* json.c
* JSON data type support.
*
- * Portions Copyright (c) 1996-2013, PostgreSQL Global Development Group
+ * Portions Copyright (c) 1996-2019, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
#include "access/htup_details.h"
#include "access/transam.h"
-#include "catalog/pg_cast.h"
#include "catalog/pg_type.h"
#include "executor/spi.h"
+#include "funcapi.h"
#include "lib/stringinfo.h"
#include "libpq/pqformat.h"
#include "mb/pg_wchar.h"
+#include "miscadmin.h"
#include "parser/parse_coerce.h"
#include "utils/array.h"
#include "utils/builtins.h"
+#include "utils/date.h"
+#include "utils/datetime.h"
#include "utils/lsyscache.h"
#include "utils/json.h"
#include "utils/jsonapi.h"
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;
+} JsonParseContext;
+
+typedef enum /* type categories for datum_to_json */
+{
+ JSONTYPE_NULL, /* null, so we didn't bother to identify */
+ JSONTYPE_BOOL, /* boolean (built-in types only) */
+ JSONTYPE_NUMERIC, /* numeric (ditto) */
+ JSONTYPE_DATE, /* we use special formatting for datetimes */
+ JSONTYPE_TIMESTAMP,
+ JSONTYPE_TIMESTAMPTZ,
+ JSONTYPE_JSON, /* JSON itself (and JSONB) */
+ JSONTYPE_ARRAY, /* array */
+ JSONTYPE_COMPOSITE, /* composite */
+ JSONTYPE_CAST, /* something with an explicit cast to JSON */
+ JSONTYPE_OTHER /* all else */
+} JsonTypeCategory;
+
+typedef struct JsonAggState
+{
+ StringInfo str;
+ JsonTypeCategory key_category;
+ Oid key_output_func;
+ JsonTypeCategory val_category;
+ Oid val_output_func;
+} JsonAggState;
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 inline void json_lex_number(JsonLexContext *lex, char *s,
+ bool *num_err, int *total_len);
+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) pg_attribute_noreturn();
+static void report_invalid_token(JsonLexContext *lex) pg_attribute_noreturn();
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);
+ 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);
+ Datum *vals, bool *nulls, int *valcount,
+ JsonTypeCategory tcategory, Oid outfuncoid,
+ bool use_line_feeds);
static void array_to_json_internal(Datum array, StringInfo result,
- bool use_line_feeds);
+ bool use_line_feeds);
+static void json_categorize_type(Oid typoid,
+ JsonTypeCategory *tcategory,
+ Oid *outfuncoid);
+static void datum_to_json(Datum val, bool is_null, StringInfo result,
+ JsonTypeCategory tcategory, Oid outfuncoid,
+ bool key_scalar);
+static void add_json(Datum val, bool is_null, StringInfo result,
+ Oid val_type, bool key_scalar);
+static text *catenate_stringinfo_string(StringInfo buffer, const char *addon);
/* the null action object used for pure validation */
-static jsonSemAction nullSemAction =
+static JsonSemAction nullSemAction =
{
NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL
};
-static JsonSemAction NullSemAction = &nullSemAction;
/* Recursive Descent parser support routines */
lex_expect(JsonParseContext ctx, JsonLexContext *lex, JsonTokenType token)
{
if (!lex_accept(lex, token, NULL))
- report_parse_error(ctx, lex);;
+ 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) == '_' || \
IS_HIGHBIT_SET(c))
+/*
+ * Utility function to check if a string is a valid JSON number.
+ *
+ * str is of length len, and need not be null-terminated.
+ */
+bool
+IsValidJsonNumber(const char *str, int len)
+{
+ bool numeric_error;
+ int total_len;
+ JsonLexContext dummy_lex;
+
+ if (len <= 0)
+ return false;
+
+ /*
+ * 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;
+ }
+
+ json_lex_number(&dummy_lex, dummy_lex.input, &numeric_error, &total_len);
+
+ return (!numeric_error) && (total_len == dummy_lex.input_length);
+}
+
/*
* Input.
*/
/* validate it */
lex = makeJsonLexContext(result, false);
- pg_parse_json(lex, NullSemAction);
+ pg_parse_json(lex, &nullSemAction);
/* Internal representation is the same as text, for now */
PG_RETURN_TEXT_P(result);
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);
+ lex = makeJsonLexContextCstringLen(str, nbytes, false);
+ pg_parse_json(lex, &nullSemAction);
- PG_RETURN_TEXT_P(result);
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(str, nbytes));
}
/*
*
* Without is better as it makes the processing faster, so only make one
* if really required.
+ *
+ * If you already have the json as a text* value, use the first of these
+ * functions, otherwise use makeJsonLexContextCstringLen().
*/
JsonLexContext *
makeJsonLexContext(text *json, bool need_escapes)
+{
+ return makeJsonLexContextCstringLen(VARDATA_ANY(json),
+ VARSIZE_ANY_EXHDR(json),
+ need_escapes);
+}
+
+JsonLexContext *
+makeJsonLexContextCstringLen(char *json, int len, bool need_escapes)
{
JsonLexContext *lex = palloc0(sizeof(JsonLexContext));
- lex->input = lex->token_terminator = lex->line_start = VARDATA(json);
+ lex->input = lex->token_terminator = lex->line_start = json;
lex->line_number = 1;
- lex->input_length = VARSIZE(json) - VARHDRSZ;
+ lex->input_length = len;
if (need_escapes)
lex->strval = makeStringInfo();
return lex;
* 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
+ * makeJsonLexContext(). sem is a structure 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)
+pg_parse_json(JsonLexContext *lex, JsonSemAction *sem)
{
JsonTokenType tok;
parse_array(lex, sem);
break;
default:
- parse_scalar(lex, sem); /* json can be a bare scalar */
+ parse_scalar(lex, sem); /* json can be a bare scalar */
}
lex_expect(JSON_PARSE_END, lex, JSON_TOKEN_END);
}
+/*
+ * json_count_array_elements
+ *
+ * Returns number of array elements in lex context at start of array token
+ * until end of array token at same nesting level.
+ *
+ * Designed to be called from array_start routines.
+ */
+int
+json_count_array_elements(JsonLexContext *lex)
+{
+ JsonLexContext copylex;
+ int count;
+
+ /*
+ * 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(©lex, lex, sizeof(JsonLexContext));
+ copylex.strval = NULL; /* not interested in values here */
+ copylex.lex_level++;
+
+ count = 0;
+ lex_expect(JSON_PARSE_ARRAY_START, ©lex, JSON_TOKEN_ARRAY_START);
+ if (lex_peek(©lex) != JSON_TOKEN_ARRAY_END)
+ {
+ do
+ {
+ count++;
+ parse_array_element(©lex, &nullSemAction);
+ }
+ while (lex_accept(©lex, JSON_TOKEN_COMMA, NULL));
+ }
+ lex_expect(JSON_PARSE_ARRAY_NEXT, ©lex, JSON_TOKEN_ARRAY_END);
+
+ return count;
+}
+
/*
* Recursive Descent parse routines. There is one for each structural
* element in a json document:
* - object field
*/
static inline void
-parse_scalar(JsonLexContext *lex, JsonSemAction sem)
+parse_scalar(JsonLexContext *lex, JsonSemAction *sem)
{
char *val = NULL;
json_scalar_action sfunc = sem->scalar;
}
static void
-parse_object_field(JsonLexContext *lex, JsonSemAction sem)
+parse_object_field(JsonLexContext *lex, JsonSemAction *sem)
{
/*
- * an object field is "fieldname" : value where value can be a scalar,
- * object or array
+ * An object field is "fieldname" : value where value can be a scalar,
+ * object or array. Note: in user-facing docs and error messages, we
+ * generally call a field name a "key".
