From caa4056cd113c3488c107ffc5086ded94fde378d Mon Sep 17 00:00:00 2001 From: Mark Dilger Date: Wed, 22 Jan 2020 16:33:04 -0800 Subject: [PATCH 10/11] Moving json parsing into src/common This commit moves the code for parsing json out of src/backend/utils and also removes the inherent dependency on the database encoding. --- src/backend/utils/adt/json.c | 920 +---------------------------- src/backend/utils/adt/jsonb.c | 2 +- src/backend/utils/adt/jsonfuncs.c | 9 +- src/common/jsonapi.c | 938 +++++++++++++++++++++++++++++- src/include/common/jsonapi.h | 9 +- 5 files changed, 951 insertions(+), 927 deletions(-) diff --git a/src/backend/utils/adt/json.c b/src/backend/utils/adt/json.c index 54075d07e3..1e514cb02c 100644 --- a/src/backend/utils/adt/json.c +++ b/src/backend/utils/adt/json.c @@ -34,31 +34,6 @@ #include "utils/syscache.h" #include "utils/typcache.h" -#define INSIST(x) \ -do { \ - JsonParseErrorType parse_result; \ - if((parse_result = (x)) != JSON_SUCCESS) \ - return parse_result; \ -} while (0) - -/* - * The context of the parser is maintained by the recursive descent - * mechanism, but is passed explicitly to the error reporting routine - * for better diagnostics. - */ -typedef enum /* contexts of JSON parser */ -{ - JSON_PARSE_VALUE, /* expecting a value */ - JSON_PARSE_STRING, /* expecting a string (for a field name) */ - JSON_PARSE_ARRAY_START, /* saw '[', expecting value or ']' */ - JSON_PARSE_ARRAY_NEXT, /* saw array element, expecting ',' or ']' */ - JSON_PARSE_OBJECT_START, /* saw '{', expecting label or '}' */ - JSON_PARSE_OBJECT_LABEL, /* saw object label, expecting ':' */ - JSON_PARSE_OBJECT_NEXT, /* saw object value, expecting ',' or '}' */ - JSON_PARSE_OBJECT_COMMA, /* saw object ',', expecting next label */ - JSON_PARSE_END /* saw the end of a document, expect nothing */ -} JsonParseContext; - typedef enum /* type categories for datum_to_json */ { JSONTYPE_NULL, /* null, so we didn't bother to identify */ @@ -83,17 +58,6 @@ typedef struct JsonAggState Oid val_output_func; } JsonAggState; -static inline JsonTokenType lex_peek(JsonLexContext *lex) __attribute__((warn_unused_result)); -static inline JsonParseErrorType lex_expect(JsonParseContext ctx, JsonLexContext *lex, JsonTokenType token) __attribute__((warn_unused_result)); -static inline JsonParseErrorType json_lex_string(JsonLexContext *lex) __attribute__((warn_unused_result)); -static inline JsonParseErrorType json_lex_number(JsonLexContext *lex, char *s, - bool *num_err, int *total_len) __attribute__((warn_unused_result)); -static inline JsonParseErrorType parse_scalar(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); -static JsonParseErrorType parse_object_field(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); -static JsonParseErrorType parse_object(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); -static JsonParseErrorType parse_array_element(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); -static JsonParseErrorType parse_array(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); -static JsonParseErrorType report_parse_error(JsonParseContext ctx, JsonLexContext *lex) __attribute__((warn_unused_result)); static void composite_to_json(Datum composite, StringInfo result, bool use_line_feeds); static void array_dim_to_json(StringInfo result, int dim, int ndims, int *dims, @@ -112,81 +76,6 @@ 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); -/* Recursive Descent parser support routines */ - -/* - * lex_peek - * - * what is the current look_ahead token? -*/ -static inline JsonTokenType -lex_peek(JsonLexContext *lex) -{ - return lex->token_type; -} - -/* - * lex_except - * - * move the lexer to the next token if the current look_ahead token matches - * the parameter token. Otherwise, report an error. - */ -static inline JsonParseErrorType -lex_expect(JsonParseContext ctx, JsonLexContext *lex, JsonTokenType token) -{ - if (lex_peek(lex) == token) - INSIST(json_lex(lex)); - else - return report_parse_error(ctx, lex); - return JSON_SUCCESS; -} - -/* chars to consider as part of an alphanumeric token */ -#define JSON_ALPHANUMERIC_CHAR(c) \ - (((c) >= 'a' && (c) <= 'z') || \ - ((c) >= 'A' && (c) <= 'Z') || \ - ((c) >= '0' && (c) <= '9') || \ - (c) == '_' || \ - IS_HIGHBIT_SET(c)) - -/* - * 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; - } - - if (JSON_SUCCESS != json_lex_number(&dummy_lex, dummy_lex.input, &numeric_error, &total_len)) - return false; - - return (!numeric_error) && (total_len == dummy_lex.input_length); -} - /* * Input. */ @@ -245,819 +134,12 @@ json_recv(PG_FUNCTION_ARGS) str = pq_getmsgtext(buf, buf->len - buf->cursor, &nbytes); /* Validate it. */ - lex = makeJsonLexContextCstringLen(str, nbytes, false); + lex = makeJsonLexContextCstringLen(str, nbytes, GetDatabaseEncoding(), false); pg_parse_json_or_throw(lex, &nullSemAction); PG_RETURN_TEXT_P(cstring_to_text_with_len(str, nbytes)); } -/* - * pg_parse_json - * - * Publicly visible entry point for the JSON parser. - * - * lex is a lexing context, set up for the json to be processed by calling - * makeJsonLexContext(). sem is a 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. - */ -JsonParseErrorType -pg_parse_json(JsonLexContext *lex, const JsonSemAction *sem) -{ - JsonTokenType tok; - - /* get the initial token */ - INSIST(json_lex(lex)); - - tok = lex_peek(lex); - - /* parse by recursive descent */ - switch (tok) - { - case