diff options
Diffstat (limited to 'src/interfaces/ecpg/preproc/pgc.l')
| -rw-r--r-- | src/interfaces/ecpg/preproc/pgc.l | 128 |
1 files changed, 84 insertions, 44 deletions
diff --git a/src/interfaces/ecpg/preproc/pgc.l b/src/interfaces/ecpg/preproc/pgc.l index 27ba1c9375a..3bc4a0aa722 100644 --- a/src/interfaces/ecpg/preproc/pgc.l +++ b/src/interfaces/ecpg/preproc/pgc.l @@ -12,7 +12,7 @@ * * * IDENTIFICATION - * $Header: /cvsroot/pgsql/src/interfaces/ecpg/preproc/pgc.l,v 1.90 2002/04/05 11:39:45 momjian Exp $ + * $Header: /cvsroot/pgsql/src/interfaces/ecpg/preproc/pgc.l,v 1.91 2002/05/19 20:00:53 meskes Exp $ * *------------------------------------------------------------------------- */ @@ -49,7 +49,10 @@ static int literalalloc; /* current allocated buffer size */ #define startlit() (literalbuf[0] = '\0', literallen = 0) static void addlit(char *ytext, int yleng); +static void addlitchar (unsigned char); +static unsigned char unescape_single_char(unsigned char); +static char *token_start; int state_before; struct _yy_buffer @@ -131,10 +134,11 @@ xqstart {quote} xqstop {quote} xqdouble {quote}{quote} xqinside [^\\']+ -xqliteral [\\](.|\n) +xqescape [\\][^0-7] +xqoctesc [\\][0-7]{1,3} xqcat {quote}{whitespace_with_newline}{quote} -/* Delimited quote +/* Double quote * Allows embedded spaces and other special characters into identifiers. */ dquote \" @@ -229,7 +233,7 @@ non_newline [^\n\r] comment ("--"{non_newline}*) -whitespace ({space}|{comment}) +whitespace ({space}+|{comment}) /* * SQL92 requires at least one newline in the whitespace separating @@ -262,10 +266,7 @@ ip {ipdigit}\.{ipdigit}\.{ipdigit}\.{ipdigit} /* Take care of cpp continuation lines */ cppline {space}*#(.*\\{space})*.* -/* DO NOT PUT ANY COMMENTS IN THE FOLLOWING SECTION. - * AT&T lex does not properly handle C-style comments in this second lex block. - * So, put comments here. thomas - 1997-09-08 - * +/* * Quoted strings must allow some special characters such as single-quote * and newline. * Embedded single-quotes are implemented both in the SQL92-standard @@ -278,9 +279,16 @@ cppline {space}*#(.*\\{space})*.* */ %% + +%{ + /* code to execute during start of each call of yylex() */ + token_start = NULL; +%} + <SQL>{whitespace} { /* ignore */ } {xcstart} { + token_start = yytext; state_before = YYSTATE; xcdepth = 0; BEGIN(xc); @@ -298,7 +306,10 @@ cppline {space}*#(.*\\{space})*.* <xc>{xcstop} { ECHO; if (xcdepth <= 0) - BEGIN(state_before); + { + BEGIN(INITIAL); + token_start = NULL; + } else xcdepth--; } @@ -309,8 +320,10 @@ cppline {space}*#(.*\\{space})*.* <xc><<EOF>> { mmerror(PARSE_ERROR, ET_ERROR, "Unterminated /* comment"); } <SQL>{xbitstart} { + token_start = yytext; BEGIN(xbit); startlit(); + addlitchar('b'); } <xbit>{xbitstop} { BEGIN(SQL); @@ -327,6 +340,7 @@ cppline {space}*#(.*\\{space})*.* <xbit><<EOF>> { mmerror(PARSE_ERROR, ET_ERROR, "Unterminated bit string"); } <SQL>{xhstart} { + token_start = yytext; BEGIN(xh); startlit(); } @@ -351,6 +365,7 @@ cppline {space}*#(.*\\{space})*.* <xh><<EOF>> { mmerror(PARSE_ERROR, ET_ERROR, "Unterminated hexadecimal integer"); } <C,SQL>{xqstart} { + token_start = yytext; state_before = YYSTATE; BEGIN(xq); startlit(); @@ -360,9 +375,11 @@ cppline {space}*#(.*\\{space})*.* yylval.str = mm_strdup(literalbuf); return SCONST; } -<xq>{xqdouble} | -<xq>{xqinside} | -<xq>{xqliteral} { addlit(yytext, yyleng); } +<xq>{xqdouble} { addlitchar('\''); } +<xq>{xqinside} { addlit(yytext, yyleng); } +<xq>{xqescape} { addlitchar(unescape_single_char(yytext[1])); } +<xq>{xqoctesc} { unsigned char c = strtoul(yytext+1, NULL, 8); + addlitchar(c); } <xq>{xqcat} { /* ignore */ } <xq><<EOF>> { mmerror(PARSE_ERROR, ET_ERROR, "Unterminated quoted string"); } @@ -374,20 +391,13 @@ cppline {space}*#(.*\\{space})*.* } <xd>{xdstop} { BEGIN(state_before); - if (strlen(literalbuf) >= NAMEDATALEN) + if (literallen == 0) + mmerror(PARSE_ERROR, ET_ERROR, "zero-length delimited identifier"); + if (literallen >= NAMEDATALEN) { -#ifdef MULTIBYTE_NOTUSED - int len; - - len = pg_mbcliplen(literalbuf,strlen(literalbuf),NAMEDATALEN-1); - sprintf(errortext, "identifier \"%s\" will be truncated to \"%.*s\"", - literalbuf, len, literalbuf); - literalbuf[len] = '\0'; -#else sprintf(errortext, "identifier \"%s\" will be truncated to \"%.