*/
char *fname = NULL; /* keep compiler quiet */
if (oend != NULL)
(*oend) (sem->semstate, fname, isnull);
-
- if (fname != NULL)
- pfree(fname);
}
static void
-parse_object(JsonLexContext *lex, JsonSemAction sem)
+parse_object(JsonLexContext *lex, JsonSemAction *sem)
{
/*
* an object is a possibly empty sequence of object fields, separated by
- * commas and surrounde by curly braces.
+ * commas and surrounded by curly braces.
*/
json_struct_action ostart = sem->object_start;
json_struct_action oend = sem->object_end;
JsonTokenType tok;
+ check_stack_depth();
+
if (ostart != NULL)
(*ostart) (sem->semstate);
/*
- * Data inside an object at at a higher nesting level than the object
+ * Data inside an object is 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 */
+ /* we know this will succeed, just clearing the token */
lex_expect(JSON_PARSE_OBJECT_START, lex, JSON_TOKEN_OBJECT_START);
tok = lex_peek(lex);
}
static void
-parse_array_element(JsonLexContext *lex, JsonSemAction sem)
+parse_array_element(JsonLexContext *lex, JsonSemAction *sem)
{
json_aelem_action astart = sem->array_element_start;
json_aelem_action aend = sem->array_element_end;
}
static void
-parse_array(JsonLexContext *lex, JsonSemAction sem)
+parse_array(JsonLexContext *lex, JsonSemAction *sem)
{
/*
* an array is a possibly empty sequence of array elements, separated by
json_struct_action astart = sem->array_start;
json_struct_action aend = sem->array_end;
+ check_stack_depth();
+
if (astart != NULL)
(*astart) (sem->semstate);
/*
- * Data inside an array at at a higher nesting level than the array
+ * Data inside an array is 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.
break;
case '-':
/* Negative number. */
- json_lex_number(lex, s + 1);
+ json_lex_number(lex, s + 1, NULL, NULL);
lex->token_type = JSON_TOKEN_NUMBER;
break;
case '0':
case '8':
case '9':
/* Positive number. */
- json_lex_number(lex, s);
+ json_lex_number(lex, s, NULL, NULL);
lex->token_type = JSON_TOKEN_NUMBER;
break;
default:
/*
* We're not dealing with a string, number, legal
- * punctuation mark, or end of string. The only 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
+ * 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++)
+ for (p = s; p - s < lex->input_length - len && JSON_ALPHANUMERIC_CHAR(*p); p++)
/* skip */ ;
/*
{
char *s;
int len;
+ int hi_surrogate = -1;
if (lex->strval != NULL)
resetStringInfo(lex->strval);
+ Assert(lex->input_length > 0);
+ s = lex->token_start;
len = lex->token_start - lex->input;
- len++;
- for (s = lex->token_start + 1; *s != '"'; s++, len++)
+ for (;;)
{
+ s++;
+ len++;
/* Premature end of the string. */
if (len >= lex->input_length)
{
lex->token_terminator = s;
report_invalid_token(lex);
}
+ else if (*s == '"')
+ break;
else if ((unsigned char) *s < 32)
{
/* Per RFC4627, these characters MUST be escaped. */
lex->token_terminator = s;
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Character with value 0x%02x must be escaped.",
(unsigned char) *s),
report_json_context(lex)));
lex->token_terminator = s + pg_mblen(s);
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s",
+ "json"),
errdetail("\"\\u\" must be followed by four hexadecimal digits."),
report_json_context(lex)));
}
{
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);
+ if (ch >= 0xd800 && ch <= 0xdbff)
+ {
+ if (hi_surrogate != -1)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+ errmsg("invalid input syntax for type %s",
+ "json"),
+ errdetail("Unicode high surrogate must not follow a high surrogate."),
+ report_json_context(lex)));
+ hi_surrogate = (ch & 0x3ff) << 10;
+ continue;
+ }
+ else if (ch >= 0xdc00 && ch <= 0xdfff)
+ {
+ if (hi_surrogate == -1)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+ errmsg("invalid input syntax for type %s", "json"),
+ errdetail("Unicode low surrogate must follow a high surrogate."),
+ report_json_context(lex)));
+ ch = 0x10000 + hi_surrogate + (ch & 0x3ff);
+ hi_surrogate = -1;
+ }
+
+ if (hi_surrogate != -1)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+ errmsg("invalid input syntax for type %s", "json"),
+ errdetail("Unicode low surrogate must follow a high surrogate."),
+ report_json_context(lex)));
+
+ /*
+ * For UTF8, replace the escape sequence by the actual
+ * utf8 character in lex->strval. Do this also for other
+ * encodings if the escape designates an ASCII character,
+ * otherwise raise an error.
+ */
+
+ if (ch == 0)
+ {
+ /* We can't allow this, since our TEXT type doesn't */
+ ereport(ERROR,
+ (errcode(ERRCODE_UNTRANSLATABLE_CHARACTER),
+ errmsg("unsupported Unicode escape sequence"),
+ errdetail("\\u0000 cannot be converted to text."),
+ report_json_context(lex)));
+ }
+ else if (GetDatabaseEncoding() == PG_UTF8)
+ {
+ unicode_to_utf8(ch, (unsigned char *) utf8str);
+ utf8len = pg_utf_mblen((unsigned char *) utf8str);
+ appendBinaryStringInfo(lex->strval, utf8str, utf8len);
+ }
+ else if (ch <= 0x007f)
+ {
+ /*
+ * This is the only way to designate things like a
+ * form feed character in JSON, so it's useful in all
+ * encodings.
+ */
+ appendStringInfoChar(lex->strval, (char) ch);
+ }
+ else
+ {
+ ereport(ERROR,
+ (errcode(ERRCODE_UNTRANSLATABLE_CHARACTER),
+ errmsg("unsupported Unicode escape sequence"),
+ errdetail("Unicode escape values cannot be used for code point values above 007F when the server encoding is not UTF8."),
+ report_json_context(lex)));
+ }
}
}
else if (lex->strval != NULL)
{
+ if (hi_surrogate != -1)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+ errmsg("invalid input syntax for type %s",
+ "json"),
+ errdetail("Unicode low surrogate must follow a high surrogate."),
+ report_json_context(lex)));
+
switch (*s)
{
case '"':
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)),
+ errmsg("invalid input syntax for type %s",
+ "json"),
+ errdetail("Escape sequence \"\\%s\" is invalid.",
+ extract_mb_char(s)),
report_json_context(lex)));
}
}
lex->token_terminator = s + pg_mblen(s);
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Escape sequence \"\\%s\" is invalid.",
extract_mb_char(s)),
report_json_context(lex)));
}
else if (lex->strval != NULL)
{
+ if (hi_surrogate != -1)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+ errmsg("invalid input syntax for type %s", "json"),
+ errdetail("Unicode low surrogate must follow a high surrogate."),
+ report_json_context(lex)));
+
appendStringInfoChar(lex->strval, *s);
}
}
+ if (hi_surrogate != -1)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+ errmsg("invalid input syntax for type %s", "json"),
+ errdetail("Unicode low surrogate must follow a high surrogate."),
+ report_json_context(lex)));
+
/* 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:
* 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
+ * 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:
+ * 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
+ * of part 2 - i.e. the character after any optional minus sign, or the
* first character of the string if there is none.
