JSON_TOKEN_STRING,
    JSON_TOKEN_NUMBER,
    JSON_TOKEN_OBJECT_START,
    JSON_TOKEN_OBJECT_END,
    JSON_TOKEN_ARRAY_START,
    JSON_TOKEN_ARRAY_END,
    JSON_TOKEN_COMMA,
    JSON_TOKEN_COLON,
    JSON_TOKEN_TRUE,
    JSON_TOKEN_FALSE,
    JSON_TOKEN_NULL,
    JSON_TOKEN_END,
} JsonTokenType;

Function Documentation

freeJsonLexContext()

void freeJsonLexContext

(

JsonLexContext *

lex

)

Definition at line 482 of file jsonapi.c.

{
    if (lex->flags & JSONLEX_FREE_STRVAL)
        destroyStringInfo(lex->strval);
 
    if (lex->errormsg)
        destroyStringInfo(lex->errormsg);
 
    if (lex->incremental)
    {
        pfree(lex->inc_state->partial_token.data);
        pfree(lex->inc_state);
        pfree(lex->pstack->prediction);
        pfree(lex->pstack->fnames);
        pfree(lex->pstack->fnull);
        pfree(lex->pstack);
    }
 
    if (lex->flags & JSONLEX_FREE_STRUCT)
        pfree(lex);

References StringInfoData::data, destroyStringInfo(), JsonLexContext::errormsg, JsonLexContext::flags, JsonParserStack::fnames, JsonParserStack::fnull, JsonLexContext::inc_state, JsonLexContext::incremental, JSONLEX_FREE_STRUCT, JSONLEX_FREE_STRVAL, JsonIncrementalState::partial_token, pfree(), JsonParserStack::prediction, JsonLexContext::pstack, and JsonLexContext::strval.

Referenced by datum_to_jsonb_internal(), each_worker(), elements_worker(), get_json_object_as_hash(), get_worker(), iterate_json_values(), json_object_keys(), json_parse_manifest(), json_parse_manifest_incremental_shutdown(), json_validate(), main(), populate_array_json(), populate_recordset_worker(), and transform_json_string_values().

IsValidJsonNumber()

bool IsValidJsonNumber	(	const char *	str,
		int	len
	)

Definition at line 272 of file jsonapi.c.

{
    bool        numeric_error;
    int         total_len;
    JsonLexContext dummy_lex;
 
    if (len <= 0)
        return false;
 
    dummy_lex.incremental = false;
    dummy_lex.inc_state = NULL;
    dummy_lex.pstack = NULL;
 
    /*
     * json_lex_number expects a leading  '-' to have been eaten already.
     *
     * having to cast away the constness of str is ugly, but there's not much
     * easy alternative.
     */
    if (*str == '-')
    {
        dummy_lex.input = unconstify(char *, str) + 1;
        dummy_lex.input_length = len - 1;
    }
    else
    {
        dummy_lex.input = unconstify(char *, str);
        dummy_lex.input_length = len;
    }
 
    dummy_lex.token_start = dummy_lex.input;
 
    json_lex_number(&dummy_lex, dummy_lex.input, &numeric_error, &total_len);
 
    return (!numeric_error) && (total_len == dummy_lex.input_length);
}

References JsonLexContext::inc_state, JsonLexContext::incremental, JsonLexContext::input, JsonLexContext::input_length, json_lex_number(), len, JsonLexContext::pstack, str, JsonLexContext::token_start, and unconstify.

Referenced by hstore_to_json_loose(), and hstore_to_jsonb_loose().

json_count_array_elements()

JsonParseErrorType json_count_array_elements	(	JsonLexContext *	lex,
		int *	elements
	)

Definition at line 587 of file jsonapi.c.

{
    JsonLexContext copylex;
    int         count;
    JsonParseErrorType result;
 
    /*
     * It's safe to do this with a shallow copy because the lexical routines
     * don't scribble on the input. They do scribble on the other pointers
     * etc, so doing this with a copy makes that safe.
     */
    memcpy(&copylex, lex, sizeof(JsonLexContext));
    copylex.strval = NULL;      /* not interested in values here */
    copylex.lex_level++;
 
    count = 0;
    result = lex_expect(JSON_PARSE_ARRAY_START, &copylex,
                        JSON_TOKEN_ARRAY_START);
    if (result != JSON_SUCCESS)
        return result;
    if (lex_peek(&copylex) != JSON_TOKEN_ARRAY_END)
    {
        while (1)
        {
            count++;
            result = parse_array_element(&copylex, &nullSemAction);
            if (result != JSON_SUCCESS)
                return result;
            if (copylex.token_type != JSON_TOKEN_COMMA)
                break;
            result = json_lex(&copylex);
            if (result != JSON_SUCCESS)
                return result;
        }
    }
    result = lex_expect(JSON_PARSE_ARRAY_NEXT, &copylex,
                        JSON_TOKEN_ARRAY_END);
    if (result != JSON_SUCCESS)
        return result;
 
    *elements = count;
    return JSON_SUCCESS;

References json_lex(), JSON_PARSE_ARRAY_NEXT, JSON_PARSE_ARRAY_START, JSON_SUCCESS, JSON_TOKEN_ARRAY_END, JSON_TOKEN_ARRAY_START, JSON_TOKEN_COMMA, lex_expect(), JsonLexContext::lex_level, lex_peek(), nullSemAction, parse_array_element(), JsonLexContext::strval, and JsonLexContext::token_type.

