Get rid of backtracking in jsonpath_scan.l

Non-backtracking flex parsers work faster than backtracking ones.  So, this
commit gets rid of backtracking in jsonpath_scan.l.  That required explicit
handling of some cases as well as manual backtracking for some cases.  More
regression tests for numerics are added.

Discussion: https://mail.google.com/mail/u/0?ik=a20b091faa&view=om&permmsgid=msg-f%3A1628425344167939063
Author: John Naylor, Nikita Gluknov, Alexander Korotkov

Author: akorotkov

src/backend/utils/adt/Makefile

@@ -34,6 +34,7 @@ OBJS = acl.o amutils.o arrayfuncs.o array_expanded.o array_selfuncs.o \
 	windowfuncs.o xid.o xml.o
 
 jsonpath_scan.c: FLEXFLAGS = -CF -p -p
+jsonpath_scan.c: FLEX_NO_BACKUP=yes
 
 # Force these dependencies to be known even without dependency info built:
 jsonpath_gram.o: jsonpath_scan.c

src/backend/utils/adt/jsonpath_scan.l

@@ -31,7 +31,7 @@ static void addstring(bool init, char *s, int l);
 static void addchar(bool init, char s);
 static enum yytokentype checkKeyword(void);
 static void parseUnicode(char *s, int l);
-static void parseHexChars(char *s, int l);
+static void parseHexChar(char *s);
 
 /* Avoid exit() on fatal scanner errors (a bit ugly -- see yy_fatal_error) */
 #undef fprintf
@@ -78,9 +78,20 @@ fprintf_to_ereport(const char *fmt, const char *msg)
 special		 [\?\%\$\.\[\]\{\}\(\)\|\&\!\=\<\>\@\#\,\*:\-\+\/]
 any			[^\?\%\$\.\[\]\{\}\(\)\|\&\!\=\<\>\@\#\,\*:\-\+\/\\\"\' \t\n\r\f]
 blank		[ \t\n\r\f]
+
+digit		[0-9]
+integer		{digit}+
+decimal		{digit}*\.{digit}+
+decimalfail	{digit}+\.
+real		({integer}|{decimal})[Ee][-+]?{digit}+
+realfail1	({integer}|{decimal})[Ee]
+realfail2	({integer}|{decimal})[Ee][-+]
+
 hex_dig		[0-9A-Fa-f]
 unicode		\\u({hex_dig}{4}|\{{hex_dig}{1,6}\})
+unicodefail	\\u({hex_dig}{0,3}|\{{hex_dig}{0,6})
 hex_char	\\x{hex_dig}{2}
+hex_fail	\\x{hex_dig}{0,1}
 
 %%
 
@@ -129,11 +140,17 @@ hex_char	\\x{hex_dig}{2}
 
 <xnq,xq,xvq,xsq>{unicode}+		{ parseUnicode(yytext, yyleng); }
 
-<xnq,xq,xvq,xsq>{hex_char}+		{ parseHexChars(yytext, yyleng); }
+<xnq,xq,xvq,xsq>{hex_char}		{ parseHexChar(yytext); }
+
+<xnq,xq,xvq,xsq>{unicode}*{unicodefail}	{ yyerror(NULL, "Unicode sequence is invalid"); }
 
-<xnq,xq,xvq,xsq>\\x				{ yyerror(NULL, "Hex character sequence is invalid"); }
+<xnq,xq,xvq,xsq>{hex_fail}		{ yyerror(NULL, "Hex character sequence is invalid"); }
 
-<xnq,xq,xvq,xsq>\\u				{ yyerror(NULL, "Unicode sequence is invalid"); }
+<xnq,xq,xvq,xsq>{unicode}+\\	{
+									/* throw back the \\, and treat as unicode */
+									yyless(yyleng - 1);
+									parseUnicode(yytext, yyleng);
+								}
 
 <xnq,xq,xvq,xsq>\\.				{ yyerror(NULL, "Escape sequence is invalid"); }
 
@@ -214,34 +231,38 @@ hex_char	\\x{hex_dig}{2}
 									BEGIN xc;
 								}
 
-[0-9]+(\.[0-9]+)?[eE][+-]?[0-9]+ { /* float */
+{real}							{
 									addstring(true, yytext, yyleng);
 									addchar(false, '\0');
 									yylval->str = scanstring;
 									return NUMERIC_P;
 								}
 
-\.[0-9]+[eE][+-]?[0-9]+			{ /* float */
+{decimal}						{
 									addstring(true, yytext, yyleng);
 									addchar(false, '\0');
 									yylval->str = scanstring;
 									return NUMERIC_P;
 								}
 
-([0-9]+)?\.[0-9]+				{
+{integer}						{
 									addstring(true, yytext, yyleng);
 									addchar(false, '\0');
 									yylval->str = scanstring;
-									return NUMERIC_P;
+									return INT_P;
 								}
 
-[0-9]+							{
+{decimalfail}					{
+									/* throw back the ., and treat as integer */
+									yyless(yyleng - 1);
 									addstring(true, yytext, yyleng);
 									addchar(false, '\0');
 									yylval->str = scanstring;
 									return INT_P;
 								}
 
+({realfail1}|{realfail2})		{ yyerror(NULL, "Floating point number is invalid"); }
+
 {any}+							{
 									addstring(true, yytext, yyleng);
 									BEGIN xnq;
@@ -571,7 +592,7 @@ addUnicode(int ch, int *hi_surrogate)
 static void
 parseUnicode(char *s, int l)
 {
-	int			i;
+	int			i = 2;
 	int			hi_surrogate = -1;
 
 	for (i = 2; i < l; i += 2)	/* skip '\u' */
@@ -606,19 +627,12 @@ parseUnicode(char *s, int l)
 