*
- *-------------------------------------------------------------------------
+ * If num_err is not NULL, we return an error flag to *num_err rather than
+ * raising an error for a badly-formed number. Also, if total_len is not NULL
+ * the distance from lex->input to the token end+1 is returned to *total_len.
*/
static inline void
-json_lex_number(JsonLexContext *lex, char *s)
+json_lex_number(JsonLexContext *lex, char *s,
+ bool *num_err, int *total_len)
{
bool error = false;
- char *p;
- int len;
+ int len = s - lex->input;
- 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')
+ if (len < lex->input_length && *s == '0')
{
s++;
len++;
}
- else if (*s >= '1' && *s <= '9')
+ else if (len < lex->input_length && *s >= '1' && *s <= '9')
{
do
{
s++;
len++;
- } while (*s >= '0' && *s <= '9' && len < lex->input_length);
+ } while (len < lex->input_length && *s >= '0' && *s <= '9');
}
else
error = true;
{
s++;
len++;
- } while (*s >= '0' && *s <= '9' && len < lex->input_length);
+ } while (len < lex->input_length && *s >= '0' && *s <= '9');
}
}
}
/*
- * Check for trailing garbage. As in json_lex(), any alphanumeric stuff
+ * 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++)
+ for (; len < lex->input_length && JSON_ALPHANUMERIC_CHAR(*s); s++, len++)
error = true;
- lex->prev_token_terminator = lex->token_terminator;
- lex->token_terminator = p;
- if (error)
- report_invalid_token(lex);
+
+ if (total_len != NULL)
+ *total_len = len;
+
+ if (num_err != NULL)
+ {
+ /* let the caller handle any error */
+ *num_err = error;
+ }
+ else
+ {
+ /* return token endpoint */
+ lex->prev_token_terminator = lex->token_terminator;
+ lex->token_terminator = s;
+ /* handle error if any */
+ if (error)
+ report_invalid_token(lex);
+ }
}
/*
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"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("The input string ended unexpectedly."),
report_json_context(lex)));
if (ctx == JSON_PARSE_END)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Expected end of input, but found \"%s\".",
token),
report_json_context(lex)));
case JSON_PARSE_VALUE:
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Expected JSON value, but found \"%s\".",
token),
report_json_context(lex)));
case JSON_PARSE_STRING:
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Expected string, but found \"%s\".",
token),
report_json_context(lex)));
case JSON_PARSE_ARRAY_START:
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Expected array element or \"]\", but found \"%s\".",
token),
report_json_context(lex)));
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),
+ errmsg("invalid input syntax for type %s", "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),
+ errmsg("invalid input syntax for type %s", "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"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Expected \":\", but found \"%s\".",
token),
report_json_context(lex)));
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),
+ errmsg("invalid input syntax for type %s", "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"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Expected string, but found \"%s\".",
token),
report_json_context(lex)));
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
- errmsg("invalid input syntax for type json"),
+ errmsg("invalid input syntax for type %s", "json"),
errdetail("Token \"%s\" is invalid.", token),
report_json_context(lex)));
}
line_number = 1;
for (;;)
{
- /* Always advance over newlines (context_end test is just paranoia) */
- if (*context_start == '\n' && context_start < context_end)
+ /* Always advance over newlines */
+ if (context_start < context_end && *context_start == '\n')
{
context_start++;
line_start = context_start;
}
/*
- * Turn a scalar Datum into JSON, appending the string to "result".
+ * Determine how we want to print values of a given type in datum_to_json.
+ *
+ * Given the datatype OID, return its JsonTypeCategory, as well as the type's
+ * output function OID. If the returned category is JSONTYPE_CAST, we
+ * return the OID of the type->JSON cast function instead.
+ */
+static void
+json_categorize_type(Oid typoid,
+ JsonTypeCategory *tcategory,
+ Oid *outfuncoid)
+{
+ bool typisvarlena;
+
+ /* Look through any domain */
+ typoid = getBaseType(typoid);
+
+ *outfuncoid = InvalidOid;
+
+ /*
+ * We need to get the output function for everything except date and
+ * timestamp types, array and composite types, booleans, and non-builtin
+ * types where there's a cast to json.
+ */
+
+ switch (typoid)
+ {
+ case BOOLOID:
+ *tcategory = JSONTYPE_BOOL;
+ break;
+
+ case INT2OID:
+ case INT4OID:
+ case INT8OID:
+ case FLOAT4OID:
+ case FLOAT8OID:
+ case NUMERICOID:
+ getTypeOutputInfo(typoid, outfuncoid, &typisvarlena);
+ *tcategory = JSONTYPE_NUMERIC;
+ break;
+
+ case DATEOID:
+ *tcategory = JSONTYPE_DATE;
+ break;
+
+ case TIMESTAMPOID:
+ *tcategory = JSONTYPE_TIMESTAMP;
+ break;
+
+ case TIMESTAMPTZOID:
+ *tcategory = JSONTYPE_TIMESTAMPTZ;
+ break;
+
+ case JSONOID:
+ case JSONBOID:
+ getTypeOutputInfo(typoid, outfuncoid, &typisvarlena);
+ *tcategory = JSONTYPE_JSON;
+ break;
+
+ default:
+ /* Check for arrays and composites */
+ if (OidIsValid(get_element_type(typoid)) || typoid == ANYARRAYOID
+ || typoid == RECORDARRAYOID)
+ *tcategory = JSONTYPE_ARRAY;
+ else if (type_is_rowtype(typoid)) /* includes RECORDOID */
+ *tcategory = JSONTYPE_COMPOSITE;
+ else
+ {
+ /* It's probably the general case ... */
+ *tcategory = JSONTYPE_OTHER;
+ /* but let's look for a cast to json, if it's not built-in */
+ if (typoid >= FirstNormalObjectId)
+ {
+ Oid castfunc;
+ CoercionPathType ctype;
+
+ ctype = find_coercion_pathway(JSONOID, typoid,
+ COERCION_EXPLICIT,
+ &castfunc);
+ if (ctype == COERCION_PATH_FUNC && OidIsValid(castfunc))
+ {
+ *tcategory = JSONTYPE_CAST;
+ *outfuncoid = castfunc;
+ }
+ else
+ {
+ /* non builtin type with no cast */
+ getTypeOutputInfo(typoid, outfuncoid, &typisvarlena);
+ }
+ }
+ else
+ {
+ /* any other builtin type */
+ getTypeOutputInfo(typoid, outfuncoid, &typisvarlena);
+ }
+ }
+ break;
+ }
+}
+
+/*
+ * Turn a Datum into JSON text, appending the string to "result".