Referenced by get_array_start().

json_errdetail()

char* json_errdetail	(	JsonParseErrorType	error,
		JsonLexContext *	lex
	)

Definition at line 2096 of file jsonapi.c.

{
    if (lex->errormsg)
        resetStringInfo(lex->errormsg);
    else
        lex->errormsg = makeStringInfo();
 
    /*
     * A helper for error messages that should print the current token. The
     * format must contain exactly one %.*s specifier.
     */
#define token_error(lex, format) \
    appendStringInfo((lex)->errormsg, _(format), \
                     (int) ((lex)->token_terminator - (lex)->token_start), \
                     (lex)->token_start);
 
    switch (error)
    {
        case JSON_INCOMPLETE:
        case JSON_SUCCESS:
            /* fall through to the error code after switch */
            break;
        case JSON_INVALID_LEXER_TYPE:
            if (lex->incremental)
                return _("Recursive descent parser cannot use incremental lexer");
            else
                return _("Incremental parser requires incremental lexer");
        case JSON_NESTING_TOO_DEEP:
            return (_("JSON nested too deep, maximum permitted depth is 6400"));
        case JSON_ESCAPING_INVALID:
            token_error(lex, "Escape sequence \"\\%.*s\" is invalid.");
            break;
        case JSON_ESCAPING_REQUIRED:
            appendStringInfo(lex->errormsg,
                             _("Character with value 0x%02x must be escaped."),
                             (unsigned char) *(lex->token_terminator));
            break;
        case JSON_EXPECTED_END:
            token_error(lex, "Expected end of input, but found \"%.*s\".");
            break;
        case JSON_EXPECTED_ARRAY_FIRST:
            token_error(lex, "Expected array element or \"]\", but found \"%.*s\".");
            break;
        case JSON_EXPECTED_ARRAY_NEXT:
            token_error(lex, "Expected \",\" or \"]\", but found \"%.*s\".");
            break;
        case JSON_EXPECTED_COLON:
            token_error(lex, "Expected \":\", but found \"%.*s\".");
            break;
        case JSON_EXPECTED_JSON:
            token_error(lex, "Expected JSON value, but found \"%.*s\".");
            break;
        case JSON_EXPECTED_MORE:
            return _("The input string ended unexpectedly.");
        case JSON_EXPECTED_OBJECT_FIRST:
            token_error(lex, "Expected string or \"}\", but found \"%.*s\".");
            break;
        case JSON_EXPECTED_OBJECT_NEXT:
            token_error(lex, "Expected \",\" or \"}\", but found \"%.*s\".");
            break;
        case JSON_EXPECTED_STRING:
            token_error(lex, "Expected string, but found \"%.*s\".");
            break;
        case JSON_INVALID_TOKEN:
            token_error(lex, "Token \"%.*s\" is invalid.");
            break;
        case JSON_UNICODE_CODE_POINT_ZERO:
            return _("\\u0000 cannot be converted to text.");
        case JSON_UNICODE_ESCAPE_FORMAT:
            return _("\"\\u\" must be followed by four hexadecimal digits.");
        case JSON_UNICODE_HIGH_ESCAPE:
            /* note: this case is only reachable in frontend not backend */
            return _("Unicode escape values cannot be used for code point values above 007F when the encoding is not UTF8.");
        case JSON_UNICODE_UNTRANSLATABLE:
 
            /*
             * Note: this case is only reachable in backend and not frontend.
             * #ifdef it away so the frontend doesn't try to link against
             * backend functionality.
             */
#ifndef FRONTEND
            return psprintf(_("Unicode escape value could not be translated to the server's encoding %s."),
                            GetDatabaseEncodingName());
#else
            Assert(false);
            break;
#endif
        case JSON_UNICODE_HIGH_SURROGATE:
            return _("Unicode high surrogate must not follow a high surrogate.");
        case JSON_UNICODE_LOW_SURROGATE:
            return _("Unicode low surrogate must follow a high surrogate.");
        case JSON_SEM_ACTION_FAILED:
            /* fall through to the error code after switch */
            break;
    }
#undef token_error
 
    /*
     * We don't use a default: case, so that the compiler will warn about
     * unhandled enum values.  But this needs to be here anyway to cover the
     * possibility of an incorrect input.
     */
    if (lex->errormsg->len == 0)
        appendStringInfo(lex->errormsg,
                         _("unexpected json parse error type: %d"),
                         (int) error);
 
    return lex->errormsg->data;

