Skip to content

P
Postgres FD Implementation
Commit 8b17298f authored by Alexander Korotkov's avatar Alexander Korotkov
Browse files
Options

Cosmetic changes for jsonpath_gram.y and jsonpath_scan.l 

This commit include formatting improvements, renamings and comments.  Also,
it makes jsonpath_scan.l be more uniform with other our lexers.  Firstly,
states names are renamed to more short alternatives.  Secondly, <INITIAL>
prefix removed from the rules.  Corresponding rules are moved to the tail, so
they would anyway work only in initial state.

Author: Alexander Korotkov
Reviewed-by: John Naylor
parent d303122e
Hide whitespace changes
Inline Side-by-side
Showing with 223 additions and 189 deletions
src/backend/utils/adt/jsonpath_gram.y
View file @ 8b17298f
...@@ -4,6 +4,8 @@ ...@@ -4,6 +4,8 @@
* jsonpath_gram.y * jsonpath_gram.y
* Grammar definitions for jsonpath datatype * Grammar definitions for jsonpath datatype
* *
* Transforms tokenized jsonpath into tree of JsonPathParseItem structs.
*
* Copyright (c) 2019, PostgreSQL Global Development Group * Copyright (c) 2019, PostgreSQL Global Development Group
* *
* IDENTIFICATION * IDENTIFICATION
...@@ -37,15 +39,17 @@ int jsonpath_yylex(union YYSTYPE *yylval_param); ...@@ -37,15 +39,17 @@ int jsonpath_yylex(union YYSTYPE *yylval_param);
int jsonpath_yyparse(JsonPathParseResult **result); int jsonpath_yyparse(JsonPathParseResult **result);
void jsonpath_yyerror(JsonPathParseResult **result, const char *message); void jsonpath_yyerror(JsonPathParseResult **result, const char *message);
static JsonPathParseItem *makeItemType(int type); static JsonPathParseItem *makeItemType(JsonPathItemType type);
static JsonPathParseItem *makeItemString(JsonPathString *s); static JsonPathParseItem *makeItemString(JsonPathString *s);
static JsonPathParseItem *makeItemVariable(JsonPathString *s); static JsonPathParseItem *makeItemVariable(JsonPathString *s);
static JsonPathParseItem *makeItemKey(JsonPathString *s); static JsonPathParseItem *makeItemKey(JsonPathString *s);
static JsonPathParseItem *makeItemNumeric(JsonPathString *s); static JsonPathParseItem *makeItemNumeric(JsonPathString *s);
static JsonPathParseItem *makeItemBool(bool val); static JsonPathParseItem *makeItemBool(bool val);
static JsonPathParseItem *makeItemBinary(int type, JsonPathParseItem *la, static JsonPathParseItem *makeItemBinary(JsonPathItemType type,
JsonPathParseItem *la,
JsonPathParseItem *ra); JsonPathParseItem *ra);
static JsonPathParseItem *makeItemUnary(int type, JsonPathParseItem *a); static JsonPathParseItem *makeItemUnary(JsonPathItemType type,
JsonPathParseItem *a);
static JsonPathParseItem *makeItemList(List *list); static JsonPathParseItem *makeItemList(List *list);
static JsonPathParseItem *makeIndexArray(List *list); static JsonPathParseItem *makeIndexArray(List *list);