*
- * Hand off a non-scalar datum to composite_to_json or array_to_json_internal
- * as appropriate.
+ * tcategory and outfuncoid are from a previous call to json_categorize_type,
+ * except that if is_null is true then they can be invalid.
+ *
+ * If key_scalar is true, the value is being printed as a key, so insist
+ * it's of an acceptable type, and force it to be quoted.
*/
static void
datum_to_json(Datum val, bool is_null, StringInfo result,
- TYPCATEGORY tcategory, Oid typoutputfunc)
+ JsonTypeCategory tcategory, Oid outfuncoid,
+ bool key_scalar)
{
char *outputstr;
text *jsontext;
+ check_stack_depth();
+
+ /* callers are expected to ensure that null keys are not passed in */
+ Assert(!(key_scalar && is_null));
+
if (is_null)
{
appendStringInfoString(result, "null");
return;
}
+ if (key_scalar &&
+ (tcategory == JSONTYPE_ARRAY ||
+ tcategory == JSONTYPE_COMPOSITE ||
+ tcategory == JSONTYPE_JSON ||
+ tcategory == JSONTYPE_CAST))
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("key value must be scalar, not array, composite, or json")));
+
switch (tcategory)
{
- case TYPCATEGORY_ARRAY:
+ case JSONTYPE_ARRAY:
array_to_json_internal(val, result, false);
break;
- case TYPCATEGORY_COMPOSITE:
+ case JSONTYPE_COMPOSITE:
composite_to_json(val, result, false);
break;
- case TYPCATEGORY_BOOLEAN:
- if (DatumGetBool(val))
- appendStringInfoString(result, "true");
+ case JSONTYPE_BOOL:
+ outputstr = DatumGetBool(val) ? "true" : "false";
+ if (key_scalar)
+ escape_json(result, outputstr);
else
- appendStringInfoString(result, "false");
+ appendStringInfoString(result, outputstr);
break;
- case TYPCATEGORY_NUMERIC:
- outputstr = OidOutputFunctionCall(typoutputfunc, val);
+ case JSONTYPE_NUMERIC:
+ outputstr = OidOutputFunctionCall(outfuncoid, 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.
+ * Don't call escape_json for a non-key if it's a valid JSON
+ * number.
*/
- if (strpbrk(outputstr, NON_NUMERIC_LETTER) == NULL)
+ if (!key_scalar && IsValidJsonNumber(outputstr, strlen(outputstr)))
appendStringInfoString(result, outputstr);
else
escape_json(result, outputstr);
pfree(outputstr);
break;
- case TYPCATEGORY_JSON:
- /* JSON will already be escaped */
- outputstr = OidOutputFunctionCall(typoutputfunc, val);
+ case JSONTYPE_DATE:
+ {
+ char buf[MAXDATELEN + 1];
+
+ JsonEncodeDateTime(buf, val, DATEOID, NULL);
+ appendStringInfo(result, "\"%s\"", buf);
+ }
+ break;
+ case JSONTYPE_TIMESTAMP:
+ {
+ char buf[MAXDATELEN + 1];
+
+ JsonEncodeDateTime(buf, val, TIMESTAMPOID, NULL);
+ appendStringInfo(result, "\"%s\"", buf);
+ }
+ break;
+ case JSONTYPE_TIMESTAMPTZ:
+ {
+ char buf[MAXDATELEN + 1];
+
+ JsonEncodeDateTime(buf, val, TIMESTAMPTZOID, NULL);
+ appendStringInfo(result, "\"%s\"", buf);
+ }
+ break;
+ case JSONTYPE_JSON:
+ /* JSON and JSONB output will already be escaped */
+ outputstr = OidOutputFunctionCall(outfuncoid, val);
appendStringInfoString(result, outputstr);
pfree(outputstr);
break;
- case TYPCATEGORY_JSON_CAST:
- jsontext = DatumGetTextP(OidFunctionCall1(typoutputfunc, val));
+ case JSONTYPE_CAST:
+ /* outfuncoid refers to a cast function, not an output function */
+ jsontext = DatumGetTextPP(OidFunctionCall1(outfuncoid, val));
outputstr = text_to_cstring(jsontext);
appendStringInfoString(result, outputstr);
pfree(outputstr);
pfree(jsontext);
break;
default:
- outputstr = OidOutputFunctionCall(typoutputfunc, val);
+ outputstr = OidOutputFunctionCall(outfuncoid, val);
escape_json(result, outputstr);
pfree(outputstr);
break;
}
}
+/*
+ * Encode 'value' of datetime type 'typid' into JSON string in ISO format using
+ * optionally preallocated buffer 'buf'. Optional 'tzp' determines time-zone
+ * offset (in seconds) in which we want to show timestamptz.
+ */
+char *
+JsonEncodeDateTime(char *buf, Datum value, Oid typid, const int *tzp)
+{
+ if (!buf)
+ buf = palloc(MAXDATELEN + 1);
+
+ switch (typid)
+ {
+ case DATEOID:
+ {
+ DateADT date;
+ struct pg_tm tm;
+
+ date = DatumGetDateADT(value);
+
+ /* Same as date_out(), but forcing DateStyle */
+ if (DATE_NOT_FINITE(date))
+ EncodeSpecialDate(date, buf);
+ else
+ {
+ j2date(date + POSTGRES_EPOCH_JDATE,
+ &(tm.tm_year), &(tm.tm_mon), &(tm.tm_mday));
+ EncodeDateOnly(&tm, USE_XSD_DATES, buf);
+ }
+ }
+ break;
+ case TIMEOID:
+ {
+ TimeADT time = DatumGetTimeADT(value);
+ struct pg_tm tt,
+ *tm = &tt;
+ fsec_t fsec;
+
+ /* Same as time_out(), but forcing DateStyle */
+ time2tm(time, tm, &fsec);
+ EncodeTimeOnly(tm, fsec, false, 0, USE_XSD_DATES, buf);
+ }
+ break;
+ case TIMETZOID:
+ {
+ TimeTzADT *time = DatumGetTimeTzADTP(value);
+ struct pg_tm tt,
+ *tm = &tt;
+ fsec_t fsec;
+ int tz;
+
+ /* Same as timetz_out(), but forcing DateStyle */
+ timetz2tm(time, tm, &fsec, &tz);
+ EncodeTimeOnly(tm, fsec, true, tz, USE_XSD_DATES, buf);
+ }
+ break;
+ case TIMESTAMPOID:
+ {
+ Timestamp timestamp;
+ struct pg_tm tm;
+ fsec_t fsec;
+
+ timestamp = DatumGetTimestamp(value);
+ /* Same as timestamp_out(), but forcing DateStyle */
+ if (TIMESTAMP_NOT_FINITE(timestamp))
+ EncodeSpecialTimestamp(timestamp, buf);
+ else if (timestamp2tm(timestamp, NULL, &tm, &fsec, NULL, NULL) == 0)
+ EncodeDateTime(&tm, fsec, false, 0, NULL, USE_XSD_DATES, buf);
+ else
+ ereport(ERROR,
+ (errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
+ errmsg("timestamp out of range")));
+ }
+ break;
+ case TIMESTAMPTZOID:
+ {
+ TimestampTz timestamp;
+ struct pg_tm tm;
+ int tz;
+ fsec_t fsec;
+ const char *tzn = NULL;
+
+ timestamp = DatumGetTimestampTz(value);
+
+ /*
+ * If a time zone is specified, we apply the time-zone shift,
+ * convert timestamptz to pg_tm as if it were without a time
+ * zone, and then use the specified time zone for converting
+ * the timestamp into a string.
