xref: /OpenGrok/opengrok-indexer/src/main/jflex/analysis/lisp/LispSymbolTokenizer.lex (revision d219b4cea555a12b602d2d5518daa22134ad4879)
1 /*
2  * CDDL HEADER START
3  *
4  * The contents of this file are subject to the terms of the
5  * Common Development and Distribution License (the "License").
6  * You may not use this file except in compliance with the License.
7  *
8  * See LICENSE.txt included in this distribution for the specific
9  * language governing permissions and limitations under the License.
10  *
11  * When distributing Covered Code, include this CDDL HEADER in each
12  * file and include the License file at LICENSE.txt.
13  * If applicable, add the following below this CDDL HEADER, with the
14  * fields enclosed by brackets "[]" replaced with your own identifying
15  * information: Portions Copyright [yyyy] [name of copyright owner]
16  *
17  * CDDL HEADER END
18  */
19 
20 /*
21  * Copyright (c) 2006, 2018, Oracle and/or its affiliates. All rights reserved.
22  * Portions Copyright (c) 2017-2018, Chris Fraire <cfraire@me.com>.
23  */
24 
25 /*
26  * Gets Lisp symbols - ignores comments, strings, keywords
27  */
28 
29 package org.opengrok.indexer.analysis.lisp;
30 
31 import java.util.Locale;
32 import org.opengrok.indexer.analysis.JFlexSymbolMatcher;
33 %%
34 %public
35 %class LispSymbolTokenizer
36 %extends JFlexSymbolMatcher
37 %unicode
38 %int
39 %include ../CommonLexer.lexh
40 %char
41 
42 %{
43     private int nestedComment;
44 %}
45 
46 Identifier = [\-\+\*\!\@\$\%\&\/\?\.\,\:\{\}\=a-zA-Z0-9_\<\>]+
47 
48 %state STRING COMMENT SCOMMENT
49 
50 %%
51 
52 <YYINITIAL> {
53 {Identifier} {String id = yytext();
54               if (!Consts.kwd.contains(id.toLowerCase(Locale.ROOT))) {
55                         onSymbolMatched(id, yychar);
56                         return yystate(); }
57               }
58  \"     { yybegin(STRING); }
59 ";"     { yybegin(SCOMMENT); }
60 }
61 
62 <STRING> {
63  \"     { yybegin(YYINITIAL); }
64 \\\\ | \\\"     {}
65 }
66 
67 <YYINITIAL, COMMENT> {
68  "#|"    { yybegin(COMMENT); ++nestedComment; }
69 }
70 
71 <COMMENT> {
72 "|#"    { if (--nestedComment == 0) { yybegin(YYINITIAL); } }
73 }
74 
75 <SCOMMENT> {
76 \n      { yybegin(YYINITIAL);}
77 }
78 
79 <YYINITIAL, STRING, COMMENT, SCOMMENT> {
80 [^]    {}
81 }
82