filter-expr-parser: move operators to have dedicated tokens at the le…

…xer level Previously everything was captured as LL_IDENTIFIER, operators as well, and then mapped to tokens using the keyword mechanism. Since I'd like to make the LL_IDENTIFIER pattern stricter, that wouldn't allow typical operator characters into LL_IDENTIFIERS, so these should be recognized at the lexer level. NOTE: this does not include the spelled out operators like "lt" for "<", those remain LL_IDENTIFIER and are continued to be mapped using the keyword mechanism. Signed-off-by: Balazs Scheidler <[email protected]>
MrAnno · Jan 9, 2024 · eaa8a42 · eaa8a42
1 parent 1a5313b
commit eaa8a42
Show file tree

Hide file tree

Showing 4 changed files with 19 additions and 18 deletions.
diff --git a/lib/cfg-grammar.y b/lib/cfg-grammar.y
@@ -151,6 +151,16 @@
 /* this is a placeholder for unit tests, must be the latest & largest */
 %token LL_CONTEXT_MAX                 23
 
+/* operators in the filter language, the order of this determines precedence */
+%right KW_ASSIGN 9000
+%left  KW_OR 9001
+%left  KW_AND 9002
+%left  KW_STR_EQ 9003 KW_STR_NE 9004, KW_TA_EQ 9005, KW_TA_NE 9006, KW_TAV_EQ 9007, KW_TAV_NE 9008
+%left  KW_STR_LT 9009, KW_STR_LE 9010, KW_STR_GE, 9011 KW_STR_GT, 9012, KW_TA_LT 9013, KW_TA_LE 9014, KW_TA_GE 9015, KW_TA_GT 9016
+
+%left  '+' '-'
+%left  '*' '/'
+%left '.' KW_NOT 9017
 
 /* statements */
 %token KW_SOURCE                      10000

diff --git a/lib/cfg-lex.l b/lib/cfg-lex.l
@@ -305,6 +305,15 @@ word	[^ \#'"\(\)\{\}\[\]\\;\r\n\t,|\.@:]
 <INITIAL,filterx>\.\.      { return LL_DOTDOT; }
 <INITIAL,filterx>\.\.\.    { return LL_DOTDOTDOT; }
 <INITIAL,filterx>=>        { return LL_ARROW; }
+<INITIAL,filterx><	   { return KW_TA_LT; }
+<INITIAL,filterx><=        { return KW_TA_LE; }
+<INITIAL,filterx>==        { return KW_TA_EQ; }
+<INITIAL,filterx>!=        { return KW_TA_NE; }
+<INITIAL,filterx>>=        { return KW_TA_GE; }
+<INITIAL,filterx>>         { return KW_TA_GT; }
+<INITIAL,filterx>===       { return KW_TAV_EQ; }
+<INITIAL,filterx>!==       { return KW_TAV_NE; }
+
 <INITIAL,filterx>(-|\+)?{digit}+\.{digit}+  { yylval->fnum = strtod(yytext, NULL); return LL_FLOAT; }
 <INITIAL,filterx>0x{xdigit}+ {
                              if (!parse_int64_base16(yytext, &yylval->num))

diff --git a/lib/filter/filter-expr-grammar.ym b/lib/filter/filter-expr-grammar.ym
@@ -113,12 +113,6 @@ _translate_number_literals(CfgLexer *lexer, gint compare_mode, LogTemplate *expr
 %token KW_IN_LIST
 
 %left   ';'
-%left	KW_OR
-%left	KW_AND
-%left   KW_NOT
-%left   KW_STR_LT KW_STR_LE KW_STR_EQ KW_STR_NE KW_STR_GE KW_STR_GT
-%left   KW_TA_LT KW_TA_LE KW_TA_EQ KW_TA_NE KW_TA_GE KW_TA_GT
-%left   KW_TAV_EQ KW_TAV_NE
 
 %type	<node> filter_expr
 %type	<node> filter_simple_expr

diff --git a/lib/filter/filter-expr-parser.c b/lib/filter/filter-expr-parser.c
@@ -41,18 +41,6 @@ static CfgLexerKeyword filter_expr_keywords[] =
   { "ge",                 KW_STR_GE },
   { "gt",                 KW_STR_GT },
 
-  /* type aware comparisons */
-  { "<",                  KW_TA_LT },
-  { "<=",                 KW_TA_LE },
-  { "==",                 KW_TA_EQ },
-  { "!=",                 KW_TA_NE },
-  { ">=",                 KW_TA_GE },
-  { ">",                  KW_TA_GT },
-
-  /* equal type and value */
-  { "===",                KW_TAV_EQ },
-  { "!==",                KW_TAV_NE },
-
   /* filter expressions */
   { "severity",           KW_SEVERITY },
   { "level",              KW_SEVERITY },