+ */
+ if (tzp)
+ {
+ tz = *tzp;
+ timestamp -= (TimestampTz) tz * USECS_PER_SEC;
+ }
+
+ /* Same as timestamptz_out(), but forcing DateStyle */
+ if (TIMESTAMP_NOT_FINITE(timestamp))
+ EncodeSpecialTimestamp(timestamp, buf);
+ else if (timestamp2tm(timestamp, tzp ? NULL : &tz, &tm, &fsec,
+ tzp ? NULL : &tzn, NULL) == 0)
+ {
+ if (tzp)
+ tm.tm_isdst = 1; /* set time-zone presence flag */
+
+ EncodeDateTime(&tm, fsec, true, tz, tzn, USE_XSD_DATES, buf);
+ }
+ else
+ ereport(ERROR,
+ (errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
+ errmsg("timestamp out of range")));
+ }
+ break;
+ default:
+ elog(ERROR, "unknown jsonb value datetime type oid %d", typid);
+ return NULL;
+ }
+
+ return buf;
+}
+
/*
* Process a single dimension of an array.
* If it's the innermost dimension, output the values, otherwise call
*/
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)
+ bool *nulls, int *valcount, JsonTypeCategory tcategory,
+ Oid outfuncoid, bool use_line_feeds)
{
int i;
const char *sep;
if (dim + 1 == ndims)
{
datum_to_json(vals[*valcount], nulls[*valcount], result, tcategory,
- typoutputfunc);
+ outfuncoid, false);
(*valcount)++;
}
else
* we'll say no.
*/
array_dim_to_json(result, dim + 1, ndims, dims, vals, nulls,
- valcount, tcategory, typoutputfunc, false);
+ valcount, tcategory, outfuncoid, false);
}
}
bool *nulls;
int16 typlen;
bool typbyval;
- char typalign,
- typdelim;
- Oid typioparam;
- Oid typoutputfunc;
- TYPCATEGORY tcategory;
- Oid castfunc = InvalidOid;
+ char typalign;
+ JsonTypeCategory tcategory;
+ Oid outfuncoid;
ndim = ARR_NDIM(v);
dim = ARR_DIMS(v);
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;
+ get_typlenbyvalalign(element_type,
+ &typlen, &typbyval, &typalign);
- ReleaseSysCache(tuple);
- }
- }
+ json_categorize_type(element_type,
+ &tcategory, &outfuncoid);
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);
+ outfuncoid, use_line_feeds);
pfree(elements);
pfree(nulls);
for (i = 0; i < tupdesc->natts; i++)
{
- Datum val,
- origval;
+ Datum val;
bool isnull;
char *attname;
- TYPCATEGORY tcategory;
- Oid typoutput;
- bool typisvarlena;
- Oid castfunc = InvalidOid;
+ JsonTypeCategory tcategory;
+ Oid outfuncoid;
+ Form_pg_attribute att = TupleDescAttr(tupdesc, i);
- if (tupdesc->attrs[i]->attisdropped)
+ if (att->attisdropped)
continue;
if (needsep)
appendStringInfoString(result, sep);
needsep = true;
- attname = NameStr(tupdesc->attrs[i]->attname);
+ attname = NameStr(att->attname);
escape_json(result, attname);
appendStringInfoChar(result, ':');
- origval = heap_getattr(tuple, i + 1, tupdesc, &isnull);
-
- getTypeOutputInfo(tupdesc->attrs[i]->atttypid,
- &typoutput, &typisvarlena);
+ val = heap_getattr(tuple, i + 1, tupdesc, &isnull);
- if (tupdesc->attrs[i]->atttypid > FirstNormalObjectId)
+ if (isnull)
{
- 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);
- }
+ tcategory = JSONTYPE_NULL;
+ outfuncoid = InvalidOid;
}
-
- 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;
+ json_categorize_type(att->atttypid, &tcategory, &outfuncoid);
- datum_to_json(val, isnull, result, tcategory, typoutput);
-
- /* Clean up detoasted copy, if any */
- if (val != origval)
- pfree(DatumGetPointer(val));
+ datum_to_json(val, isnull, result, tcategory, outfuncoid, false);
}
appendStringInfoChar(result, '}');
ReleaseTupleDesc(tupdesc);
}
+/*
+ * Append JSON text for "val" to "result".
+ *
+ * This is just a thin wrapper around datum_to_json. If the same type will be
+ * printed many times, avoid using this; better to do the json_categorize_type
+ * lookups only once.
+ */
+static void
+add_json(Datum val, bool is_null, StringInfo result,
+ Oid val_type, bool key_scalar)
+{
+ JsonTypeCategory tcategory;
+ Oid outfuncoid;
+
+ if (val_type == InvalidOid)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("could not determine input data type")));
+
+ if (is_null)
+ {
+ tcategory = JSONTYPE_NULL;
+ outfuncoid = InvalidOid;
+ }
+ else
+ json_categorize_type(val_type,
+ &tcategory, &outfuncoid);
+
+ datum_to_json(val, is_null, result, tcategory, outfuncoid, key_scalar);
+}
+
/*
* SQL function array_to_json(row)
*/
-extern Datum
+Datum
array_to_json(PG_FUNCTION_ARGS)
{
Datum array = PG_GETARG_DATUM(0);
array_to_json_internal(array, result, false);
- PG_RETURN_TEXT_P(cstring_to_text(result->data));
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(result->data, result->len));
}
/*
* SQL function array_to_json(row, prettybool)
*/
-extern Datum
+Datum
array_to_json_pretty(PG_FUNCTION_ARGS)
{
Datum array = PG_GETARG_DATUM(0);
array_to_json_internal(array, result, use_line_feeds);
- PG_RETURN_TEXT_P(cstring_to_text(result->data));
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(result->data, result->len));
}
/*
* SQL function row_to_json(row)
*/
-extern Datum
+Datum
row_to_json(PG_FUNCTION_ARGS)
{
Datum array = PG_GETARG_DATUM(0);
composite_to_json(array, result, false);
- PG_RETURN_TEXT_P(cstring_to_text(result->data));
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(result->data, result->len));
}
/*
* SQL function row_to_json(row, prettybool)
*/
-extern Datum
+Datum
row_to_json_pretty(PG_FUNCTION_ARGS)
{
Datum array = PG_GETARG_DATUM(0);
composite_to_json(array, result, use_line_feeds);
- PG_RETURN_TEXT_P(cstring_to_text(result->data));
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(result->data, result->len));
}
/*
Datum
to_json(PG_FUNCTION_ARGS)
{
+ Datum val = PG_GETARG_DATUM(0);
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;
+ JsonTypeCategory tcategory;
+ Oid outfuncoid;
if (val_type == InvalidOid)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("could not determine input data type")));
+ json_categorize_type(val_type,
+ &tcategory, &outfuncoid);
result = makeStringInfo();
- orig_val = PG_ARGISNULL(0) ? (Datum) 0 : PG_GETARG_DATUM(0);
+ datum_to_json(val, false, result, tcategory, outfuncoid, false);
- getTypeOutputInfo(val_type, &typoutput, &typisvarlena);
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(result->data, result->len));
+}
- if (val_type > FirstNormalObjectId)
+/*
+ * json_agg transition function
+ *
+ * aggregate input column as a json array value.