JSON_TOKEN_OBJECT_START: - INSIST(parse_object(lex, sem)); - break; - case JSON_TOKEN_ARRAY_START: - INSIST(parse_array(lex, sem)); - break; - default: - INSIST(parse_scalar(lex, sem)); /* json can be a bare scalar */ - } - - INSIST(lex_expect(JSON_PARSE_END, lex, JSON_TOKEN_END)); - return JSON_SUCCESS; -} - -/* - * 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; - INSIST(lex_expect(JSON_PARSE_ARRAY_START, ©lex, JSON_TOKEN_ARRAY_START)); - if (lex_peek(©lex) != JSON_TOKEN_ARRAY_END) - { - while (1) - { - count++; - if (JSON_SUCCESS != parse_array_element(©lex, &nullSemAction)) - break; - if (copylex.token_type != JSON_TOKEN_COMMA) - break; - INSIST(json_lex(©lex)); - } - } - INSIST(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: - * - scalar (string, number, true, false, null) - * - array ( [ ] ) - * - array element - * - object ( { } ) - * - object field - */ -static inline JsonParseErrorType -parse_scalar(JsonLexContext *lex, const JsonSemAction *sem) -{ - char *val = NULL; - json_scalar_action sfunc = sem->scalar; - JsonTokenType tok = lex_peek(lex); - - /* a scalar must be a string, a number, true, false, or null */ - switch (tok) - { - case JSON_TOKEN_STRING: - case JSON_TOKEN_NUMBER: - case JSON_TOKEN_TRUE: - case JSON_TOKEN_FALSE: - case JSON_TOKEN_NULL: - break; - default: - return report_parse_error(JSON_PARSE_VALUE, lex); - } - - /* if no semantic function, just consume the token */ - if (sfunc == NULL) - { - INSIST(json_lex(lex)); - return JSON_SUCCESS; - } - - /* extract the de-escaped string value, or the raw lexeme */ - if (lex_peek(lex) == JSON_TOKEN_STRING) - { - if (lex->strval != NULL) - val = pstrdup(lex->strval->data); - } - else - { - int len = (lex->token_terminator - lex->token_start); - - val = palloc(len + 1); - memcpy(val, lex->token_start, len); - val[len] = '\0'; - } - - /* consume the token */ - INSIST(json_lex(lex)); - - /* invoke the callback */ - (*sfunc) (sem->semstate, val, tok); - return JSON_SUCCESS; -} - -static JsonParseErrorType -parse_object_field(JsonLexContext *lex, const JsonSemAction *sem) -{ - /* - * 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 */ - json_ofield_action ostart = sem->object_field_start; - json_ofield_action oend = sem->object_field_end; - bool isnull; - JsonTokenType tok; - - if (lex_peek(lex) != JSON_TOKEN_STRING) - return report_parse_error(JSON_PARSE_STRING, lex); - if ((ostart != NULL || oend != NULL) && lex->strval != NULL) - fname = pstrdup(lex->strval->data); - INSIST(json_lex(lex)); - - INSIST(lex_expect(JSON_PARSE_OBJECT_LABEL, lex, JSON_TOKEN_COLON)); - - tok = lex_peek(lex); - isnull = tok == JSON_TOKEN_NULL; - - if (ostart != NULL) - (*ostart) (sem->semstate, fname, isnull); - - switch (tok) - { - case JSON_TOKEN_OBJECT_START: - INSIST(parse_object(lex, sem)); - break; - case JSON_TOKEN_ARRAY_START: - INSIST(parse_array(lex, sem)); - break; - default: - INSIST(parse_scalar(lex, sem)); - } - - if (oend != NULL) - (*oend) (sem->semstate, fname, isnull); - return JSON_SUCCESS; -} - -static JsonParseErrorType -parse_object(JsonLexContext *lex, const JsonSemAction *sem) -{ - /* - * an object is a possibly empty sequence of object fields, separated by - * 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 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++; - - Assert(lex_peek(lex) == JSON_TOKEN_OBJECT_START); - INSIST(json_lex(lex)); - - tok = lex_peek(lex); - switch (tok) - { - case JSON_TOKEN_STRING: - INSIST(parse_object_field(lex, sem)); - while (lex_peek(lex) == JSON_TOKEN_COMMA) - { - INSIST(json_lex(lex)); - INSIST(parse_object_field(lex, sem)); - } - break; - case JSON_TOKEN_OBJECT_END: - break; - default: - /* case of an invalid initial token inside the object */ - return report_parse_error(JSON_PARSE_OBJECT_START, lex); - } - - INSIST(lex_expect(JSON_PARSE_OBJECT_NEXT, lex, JSON_TOKEN_OBJECT_END)); - - lex->lex_level--; - - if (oend != NULL) - (*oend) (sem->semstate); - return JSON_SUCCESS; -} - -static JsonParseErrorType -parse_array_element(JsonLexContext *lex, const JsonSemAction *sem) -{ - json_aelem_action astart = sem->array_element_start; - json_aelem_action aend = sem->array_element_end; - JsonTokenType tok = lex_peek(lex); - - bool isnull; - - isnull = tok == JSON_TOKEN_NULL; - - if (astart != NULL) - (*astart) (sem->semstate, isnull); - - /* an array element is any object, array or scalar */ - switch (tok) - { - case JSON_TOKEN_OBJECT_START: - INSIST(parse_object(lex, sem)); - break; - case JSON_TOKEN_ARRAY_START: - INSIST(parse_array(lex, sem)); - break; - default: - INSIST(parse_scalar(lex, sem)); - } - - if (aend != NULL) - (*aend) (sem->semstate, isnull); - return JSON_SUCCESS; -} - -static JsonParseErrorType -parse_array(JsonLexContext *lex, const JsonSemAction *sem) -{ - /* - * an array is a possibly empty sequence of array elements, separated by - * commas and surrounded by square brackets. - */ - json_struct_action astart = sem->array_start; - json_struct_action aend = sem->array_end; - - check_stack_depth(); - - if (astart != NULL) - (*astart) (sem->semstate); - - /* - * 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. - */ - lex->lex_level++; - - INSIST(lex_expect(JSON_PARSE_ARRAY_START, lex, JSON_TOKEN_ARRAY_START)); - if (lex_peek(lex) != JSON_TOKEN_ARRAY_END) - { - - INSIST(parse_array_element(lex, sem)); - - while (lex_peek(lex) == JSON_TOKEN_COMMA) - { - INSIST(json_lex(lex)); - INSIST(parse_array_element(lex, sem)); - } - } - - INSIST(lex_expect(JSON_PARSE_ARRAY_NEXT, lex, JSON_TOKEN_ARRAY_END)); - - lex->lex_level--; - - if (aend != NULL) - (*aend) (sem->semstate); - return JSON_SUCCESS; -} - -/* - * Lex one token from the input stream. - */ -JsonParseErrorType -json_lex(JsonLexContext *lex) -{ - char *s; - int len; - - /* Skip leading whitespace. */ - s = lex->token_terminator; - len = s - lex->input; - while (len < lex->input_length && - (*s == ' ' || *s == '\t' || *s == '\n' || *s == '\r')) - { - if (*s == '\n') - ++lex->line_number; - ++s; - ++len; - } - lex->token_start = s; - - /* Determine token type. */ - if (len >= lex->input_length) - { - lex->token_start = NULL; - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s; - lex->token_type = JSON_TOKEN_END; - } - else - switch (*s) - { - /* Single-character token, some kind of punctuation mark. */ - case '{': - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s + 1; - lex->token_type = JSON_TOKEN_OBJECT_START; - break; - case '}': - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s + 1; - lex->token_type = JSON_TOKEN_OBJECT_END; - break; - case '[': - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s + 1; - lex->token_type = JSON_TOKEN_ARRAY_START; - break; - case ']': - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s + 1; - lex->token_type = JSON_TOKEN_ARRAY_END; - break; - case ',': - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s + 1; - lex->token_type = JSON_TOKEN_COMMA; - break; - case ':': - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s + 1; - lex->token_type = JSON_TOKEN_COLON; - break; - case '"': - /* string */ - INSIST(json_lex_string(lex)); - lex->token_type = JSON_TOKEN_STRING; - break; - case '-': - /* Negative number. */ - INSIST(json_lex_number(lex, s + 1, NULL, NULL)); - 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. */ - INSIST(json_lex_number(lex, s, NULL, NULL)); - 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 - s < lex->input_length - len && 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; - } - - /* - * 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 */ - return JSON_SUCCESS; -} - -/* - * The next token in the input stream is known to be a string; lex it. - */ -static inline JsonParseErrorType -json_lex_string(JsonLexContext *lex) -{ - 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; - for (;;) - { - s++; - len++; - /* Premature end of the string. */ - if (len >= lex->input_length) - { - lex->token_terminator = s; - return JSON_INVALID_TOKEN; - } - else if (*s == '"') - break; - else if ((unsigned char) *s < 32) - { - /* Per RFC4627, these characters MUST be escaped. */ - /* Since *s isn't printable, exclude it from the context string */ - lex->token_terminator = s; - return JSON_ESCAPING_REQUIRED; - } - else if (*s == '\\') - { - /* OK, we have an escape character. */ - s++; - len++; - if (len >= lex->input_length) - { - lex->token_terminator = s; - return JSON_INVALID_TOKEN; - } - else if (*s == 'u') - { - int i; - int ch = 0; - - for (i = 1; i <= 4; i++) - { - s++; - len++; - if (len >= lex->input_length) - { - lex->token_terminator = s; - return JSON_INVALID_TOKEN; - } - else if (*s >= '0' && *s <= '9') - ch = (ch * 16) + (*s - '0'); - else if (*s >= 'a' && *s <= 'f') - ch = (ch * 16) + (*s - 'a') + 10; - else if (*s >= 'A' && *s <= 'F') - ch = (ch * 16) + (*s - 'A') + 10; - else - { - lex->token_terminator = s + pg_mblen(s); - return JSON_UNICODE_ESCAPE_FORMAT; - } - } - if (lex->strval != NULL) - { - char utf8str[5]; - int utf8len; - - if (ch >= 0xd800 && ch <= 0xdbff) - { - if (hi_surrogate != -1) - return JSON_UNICODE_HIGH_SURROGATE; - hi_surrogate = (ch & 0x3ff) << 10; - continue; - } - else if (ch >= 0xdc00 && ch <= 0xdfff) - { - if (hi_surrogate == -1) - return JSON_UNICODE_LOW_SURROGATE; - ch = 0x10000 + hi_surrogate + (ch & 0x3ff); - hi_surrogate = -1; - } - - if (hi_surrogate != -1) - return JSON_UNICODE_LOW_SURROGATE; - - /* - * 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 */ - return JSON_UNICODE_CODE_POINT_ZERO; - } - 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 - return JSON_UNICODE_HIGH_ESCAPE; - - } - } - else if (lex->strval != NULL) - { - if (hi_surrogate != -1) - return JSON_UNICODE_LOW_SURROGATE; - - switch (*s) - { - case '"': - case '\\': - case '/': - appendStringInfoChar(lex->strval, *s); - break; - case 'b': - appendStringInfoChar(lex->strval, '\b'); - break; - case 'f': - appendStringInfoChar(lex->strval, '\f'); - break; - case 'n': - appendStringInfoChar(lex->strval, '\n'); - break; - case 'r': - appendStringInfoChar(lex->strval, '\r'); - break; - case 't': - appendStringInfoChar(lex->strval, '\t'); - break; - default: - /* Not a valid string escape, so error out. */ - lex->token_terminator = s + pg_mblen(s); - return JSON_ESCAPING_INVALID; - } - } - else if (strchr("\"\\/bfnrt", *s) == NULL) - { - /* - * Simpler processing if we're not bothered about de-escaping - * - * It's very tempting to remove the strchr() call here and - * replace it with a switch statement, but testing so far has - * shown it's not a performance win. - */ - lex->token_terminator = s + pg_mblen(s); - return JSON_ESCAPING_INVALID; - } - - } - else if (lex->strval != NULL) - { - if (hi_surrogate != -1) - return JSON_UNICODE_LOW_SURROGATE; - appendStringInfoChar(lex->strval, *s); - } - - } - - if (hi_surrogate != -1) - return