static JsonPathParseItem *makeAny(int first, int last); static JsonPathParseItem *makeAny(int first, int last);
...@@ -75,9 +79,9 @@ static JsonPathParseItem *makeItemLikeRegex(JsonPathParseItem *expr, ...@@ -75,9 +79,9 @@ static JsonPathParseItem *makeItemLikeRegex(JsonPathParseItem *expr,
%union { %union {
JsonPathString str; JsonPathString str;
List *elems; /* list of JsonPathParseItem */ List *elems; /* list of JsonPathParseItem */
List *indexs; /* list of integers */ List *indexs; /* list of integers */
JsonPathParseItem *value; JsonPathParseItem *value;
JsonPathParseResult *result; JsonPathParseResult *result;
JsonPathItemType optype; JsonPathItemType optype;
bool boolean; bool boolean;
...@@ -160,7 +164,7 @@ comp_op: ...@@ -160,7 +164,7 @@ comp_op:
; ;
delimited_predicate: delimited_predicate:
'(' predicate ')' { $$ = $2; } '(' predicate ')' { $$ = $2; }
| EXISTS_P '(' expr ')' { $$ = makeItemUnary(jpiExists, $3); } | EXISTS_P '(' expr ')' { $$ = makeItemUnary(jpiExists, $3); }
; ;
...@@ -170,9 +174,10 @@ predicate: ...@@ -170,9 +174,10 @@ predicate:
| predicate AND_P predicate { $$ = makeItemBinary(jpiAnd, $1, $3); } | predicate AND_P predicate { $$ = makeItemBinary(jpiAnd, $1, $3); }
| predicate OR_P predicate { $$ = makeItemBinary(jpiOr, $1, $3); } | predicate OR_P predicate { $$ = makeItemBinary(jpiOr, $1, $3); }
| NOT_P delimited_predicate { $$ = makeItemUnary(jpiNot, $2); } | NOT_P delimited_predicate { $$ = makeItemUnary(jpiNot, $2); }
| '(' predicate ')' IS_P UNKNOWN_P { $$ = makeItemUnary(jpiIsUnknown, $2); } | '(' predicate ')' IS_P UNKNOWN_P
{ $$ = makeItemUnary(jpiIsUnknown, $2); }
| expr STARTS_P WITH_P starts_with_initial | expr STARTS_P WITH_P starts_with_initial
{ $$ = makeItemBinary(jpiStartsWith, $1, $4); } { $$ = makeItemBinary(jpiStartsWith, $1, $4); }
| expr LIKE_REGEX_P STRING_P { $$ = makeItemLikeRegex($1, &$3, NULL); } | expr LIKE_REGEX_P STRING_P { $$ = makeItemLikeRegex($1, &$3, NULL); }
| expr LIKE_REGEX_P STRING_P FLAG_P STRING_P | expr LIKE_REGEX_P STRING_P FLAG_P STRING_P
{ $$ = makeItemLikeRegex($1, &$3, &$5); } { $$ = makeItemLikeRegex($1, &$3, &$5); }
...@@ -232,7 +237,8 @@ any_level: ...@@ -232,7 +237,8 @@ any_level:
any_path: any_path:
ANY_P { $$ = makeAny(0, -1); } ANY_P { $$ = makeAny(0, -1); }
| ANY_P '{' any_level '}' { $$ = makeAny($3, $3); } | ANY_P '{' any_level '}' { $$ = makeAny($3, $3); }
| ANY_P '{' any_level TO_P any_level '}' { $$ = makeAny($3, $5); } | ANY_P '{' any_level TO_P any_level '}'
{ $$ = makeAny($3, $5); }
; ;
accessor_op: accessor_op:
...@@ -285,10 +291,15 @@ method: ...@@ -285,10 +291,15 @@ method:
; ;
%% %%
static JsonPathParseItem* /*
makeItemType(int type) * The helper functions below allocate and fill JsonPathParseItem's of various
* types.