References _, appendStringInfo(), Assert, StringInfoData::data, error(), JsonLexContext::errormsg, GetDatabaseEncodingName(), JsonLexContext::incremental, JSON_ESCAPING_INVALID, JSON_ESCAPING_REQUIRED, JSON_EXPECTED_ARRAY_FIRST, JSON_EXPECTED_ARRAY_NEXT, JSON_EXPECTED_COLON, JSON_EXPECTED_END, JSON_EXPECTED_JSON, JSON_EXPECTED_MORE, JSON_EXPECTED_OBJECT_FIRST, JSON_EXPECTED_OBJECT_NEXT, JSON_EXPECTED_STRING, JSON_INCOMPLETE, JSON_INVALID_LEXER_TYPE, JSON_INVALID_TOKEN, JSON_NESTING_TOO_DEEP, JSON_SEM_ACTION_FAILED, JSON_SUCCESS, JSON_UNICODE_CODE_POINT_ZERO, JSON_UNICODE_ESCAPE_FORMAT, JSON_UNICODE_HIGH_ESCAPE, JSON_UNICODE_HIGH_SURROGATE, JSON_UNICODE_LOW_SURROGATE, JSON_UNICODE_UNTRANSLATABLE, StringInfoData::len, makeStringInfo(), psprintf(), resetStringInfo(), token_error, and JsonLexContext::token_terminator.

Referenced by json_errsave_error(), json_parse_manifest(), json_parse_manifest_incremental_chunk(), and main().

json_lex()

JsonParseErrorType json_lex

(

JsonLexContext *

lex

)

Definition at line 1308 of file jsonapi.c.

{
    char       *s;
    char       *const end = lex->input + lex->input_length;
    JsonParseErrorType result;
 
    if (lex->incremental && lex->inc_state->partial_completed)
    {
        /*
         * We just lexed a completed partial token on the last call, so reset
         * everything
         */
        resetStringInfo(&(lex->inc_state->partial_token));
        lex->token_terminator = lex->input;
        lex->inc_state->partial_completed = false;
    }
 
    s = lex->token_terminator;
 
    if (lex->incremental && lex->inc_state->partial_token.len)
    {
        /*
         * We have a partial token. Extend it and if completed lex it by a
         * recursive call
         */
        StringInfo  ptok = &(lex->inc_state->partial_token);
        int         added = 0;
        bool        tok_done = false;
        JsonLexContext dummy_lex;
        JsonParseErrorType partial_result;
 
        if (ptok->data[0] == '"')
        {
            /*
             * It's a string. Accumulate characters until we reach an
             * unescaped '"'.
             */
            int         escapes = 0;
 
            for (int i = ptok->len - 1; i > 0; i--)
            {
                /* count the trailing backslashes on the partial token */
                if (ptok->data[i] == '\\')
                    escapes++;
                else
                    break;
            }
 
            for (int i = 0; i < lex->input_length; i++)
            {
                char        c = lex->input[i];
 
                appendStringInfoCharMacro(ptok, c);
                added++;
                if (c == '"' && escapes % 2 == 0)
                {
                    tok_done = true;
                    break;
                }
                if (c == '\\')
                    escapes++;
                else
                    escapes = 0;
            }
        }
        else
        {
            /* not a string */
            char        c = ptok->data[0];
 
            if (c == '-' || (c >= '0' && c <= '9'))
            {
                /* for numbers look for possible numeric continuations */
 
                bool        numend = false;
 
                for (int i = 0; i < lex->input_length && !numend; i++)
                {
                    char        cc = lex->input[i];
 
                    switch (cc)
                    {
                        case '+':
                        case '-':
                        case 'e':
                        case 'E':
                        case '0':
                        case '1':
                        case '2':
                        case '3':
                        case '4':
                        case '5':
                        case '6':
                        case '7':
                        case '8':
                        case '9':
                            {
                                appendStringInfoCharMacro(ptok, cc);
                                added++;
                            }
                            break;
                        default:
                            numend = true;
                    }
                }
            }
 
            /*
             * Add any remaining alphanumeric chars. This takes care of the
             * {null, false, true} literals as well as any trailing
             * alphanumeric junk on non-string tokens.
             */
            for (int i = added; i < lex->input_length; i++)
            {
                char        cc = lex->input[i];
 
                if (JSON_ALPHANUMERIC_CHAR(cc))
                {
                    appendStringInfoCharMacro(ptok, cc);
                    added++;
                }
                else
                {
                    tok_done = true;
                    break;
                }
            }
            if (added == lex->input_length &&
                lex->inc_state->is_last_chunk)
            {
                tok_done = true;
            }
        }
 
        if (!tok_done)
        {
            /* We should have consumed the whole chunk in this case. */
            Assert(added == lex->input_length);
 
            if (!lex->inc_state->is_last_chunk)
                return JSON_INCOMPLETE;
 