JSON_UNICODE_LOW_SURROGATE; - - /* Hooray, we found the end of the string! */ - lex->prev_token_terminator = lex->token_terminator; - lex->token_terminator = s + 1; - return JSON_SUCCESS; -} - -/* - * The next token in the input stream is known to be a number; lex it. - * - * In JSON, a number consists of four parts: - * - * (1) An optional minus sign ('-'). - * - * (2) Either a single '0', or a string of one or more digits that does not - * begin with a '0'. - * - * (3) An optional decimal part, consisting of a period ('.') followed by - * one or more digits. (Note: While this part can be omitted - * completely, it's not OK to have only the decimal point without - * any digits afterwards.) - * - * (4) An optional exponent part, consisting of 'e' or 'E', optionally - * followed by '+' or '-', followed by one or more digits. (Note: - * As with the decimal part, if 'e' or 'E' is present, it must be - * followed by at least one digit.) - * - * The 's' argument to this function points to the ostensible beginning - * of part 2 - i.e. the character after any optional minus sign, 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 JsonParseErrorType -json_lex_number(JsonLexContext *lex, char *s, - bool *num_err, int *total_len) -{ - bool error = false; - int 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 (len < lex->input_length && *s == '0') - { - s++; - len++; - } - else if (len < lex->input_length && *s >= '1' && *s <= '9') - { - do - { - s++; - len++; - } while (len < lex->input_length && *s >= '0' && *s <= '9'); - } - else - error = true; - - /* Part (3): parse optional decimal portion. */ - if (len < lex->input_length && *s == '.') - { - s++; - len++; - if (len == lex->input_length || *s < '0' || *s > '9') - error = true; - else - { - do - { - s++; - len++; - } while (len < lex->input_length && *s >= '0' && *s <= '9'); - } - } - - /* Part (4): parse optional exponent. */ - if (len < lex->input_length && (*s == 'e' || *s == 'E')) - { - s++; - len++; - if (len < lex->input_length && (*s == '+' || *s == '-')) - { - s++; - len++; - } - if (len == lex->input_length || *s < '0' || *s > '9') - error = true; - else - { - do - { - s++; - len++; - } while (len < lex->input_length && *s >= '0' && *s <= '9'); - } - } - - /* - * Check for trailing garbage. As in json_lex(), any alphanumeric stuff - * here should be considered part of the token for error-reporting - * purposes. - */ - for (; len < lex->input_length && JSON_ALPHANUMERIC_CHAR(*s); s++, len++) - error = true; - - 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) - return JSON_INVALID_TOKEN; - } - return JSON_SUCCESS; -} - -/* - * Report a parse error. - * - * lex->token_start and lex->token_terminator must identify the current token. - */ -static JsonParseErrorType -report_parse_error(JsonParseContext ctx, JsonLexContext *lex) -{ - /* Handle case where the input ended prematurely. */ - if (lex->token_start == NULL || lex->token_type == JSON_TOKEN_END) - return JSON_EXPECTED_MORE; - - switch (ctx) - { - case JSON_PARSE_END: - return JSON_EXPECTED_END; - case JSON_PARSE_VALUE: - return JSON_EXPECTED_JSON; - case JSON_PARSE_STRING: - return JSON_EXPECTED_STRING; - case JSON_PARSE_ARRAY_START: - return JSON_EXPECTED_ARRAY_FIRST; - case JSON_PARSE_ARRAY_NEXT: - return JSON_EXPECTED_ARRAY_NEXT; - case JSON_PARSE_OBJECT_START: - return JSON_EXPECTED_OBJECT_FIRST; - case JSON_PARSE_OBJECT_LABEL: - return JSON_EXPECTED_COLON; - case JSON_PARSE_OBJECT_NEXT: - return JSON_EXPECTED_OBJECT_NEXT; - case JSON_PARSE_OBJECT_COMMA: - return JSON_EXPECTED_STRING; - default: - return JSON_BAD_PARSER_STATE;; - } -} - /* * Determine how we want to print values of a given type in datum_to_json. * diff --git a/src/backend/utils/adt/jsonb.c b/src/backend/utils/adt/jsonb.c index 63072f616e..a5f2f3eeca 100644 --- a/src/backend/utils/adt/jsonb.c +++ b/src/backend/utils/adt/jsonb.c @@ -261,7 +261,7 @@ jsonb_from_cstring(char *json, int len) memset(&state, 0, sizeof(state)); memset(&sem, 0, sizeof(sem)); - lex = makeJsonLexContextCstringLen(json, len, true); + lex = makeJsonLexContextCstringLen(json, len, GetDatabaseEncoding(), true); sem.semstate = (void *) &state; diff --git a/src/backend/utils/adt/jsonfuncs.c b/src/backend/utils/adt/jsonfuncs.c index be5d30239d..bb22fd23ea 100644 --- a/src/backend/utils/adt/jsonfuncs.c +++ b/src/backend/utils/adt/jsonfuncs.c @@ -521,6 +521,7 @@ makeJsonLexContext(text *json, bool need_escapes) { return makeJsonLexContextCstringLen(VARDATA_ANY(json), VARSIZE_ANY_EXHDR(json), + GetDatabaseEncoding(), need_escapes); } @@ -1260,7 +1261,9 @@ get_array_start(void *state) _state->path_indexes[lex_level] != INT_MIN) { /* Negative subscript -- convert to positive-wise subscript */ - int nelements = json_count_array_elements(_state->lex); + int nelements; + + PARSE_OR_THROW(json_count_array_elements(_state->lex, &nelements), _state->lex); if (-_state->path_indexes[lex_level] <= nelements) _state->path_indexes[lex_level] += nelements; @@ -2725,7 +2728,7 @@ populate_array_json(PopulateArrayContext *ctx, char *json, int len) PopulateArrayState state; JsonSemAction sem; - state.lex = makeJsonLexContextCstringLen(json, len, true); + state.lex = makeJsonLexContextCstringLen(json, len, GetDatabaseEncoding(), true); state.ctx = ctx; memset(&sem, 0, sizeof(sem)); @@ -3568,7 +3571,7 @@ get_json_object_as_hash(char *json, int len, const char *funcname) HASHCTL