*/
static JsonPathParseItem *
makeItemType(JsonPathItemType type)
{ {
JsonPathParseItem* v = palloc(sizeof(*v)); JsonPathParseItem *v = palloc(sizeof(*v));
CHECK_FOR_INTERRUPTS(); CHECK_FOR_INTERRUPTS();
...@@ -298,10 +309,10 @@ makeItemType(int type) ...@@ -298,10 +309,10 @@ makeItemType(int type)
return v; return v;
} }
static JsonPathParseItem* static JsonPathParseItem *
makeItemString(JsonPathString *s) makeItemString(JsonPathString *s)
{ {
JsonPathParseItem *v; JsonPathParseItem *v;
if (s == NULL) if (s == NULL)
{ {
...@@ -320,7 +331,7 @@ makeItemString(JsonPathString *s) ...@@ -320,7 +331,7 @@ makeItemString(JsonPathString *s)
static JsonPathParseItem * static JsonPathParseItem *
makeItemVariable(JsonPathString *s) makeItemVariable(JsonPathString *s)
{ {
JsonPathParseItem *v; JsonPathParseItem *v;
v = makeItemType(jpiVariable); v = makeItemType(jpiVariable);
v->value.string.val = s->val; v->value.string.val = s->val;
...@@ -332,7 +343,7 @@ makeItemVariable(JsonPathString *s) ...@@ -332,7 +343,7 @@ makeItemVariable(JsonPathString *s)
static JsonPathParseItem * static JsonPathParseItem *
makeItemKey(JsonPathString *s) makeItemKey(JsonPathString *s)
{ {
JsonPathParseItem *v; JsonPathParseItem *v;
v = makeItemString(s); v = makeItemString(s);
v->type = jpiKey; v->type = jpiKey;
...@@ -343,7 +354,7 @@ makeItemKey(JsonPathString *s) ...@@ -343,7 +354,7 @@ makeItemKey(JsonPathString *s)
static JsonPathParseItem * static JsonPathParseItem *
makeItemNumeric(JsonPathString *s) makeItemNumeric(JsonPathString *s)
{ {
JsonPathParseItem *v; JsonPathParseItem *v;
v = makeItemType(jpiNumeric); v = makeItemType(jpiNumeric);
v->value.numeric = v->value.numeric =
...@@ -356,7 +367,7 @@ makeItemNumeric(JsonPathString *s) ...@@ -356,7 +367,7 @@ makeItemNumeric(JsonPathString *s)
static JsonPathParseItem * static JsonPathParseItem *
makeItemBool(bool val) makeItemBool(bool val)
{ {
JsonPathParseItem *v = makeItemType(jpiBool); JsonPathParseItem *v = makeItemType(jpiBool);
v->value.boolean = val; v->value.boolean = val;
...@@ -364,7 +375,7 @@ makeItemBool(bool val) ...@@ -364,7 +375,7 @@ makeItemBool(bool val)
} }
static JsonPathParseItem * static JsonPathParseItem *
makeItemBinary(int type, JsonPathParseItem* la, JsonPathParseItem *ra) makeItemBinary(JsonPathItemType type, JsonPathParseItem *la, JsonPathParseItem *ra)
{ {
JsonPathParseItem *v = makeItemType(type); JsonPathParseItem *v = makeItemType(type);
...@@ -375,7 +386,7 @@ makeItemBinary(int type, JsonPathParseItem* la, JsonPathParseItem *ra) ...@@ -375,7 +386,7 @@ makeItemBinary(int type, JsonPathParseItem* la, JsonPathParseItem *ra)
} }
static JsonPathParseItem * static JsonPathParseItem *
makeItemUnary(int type, JsonPathParseItem* a) makeItemUnary(JsonPathItemType type, JsonPathParseItem *a)
{ {
JsonPathParseItem *v; JsonPathParseItem *v;
...@@ -401,8 +412,9 @@ makeItemUnary(int type, JsonPathParseItem* a) ...@@ -401,8 +412,9 @@ makeItemUnary(int type, JsonPathParseItem* a)
static JsonPathParseItem * static JsonPathParseItem *
makeItemList(List *list) makeItemList(List *list)
{ {
JsonPathParseItem *head, *end; JsonPathParseItem *head,
ListCell *cell = list_head(list); *end;
ListCell *cell = list_head(list);
head = end = (JsonPathParseItem *) lfirst(cell); head = end = (JsonPathParseItem *) lfirst(cell);
...@@ -427,8 +439,8 @@ makeItemList(List *list) ...@@ -427,8 +439,8 @@ makeItemList(List *list)
static JsonPathParseItem * static JsonPathParseItem *
makeIndexArray(List *list) makeIndexArray(List *list)
{ {
JsonPathParseItem *v = makeItemType(jpiIndexArray); JsonPathParseItem *v = makeItemType(jpiIndexArray);
ListCell *cell; ListCell *cell;
int i = 0; int i = 0;
Assert(list_length(list) > 0); Assert(list_length(list) > 0);
...@@ -439,7 +451,7 @@ makeIndexArray(List *list) ...@@ -439,7 +451,7 @@ makeIndexArray(List *list)
