parse.y: parse_qmark

* parse.y (parse_qmark): extract from parser_yylex().

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46845 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

1 files changed, 86 insertions, 78 deletions

diff --git a/parse.y b/parse.y
index 081cb3e291..2f864e597d 100644
--- a/parse.y
+++ b/parse.y
@@ -7185,13 +7185,97 @@ parse_numeric(struct parser_params *parser, int c)
 }
 
 static int
+parse_qmark(struct parser_params *parser)
+{
+    rb_encoding *enc;
+    register int c;
+
+    if (IS_END()) {
+	lex_state = EXPR_VALUE;
+	return '?';
+    }
+    c = nextc();
+    if (c == -1) {
+	compile_error(PARSER_ARG "incomplete character syntax");
+	return 0;
+    }
+    if (rb_enc_isspace(c, current_enc)) {
+	if (!IS_ARG()) {
+	    int c2 = 0;
+	    switch (c) {
+	      case ' ':
+		c2 = 's';
+		break;
+	      case '\n':
+		c2 = 'n';
+		break;
+	      case '\t':
+		c2 = 't';
+		break;
+	      case '\v':
+		c2 = 'v';
+		break;
+	      case '\r':
+		c2 = 'r';
+		break;
+	      case '\f':
+		c2 = 'f';
+		break;
+	    }
+	    if (c2) {
+		rb_warnI("invalid character syntax; use ?\\%c", c2);
+	    }
+	}
+      ternary:
+	pushback(c);
+	lex_state = EXPR_VALUE;
+	return '?';
+    }
+    newtok();
+    enc = current_enc;
+    if (!parser_isascii()) {
+	if (tokadd_mbchar(c) == -1) return 0;
+    }
+    else if ((rb_enc_isalnum(c, current_enc) || c == '_') &&
+	     lex_p < lex_pend && is_identchar(lex_p, lex_pend, current_enc)) {
+	goto ternary;
+    }
+    else if (c == '\\') {
+	if (peek('u')) {
+	    nextc();
+	    c = parser_tokadd_utf8(parser, &enc, 0, 0, 0);
+	    if (0x80 <= c) {
+		tokaddmbc(c, enc);
+	    }
+	    else {
+		tokadd(c);
+	    }
+	}
+	else if (!lex_eol_p() && !(c = *lex_p, ISASCII(c))) {
+	    nextc();
+	    if (tokadd_mbchar(c) == -1) return 0;
+	}
+	else {
+	    c = read_escape(0, &enc);
+	    tokadd(c);
+	}
+    }
+    else {
+	tokadd(c);
+    }
+    tokfix();
+    set_yylval_str(STR_NEW3(tok(), toklen(), enc, 0));
+    lex_state = EXPR_END;
+    return tCHAR;
+}
+
+static int
 parser_yylex(struct parser_params *parser)
 {
     register int c;
     int space_seen = 0;
     int cmd_state;
     enum lex_state_e last_state;
-    rb_encoding *enc;
     int mb;
 #ifdef RIPPER
     int fallthru = FALSE;
@@ -7499,83 +7583,7 @@ parser_yylex(struct parser_params *parser)
 	return tSTRING_BEG;
 
       case '?':
-	if (IS_END()) {
-	    lex_state = EXPR_VALUE;
-	    return '?';
-	}
-	c = nextc();
-	if (c == -1) {
-	    compile_error(PARSER_ARG "incomplete character syntax");
-	    return 0;
-	}
-	if (rb_enc_isspace(c, current_enc)) {
-	    if (!IS_ARG()) {
-		int c2 = 0;
-		switch (c) {
-		  case ' ':
-		    c2 = 's';
-		    break;
-		  case '\n':
-		    c2 = 'n';
-		    break;
-		  case '\t':
-		    c2 = 't';
-		    break;
-		  case '\v':
-		    c2 = 'v';
-		    break;
-		  case '\r':
-		    c2 = 'r';
-		    break;
-		  case '\f':
-		    c2 = 'f';
-		    break;
-		}
-		if (c2) {
-		    rb_warnI("invalid character syntax; use ?\\%c", c2);
-		}
-	    }
-	  ternary:
-	    pushback(c);
-	    lex_state = EXPR_VALUE;
-	    return '?';
-	}
-	newtok();
-	enc = current_enc;
-	if (!parser_isascii()) {
-	    if (tokadd_mbchar(c) == -1) return 0;
-	}
-	else if ((rb_enc_isalnum(c, current_enc) || c == '_') &&
-		 lex_p < lex_pend && is_identchar(lex_p, lex_pend, current_enc)) {
-	    goto ternary;
-	}
-        else if (c == '\\') {
-            if (peek('u')) {
-                nextc();
-                c = parser_tokadd_utf8(parser, &enc, 0, 0, 0);
-                if (0x80 <= c) {
-                    tokaddmbc(c, enc);
-                }
-                else {
-                    tokadd(c);
-                }
-            }
-            else if (!lex_eol_p() && !(c = *lex_p, ISASCII(c))) {
-                nextc();
-                if (tokadd_mbchar(c) == -1) return 0;
-            }
-            else {
-                c = read_escape(0, &enc);
-                tokadd(c);
-            }
-        }
-        else {
-	    tokadd(c);
-        }
-	tokfix();
-	set_yylval_str(STR_NEW3(tok(), toklen(), enc, 0));
-	lex_state = EXPR_END;
-	return tCHAR;
+	return parse_qmark(parser);
 
       case '&':
 	if ((c = nextc()) == '&') {