ctl; HTAB *tab; JHashState *state; - JsonLexContext *lex = makeJsonLexContextCstringLen(json, len, true); + JsonLexContext *lex = makeJsonLexContextCstringLen(json, len, GetDatabaseEncoding(), true); JsonSemAction *sem; memset(&ctl, 0, sizeof(ctl)); diff --git a/src/common/jsonapi.c b/src/common/jsonapi.c index c14d6ff4f2..585d4842b2 100644 --- a/src/common/jsonapi.c +++ b/src/common/jsonapi.c @@ -10,14 +10,44 @@ * *------------------------------------------------------------------------- */ -#include "c.h" +#ifndef FRONTEND +#include "postgres.h" +#else +#include "postgres_fe.h" +#endif +#include "c.h" #include "common/jsonapi.h" #ifndef FRONTEND -#include "utils/palloc.h" +#include "miscadmin.h" #endif +#define INSIST(x) \ +do { \ + JsonParseErrorType parse_result; \ + if((parse_result = (x)) != JSON_SUCCESS) \ + return parse_result; \ +} while (0) + +/* + * The context of the parser is maintained by the recursive descent + * mechanism, but is passed explicitly to the error reporting routine + * for better diagnostics. + */ +typedef enum /* contexts of JSON parser */ +{ + JSON_PARSE_VALUE, /* expecting a value */ + JSON_PARSE_STRING, /* expecting a string (for a field name) */ + JSON_PARSE_ARRAY_START, /* saw '[', expecting value or ']' */ + JSON_PARSE_ARRAY_NEXT, /* saw array element, expecting ',' or ']' */ + JSON_PARSE_OBJECT_START, /* saw '{', expecting label or '}' */ + JSON_PARSE_OBJECT_LABEL, /* saw object label, expecting ':' */ + JSON_PARSE_OBJECT_NEXT, /* saw object value, expecting ',' or '}' */ + JSON_PARSE_OBJECT_COMMA, /* saw object ',', expecting next label */ + JSON_PARSE_END /* saw the end of a document, expect nothing */ +} JsonParseContext; + /* the null action object used for pure validation */ const JsonSemAction nullSemAction = { @@ -25,8 +55,19 @@ const JsonSemAction nullSemAction = NULL, NULL, NULL, NULL, NULL }; +static inline JsonParseErrorType lex_expect(JsonParseContext ctx, JsonLexContext *lex, JsonTokenType token) __attribute__((warn_unused_result)); +static inline JsonParseErrorType json_lex_string(JsonLexContext *lex) __attribute__((warn_unused_result)); +static inline JsonParseErrorType json_lex_number(JsonLexContext *lex, char *s, + bool *num_err, int *total_len) __attribute__((warn_unused_result)); +static inline JsonParseErrorType parse_scalar(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); +static JsonParseErrorType parse_object_field(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); +static JsonParseErrorType parse_object(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); +static JsonParseErrorType parse_array_element(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); +static JsonParseErrorType parse_array(JsonLexContext *lex, const JsonSemAction *sem) __attribute__((warn_unused_result)); +static JsonParseErrorType report_parse_error(JsonParseContext ctx, JsonLexContext *lex) __attribute__((warn_unused_result)); + JsonLexContext * -makeJsonLexContextCstringLen(char *json, int len, bool need_escapes) +makeJsonLexContextCstringLen(char *json, int len, pg_enc encoding, bool need_escapes) { JsonLexContext *lex; @@ -40,9 +81,900 @@ makeJsonLexContextCstringLen(char *json, int len, bool need_escapes) 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(); return lex; } +/* + * lex_peek + * + * what is the current look_ahead token? +*/ +JsonTokenType +lex_peek(JsonLexContext *lex) +{ + return lex->token_type; +} + +/* + * lex_except + * + * move the lexer to the next token if the current look_ahead token matches + * the parameter token. Otherwise, report an error. + */ +static inline JsonParseErrorType +lex_expect(JsonParseContext ctx, JsonLexContext *lex, JsonTokenType token) +{ + if (lex_peek(lex) == token) + INSIST(json_lex(lex)); + else + return report_parse_error(ctx, lex); + return JSON_SUCCESS; +} + +/* chars to consider as part of an alphanumeric token */ +#define JSON_ALPHANUMERIC_CHAR(c) \ + (((c) >= 'a' && (c) <= 'z') || \ + ((c) >= 'A' && (c) <= 'Z') || \ + ((c) >= '0' && (c) <= '9') || \ + (c) == '_' || \ + IS_HIGHBIT_SET(c)) + +/* + * 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; + } + + if (JSON_SUCCESS != json_lex_number(&dummy_lex, dummy_lex.input, &numeric_error, &total_len)) + return false; + + return (!numeric_error) && (total_len == dummy_lex.input_length); +} + +/* Recursive Descent parser support routines */ + +/* + * pg_parse_json + * + * Publicly visible entry point for the JSON parser. + * + * lex is a lexing context, set up for the json to be processed by calling + * makeJsonLexContext(). sem is a 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. + */ +JsonParseErrorType +pg_parse_json(JsonLexContext *lex, const JsonSemAction *sem) +{ + JsonTokenType tok; + + /* get the initial token */ + INSIST(json_lex(lex)); + + tok = lex_peek(lex); + + /* parse by recursive descent */ + switch (tok) + { + case JSON_TOKEN_OBJECT_START: + INSIST(parse_object(lex, sem)); + break; + case JSON_TOKEN_ARRAY_START: + INSIST(parse_array(lex, sem)); + break; + default: + INSIST(parse_scalar(lex, sem)); /* json can be a bare scalar */ + } + + INSIST(lex_expect(JSON_PARSE_END, lex, JSON_TOKEN_END)); + return JSON_SUCCESS; +} + +/* + * 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. + */ +JsonParseErrorType +json_count_array_elements(JsonLexContext *lex, int *elements) +{ + 