foreach(cell, list) foreach(cell, list)
{ {
JsonPathParseItem *jpi = lfirst(cell); JsonPathParseItem *jpi = lfirst(cell);
Assert(jpi->type == jpiSubscript); Assert(jpi->type == jpiSubscript);
...@@ -453,7 +465,7 @@ makeIndexArray(List *list) ...@@ -453,7 +465,7 @@ makeIndexArray(List *list)
static JsonPathParseItem * static JsonPathParseItem *
makeAny(int first, int last) makeAny(int first, int last)
{ {
JsonPathParseItem *v = makeItemType(jpiAny); JsonPathParseItem *v = makeItemType(jpiAny);
v->value.anybounds.first = (first >= 0) ? first : PG_UINT32_MAX; v->value.anybounds.first = (first >= 0) ? first : PG_UINT32_MAX;
v->value.anybounds.last = (last >= 0) ? last : PG_UINT32_MAX; v->value.anybounds.last = (last >= 0) ? last : PG_UINT32_MAX;
...@@ -465,9 +477,9 @@ static JsonPathParseItem * ...@@ -465,9 +477,9 @@ static JsonPathParseItem *
makeItemLikeRegex(JsonPathParseItem *expr, JsonPathString *pattern, makeItemLikeRegex(JsonPathParseItem *expr, JsonPathString *pattern,
JsonPathString *flags) JsonPathString *flags)
{ {
JsonPathParseItem *v = makeItemType(jpiLikeRegex); JsonPathParseItem *v = makeItemType(jpiLikeRegex);
int i; int i;
int cflags = REG_ADVANCED; int cflags = REG_ADVANCED;
v->value.like_regex.expr = expr; v->value.like_regex.expr = expr;
v->value.like_regex.pattern = pattern->val; v->value.like_regex.pattern = pattern->val;
...@@ -510,4 +522,12 @@ makeItemLikeRegex(JsonPathParseItem *expr, JsonPathString *pattern, ...@@ -510,4 +522,12 @@ makeItemLikeRegex(JsonPathParseItem *expr, JsonPathString *pattern,
return v; return v;
} }
/*
* jsonpath_scan.l is compiled as part of jsonpath_gram.y. Currently, this is
* unavoidable because jsonpath_gram does not create a .h file to export its
* token symbols. If these files ever grow large enough to be worth compiling
* separately, that could be fixed; but for now it seems like useless
* complication.
*/
#include "jsonpath_scan.c" #include "jsonpath_scan.c"
src/backend/utils/adt/jsonpath_scan.l
View file @ 8b17298f
...@@ -4,6 +4,9 @@ ...@@ -4,6 +4,9 @@
* jsonpath_scan.l * jsonpath_scan.l
* Lexical parser for jsonpath datatype * Lexical parser for jsonpath datatype
* *
* Splits jsonpath string into tokens represented as JsonPathString structs.
* Decodes unicode and hex escaped strings.
*
* Copyright (c) 2019, PostgreSQL Global Development Group * Copyright (c) 2019, PostgreSQL Global Development Group
* *
* IDENTIFICATION * IDENTIFICATION
...@@ -19,9 +22,6 @@ ...@@ -19,9 +22,6 @@
static JsonPathString scanstring; static JsonPathString scanstring;
/* No reason to constrain amount of data slurped */
/* #define YY_READ_BUF_SIZE 16777216 */
/* Handles to the buffer that the lexer uses internally */ /* Handles to the buffer that the lexer uses internally */
static YY_BUFFER_STATE scanbufhandle; static YY_BUFFER_STATE scanbufhandle;
static char *scanbuf; static char *scanbuf;
...@@ -29,9 +29,7 @@ static int scanbuflen; ...@@ -29,9 +29,7 @@ static int scanbuflen;
static void addstring(bool init, char *s, int l); static void addstring(bool init, char *s, int l);
static void addchar(bool init, char s); static void addchar(bool init, char s);
static int checkSpecialVal(void); /* examine scanstring for the special static enum yytokentype checkKeyword(void);
* value */
static void parseUnicode(char *s, int l); static void parseUnicode(char *s, int l);
static void parseHexChars(char *s, int l); static void parseHexChars(char *s, int l);
...@@ -60,11 +58,22 @@ fprintf_to_ereport(const char *fmt, const char *msg) ...@@ -60,11 +58,22 @@ fprintf_to_ereport(const char *fmt, const char *msg)
%option noyyrealloc %option noyyrealloc
%option noyyfree %option noyyfree
%x xQUOTED /*
%x xNONQUOTED * We use exclusive states for quoted, signle-quoted and non-quoted strings,
%x xVARQUOTED * quoted variable names and C-tyle comments.