+ */
+Datum
+json_agg_transfn(PG_FUNCTION_ARGS)
+{
+ MemoryContext aggcontext,
+ oldcontext;
+ JsonAggState *state;
+ Datum val;
+
+ if (!AggCheckCallContext(fcinfo, &aggcontext))
{
- HeapTuple tuple;
- Form_pg_cast castForm;
+ /* cannot be called directly because of internal-type argument */
+ elog(ERROR, "json_agg_transfn called in non-aggregate context");
+ }
- tuple = SearchSysCache2(CASTSOURCETARGET,
- ObjectIdGetDatum(val_type),
- ObjectIdGetDatum(JSONOID));
- if (HeapTupleIsValid(tuple))
- {
- castForm = (Form_pg_cast) GETSTRUCT(tuple);
+ if (PG_ARGISNULL(0))
+ {
+ Oid arg_type = get_fn_expr_argtype(fcinfo->flinfo, 1);
- if (castForm->castmethod == COERCION_METHOD_FUNCTION)
- castfunc = typoutput = castForm->castfunc;
+ if (arg_type == InvalidOid)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("could not determine input data type")));
- ReleaseSysCache(tuple);
- }
- }
+ /*
+ * Make this state object in a context where it will persist for the
+ * duration of the aggregate call. MemoryContextSwitchTo is only
+ * needed the first time, as the StringInfo routines make sure they
+ * use the right context to enlarge the object if necessary.
+ */
+ oldcontext = MemoryContextSwitchTo(aggcontext);
+ state = (JsonAggState *) palloc(sizeof(JsonAggState));
+ state->str = makeStringInfo();
+ MemoryContextSwitchTo(oldcontext);
- 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;
+ appendStringInfoChar(state->str, '[');
+ json_categorize_type(arg_type, &state->val_category,
+ &state->val_output_func);
+ }
else
- tcategory = TypeCategory(val_type);
+ {
+ state = (JsonAggState *) PG_GETARG_POINTER(0);
+ appendStringInfoString(state->str, ", ");
+ }
+
+ /* fast path for NULLs */
+ if (PG_ARGISNULL(1))
+ {
+ datum_to_json((Datum) 0, true, state->str, JSONTYPE_NULL,
+ InvalidOid, false);
+ PG_RETURN_POINTER(state);
+ }
+
+ val = PG_GETARG_DATUM(1);
+
+ /* add some whitespace if structured type and not first item */
+ if (!PG_ARGISNULL(0) &&
+ (state->val_category == JSONTYPE_ARRAY ||
+ state->val_category == JSONTYPE_COMPOSITE))
+ {
+ appendStringInfoString(state->str, "\n ");
+ }
+
+ datum_to_json(val, false, state->str, state->val_category,
+ state->val_output_func, false);
/*
- * If we have a toasted datum, forcibly detoast it here to avoid memory
- * leakage inside the type's output routine.
+ * The transition type for json_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 JsonAggState pointer through nodeAgg.c's machinations.
*/
- if (typisvarlena && orig_val != (Datum) 0)
- val = PointerGetDatum(PG_DETOAST_DATUM(orig_val));
- else
- val = orig_val;
+ PG_RETURN_POINTER(state);
+}
+
+/*
+ * json_agg final function
+ */
+Datum
+json_agg_finalfn(PG_FUNCTION_ARGS)
+{
+ JsonAggState *state;
+
+ /* cannot be called directly because of internal-type argument */
+ Assert(AggCheckCallContext(fcinfo, NULL));
- datum_to_json(val, false, result, tcategory, typoutput);
+ state = PG_ARGISNULL(0) ?
+ NULL :
+ (JsonAggState *) PG_GETARG_POINTER(0);
- /* Clean up detoasted copy, if any */
- if (val != orig_val)
- pfree(DatumGetPointer(val));
+ /* NULL result for no rows in, as is standard with aggregates */
+ if (state == NULL)
+ PG_RETURN_NULL();
- PG_RETURN_TEXT_P(cstring_to_text(result->data));
+ /* Else return state with appropriate array terminator added */
+ PG_RETURN_TEXT_P(catenate_stringinfo_string(state->str, "]"));
}
/*
- * json_agg transition function
+ * json_object_agg transition function.
+ *
+ * aggregate two input columns as a single json object value.
*/
Datum
-json_agg_transfn(PG_FUNCTION_ARGS)
+json_object_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")));
+ JsonAggState *state;
+ Datum arg;
if (!AggCheckCallContext(fcinfo, &aggcontext))
{
/* cannot be called directly because of internal-type argument */
- elog(ERROR, "json_agg_transfn called in non-aggregate context");
+ elog(ERROR, "json_object_agg_transfn called in non-aggregate context");
}
if (PG_ARGISNULL(0))
{
+ Oid arg_type;
+
/*
- * 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.
+ * Make the StringInfo in a context where it will persist for the
+ * duration of the aggregate call. Switching context is only needed
+ * for this initial step, as the StringInfo routines make sure they
+ * use the right context to enlarge the object if necessary.
*/
oldcontext = MemoryContextSwitchTo(aggcontext);
- state = makeStringInfo();
+ state = (JsonAggState *) palloc(sizeof(JsonAggState));
+ state->str = makeStringInfo();
MemoryContextSwitchTo(oldcontext);
- appendStringInfoChar(state, '[');
+ arg_type = get_fn_expr_argtype(fcinfo->flinfo, 1);
+
+ if (arg_type == InvalidOid)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("could not determine data type for argument %d", 1)));
+
+ json_categorize_type(arg_type, &state->key_category,
+ &state->key_output_func);
+
+ arg_type = get_fn_expr_argtype(fcinfo->flinfo, 2);
+
+ if (arg_type == InvalidOid)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("could not determine data type for argument %d", 2)));
+
+ json_categorize_type(arg_type, &state->val_category,
+ &state->val_output_func);
+
+ appendStringInfoString(state->str, "{ ");
}
else
{
- state = (StringInfo) PG_GETARG_POINTER(0);
- appendStringInfoString(state, ", ");
+ state = (JsonAggState *) PG_GETARG_POINTER(0);
+ appendStringInfoString(state->str, ", ");
}
- /* fast path for NULLs */
+ /*
+ * Note: since json_object_agg() is declared as taking type "any", the
+ * parser will not do any type conversion on unknown-type literals (that
+ * is, undecorated strings or NULLs). Such values will arrive here as
+ * type UNKNOWN, which fortunately does not matter to us, since
+ * unknownout() works fine.