*s\"", literalbuf, NAMEDATALEN-1, literalbuf); literalbuf[NAMEDATALEN-1] = '\0'; -#endif mmerror(PARSE_ERROR, ET_WARNING, errortext); } @@ -399,7 +409,7 @@ cppline {space}*#(.*\\{space})*.* yylval.str = mm_strdup(literalbuf); return CSTRING; } -<xd>{xddouble} { addlit(yytext, yyleng-1); } +<xd>{xddouble} { addlitchar('"'); } <xd>{xdinside} { addlit(yytext, yyleng); } <xd,xdc><<EOF>> { mmerror(PARSE_ERROR, ET_ERROR, "Unterminated quoted identifier"); } <C,SQL>{xdstart} { @@ -426,8 +436,8 @@ cppline {space}*#(.*\\{space})*.* * character will match a prior rule, not this one. */ int nchars = yyleng; - char *slashstar = strstr((char*)yytext, "/*"); - char *dashdash = strstr((char*)yytext, "--"); + char *slashstar = strstr(yytext, "/*"); + char *dashdash = strstr(yytext, "--"); if (slashstar && dashdash) { @@ -438,7 +448,7 @@ cppline {space}*#(.*\\{space})*.* else if (!slashstar) slashstar = dashdash; if (slashstar) - nchars = slashstar - ((char*)yytext); + nchars = slashstar - yytext; /* * For SQL92 compatibility, '+' and '-' cannot be the @@ -480,14 +490,14 @@ cppline {space}*#(.*\\{space})*.* } /* Convert "!=" operator to "<>" for compatibility */ - if (strcmp((char*)yytext, "!=") == 0) + if (strcmp(yytext, "!=") == 0) yylval.str = mm_strdup("<>"); else - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return Op; } <SQL>{param} { - yylval.ival = atol((char*)&yytext[1]); + yylval.ival = atol(yytext+1); return PARAM; } <C,SQL>{integer} { @@ -504,26 +514,26 @@ cppline {space}*#(.*\\{space})*.* ) { errno = 0; - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return FCONST; } yylval.ival = val; return ICONST; } <SQL>{ip} { - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return IP; } {decimal} { - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return FCONST; } <C,SQL>{real} { - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return FCONST; } <SQL>:{identifier}(("->"|\.){identifier})* { - yylval.str = mm_strdup((char*)yytext+1); + yylval.str = mm_strdup(yytext+1); return(CVARIABLE); } <SQL>{identifier} { @@ -531,12 +541,12 @@ cppline {space}*#(.*\\{space})*.* struct _defines *ptr; /* Is it an SQL keyword? */ - keyword = ScanKeywordLookup((char*) yytext); + keyword = ScanKeywordLookup(yytext); if (keyword != NULL) return keyword->value; /* Is it an ECPG keyword? */ - keyword = ScanECPGKeywordLookup((char*) yytext); + keyword = ScanECPGKeywordLookup( yytext); if (keyword != NULL) return keyword->value; @@ -571,7 +581,7 @@ cppline {space}*#(.*\\{space})*.* */ if (ptr == NULL) { - yylval.str = mm_strdup((char*) yytext); + yylval.str = mm_strdup( yytext); return IDENT; } } @@ -586,19 +596,19 @@ cppline {space}*#(.*\\{space})*.* if (*endptr != '\0' || errno == ERANGE) { errno = 0; - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return SCONST; } return ICONST; } <C>{cppline} { - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return(CPP_LINE); } <C>{identifier} { ScanKeyword *keyword; - keyword = ScanCKeywordLookup((char*)yytext); + keyword = ScanCKeywordLookup(yytext); if (keyword != NULL) { return keyword->value; } @@ -627,7 +637,7 @@ cppline {space}*#(.*\\{space})*.* } if (ptr == NULL) { - yylval.str = mm_strdup((char*)yytext); + yylval.str = mm_strdup(yytext); return IDENT; } } @@ -731,7 +741,7 @@ cppline {space}*#(.*\\{space})*.* yytext[i+1] = '\0'; for ( defptr = defines; defptr != NULL && - ( strcmp((char*)yytext, defptr->old) != 0 ); defptr = defptr->next ); + ( strcmp(yytext, defptr->old) != 0 ); defptr = defptr->next ); preproc_tos++; stacked_if_value[preproc_tos].else_branch = FALSE; @@ -911,7 +921,37 @@ addlit(char *ytext, int yleng) literalbuf[literallen] = '\0'; } -int yywrap(void) +static void +addlitchar(unsigned char ychar) { - return 1; + /* enlarge buffer if needed */ + if ((literallen+1) >= literalalloc) + { + literalalloc *= 2; + literalbuf = (char *) realloc(literalbuf, literalalloc); + } + /* append new data, add trailing null */ + literalbuf[literallen] = ychar; + literallen += 1; + literalbuf[literallen] = '\0'; +} + +unsigned char +unescape_single_char(unsigned char c) +{ + switch (c) + { + case 'b': + return '\b'; + case 'f': + return '\f'; + case 'n': + return '\n'; + case 'r': + return '\r'; + case 't': + return '\t'; + default: + return c; + } } |