%x xSINGLEQUOTED * Exclusive states:
%x xCOMMENT * <xq> - quoted strings
* <xnq> - non-quoted strings
* <xvq> - quoted variable names
* <xsq> - single-quoted strings
* <xc> - C-style comment
*/
%x xq
%x xnq
%x xvq
%x xsq
%x xc
special [\?\%\$\.\[\]\{\}\(\)\|\&\!\=\<\>\@\#\,\*:\-\+\/] special [\?\%\$\.\[\]\{\}\(\)\|\&\!\=\<\>\@\#\,\*:\-\+\/]
any [^\?\%\$\.\[\]\{\}\(\)\|\&\!\=\<\>\@\#\,\*:\-\+\/\\\"\' \t\n\r\f] any [^\?\%\$\.\[\]\{\}\(\)\|\&\!\=\<\>\@\#\,\*:\-\+\/\\\"\' \t\n\r\f]
...@@ -73,189 +82,188 @@ hex_dig [0-9A-Fa-f] ...@@ -73,189 +82,188 @@ hex_dig [0-9A-Fa-f]
unicode \\u({hex_dig}{4}|\{{hex_dig}{1,6}\}) unicode \\u({hex_dig}{4}|\{{hex_dig}{1,6}\})
hex_char \\x{hex_dig}{2} hex_char \\x{hex_dig}{2}
%% %%
<INITIAL>\&\& { return AND_P; } <xnq>{any}+ {
addstring(false, yytext, yyleng);
<INITIAL>\|\| { return OR_P; } }
<INITIAL>\! { return NOT_P; }
<INITIAL>\*\* { return ANY_P; }
<INITIAL>\< { return LESS_P; }
<INITIAL>\<\= { return LESSEQUAL_P; }
<INITIAL>\=\= { return EQUAL_P; }
<INITIAL>\<\> { return NOTEQUAL_P; }
<INITIAL>\!\= { return NOTEQUAL_P; } <xnq>{blank}+ {
yylval->str = scanstring;
BEGIN INITIAL;
return checkKeyword();
}
<INITIAL>\>\= { return GREATEREQUAL_P; }
<INITIAL>\> { return GREATER_P; } <xnq>\/\* {
yylval->str = scanstring;
BEGIN xc;
}
<INITIAL>\${any}+ { <xnq>({special}|\"|\') {
addstring(true, yytext + 1, yyleng - 1);
addchar(false, '\0');
yylval->str = scanstring; yylval->str = scanstring;
return VARIABLE_P; yyless(0);
BEGIN INITIAL;
return checkKeyword();
} }
<INITIAL>\$\" { <xnq><<EOF>> {
addchar(true, '\0'); yylval->str = scanstring;
BEGIN xVARQUOTED; BEGIN INITIAL;
return checkKeyword();
} }
<INITIAL>{special} { return *yytext; } <xnq,xq,xvq,xsq>\\[\"\'\\] { addchar(false, yytext[1]); }
<INITIAL>{blank}+ { /* ignore */ } <xnq,xq,xvq,xsq>\\b { addchar(false, '\b'); }
<INITIAL>\/\* { <xnq,xq,xvq,xsq>\\f { addchar(false, '\f'); }
addchar(true, '\0');
BEGIN xCOMMENT;
}
<INITIAL>[0-9]+(\.[0-9]+)?[eE][+-]?[0-9]+ /* float */ { <xnq,xq,xvq,xsq>\\n { addchar(false, '\n'); }
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return NUMERIC_P;
}
<INITIAL>\.[0-9]+[eE][+-]?[0-9]+ /* float */ { <xnq,xq,xvq,xsq>\\r { addchar(false, '\r'); }
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return NUMERIC_P;
}
<INITIAL>([0-9]+)?\.[0-9]+ { <xnq,xq,xvq,xsq>\\t { addchar(false, '\t'); }
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return NUMERIC_P;
}
<INITIAL>[0-9]+ { <xnq,xq,xvq,xsq>\\v { addchar(false, '\v'); }
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return INT_P;
}
<INITIAL>{any}+ { <xnq,xq,xvq,xsq>{unicode}+ { parseUnicode(yytext, yyleng); }
addstring(true, yytext, yyleng);
BEGIN xNONQUOTED;
}
<INITIAL>\" { <xnq,xq,xvq,xsq>{hex_char}+ { parseHexChars(yytext, yyleng); }
addchar(true, '\0');
BEGIN xQUOTED;
}
<INITIAL>\' { <xnq,xq,xvq,xsq>\\x { yyerror(NULL, "Hex character sequence is invalid"); }