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; + INSIST(lex_expect(JSON_PARSE_ARRAY_START, ©lex, JSON_TOKEN_ARRAY_START)); + if (lex_peek(©lex) != JSON_TOKEN_ARRAY_END) + { + while (1) + { + count++; + if (JSON_SUCCESS != parse_array_element(©lex, &nullSemAction)) + break; + if (copylex.token_type != JSON_TOKEN_COMMA) + break; + INSIST(json_lex(©lex)); + } + } + INSIST(lex_expect(JSON_PARSE_ARRAY_NEXT, ©lex, JSON_TOKEN_ARRAY_END)); + + *elements = count; + return JSON_SUCCESS; +} + +/* + * Recursive Descent parse routines. There is one for each structural + * element in a json document: + * - scalar (string, number, true, false, null) + * - array ( [ ] ) + * - array element + * - object ( { } ) + * - object field + */ +static inline JsonParseErrorType +parse_scalar(JsonLexContext *lex, const JsonSemAction *sem) +{ + char *val = NULL; + json_scalar_action sfunc = sem->scalar; + JsonTokenType tok = lex_peek(lex); + + /* a scalar must be a string, a number, true, false, or null */ + switch (tok) + { + case JSON_TOKEN_STRING: + case JSON_TOKEN_NUMBER: + case JSON_TOKEN_TRUE: + case JSON_TOKEN_FALSE: + case JSON_TOKEN_NULL: + break; + default: + return report_parse_error(JSON_PARSE_VALUE, lex); + } + + /* if no semantic function, just consume the token */ + if (sfunc == NULL) + { + INSIST(json_lex(lex)); + return JSON_SUCCESS; + } + + /* extract the de-escaped string value, or the raw lexeme */ + if (lex_peek(lex) == JSON_TOKEN_STRING) + { + if (lex->strval != NULL) + val = pstrdup(lex->strval->data); + } + else + { + int len = (lex->token_terminator - lex->token_start); + + val = palloc(len + 1); + memcpy(val, lex->token_start, len); + val[len] = '\0'; + } + + /* consume the token */ + INSIST(json_lex(lex)); + + /* invoke the callback */ + (*sfunc) (sem->semstate, val, tok); + return JSON_SUCCESS; +} + +static JsonParseErrorType +parse_object_field(JsonLexContext *lex, const JsonSemAction *sem) +{ + /* + * 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 */ + json_ofield_action ostart = sem->object_field_start; + json_ofield_action oend = sem->object_field_end; + bool isnull; + JsonTokenType tok; + + if (lex_peek(lex) != JSON_TOKEN_STRING) + return report_parse_error(JSON_PARSE_STRING, lex); + if ((ostart != NULL || oend != NULL) && lex->strval != NULL) + fname = pstrdup(lex->strval->data); + INSIST(json_lex(lex)); + + INSIST(lex_expect(JSON_PARSE_OBJECT_LABEL, lex, JSON_TOKEN_COLON)); + + tok = lex_peek(lex); + isnull = tok == JSON_TOKEN_NULL; + + if (ostart != NULL) + (*ostart) (sem->semstate, fname, isnull); + + switch (tok) + { + case JSON_TOKEN_OBJECT_START: + INSIST(parse_object(lex, sem)); + break; + case JSON_TOKEN_ARRAY_START: + INSIST(parse_array(lex, sem)); + break; + default: + INSIST(parse_scalar(lex, sem)); + } + + if (oend != NULL) + (*oend) (sem->semstate, fname, isnull); + return JSON_SUCCESS; +} + +static JsonParseErrorType +parse_object(JsonLexContext *lex, const JsonSemAction *sem) +{ + /* + * an object is a possibly empty sequence of object fields, separated by + * commas and surrounded by curly braces. + */ + json_struct_action ostart = sem->object_start; + json_struct_action oend = sem->object_end; + JsonTokenType tok; + +#ifndef FRONTEND + check_stack_depth(); +#else + /* TODO: What do we do in frontend code? */ +#endif + + if (ostart != NULL) + (*ostart) (sem->semstate); + + /* + * 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++; + + Assert(lex_peek(lex) == JSON_TOKEN_OBJECT_START); + INSIST(json_lex(lex)); + + tok = lex_peek(lex); + switch (tok) + { + case JSON_TOKEN_STRING: + INSIST(parse_object_field(lex, sem)); + while (lex_peek(lex) == JSON_TOKEN_COMMA) + { + INSIST(json_lex(lex)); + INSIST(parse_object_field(lex, sem)); + } + break; + case JSON_TOKEN_OBJECT_END: + break; + default: + /* case of an invalid initial token inside the object */ + return report_parse_error(JSON_PARSE_OBJECT_START, lex); + } + + INSIST(lex_expect(JSON_PARSE_OBJECT_NEXT, lex, JSON_TOKEN_OBJECT_END)); + + lex->lex_level--; + + if (oend != NULL) + (*oend) (sem->semstate); + return JSON_SUCCESS; +} + +static JsonParseErrorType +parse_array_element(JsonLexContext *lex, const JsonSemAction *sem) +{ + json_aelem_action astart = sem->array_element_start; + json_aelem_action aend = sem->array_element_end; + JsonTokenType tok = lex_peek(lex); + + bool isnull; + + isnull = tok == JSON_TOKEN_NULL; + + if (astart != NULL) + (*astart) (sem->semstate, isnull); + + /* an array element is any object, array or scalar */ + switch (tok) + { + case JSON_TOKEN_OBJECT_START: + INSIST(parse_object(lex, sem)); + break; + case JSON_TOKEN_ARRAY_START: + INSIST(parse_array(lex, sem)); + break; + default: + INSIST(parse_scalar(lex, sem)); + } + + if (aend != NULL) + (*aend) (sem->semstate, isnull); + return JSON_SUCCESS; +} + +static JsonParseErrorType +parse_array(JsonLexContext *lex, const JsonSemAction *sem) +{ + /* + * an array is a possibly empty sequence of array elements, separated by + * commas and surrounded by square brackets. + */ + json_struct_action astart = sem->array_start; + json_struct_action aend = sem->array_end; + +#ifndef FRONTEND + check_stack_depth(); +#else + /* TODO: What do we do in frontend code? */ +#endif + + if (astart != NULL) + (*astart) (sem->semstate); + + /* + * 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. + */ + lex->lex_level++; + + INSIST(lex_expect(JSON_PARSE_ARRAY_START, lex, JSON_TOKEN_ARRAY_START)); + if (lex_peek(lex) != JSON_TOKEN_ARRAY_END) + { + + INSIST(parse_array_element(lex, sem)); + + while (lex_peek(lex) == JSON_TOKEN_COMMA) + { + INSIST(json_lex(lex)); + INSIST(parse_array_element(lex, sem)); + } + } + + INSIST(lex_expect(JSON_PARSE_ARRAY_NEXT, lex, JSON_TOKEN_ARRAY_END)); + + lex->lex_level--; + + if (aend != NULL) + (*aend) (sem->semstate); + return JSON_SUCCESS; +} + +/* + * Lex one token from the input stream. + */ +JsonParseErrorType +json_lex(JsonLexContext *lex) +{ + char *s; + int len; + + /* Skip leading whitespace. */ + s = lex->token_terminator; + len = s - lex->input; + while (len < lex->input_length && + (*s == ' ' || *s == '\t' || *s == '\n' || *s == '\r')) + { + if (*s == '\n') + ++lex->line_number; + ++s; + ++len; + } + lex->token_start = s; + + /* Determine token type. */ + if (len >= lex->input_length) + { + lex->token_start = NULL; + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s; + lex->token_type = JSON_TOKEN_END; + } + else + switch (*s) + { + /* Single-character token, some kind of punctuation mark. */ + case '{': + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s + 1; + lex->token_type = JSON_TOKEN_OBJECT_START; + break; + case '}': + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s + 1; + lex->token_type = JSON_TOKEN_OBJECT_END; + break; + case '[': + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s + 1; + lex->token_type = JSON_TOKEN_ARRAY_START; + break; + case ']': + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s + 1; + lex->token_type = JSON_TOKEN_ARRAY_END; + break; + case ',': + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s + 1; + lex->token_type = JSON_TOKEN_COMMA; + break; + case ':': + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s + 1; + lex->token_type = JSON_TOKEN_COLON; + break; + case '"': + /* string */ + INSIST(json_lex_string(lex)); + lex->token_type = JSON_TOKEN_STRING; + break; + case '-': + /* Negative number. */ + INSIST(json_lex_number(lex, s + 1, NULL, NULL)); + 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. */ + INSIST(json_lex_number(lex, s, NULL, NULL)); + 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 - s < lex->input_length - len && 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; + } + + /* + * 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 */ + return JSON_SUCCESS; +} + +/* + * The next token in the input stream is known to be a string; lex it. + */ +static inline JsonParseErrorType +json_lex_string(JsonLexContext *lex) +{ + 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; + for (;;) + { + s++; + len++; + /* Premature end of the string. */ + if (len >= lex->input_length) + { + lex->token_terminator = s; + return JSON_INVALID_TOKEN; + } + else if (*s == '"') + break; + else if ((unsigned char) *s < 32) + { + /* Per RFC4627, these characters MUST be escaped. */ + /* Since *s isn't printable, exclude it from the context string */ + lex->token_terminator = s; + return JSON_ESCAPING_REQUIRED; + } + else if (*s == '\\') + { + /* OK, we have an escape character. */ + s++; + len++; + if (len >= lex->input_length) + { + lex->token_terminator = s; + return JSON_INVALID_TOKEN; + } + else if (*s == 'u') + { + int i; + int ch = 0; + + for (i = 1; i <= 4; i++) + { + s++; + len++; + if (len >= lex->input_length) + { + lex->token_terminator = s; + return JSON_INVALID_TOKEN; + } + else if (*s >= '0' && *s <= '9') + ch = (ch * 16) + (*s - '0'); + else if (*s >= 'a' && *s <= 'f') + ch = (ch * 16) + (*s - 'a') + 10; + else if (*s >= 'A' && *s <= 'F') + ch = (ch * 16) + (*s - 'A') + 10; + else + { + lex->token_terminator = s + pg_wchar_table[lex->input_encoding].mblen((const unsigned char *) s); + return JSON_UNICODE_ESCAPE_FORMAT; + } + } + if (lex->strval != NULL) + { + char utf8str[5]; + int utf8len; + + if (ch >= 0xd800 && ch <= 0xdbff) + { + if (hi_surrogate != -1) + return JSON_UNICODE_HIGH_SURROGATE; + hi_surrogate = (ch & 0x3ff) << 10; + continue; + } + else if (ch >= 0xdc00 && ch <= 0xdfff) + { + if (hi_surrogate == -1) + return JSON_UNICODE_LOW_SURROGATE; + ch = 0x10000 + hi_surrogate + (ch & 0x3ff); + hi_surrogate = -1; + } + + if (hi_surrogate != -1) + return JSON_UNICODE_LOW_SURROGATE; + + /* + * 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 */ + return JSON_UNICODE_CODE_POINT_ZERO; + } + else if (lex->input_encoding == 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 + return JSON_UNICODE_HIGH_ESCAPE; + + } + } + else if (lex->strval != NULL) + { + if (hi_surrogate != -1) + return JSON_UNICODE_LOW_SURROGATE; + + switch (*s) + { + case '"': + case '\\': + case '/': + appendStringInfoChar(lex->strval, *s); + break; + case 'b': + appendStringInfoChar(lex->strval, '\b'); + break; + case 'f': + appendStringInfoChar(lex->strval, '\f'); + break; + case 'n': + appendStringInfoChar(lex->strval, '\n'); + break; + case 'r': + appendStringInfoChar(lex->strval, '\r'); + break; + case 't': + appendStringInfoChar(lex->strval, '\t'); + break; + default: + /* Not a valid string escape, so error out. */ + lex->token_terminator = s + pg_wchar_table[lex->input_encoding].mblen((const unsigned char *) s); + return JSON_ESCAPING_INVALID; + } + } + else if (strchr("\"\\/bfnrt", *s) == NULL) + { + /* + * Simpler processing if we're not bothered about de-escaping + * + * It's very tempting to remove the strchr() call here and + * replace it with a switch statement, but testing so far has + * shown it's not a