            /* json_errdetail() needs access to the accumulated token. */
            lex->token_start = ptok->data;
            lex->token_terminator = ptok->data + ptok->len;
            return JSON_INVALID_TOKEN;
        }
 
        /*
         * Everything up to lex->input[added] has been added to the partial
         * token, so move the input past it.
         */
        lex->input += added;
        lex->input_length -= added;
 
        dummy_lex.input = dummy_lex.token_terminator =
            dummy_lex.line_start = ptok->data;
        dummy_lex.line_number = lex->line_number;
        dummy_lex.input_length = ptok->len;
        dummy_lex.input_encoding = lex->input_encoding;
        dummy_lex.incremental = false;
        dummy_lex.strval = lex->strval;
 
        partial_result = json_lex(&dummy_lex);
 
        /*
         * We either have a complete token or an error. In either case we need
         * to point to the partial token data for the semantic or error
         * routines. If it's not an error we'll readjust on the next call to
         * json_lex.
         */
        lex->token_type = dummy_lex.token_type;
        lex->line_number = dummy_lex.line_number;
 
        /*
         * We know the prev_token_terminator must be back in some previous
         * piece of input, so we just make it NULL.
         */
        lex->prev_token_terminator = NULL;
 
        /*
         * Normally token_start would be ptok->data, but it could be later,
         * see json_lex_string's handling of invalid escapes.
         */
        lex->token_start = dummy_lex.token_start;
        lex->token_terminator = dummy_lex.token_terminator;
        if (partial_result == JSON_SUCCESS)
        {
            /* make sure we've used all the input */
            if (lex->token_terminator - lex->token_start != ptok->len)
            {
                Assert(false);
                return JSON_INVALID_TOKEN;
            }
 
            lex->inc_state->partial_completed = true;
        }
        return partial_result;
        /* end of partial token processing */
    }
 
    /* Skip leading whitespace. */
    while (s < end && (*s == ' ' || *s == '\t' || *s == '\n' || *s == '\r'))
    {
        if (*s++ == '\n')
        {
            ++lex->line_number;
            lex->line_start = s;
        }
    }
    lex->token_start = s;
 
    /* Determine token type. */
    if (s >= end)
    {
        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 */
                result = json_lex_string(lex);
                if (result != JSON_SUCCESS)
                    return result;
                lex->token_type = JSON_TOKEN_STRING;
                break;
            case '-':
                /* Negative number. */
                result = json_lex_number(lex, s + 1, NULL, NULL);
                if (result != JSON_SUCCESS)
                    return result;
                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. */
                result = json_lex_number(lex, s, NULL, NULL);
                if (result != JSON_SUCCESS)
                    return result;
                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; p < end && JSON_ALPHANUMERIC_CHAR(*p); 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;
                        return JSON_INVALID_TOKEN;
                    }
 
                    if (lex->incremental && !lex->inc_state->is_last_chunk &&
                        p == lex->input + lex->input_length)
                    {
                        appendBinaryStringInfo(
                                               &(lex->inc_state->partial_token), s, end - s);
                        return JSON_INCOMPLETE;
                    }
 
                    /*
                     * 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
                            return JSON_INVALID_TOKEN;
                    }
                    else if (p - s == 5 && memcmp(s, "false", 5) == 0)
                        lex->token_type = JSON_TOKEN_FALSE;
                    else
                        return JSON_INVALID_TOKEN;
                }
        }                       /* end of switch */
    }
 
    if (lex->incremental && lex->token_type == JSON_TOKEN_END && !lex->inc_state->is_last_chunk)
        return JSON_INCOMPLETE;
    else
        return JSON_SUCCESS;

References appendBinaryStringInfo(), appendStringInfoCharMacro, Assert, StringInfoData::data, i, JsonLexContext::inc_state, JsonLexContext::incremental, JsonLexContext::input, JsonLexContext::input_encoding, JsonLexContext::input_length, JsonIncrementalState::is_last_chunk, JSON_ALPHANUMERIC_CHAR, JSON_INCOMPLETE, JSON_INVALID_TOKEN, json_lex_number(), json_lex_string(), JSON_SUCCESS, JSON_TOKEN_ARRAY_END, JSON_TOKEN_ARRAY_START, JSON_TOKEN_COLON, JSON_TOKEN_COMMA, JSON_TOKEN_END, JSON_TOKEN_FALSE, JSON_TOKEN_NULL, JSON_TOKEN_NUMBER, JSON_TOKEN_OBJECT_END, JSON_TOKEN_OBJECT_START, JSON_TOKEN_STRING, JSON_TOKEN_TRUE, StringInfoData::len, JsonLexContext::line_number, JsonLexContext::line_start, JsonIncrementalState::partial_completed, JsonIncrementalState::partial_token, JsonLexContext::prev_token_terminator, resetStringInfo(), JsonLexContext::strval, JsonLexContext::token_start, JsonLexContext::token_terminator, and JsonLexContext::token_type.