addchar(true, '\0');
BEGIN xSINGLEQUOTED;
}
<INITIAL>\\ { <xnq,xq,xvq,xsq>\\u { yyerror(NULL, "Unicode sequence is invalid"); }
yyless(0);
addchar(true, '\0');
BEGIN xNONQUOTED;
}
<xNONQUOTED>{any}+ { <xnq,xq,xvq,xsq>\\. { yyerror(NULL, "Escape sequence is invalid"); }
addstring(false, yytext, yyleng);
}
<xNONQUOTED>{blank}+ { <xnq,xq,xvq,xsq>\\ { yyerror(NULL, "Unexpected end after backslash"); }
yylval->str = scanstring;
BEGIN INITIAL;
return checkSpecialVal();
}
<xq,xvq,xsq><<EOF>> { yyerror(NULL, "Unexpected end of quoted string"); }
<xNONQUOTED>\/\* { <xq>\" {
yylval->str = scanstring; yylval->str = scanstring;
BEGIN xCOMMENT; BEGIN INITIAL;
return STRING_P;
} }
<xNONQUOTED>({special}|\"|\') { <xvq>\" {
yylval->str = scanstring; yylval->str = scanstring;
yyless(0);
BEGIN INITIAL; BEGIN INITIAL;
return checkSpecialVal(); return VARIABLE_P;
} }
<xNONQUOTED><<EOF>> { <xsq>\' {
yylval->str = scanstring; yylval->str = scanstring;
BEGIN INITIAL; BEGIN INITIAL;
return checkSpecialVal(); return STRING_P;
} }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\[\"\'\\] { addchar(false, yytext[1]); } <xq,xvq>[^\\\"]+ { addstring(false, yytext, yyleng); }
<xsq>[^\\\']+ { addstring(false, yytext, yyleng); }
<xc>\*\/ { BEGIN INITIAL; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\b { addchar(false, '\b'); } <xc>[^\*]+ { }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\f { addchar(false, '\f'); } <xc>\* { }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\n { addchar(false, '\n'); } <xc><<EOF>> { yyerror(NULL, "Unexpected end of comment"); }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\r { addchar(false, '\r'); } \&\& { return AND_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\t { addchar(false, '\t'); } \|\| { return OR_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\v { addchar(false, '\v'); } \! { return NOT_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>{unicode}+ { parseUnicode(yytext, yyleng); } \*\* { return ANY_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>{hex_char}+ { parseHexChars(yytext, yyleng); } \< { return LESS_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\x { yyerror(NULL, "Hex character sequence is invalid"); } \<\= { return LESSEQUAL_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\u { yyerror(NULL, "Unicode sequence is invalid"); } \=\= { return EQUAL_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\. { yyerror(NULL, "Escape sequence is invalid"); } \<\> { return NOTEQUAL_P; }
<xNONQUOTED,xQUOTED,xVARQUOTED,xSINGLEQUOTED>\\ { yyerror(NULL, "Unexpected end after backslash"); } \!\= { return NOTEQUAL_P; }
<xQUOTED,xVARQUOTED,xSINGLEQUOTED><<EOF>> { yyerror(NULL, "Unexpected end of quoted string"); } \>\= { return GREATEREQUAL_P; }
<xQUOTED>\" { \> { return GREATER_P; }
\${any}+ {
addstring(true, yytext + 1, yyleng - 1);