+ */
+
if (PG_ARGISNULL(1))
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("field name must not be null")));
+
+ arg = PG_GETARG_DATUM(1);
+
+ datum_to_json(arg, false, state->str, state->key_category,
+ state->key_output_func, true);
+
+ appendStringInfoString(state->str, " : ");
+
+ if (PG_ARGISNULL(2))
+ arg = (Datum) 0;
+ else
+ arg = PG_GETARG_DATUM(2);
+
+ datum_to_json(arg, PG_ARGISNULL(2), state->str, state->val_category,
+ state->val_output_func, false);
+
+ PG_RETURN_POINTER(state);
+}
+
+/*
+ * json_object_agg final function.
+ */
+Datum
+json_object_agg_finalfn(PG_FUNCTION_ARGS)
+{
+ JsonAggState *state;
+
+ /* cannot be called directly because of internal-type argument */
+ Assert(AggCheckCallContext(fcinfo, NULL));
+
+ state = PG_ARGISNULL(0) ? NULL : (JsonAggState *) PG_GETARG_POINTER(0);
+
+ /* NULL result for no rows in, as is standard with aggregates */
+ if (state == NULL)
+ PG_RETURN_NULL();
+
+ /* Else return state with appropriate object terminator added */
+ PG_RETURN_TEXT_P(catenate_stringinfo_string(state->str, " }"));
+}
+
+/*
+ * Helper function for aggregates: return given StringInfo's contents plus
+ * specified trailing string, as a text datum. We need this because aggregate
+ * final functions are not allowed to modify the aggregate state.
+ */
+static text *
+catenate_stringinfo_string(StringInfo buffer, const char *addon)
+{
+ /* custom version of cstring_to_text_with_len */
+ int buflen = buffer->len;
+ int addlen = strlen(addon);
+ text *result = (text *) palloc(buflen + addlen + VARHDRSZ);
+
+ SET_VARSIZE(result, buflen + addlen + VARHDRSZ);
+ memcpy(VARDATA(result), buffer->data, buflen);
+ memcpy(VARDATA(result) + buflen, addon, addlen);
+
+ return result;
+}
+
+/*
+ * SQL function json_build_object(variadic "any")
+ */
+Datum
+json_build_object(PG_FUNCTION_ARGS)
+{
+ int nargs = PG_NARGS();
+ int i;
+ const char *sep = "";
+ StringInfo result;
+ Datum *args;
+ bool *nulls;
+ Oid *types;
+
+ /* fetch argument values to build the object */
+ nargs = extract_variadic_args(fcinfo, 0, false, &args, &types, &nulls);
+
+ if (nargs < 0)
+ PG_RETURN_NULL();
+
+ if (nargs % 2 != 0)
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("argument list must have even number of elements"),
+ /* translator: %s is a SQL function name */
+ errhint("The arguments of %s must consist of alternating keys and values.",
+ "json_build_object()")));
+
+ result = makeStringInfo();
+
+ appendStringInfoChar(result, '{');
+
+ for (i = 0; i < nargs; i += 2)
{
- orig_val = (Datum) 0;
- datum_to_json(orig_val, true, state, 0, InvalidOid);
- PG_RETURN_POINTER(state);
+ appendStringInfoString(result, sep);
+ sep = ", ";
+
+ /* process key */
+ if (nulls[i])
+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("argument %d cannot be null", i + 1),
+ errhint("Object keys should be text.")));
+
+ add_json(args[i], false, result, types[i], true);
+
+ appendStringInfoString(result, " : ");
+
+ /* process value */
+ add_json(args[i + 1], nulls[i + 1], result, types[i + 1], false);
}
+ appendStringInfoChar(result, '}');
+
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(result->data, result->len));
+}
+
+/*
+ * degenerate case of json_build_object where it gets 0 arguments.
+ */
+Datum
+json_build_object_noargs(PG_FUNCTION_ARGS)
+{
+ PG_RETURN_TEXT_P(cstring_to_text_with_len("{}", 2));
+}
+
+/*
+ * SQL function json_build_array(variadic "any")
+ */
+Datum
+json_build_array(PG_FUNCTION_ARGS)
+{
+ int nargs;
+ int i;
+ const char *sep = "";
+ StringInfo result;
+ Datum *args;
+ bool *nulls;
+ Oid *types;
+
+ /* fetch argument values to build the array */
+ nargs = extract_variadic_args(fcinfo, 0, false, &args, &types, &nulls);
+
+ if (nargs < 0)
+ PG_RETURN_NULL();
- orig_val = PG_GETARG_DATUM(1);
+ result = makeStringInfo();
- getTypeOutputInfo(val_type, &typoutput, &typisvarlena);
+ appendStringInfoChar(result, '[');
- if (val_type > FirstNormalObjectId)
+ for (i = 0; i < nargs; i++)
{
- HeapTuple tuple;
- Form_pg_cast castForm;
+ appendStringInfoString(result, sep);
+ sep = ", ";
+ add_json(args[i], nulls[i], result, types[i], false);
+ }
- tuple = SearchSysCache2(CASTSOURCETARGET,
- ObjectIdGetDatum(val_type),
- ObjectIdGetDatum(JSONOID));
- if (HeapTupleIsValid(tuple))
- {
- castForm = (Form_pg_cast) GETSTRUCT(tuple);
+ appendStringInfoChar(result, ']');
- if (castForm->castmethod == COERCION_METHOD_FUNCTION)
- castfunc = typoutput = castForm->castfunc;
+ PG_RETURN_TEXT_P(cstring_to_text_with_len(result->data, result->len));
+}
- ReleaseSysCache(tuple);
- }
+/*
+ * degenerate case of json_build_array where it gets 0 arguments.
+ */
+Datum
+json_build_array_noargs(PG_FUNCTION_ARGS)
+{
+ PG_RETURN_TEXT_P(cstring_to_text_with_len("[]", 2));
+}
+
+/*
+ * SQL function json_object(text[])
+ *
+ * take a one or two dimensional array of text as key/value pairs
+ * for a json object.