Referenced by json_count_array_elements(), json_get_first_token(), json_typeof(), lex_expect(), parse_array(), parse_object(), parse_object_field(), parse_scalar(), pg_parse_json(), and pg_parse_json_incremental().

makeJsonLexContextCstringLen()

JsonLexContext* makeJsonLexContextCstringLen	(	JsonLexContext *	lex,
		char *	json,
		int	len,
		int	encoding,
		bool	need_escapes
	)

Definition at line 326 of file jsonapi.c.

{
    if (lex == NULL)
    {
        lex = palloc0(sizeof(JsonLexContext));
        lex->flags |= JSONLEX_FREE_STRUCT;
    }
    else
        memset(lex, 0, sizeof(JsonLexContext));
 
    lex->errormsg = NULL;
    lex->input = lex->token_terminator = lex->line_start = json;
    lex->line_number = 1;
    lex->input_length = len;
    lex->input_encoding = encoding;
    if (need_escapes)
    {
        lex->strval = makeStringInfo();
        lex->flags |= JSONLEX_FREE_STRVAL;
    }
 
    return lex;
}

References encoding, JsonLexContext::errormsg, JsonLexContext::flags, JsonLexContext::input, JsonLexContext::input_encoding, JsonLexContext::input_length, JSONLEX_FREE_STRUCT, JSONLEX_FREE_STRVAL, len, JsonLexContext::line_number, JsonLexContext::line_start, makeStringInfo(), palloc0(), JsonLexContext::strval, and JsonLexContext::token_terminator.

Referenced by get_json_object_as_hash(), json_parse_manifest(), json_recv(), jsonb_from_cstring(), main(), makeJsonLexContext(), and populate_array_json().

makeJsonLexContextIncremental()

JsonLexContext* makeJsonLexContextIncremental	(	JsonLexContext *	lex,
		int	encoding,
		bool	need_escapes
	)

Definition at line 366 of file jsonapi.c.

{
    if (lex == NULL)
    {
        lex = palloc0(sizeof(JsonLexContext));
        lex->flags |= JSONLEX_FREE_STRUCT;
    }
    else
        memset(lex, 0, sizeof(JsonLexContext));
 
    lex->line_number = 1;
    lex->input_encoding = encoding;
    lex->incremental = true;
    lex->inc_state = palloc0(sizeof(JsonIncrementalState));
    initStringInfo(&(lex->inc_state->partial_token));
    lex->pstack = palloc(sizeof(JsonParserStack));
    lex->pstack->stack_size = JS_STACK_CHUNK_SIZE;
    lex->pstack->prediction = palloc(JS_STACK_CHUNK_SIZE * JS_MAX_PROD_LEN);
    lex->pstack->pred_index = 0;
    lex->pstack->fnames = palloc(JS_STACK_CHUNK_SIZE * sizeof(char *));
    lex->pstack->fnull = palloc(JS_STACK_CHUNK_SIZE * sizeof(bool));
    if (need_escapes)
    {
        lex->strval = makeStringInfo();
        lex->flags |= JSONLEX_FREE_STRVAL;
    }
    return lex;

References encoding, JsonLexContext::flags, JsonParserStack::fnames, JsonParserStack::fnull, JsonLexContext::inc_state, JsonLexContext::incremental, initStringInfo(), JsonLexContext::input_encoding, JS_MAX_PROD_LEN, JS_STACK_CHUNK_SIZE, JSONLEX_FREE_STRUCT, JSONLEX_FREE_STRVAL, JsonLexContext::line_number, makeStringInfo(), palloc(), palloc0(), JsonIncrementalState::partial_token, JsonParserStack::pred_index, JsonParserStack::prediction, JsonLexContext::pstack, JsonParserStack::stack_size, and JsonLexContext::strval.

Referenced by json_parse_manifest_incremental_init(), and main().

pg_parse_json()

JsonParseErrorType pg_parse_json	(	JsonLexContext *	lex,
		JsonSemAction *	sem
	)

Definition at line 521 of file jsonapi.c.

{
#ifdef FORCE_JSON_PSTACK
 
    lex->incremental = true;
    lex->inc_state = palloc0(sizeof(JsonIncrementalState));
 
    /*
     * We don't need partial token processing, there is only one chunk. But we
     * still need to init the partial token string so that freeJsonLexContext
     * works.
     */
    initStringInfo(&(lex->inc_state->partial_token));
    lex->pstack = palloc(sizeof(JsonParserStack));
    lex->pstack->stack_size = JS_STACK_CHUNK_SIZE;
    lex->pstack->prediction = palloc(JS_STACK_CHUNK_SIZE * JS_MAX_PROD_LEN);
    lex->pstack->pred_index = 0;
    lex->pstack->fnames = palloc(JS_STACK_CHUNK_SIZE * sizeof(char *));
    lex->pstack->fnull = palloc(JS_STACK_CHUNK_SIZE * sizeof(bool));
 
    return pg_parse_json_incremental(lex, sem, lex->input, lex->input_length, true);
 