 /* Parse sequence of hex-encoded characters */
 static void
-parseHexChars(char *s, int l)
+parseHexChar(char *s)
 {
-	int i;
-
-	Assert(l % 4 /* \xXX */ == 0);
-
-	for (i = 0; i < l / 4; i++)
-	{
-		int			ch = (hexval(s[i * 4 + 2]) << 4) |
-						  hexval(s[i * 4 + 3]);
+	int			ch = (hexval(s[2]) << 4) |
+					  hexval(s[3]);
 
-		addUnicodeChar(ch);
-	}
+	addUnicodeChar(ch);
 }
 
 /*

src/test/regress/expected/jsonb_jsonpath.out

@@ -1297,7 +1297,7 @@ select jsonb_path_query('null', 'true.type()');
  "boolean"
 (1 row)
 
-select jsonb_path_query('null', '123.type()');
+select jsonb_path_query('null', '(123).type()');
  jsonb_path_query 
 ------------------
  "number"

src/test/regress/expected/jsonpath.out

@@ -365,6 +365,18 @@ select '1.type()'::jsonpath;
  1.type()
 (1 row)
 
+select '(1).type()'::jsonpath;
+ jsonpath 
+----------
+ 1.type()
+(1 row)
+
+select '1.2.type()'::jsonpath;
+  jsonpath  
+------------
+ 1.2.type()
+(1 row)
+
 select '"aaa".type()'::jsonpath;
    jsonpath   
 --------------
@@ -804,3 +816,159 @@ select '$ ? (@.a < +10.1e+1)'::jsonpath;
  $?(@."a" < 101)
 (1 row)
 
+select '0'::jsonpath;
+ jsonpath 
+----------
+ 0
+(1 row)
+
+select '00'::jsonpath;
+ jsonpath 
+----------
+ 0
+(1 row)
+
+select '0.0'::jsonpath;
+ jsonpath 
+----------
+ 0.0
+(1 row)
+
+select '0.000'::jsonpath;
+ jsonpath 
+----------
+ 0.000
+(1 row)
+
+select '0.000e1'::jsonpath;
+ jsonpath 
+----------
+ 0.00
+(1 row)
+
+select '0.000e2'::jsonpath;
+ jsonpath 
+----------
+ 0.0
+(1 row)
+
+select '0.000e3'::jsonpath;
+ jsonpath 
+----------
+ 0
+(1 row)
+
+select '0.0010'::jsonpath;
+ jsonpath 
+----------
+ 0.0010
+(1 row)
+
+select '0.0010e-1'::jsonpath;
+ jsonpath 
+----------
+ 0.00010
+(1 row)
+
+select '0.0010e+1'::jsonpath;
+ jsonpath 
+----------
+ 0.010
+(1 row)
+
+select '0.0010e+2'::jsonpath;
+ jsonpath 
+----------
+ 0.10
+(1 row)
+
+select '1e'::jsonpath;
+ERROR:  bad jsonpath representation
+LINE 1: select '1e'::jsonpath;
+               ^
+DETAIL:  Floating point number is invalid at or near "1e"
+select '1.e'::jsonpath;
+ jsonpath 
+----------
+ 1."e"
+(1 row)
+
+select '1.2e'::jsonpath;
+ERROR:  bad jsonpath representation
+LINE 1: select '1.2e'::jsonpath;
+               ^
+DETAIL:  Floating point number is invalid at or near "1.2e"
+select '1.2.e'::jsonpath;
+ jsonpath 
+----------
+ 1.2."e"
+(1 row)
+
+select '(1.2).e'::jsonpath;
+ jsonpath 
+----------
+ 1.2."e"
+(1 row)
+
+select '1e3'::jsonpath;
+ jsonpath 
+----------
+ 1000
+(1 row)
+
+select '1.e3'::jsonpath;
+ jsonpath 
+----------
+ 1."e3"
+(1 row)
+
+select '1.e3.e'::jsonpath;
+  jsonpath  
+------------
+ 1."e3"."e"
+(1 row)
+
+select '1.e3.e4'::jsonpath;
+  jsonpath   
+-------------
+ 1."e3"."e4"
+(1 row)
+
+select '1.2e3'::jsonpath;
+ jsonpath 
+----------
+ 1200
+(1 row)
+
+select '1.2.e3'::jsonpath;
+ jsonpath 
+----------
+ 1.2."e3"
+(1 row)
+
+select '(1.2).e3'::jsonpath;
+ jsonpath 
+----------
+ 1.2."e3"
+(1 row)
+
+select '1..e'::jsonpath;
+ERROR:  bad jsonpath representation
+LINE 1: select '1..e'::jsonpath;
+               ^
+DETAIL:  syntax error, unexpected '.' at or near "."
+select '1..e3'::jsonpath;
+ERROR:  bad jsonpath representation
+LINE 1: select '1..e3'::jsonpath;
+               ^
+DETAIL:  syntax error, unexpected '.' at or near "."
+select '(1.).e'::jsonpath;
+ERROR:  bad jsonpath representation
+LINE 1: select '(1.).e'::jsonpath;
+               ^
+DETAIL:  syntax error, unexpected ')' at or near ")"
+select '(1.).e3'::jsonpath;
+ERROR:  bad jsonpath representation
+LINE 1: select '(1.).e3'::jsonpath;
+               ^
+DETAIL:  syntax error, unexpected ')' at or near ")"