performance win. + */ + lex->token_terminator = s + pg_wchar_table[lex->input_encoding].mblen((const unsigned char *) s); + return JSON_ESCAPING_INVALID; + } + + } + else if (lex->strval != NULL) + { + if (hi_surrogate != -1) + return JSON_UNICODE_LOW_SURROGATE; + appendStringInfoChar(lex->strval, *s); + } + + } + + if (hi_surrogate != -1) + return JSON_UNICODE_LOW_SURROGATE; + + /* Hooray, we found the end of the string! */ + lex->prev_token_terminator = lex->token_terminator; + lex->token_terminator = s + 1; + return JSON_SUCCESS; +} + +/* + * The next token in the input stream is known to be a number; lex it. + * + * In JSON, a number consists of four parts: + * + * (1) An optional minus sign ('-'). + * + * (2) Either a single '0', or a string of one or more digits that does not + * begin with a '0'. + * + * (3) An optional decimal part, consisting of a period ('.') followed by + * one or more digits. (Note: While this part can be omitted + * completely, it's not OK to have only the decimal point without + * any digits afterwards.) + * + * (4) An optional exponent part, consisting of 'e' or 'E', optionally + * followed by '+' or '-', followed by one or more digits. (Note: + * As with the decimal part, if 'e' or 'E' is present, it must be + * followed by at least one digit.) + * + * The 's' argument to this function points to the ostensible beginning + * of part 2 - i.e. the character after any optional minus sign, 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 JsonParseErrorType +json_lex_number(JsonLexContext *lex, char *s, + bool *num_err, int *total_len) +{ + bool error = false; + int 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 (len < lex->input_length && *s == '0') + { + s++; + len++; + } + else if (len < lex->input_length && *s >= '1' && *s <= '9') + { + do + { + s++; + len++; + } while (len < lex->input_length && *s >= '0' && *s <= '9'); + } + else + error = true; + + /* Part (3): parse optional decimal portion. */ + if (len < lex->input_length && *s == '.') + { + s++; + len++; + if (len == lex->input_length || *s < '0' || *s > '9') + error = true; + else + { + do + { + s++; + len++; + } while (len < lex->input_length && *s >= '0' && *s <= '9'); + } + } + + /* Part (4): parse optional exponent. */ + if (len < lex->input_length && (*s == 'e' || *s == 'E')) + { + s++; + len++; + if (len < lex->input_length && (*s == '+' || *s == '-')) + { + s++; + len++; + } + if (len == lex->input_length || *s < '0' || *s > '9') + error = true; + else + { + do + { + s++; + len++; + } while (len < lex->input_length && *s >= '0' && *s <= '9'); + } + } + + /* + * Check for trailing garbage. As in json_lex(), any alphanumeric stuff + * here should be considered part of the token for error-reporting + * purposes. + */ + for (; len < lex->input_length && JSON_ALPHANUMERIC_CHAR(*s); s++, len++) + error = true; + + 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) + return JSON_INVALID_TOKEN; + } + return JSON_SUCCESS; +} + +/* + * Report a parse error. + * + * lex->token_start and lex->token_terminator must identify the current token. + */ +static JsonParseErrorType +report_parse_error(JsonParseContext ctx, JsonLexContext *lex) +{ + /* Handle case where the input ended prematurely. */ + if (lex->token_start == NULL || lex->token_type == JSON_TOKEN_END) + return JSON_EXPECTED_MORE; + + switch (ctx) + { + case JSON_PARSE_END: + return JSON_EXPECTED_END; + case JSON_PARSE_VALUE: + return JSON_EXPECTED_JSON; + case JSON_PARSE_STRING: + return JSON_EXPECTED_STRING; + case JSON_PARSE_ARRAY_START: + return JSON_EXPECTED_ARRAY_FIRST; + case JSON_PARSE_ARRAY_NEXT: + return JSON_EXPECTED_ARRAY_NEXT; + case JSON_PARSE_OBJECT_START: + return JSON_EXPECTED_OBJECT_FIRST; + case JSON_PARSE_OBJECT_LABEL: + return JSON_EXPECTED_COLON; + case JSON_PARSE_OBJECT_NEXT: + return JSON_EXPECTED_OBJECT_NEXT; + case JSON_PARSE_OBJECT_COMMA: + return JSON_EXPECTED_STRING; + default: + return JSON_BAD_PARSER_STATE;; + } +} diff --git a/src/include/common/jsonapi.h b/src/include/common/jsonapi.h index 162437193a..614764bcca 100644 --- a/src/include/common/jsonapi.h +++ b/src/include/common/jsonapi.h @@ -14,6 +14,7 @@ #ifndef JSONAPI_H #define JSONAPI_H +#include "common/pg_wchar.h" #include "lib/stringinfo.h" typedef enum @@ -73,6 +74,7 @@ typedef struct JsonLexContext { char *input; int input_length; + pg_enc input_encoding; char *token_start; char *token_terminator; char *prev_token_terminator; @@ -133,13 +135,17 @@ extern JsonParseErrorType pg_parse_json(JsonLexContext *lex, const JsonSemAction */ extern JsonParseErrorType json_lex(JsonLexContext *lex) __attribute__((warn_unused_result)); +/* + * Get the current look_ahead token. +*/ +extern JsonTokenType lex_peek(JsonLexContext *lex) __attribute__((warn_unused_result)); /* * json_count_array_elements performs a fast secondary parse to determine the * number of elements in passed array lex context. It should be called from an * array_start action. */ -extern int json_count_array_elements(JsonLexContext *lex); +extern JsonParseErrorType json_count_array_elements(JsonLexContext *lex, int *elements); /* * constructors for JsonLexContext, with or without strval element. @@ -152,6 +158,7 @@ extern int json_count_array_elements(JsonLexContext *lex); */ extern JsonLexContext *makeJsonLexContextCstringLen(char *json, int len, + pg_enc encoding, bool need_escapes); /* -- 2.21.1 (Apple Git-122.3)