addchar(false, '\0');
yylval->str = scanstring; yylval->str = scanstring;
BEGIN INITIAL; return VARIABLE_P;
return STRING_P; }
\$\" {
addchar(true, '\0');
BEGIN xvq;
} }
<xVARQUOTED>\" { {special} { return *yytext; }
{blank}+ { /* ignore */ }
\/\* {
addchar(true, '\0');
BEGIN xc;
}
[0-9]+(\.[0-9]+)?[eE][+-]?[0-9]+ { /* float */
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring; yylval->str = scanstring;
BEGIN INITIAL; return NUMERIC_P;
return VARIABLE_P;
} }
<xSINGLEQUOTED>\' { \.[0-9]+[eE][+-]?[0-9]+ { /* float */
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring; yylval->str = scanstring;
BEGIN INITIAL; return NUMERIC_P;
return STRING_P;
} }
<xQUOTED,xVARQUOTED>[^\\\"]+ { addstring(false, yytext, yyleng); } ([0-9]+)?\.[0-9]+ {
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return NUMERIC_P;
}
<xSINGLEQUOTED>[^\\\']+ { addstring(false, yytext, yyleng); } [0-9]+ {
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return INT_P;
}
<INITIAL><<EOF>> { yyterminate(); } {any}+ {
addstring(true, yytext, yyleng);
BEGIN xnq;
}
<xCOMMENT>\*\/ { BEGIN INITIAL; } \" {
addchar(true, '\0');
BEGIN xq;
}
<xCOMMENT>[^\*]+ { } \' {
addchar(true, '\0');
BEGIN xsq;
}
<xCOMMENT>\* { } \\ {
yyless(0);
addchar(true, '\0');
BEGIN xnq;
}
<xCOMMENT><<EOF>> { yyerror(NULL, "Unexpected end of comment"); } <<EOF>> { yyterminate(); }
%% %%
...@@ -292,7 +300,6 @@ typedef struct JsonPathKeyword ...@@ -292,7 +300,6 @@ typedef struct JsonPathKeyword
* Array of key words should be sorted by length and then * Array of key words should be sorted by length and then
* alphabetical order * alphabetical order
*/ */
static const JsonPathKeyword keywords[] = { static const JsonPathKeyword keywords[] = {
{ 2, false, IS_P, "is"}, { 2, false, IS_P, "is"},
{ 2, false, TO_P, "to"}, { 2, false, TO_P, "to"},
...@@ -317,8 +324,9 @@ static const JsonPathKeyword keywords[] = { ...@@ -317,8 +324,9 @@ static const JsonPathKeyword keywords[] = {
{ 10,false, LIKE_REGEX_P, "like_regex"}, { 10,false, LIKE_REGEX_P, "like_regex"},
}; };
static int /* Check if current scanstring value is a keyword */
checkSpecialVal() static enum yytokentype
checkKeyword()
{ {
int res = IDENT_P; int res = IDENT_P;
int diff; int diff;
...@@ -329,7 +337,7 @@ checkSpecialVal() ...@@ -329,7 +337,7 @@ checkSpecialVal()
if (scanstring.len > keywords[lengthof(keywords) - 1].len) if (scanstring.len > keywords[lengthof(keywords) - 1].len)
return res; return res;
while(StopLow < StopHigh) while (StopLow < StopHigh)
{ {
StopMiddle = StopLow + ((StopHigh - StopLow) >> 1); StopMiddle = StopLow + ((StopHigh - StopLow) >> 1);
...@@ -397,49 +405,50 @@ jsonpath_scanner_finish(void) ...@@ -397,49 +405,50 @@ jsonpath_scanner_finish(void)
pfree(scanbuf); pfree(scanbuf);
} }
/*
* Resize scanstring so that it can append string of given length.
* Reinitialize if required.