#else
 
    JsonTokenType tok;
    JsonParseErrorType result;
 
    if (lex->incremental)
        return JSON_INVALID_LEXER_TYPE;
 
    /* get the initial token */
    result = json_lex(lex);
    if (result != JSON_SUCCESS)
        return result;
 
    tok = lex_peek(lex);
 
    /* parse by recursive descent */
    switch (tok)
    {
        case JSON_TOKEN_OBJECT_START:
            result = parse_object(lex, sem);
            break;
        case JSON_TOKEN_ARRAY_START:
            result = parse_array(lex, sem);
            break;
        default:
            result = parse_scalar(lex, sem);    /* json can be a bare scalar */
    }
 
    if (result == JSON_SUCCESS)
        result = lex_expect(JSON_PARSE_END, lex, JSON_TOKEN_END);
 
    return result;
#endif

References JsonParserStack::fnames, JsonParserStack::fnull, JsonLexContext::inc_state, JsonLexContext::incremental, initStringInfo(), JsonLexContext::input, JsonLexContext::input_length, JS_MAX_PROD_LEN, JS_STACK_CHUNK_SIZE, JSON_INVALID_LEXER_TYPE, json_lex(), JSON_PARSE_END, JSON_SUCCESS, JSON_TOKEN_ARRAY_START, JSON_TOKEN_END, JSON_TOKEN_OBJECT_START, lex_expect(), lex_peek(), palloc(), palloc0(), parse_array(), parse_object(), parse_scalar(), JsonIncrementalState::partial_token, pg_parse_json_incremental(), JsonParserStack::pred_index, JsonParserStack::prediction, JsonLexContext::pstack, sem, and JsonParserStack::stack_size.

Referenced by json_parse_manifest(), json_validate(), main(), and pg_parse_json_or_errsave().

pg_parse_json_incremental()

JsonParseErrorType pg_parse_json_incremental	(	JsonLexContext *	lex,
		JsonSemAction *	sem,
		char *	json,
		int	len,
		bool	is_last
	)

Definition at line 649 of file jsonapi.c.

{
    JsonTokenType tok;
    JsonParseErrorType result;
    JsonParseContext ctx = JSON_PARSE_VALUE;
    JsonParserStack *pstack = lex->pstack;
 
 
    if (!lex->incremental)
        return JSON_INVALID_LEXER_TYPE;
 
    lex->input = lex->token_terminator = lex->line_start = json;
    lex->input_length = len;
    lex->inc_state->is_last_chunk = is_last;
 
    /* get the initial token */
    result = json_lex(lex);
    if (result != JSON_SUCCESS)
        return result;
 
    tok = lex_peek(lex);
 
    /* use prediction stack for incremental parsing */
 
    if (!have_prediction(pstack))
    {
        td_entry    goal = TD_ENTRY(JSON_PROD_GOAL);
 
        push_prediction(pstack, goal);
    }
 
    while (have_prediction(pstack))
    {
        char        top = pop_prediction(pstack);
        td_entry    entry;
 
        /*
         * these first two branches are the guts of the Table Driven method
         */
        if (top == tok)
        {
            /*
             * tok can only be a terminal symbol, so top must be too. the
             * token matches the top of the stack, so get the next token.
             */
            if (tok < JSON_TOKEN_END)
            {
                result = json_lex(lex);
                if (result != JSON_SUCCESS)
                    return result;
                tok = lex_peek(lex);
            }
        }
        else if (IS_NT(top) && (entry = td_parser_table[OFS(top)][tok]).prod != NULL)
        {
            /*
             * the token is in the director set for a production of the
             * non-terminal at the top of the stack, so push the reversed RHS
             * of the production onto the stack.
             */
            push_prediction(pstack, entry);
        }
        else if (IS_SEM(top))
        {
            /*
             * top is a semantic action marker, so take action accordingly.
             * It's important to have these markers in the prediction stack
             * before any token they might need so we don't advance the token
             * prematurely. Note in a couple of cases we need to do something
             * both before and after the token.
             */
            switch (top)
            {
                case JSON_SEM_OSTART:
                    {
                        json_struct_action ostart = sem->object_start;
 
                        if (lex->lex_level >= JSON_TD_MAX_STACK)
                            return JSON_NESTING_TOO_DEEP;
 
                        if (ostart != NULL)
                        {
                            result = (*ostart) (sem->semstate);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                        inc_lex_level(lex);
                    }
                    break;
                case JSON_SEM_OEND:
                    {
                        json_struct_action oend = sem->object_end;
 
                        dec_lex_level(lex);
                        if (oend != NULL)
                        {
                            result = (*oend) (sem->semstate);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                    }
                    break;
                case JSON_SEM_ASTART:
                    {
                        json_struct_action astart = sem->array_start;
 
