xref: /OpenGrok/opengrok-indexer/src/main/jflex/analysis/document/TroffFullTokenizer.lex (revision d219b4cea555a12b602d2d5518daa22134ad4879)
1*d219b4ceSAdam Hornacek /*
2*d219b4ceSAdam Hornacek  * CDDL HEADER START
3*d219b4ceSAdam Hornacek  *
4*d219b4ceSAdam Hornacek  * The contents of this file are subject to the terms of the
5*d219b4ceSAdam Hornacek  * Common Development and Distribution License (the "License").
6*d219b4ceSAdam Hornacek  * You may not use this file except in compliance with the License.
7*d219b4ceSAdam Hornacek  *
8*d219b4ceSAdam Hornacek  * See LICENSE.txt included in this distribution for the specific
9*d219b4ceSAdam Hornacek  * language governing permissions and limitations under the License.
10*d219b4ceSAdam Hornacek  *
11*d219b4ceSAdam Hornacek  * When distributing Covered Code, include this CDDL HEADER in each
12*d219b4ceSAdam Hornacek  * file and include the License file at LICENSE.txt.
13*d219b4ceSAdam Hornacek  * If applicable, add the following below this CDDL HEADER, with the
14*d219b4ceSAdam Hornacek  * fields enclosed by brackets "[]" replaced with your own identifying
15*d219b4ceSAdam Hornacek  * information: Portions Copyright [yyyy] [name of copyright owner]
16*d219b4ceSAdam Hornacek  *
17*d219b4ceSAdam Hornacek  * CDDL HEADER END
18*d219b4ceSAdam Hornacek  */
19*d219b4ceSAdam Hornacek 
20*d219b4ceSAdam Hornacek /*
21*d219b4ceSAdam Hornacek  * Copyright (c) 2005, 2021, Oracle and/or its affiliates. All rights reserved.
22*d219b4ceSAdam Hornacek  * Portions Copyright (c) 2017-2018, Chris Fraire <cfraire@me.com>.
23*d219b4ceSAdam Hornacek  */
24*d219b4ceSAdam Hornacek 
25*d219b4ceSAdam Hornacek package org.opengrok.indexer.analysis.document;
26*d219b4ceSAdam Hornacek 
27*d219b4ceSAdam Hornacek import java.util.Locale;
28*d219b4ceSAdam Hornacek import org.opengrok.indexer.analysis.JFlexSymbolMatcher;
29*d219b4ceSAdam Hornacek %%
30*d219b4ceSAdam Hornacek 
31*d219b4ceSAdam Hornacek %public
32*d219b4ceSAdam Hornacek %class TroffFullTokenizer
33*d219b4ceSAdam Hornacek %extends JFlexSymbolMatcher
34*d219b4ceSAdam Hornacek %unicode
35*d219b4ceSAdam Hornacek %int
36*d219b4ceSAdam Hornacek %include ../CommonLexer.lexh
37*d219b4ceSAdam Hornacek %caseless
38*d219b4ceSAdam Hornacek %char
39*d219b4ceSAdam Hornacek 
40*d219b4ceSAdam Hornacek //WhiteSpace     = [ \t\f\r]+|\n
41*d219b4ceSAdam Hornacek Identifier = [a-zA-Z_] [a-zA-Z0-9_]*
42*d219b4ceSAdam Hornacek Number = [0-9]+|[0-9]+\.[0-9]+| "0[xX]" [0-9a-fA-F]+
43*d219b4ceSAdam Hornacek Printable = [\@\$\%\^\&\-+=\?\.\:]
44*d219b4ceSAdam Hornacek 
45*d219b4ceSAdam Hornacek %%
46*d219b4ceSAdam Hornacek ^\.(EQ|in|sp|ne|rt|br|pn|ds|de|if|ig|el|ft|hy|ie|ll|ps|rm|ta|ti|NH|DT|EE)[^\n]* {}
47*d219b4ceSAdam Hornacek ^.[a-zA-Z]{1,2} {}
48*d219b4ceSAdam Hornacek \\f[ABCIR]  {}
49*d219b4ceSAdam Hornacek ^"...\\\"" {}
50*d219b4ceSAdam Hornacek 
51*d219b4ceSAdam Hornacek \\&.        {
52*d219b4ceSAdam Hornacek     onSymbolMatched(".", yychar); return yystate();
53*d219b4ceSAdam Hornacek }
54*d219b4ceSAdam Hornacek 
55*d219b4ceSAdam Hornacek {Identifier}|{Number}|{Printable} {
56*d219b4ceSAdam Hornacek     onSymbolMatched(yytext().toLowerCase(Locale.ROOT), yychar);
57*d219b4ceSAdam Hornacek     return yystate();
58*d219b4ceSAdam Hornacek }
59*d219b4ceSAdam Hornacek 
60*d219b4ceSAdam Hornacek [^]    {}
61