*/
static void static void
addstring(bool init, char *s, int l) resizeString(bool init, int appendLen)
{ {
if (init) if (init)
{ {
scanstring.total = 32; scanstring.total = Max(32, appendLen);
scanstring.val = palloc(scanstring.total); scanstring.val = (char *) palloc(scanstring.total);
scanstring.len = 0; scanstring.len = 0;
} }
else
if (s && l)
{ {
while(scanstring.len + l + 1 >= scanstring.total) if (scanstring.len + appendLen >= scanstring.total)
{ {
scanstring.total *= 2; while (scanstring.len + appendLen >= scanstring.total)
scanstring.total *= 2;
scanstring.val = repalloc(scanstring.val, scanstring.total); scanstring.val = repalloc(scanstring.val, scanstring.total);
} }
memcpy(scanstring.val + scanstring.len, s, l);
scanstring.len += l;
} }
} }
/* Add set of bytes at "s" of length "l" to scanstring */
static void static void
addchar(bool init, char s) addstring(bool init, char *s, int l)
{ {
if (init) resizeString(init, l + 1);
{ memcpy(scanstring.val + scanstring.len, s, l);
scanstring.total = 32; scanstring.len += l;
scanstring.val = palloc(scanstring.total); }
scanstring.len = 0;
}
else if(scanstring.len + 1 >= scanstring.total)
{
scanstring.total *= 2;
scanstring.val = repalloc(scanstring.val, scanstring.total);
}
scanstring.val[ scanstring.len ] = s; /* Add single byte "c" to scanstring */
if (s != '\0') static void
addchar(bool init, char c)
{
resizeString(init, 1);
scanstring.val[scanstring.len] = c;
if (c != '\0')
scanstring.len++; scanstring.len++;
} }
/* Interface to jsonpath parser */
JsonPathParseResult * JsonPathParseResult *
parsejsonpath(const char *str, int len) parsejsonpath(const char *str, int len)
{ {
...@@ -447,7 +456,7 @@ parsejsonpath(const char *str, int len) ...@@ -447,7 +456,7 @@ parsejsonpath(const char *str, int len)
jsonpath_scanner_init(str, len); jsonpath_scanner_init(str, len);
if (jsonpath_yyparse((void*)&parseresult) != 0) if (jsonpath_yyparse((void *) &parseresult) != 0)
jsonpath_yyerror(NULL, "bugus input"); jsonpath_yyerror(NULL, "bugus input");
jsonpath_scanner_finish(); jsonpath_scanner_finish();
...@@ -455,6 +464,7 @@ parsejsonpath(const char *str, int len) ...@@ -455,6 +464,7 @@ parsejsonpath(const char *str, int len)
return parseresult; return parseresult;
} }
/* Turn hex character into integer */
static int static int
hexval(char c) hexval(char c)
{ {
...@@ -468,6 +478,7 @@ hexval(char c) ...@@ -468,6 +478,7 @@ hexval(char c)
return 0; /* not reached */ return 0; /* not reached */
} }
/* Add given unicode character to scanstring */
static void static void
addUnicodeChar(int ch) addUnicodeChar(int ch)
{ {
...@@ -515,6 +526,7 @@ addUnicodeChar(int ch) ...@@ -515,6 +526,7 @@ addUnicodeChar(int ch)
} }
} }
/* Add unicode character and process its hi surrogate */
static void static void
addUnicode(int ch, int *hi_surrogate) addUnicode(int ch, int *hi_surrogate)
{ {
...@@ -592,6 +604,7 @@ parseUnicode(char *s, int l) ...@@ -592,6 +604,7 @@ parseUnicode(char *s, int l)
} }
} }
/* Parse sequence of hex-encoded characters */
static void static void
parseHexChars(char *s, int l) parseHexChars(char *s, int l)
{ {
...@@ -601,7 +614,8 @@ parseHexChars(char *s, int l) ...@@ -601,7 +614,8 @@ parseHexChars(char *s, int l)
for (i = 0; i < l / 4; i++) for (i = 0; i < l / 4; i++)
{ {
int ch = (hexval(s[i * 4 + 2]) << 4) | hexval(s[i * 4 + 3]); int ch = (hexval(s[i * 4 + 2]) << 4) |
hexval(s[i * 4 + 3]);
addUnicodeChar(ch); addUnicodeChar(ch);
} }
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment