database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-07-15 19:21:59 +03:00
Code Activity

Get rid of backtracking in jsonpath_scan.l

Browse Source 
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
This commit is contained in:
Alexander Korotkov
2019-03-25 15:43:56 +03:00
parent 8b17298f0b
commit 1d88a75c42
11 changed files with 800 additions and 29 deletions
Download Patch File Download Diff File Expand all files Collapse all files

66
src/backend/utils/adt/jsonpath_scan.l
View File

@ -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>\\x { yyerror(NULL, "Hex character sequence is invalid"); }
<xnq,xq,xvq,xsq>{unicode}*{unicodefail} { yyerror(NULL, "Unicode sequence is invalid"); }
<xnq,xq,xvq,xsq>\\u { yyerror(NULL, "Unicode sequence is invalid"); }
<xnq,xq,xvq,xsq>{hex_fail} { yyerror(NULL, "Hex character 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]+ {
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return NUMERIC_P;
}
[0-9]+ {
{integer} {
addstring(true, yytext, yyleng);
addchar(false, '\0');
yylval->str = scanstring;
return INT_P;
}
{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;
int ch = (hexval(s[2]) << 4) |
hexval(s[3]);
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]);
addUnicodeChar(ch);
}
addUnicodeChar(ch);
}
/*