                        if (lex->lex_level >= JSON_TD_MAX_STACK)
                            return JSON_NESTING_TOO_DEEP;
 
                        if (astart != NULL)
                        {
                            result = (*astart) (sem->semstate);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                        inc_lex_level(lex);
                    }
                    break;
                case JSON_SEM_AEND:
                    {
                        json_struct_action aend = sem->array_end;
 
                        dec_lex_level(lex);
                        if (aend != NULL)
                        {
                            result = (*aend) (sem->semstate);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                    }
                    break;
                case JSON_SEM_OFIELD_INIT:
                    {
                        /*
                         * all we do here is save out the field name. We have
                         * to wait to get past the ':' to see if the next
                         * value is null so we can call the semantic routine
                         */
                        char       *fname = NULL;
                        json_ofield_action ostart = sem->object_field_start;
                        json_ofield_action oend = sem->object_field_end;
 
                        if ((ostart != NULL || oend != NULL) && lex->strval != NULL)
                        {
                            fname = pstrdup(lex->strval->data);
                        }
                        set_fname(lex, fname);
                    }
                    break;
                case JSON_SEM_OFIELD_START:
                    {
                        /*
                         * the current token should be the first token of the
                         * value
                         */
                        bool        isnull = tok == JSON_TOKEN_NULL;
                        json_ofield_action ostart = sem->object_field_start;
 
                        set_fnull(lex, isnull);
 
                        if (ostart != NULL)
                        {
                            char       *fname = get_fname(lex);
 
                            result = (*ostart) (sem->semstate, fname, isnull);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                    }
                    break;
                case JSON_SEM_OFIELD_END:
                    {
                        json_ofield_action oend = sem->object_field_end;
 
                        if (oend != NULL)
                        {
                            char       *fname = get_fname(lex);
                            bool        isnull = get_fnull(lex);
 
                            result = (*oend) (sem->semstate, fname, isnull);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                    }
                    break;
                case JSON_SEM_AELEM_START:
                    {
                        json_aelem_action astart = sem->array_element_start;
                        bool        isnull = tok == JSON_TOKEN_NULL;
 
                        set_fnull(lex, isnull);
 
                        if (astart != NULL)
                        {
                            result = (*astart) (sem->semstate, isnull);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                    }
                    break;
                case JSON_SEM_AELEM_END:
                    {
                        json_aelem_action aend = sem->array_element_end;
 
                        if (aend != NULL)
                        {
                            bool        isnull = get_fnull(lex);
 
                            result = (*aend) (sem->semstate, isnull);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                    }
                    break;
                case JSON_SEM_SCALAR_INIT:
                    {
                        json_scalar_action sfunc = sem->scalar;
 
                        pstack->scalar_val = NULL;
 
                        if (sfunc != NULL)
                        {
                            /*
                             * extract the de-escaped string value, or the raw
                             * lexeme
                             */
                            /*
                             * XXX copied from RD parser but looks like a
                             * buglet
                             */
                            if (tok == JSON_TOKEN_STRING)
                            {
                                if (lex->strval != NULL)
                                    pstack->scalar_val = pstrdup(lex->strval->data);
                            }
                            else
                            {
                                int         tlen = (lex->token_terminator - lex->token_start);
 
                                pstack->scalar_val = palloc(tlen + 1);
                                memcpy(pstack->scalar_val, lex->token_start, tlen);
                                pstack->scalar_val[tlen] = '\0';
                            }
                            pstack->scalar_tok = tok;
                        }
                    }
                    break;
                case JSON_SEM_SCALAR_CALL:
                    {
                        /*
                         * We'd like to be able to get rid of this business of
                         * two bits of scalar action, but we can't. It breaks
                         * certain semantic actions which expect that when
                         * called the lexer has consumed the item. See for
                         * example get_scalar() in jsonfuncs.c.
                         */
                        json_scalar_action sfunc = sem->scalar;
 
                        if (sfunc != NULL)
                        {
                            result = (*sfunc) (sem->semstate, pstack->scalar_val, pstack->scalar_tok);
                            if (result != JSON_SUCCESS)
                                return result;
                        }
                    }
                    break;
                default:
                    /* should not happen */
                    break;
            }
        }
        else
        {
            /*
             * The token didn't match the stack top if it's a terminal nor a
             * production for the stack top if it's a non-terminal.
             *
             * Various cases here are Asserted to be not possible, as the
             * token would not appear at the top of the prediction stack
             * unless the lookahead matched.
             */
            switch (top)
            {
                case JSON_TOKEN_STRING:
                    if (next_prediction(pstack) == JSON_TOKEN_COLON)
                        ctx = JSON_PARSE_STRING;
                    else
                    {
                        Assert(false);
                        ctx = JSON_PARSE_VALUE;
                    }
                    break;
                case JSON_TOKEN_NUMBER:
                case JSON_TOKEN_TRUE:
                case JSON_TOKEN_FALSE:
                case JSON_TOKEN_NULL:
                case JSON_TOKEN_ARRAY_START:
                case JSON_TOKEN_OBJECT_START:
                    Assert(false);
                    ctx = JSON_PARSE_VALUE;
                    break;
                case JSON_TOKEN_ARRAY_END:
                    Assert(false);
                    ctx = JSON_PARSE_ARRAY_NEXT;
                    break;
                case JSON_TOKEN_OBJECT_END:
                    Assert(false);
                    ctx = JSON_PARSE_OBJECT_NEXT;
                    break;
                case JSON_TOKEN_COMMA:
                    Assert(false);
                    if (next_prediction(pstack) == JSON_TOKEN_STRING)
                        ctx = JSON_PARSE_OBJECT_NEXT;
                    else
                        ctx = JSON_PARSE_ARRAY_NEXT;
                    break;
                case JSON_TOKEN_COLON:
                    ctx = JSON_PARSE_OBJECT_LABEL;
                    break;
                case JSON_TOKEN_END:
                    ctx = JSON_PARSE_END;
                    break;
                case JSON_NT_MORE_ARRAY_ELEMENTS:
                    ctx = JSON_PARSE_ARRAY_NEXT;
                    break;
                case JSON_NT_ARRAY_ELEMENTS:
                    ctx = JSON_PARSE_ARRAY_START;
                    break;
                case JSON_NT_MORE_KEY_PAIRS:
                    ctx = JSON_PARSE_OBJECT_NEXT;
                    break;
                case JSON_NT_KEY_PAIRS:
                    ctx = JSON_PARSE_OBJECT_START;
                    break;
                default:
                    ctx = JSON_PARSE_VALUE;
            }
            return report_parse_error(ctx, lex);
        }
    }
 
    return JSON_SUCCESS;

References JsonSemAction::array_element_end, JsonSemAction::array_element_start, JsonSemAction::array_end, JsonSemAction::array_start, Assert, StringInfoData::data, dec_lex_level(), get_fname(), get_fnull(), have_prediction(), inc_lex_level(), JsonLexContext::inc_state, JsonLexContext::incremental, JsonLexContext::input, JsonLexContext::input_length, JsonIncrementalState::is_last_chunk, IS_NT, IS_SEM, JSON_INVALID_LEXER_TYPE, json_lex(), JSON_NESTING_TOO_DEEP, JSON_NT_ARRAY_ELEMENTS, JSON_NT_KEY_PAIRS, JSON_NT_MORE_ARRAY_ELEMENTS, JSON_NT_MORE_KEY_PAIRS, JSON_PARSE_ARRAY_NEXT, JSON_PARSE_ARRAY_START, JSON_PARSE_END, JSON_PARSE_OBJECT_LABEL, JSON_PARSE_OBJECT_NEXT, JSON_PARSE_OBJECT_START, JSON_PARSE_STRING, JSON_PARSE_VALUE, JSON_PROD_GOAL, JSON_SEM_AELEM_END, JSON_SEM_AELEM_START, JSON_SEM_AEND, JSON_SEM_ASTART, JSON_SEM_OEND, JSON_SEM_OFIELD_END, JSON_SEM_OFIELD_INIT, JSON_SEM_OFIELD_START, JSON_SEM_OSTART, JSON_SEM_SCALAR_CALL, JSON_SEM_SCALAR_INIT, JSON_SUCCESS, JSON_TD_MAX_STACK, JSON_TOKEN_ARRAY_END, JSON_TOKEN_ARRAY_START, JSON_TOKEN_COLON, JSON_TOKEN_COMMA, JSON_TOKEN_END, JSON_TOKEN_FALSE, JSON_TOKEN_NULL, JSON_TOKEN_NUMBER, JSON_TOKEN_OBJECT_END, JSON_TOKEN_OBJECT_START, JSON_TOKEN_STRING, JSON_TOKEN_TRUE, len, JsonLexContext::lex_level, lex_peek(), JsonLexContext::line_start, next_prediction(), JsonSemAction::object_end, JsonSemAction::object_field_end, JsonSemAction::object_field_start, JsonSemAction::object_start, OFS, palloc(), pop_prediction(), JsonLexContext::pstack, pstrdup(), push_prediction(), report_parse_error(), JsonSemAction::scalar, JsonParserStack::scalar_tok, JsonParserStack::scalar_val, sem, JsonSemAction::semstate, set_fname(), set_fnull(), JsonLexContext::strval, TD_ENTRY, td_parser_table, JsonLexContext::token_start, and JsonLexContext::token_terminator.

Referenced by json_parse_manifest_incremental_chunk(), main(), and pg_parse_json().

Variable Documentation

nullSemAction

PGDLLIMPORT JsonSemAction nullSemAction

extern

Definition at line 224 of file jsonapi.c.

Referenced by json_count_array_elements(), json_in(), json_recv(), json_validate(), and main().