+ */
+Datum
+json_object(PG_FUNCTION_ARGS)
+{
+ ArrayType *in_array = PG_GETARG_ARRAYTYPE_P(0);
+ int ndims = ARR_NDIM(in_array);
+ StringInfoData result;
+ Datum *in_datums;
+ bool *in_nulls;
+ int in_count,
+ count,
+ i;
+ text *rval;
+ char *v;
+
+ switch (ndims)
+ {
+ case 0:
+ PG_RETURN_DATUM(CStringGetTextDatum("{}"));
+ break;
+
+ case 1:
+ if ((ARR_DIMS(in_array)[0]) % 2)
+ ereport(ERROR,
+ (errcode(ERRCODE_ARRAY_SUBSCRIPT_ERROR),
+ errmsg("array must have even number of elements")));
+ break;
+
+ case 2:
+ if ((ARR_DIMS(in_array)[1]) != 2)
+ ereport(ERROR,
+ (errcode(ERRCODE_ARRAY_SUBSCRIPT_ERROR),
+ errmsg("array must have two columns")));
+ break;
+
+ default:
+ ereport(ERROR,
+ (errcode(ERRCODE_ARRAY_SUBSCRIPT_ERROR),
+ errmsg("wrong number of array subscripts")));
}
- 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);
+ deconstruct_array(in_array,
+ TEXTOID, -1, false, 'i',
+ &in_datums, &in_nulls, &in_count);
- /*
- * 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;
+ count = in_count / 2;
- if (!PG_ARGISNULL(0) &&
- (tcategory == TYPCATEGORY_ARRAY || tcategory == TYPCATEGORY_COMPOSITE))
+ initStringInfo(&result);
+
+ appendStringInfoChar(&result, '{');
+
+ for (i = 0; i < count; ++i)
{
- appendStringInfoString(state, "\n ");
+ if (in_nulls[i * 2])
+ ereport(ERROR,
+ (errcode(ERRCODE_NULL_VALUE_NOT_ALLOWED),
+ errmsg("null value not allowed for object key")));
+
+ v = TextDatumGetCString(in_datums[i * 2]);
+ if (i > 0)
+ appendStringInfoString(&result, ", ");
+ escape_json(&result, v);
+ appendStringInfoString(&result, " : ");
+ pfree(v);
+ if (in_nulls[i * 2 + 1])
+ appendStringInfoString(&result, "null");
+ else
+ {
+ v = TextDatumGetCString(in_datums[i * 2 + 1]);
+ escape_json(&result, v);
+ pfree(v);
+ }
}
- datum_to_json(val, false, state, tcategory, typoutput);
+ appendStringInfoChar(&result, '}');
- /* Clean up detoasted copy, if any */
- if (val != orig_val)
- pfree(DatumGetPointer(val));
+ pfree(in_datums);
+ pfree(in_nulls);
+
+ rval = cstring_to_text_with_len(result.data, result.len);
+ pfree(result.data);
+
+ PG_RETURN_TEXT_P(rval);
- /*
- * 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
+ * SQL function json_object(text[], text[])
+ *
+ * take separate key and value arrays of text to construct a json object
+ * pairwise.
*/
Datum
-json_agg_finalfn(PG_FUNCTION_ARGS)
+json_object_two_arg(PG_FUNCTION_ARGS)
{
- StringInfo state;
+ ArrayType *key_array = PG_GETARG_ARRAYTYPE_P(0);
+ ArrayType *val_array = PG_GETARG_ARRAYTYPE_P(1);
+ int nkdims = ARR_NDIM(key_array);
+ int nvdims = ARR_NDIM(val_array);
+ StringInfoData result;
+ Datum *key_datums,
+ *val_datums;
+ bool *key_nulls,
+ *val_nulls;
+ int key_count,
+ val_count,
+ i;
+ text *rval;
+ char *v;
+
+ if (nkdims > 1 || nkdims != nvdims)
+ ereport(ERROR,
+ (errcode(ERRCODE_ARRAY_SUBSCRIPT_ERROR),
+ errmsg("wrong number of array subscripts")));
- /* cannot be called directly because of internal-type argument */
- Assert(AggCheckCallContext(fcinfo, NULL));
+ if (nkdims == 0)
+ PG_RETURN_DATUM(CStringGetTextDatum("{}"));
- state = PG_ARGISNULL(0) ? NULL : (StringInfo) PG_GETARG_POINTER(0);
+ deconstruct_array(key_array,
+ TEXTOID, -1, false, 'i',
+ &key_datums, &key_nulls, &key_count);
- if (state == NULL)
- PG_RETURN_NULL();
+ deconstruct_array(val_array,
+ TEXTOID, -1, false, 'i',
+ &val_datums, &val_nulls, &val_count);
+
+ if (key_count != val_count)
+ ereport(ERROR,
+ (errcode(ERRCODE_ARRAY_SUBSCRIPT_ERROR),
+ errmsg("mismatched array dimensions")));
+
+ initStringInfo(&result);
+
+ appendStringInfoChar(&result, '{');
+
+ for (i = 0; i < key_count; ++i)
+ {
+ if (key_nulls[i])
+ ereport(ERROR,
+ (errcode(ERRCODE_NULL_VALUE_NOT_ALLOWED),
+ errmsg("null value not allowed for object key")));
+
+ v = TextDatumGetCString(key_datums[i]);
+ if (i > 0)
+ appendStringInfoString(&result, ", ");
+ escape_json(&result, v);
+ appendStringInfoString(&result, " : ");
+ pfree(v);
+ if (val_nulls[i])
+ appendStringInfoString(&result, "null");
+ else
+ {
+ v = TextDatumGetCString(val_datums[i]);
+ escape_json(&result, v);
+ pfree(v);
+ }
+ }
- appendStringInfoChar(state, ']');
+ appendStringInfoChar(&result, '}');
- PG_RETURN_TEXT_P(cstring_to_text(state->data));
+ pfree(key_datums);
+ pfree(key_nulls);
+ pfree(val_datums);
+ pfree(val_nulls);
+
+ rval = cstring_to_text_with_len(result.data, result.len);
+ pfree(result.data);
+
+ PG_RETURN_TEXT_P(rval);
}
+
/*
* Produce a JSON string literal, properly escaping characters in the text.
*/
{
const char *p;
- appendStringInfoCharMacro(buf, '\"');
+ appendStringInfoCharMacro(buf, '"');
for (p = str; *p; p++)
{
switch (*p)
break;
}
}
- appendStringInfoCharMacro(buf, '\"');
+ appendStringInfoCharMacro(buf, '"');
+}
+
+/*
+ * SQL function json_typeof(json) -> text
+ *
+ * Returns the type of the outermost JSON value as TEXT. Possible types are
+ * "object", "array", "string", "number", "boolean", and "null".
+ *
+ * Performs a single call to json_lex() to get the first token of the supplied
+ * value. This initial token uniquely determines the value's type. As our
+ * input must already have been validated by json_in() or json_recv(), the
+ * initial token should never be JSON_TOKEN_OBJECT_END, JSON_TOKEN_ARRAY_END,
+ * JSON_TOKEN_COLON, JSON_TOKEN_COMMA, or JSON_TOKEN_END.
+ */
+Datum
+json_typeof(PG_FUNCTION_ARGS)
+{
+ text *json;
+
+ JsonLexContext *lex;
+ JsonTokenType tok;
+ char *type;
+
+ json = PG_GETARG_TEXT_PP(0);
+ lex = makeJsonLexContext(json, false);
+
+ /* Lex exactly one token from the input and check its type. */
+ json_lex(lex);
+ tok = lex_peek(lex);
+ switch (tok)
+ {
+ case JSON_TOKEN_OBJECT_START:
+ type = "object";
+ break;
+ case JSON_TOKEN_ARRAY_START:
+ type = "array";
+ break;
+ case JSON_TOKEN_STRING:
+ type = "string";
+ break;
+ case JSON_TOKEN_NUMBER:
+ type = "number";
+ break;
+ case JSON_TOKEN_TRUE:
+ case JSON_TOKEN_FALSE:
+ type = "boolean";
+ break;
+ case JSON_TOKEN_NULL:
+ type = "null";
+ break;
+ default:
+ elog(ERROR, "unexpected json token: %d", tok);
+ }
+
+ PG_RETURN_TEXT_P(cstring_to_text(type));
}
projects / postgresql / blobdiff