1 |
/* DLexer.h (formerly DLexer.cpp) |
2 |
* |
3 |
* This was renamed because the use of the .cpp extension caused problems |
4 |
* with IDEs. |
5 |
* |
6 |
* SOFTWARE RIGHTS |
7 |
* |
8 |
* We reserve no LEGAL rights to the Purdue Compiler Construction Tool |
9 |
* Set (PCCTS) -- PCCTS is in the public domain. An individual or |
10 |
* company may do whatever they wish with source code distributed with |
11 |
* PCCTS or the code generated by PCCTS, including the incorporation of |
12 |
* PCCTS, or its output, into commerical software. |
13 |
* |
14 |
* We encourage users to develop software with PCCTS. However, we do ask |
15 |
* that credit is given to us for developing PCCTS. By "credit", |
16 |
* we mean that if you incorporate our source code into one of your |
17 |
* programs (commercial product, research project, or otherwise) that you |
18 |
* acknowledge this fact somewhere in the documentation, research report, |
19 |
* etc... If you like PCCTS and have developed a nice tool with the |
20 |
* output, please mention that you developed it using PCCTS. In |
21 |
* addition, we ask that this header remain intact in our source code. |
22 |
* As long as these guidelines are kept, we expect to continue enhancing |
23 |
* this system and expect to make other tools available as they are |
24 |
* completed. |
25 |
* |
26 |
* ANTLR 1.33 |
27 |
* Terence Parr |
28 |
* Parr Research Corporation |
29 |
* with Purdue University and AHPCRC, University of Minnesota |
30 |
* 1989-2000 |
31 |
*/ |
32 |
|
33 |
#define ZZINC {if ( track_columns ) (++_endcol);} |
34 |
|
35 |
#define ZZGETC {ch = input->nextChar(); cl = ZZSHIFT(ch);} |
36 |
|
37 |
#define ZZNEWSTATE (newstate = dfa[state][cl]) |
38 |
|
39 |
#ifndef ZZCOPY |
40 |
#define ZZCOPY \ |
41 |
/* Truncate matching buffer to size (not an error) */ \ |
42 |
if (nextpos < lastpos){ \ |
43 |
*(nextpos++) = ch; \ |
44 |
}else{ \ |
45 |
bufovf = 1; \ |
46 |
} |
47 |
#endif |
48 |
|
49 |
void DLGLexer:: |
50 |
mode( int m ) |
51 |
{ |
52 |
/* points to base of dfa table */ |
53 |
if (m<MAX_MODE){ |
54 |
automaton = m; |
55 |
/* have to redo class since using different compression */ |
56 |
cl = ZZSHIFT(ch); |
57 |
}else{ |
58 |
sprintf((char *)ebuf,"Invalid automaton mode = %d ",m); |
59 |
errstd(ebuf); |
60 |
} |
61 |
} |
62 |
|
63 |
ANTLRTokenType DLGLexer:: |
64 |
nextTokenType(void) |
65 |
{ |
66 |
register int state, newstate; |
67 |
/* last space reserved for the null char */ |
68 |
register DLGChar *lastpos; |
69 |
ANTLRTokenType tk; |
70 |
|
71 |
skip: |
72 |
bufovf = 0; |
73 |
lastpos = &_lextext[_bufsize-1]; |
74 |
nextpos = _lextext; |
75 |
_begcol = _endcol+1; |
76 |
more: |
77 |
_begexpr = nextpos; |
78 |
if ( interactive ) { |
79 |
/* interactive version of automaton */ |
80 |
/* if there is something in ch, process it */ |
81 |
state = newstate = dfa_base[automaton]; |
82 |
if (charfull){ |
83 |
ZZINC; |
84 |
ZZCOPY; |
85 |
ZZNEWSTATE; |
86 |
} |
87 |
while (alternatives[newstate]){ |
88 |
state = newstate; |
89 |
ZZGETC; |
90 |
ZZINC; |
91 |
ZZCOPY; |
92 |
ZZNEWSTATE; |
93 |
} |
94 |
/* figure out if last character really part of token */ |
95 |
if ((state != dfa_base[automaton]) && (newstate == DfaStates)){ |
96 |
charfull = 1; |
97 |
--nextpos; |
98 |
}else{ |
99 |
charfull = 0; |
100 |
state = newstate; |
101 |
} |
102 |
*(nextpos) = '\0'; |
103 |
/* Able to transition out of start state to some non err state?*/ |
104 |
if ( state == dfa_base[automaton] ){ |
105 |
/* make sure doesn't get stuck */ |
106 |
advance(); |
107 |
} |
108 |
} |
109 |
else { /* non-interactive version of automaton */ |
110 |
if (!charfull) |
111 |
advance(); |
112 |
else |
113 |
ZZINC; |
114 |
state = dfa_base[automaton]; |
115 |
while (ZZNEWSTATE != DfaStates) { |
116 |
state = newstate; |
117 |
ZZCOPY; |
118 |
ZZGETC; |
119 |
ZZINC; |
120 |
} |
121 |
charfull = 1; |
122 |
if ( state == dfa_base[automaton] ){ |
123 |
if (nextpos < lastpos){ |
124 |
*(nextpos++) = ch; |
125 |
}else{ |
126 |
bufovf = 1; |
127 |
} |
128 |
*nextpos = '\0'; |
129 |
/* make sure doesn't get stuck */ |
130 |
advance(); |
131 |
}else{ |
132 |
*nextpos = '\0'; |
133 |
} |
134 |
} |
135 |
if ( track_columns ) _endcol -= charfull; |
136 |
_endexpr = nextpos -1; |
137 |
add_erase = 0; |
138 |
#ifdef OLD |
139 |
tk = (ANTLRTokenType) |
140 |
(*actions[accepts[state]])(this); // must pass this manually |
141 |
// actions is not a [] of pointers |
142 |
// to member functions. |
143 |
#endif |
144 |
tk = (this->*actions[accepts[state]])(); |
145 |
|
146 |
// MR1 |
147 |
// MR1 11-Apr-97 Help for tracking DLG results |
148 |
// MR1 |
149 |
|
150 |
#ifdef DEBUG_LEXER |
151 |
|
152 |
/* MR1 */ if (debugLexerFlag) { |
153 |
/* MR1 */ if (parser != NULL) { |
154 |
/* MR1 */ /* MR23 */ printMessage(stdout, "\ntoken name=%s",parser->parserTokenName(tk)); |
155 |
/* MR1 */ } else { |
156 |
/* MR1 */ /* MR23 */ printMessage(stdout, "\ntoken nnumber=%d",tk); |
157 |
/* MR1 */ }; |
158 |
/* MR1 */ /* MR23 */ printMessage(stdout, " lextext=(%s) mode=%d", |
159 |
/* MR1 */ (_lextext[0]=='\n' && _lextext[1]==0) ? |
160 |
/* MR1 */ "newline" : _lextext, |
161 |
/* MR1 */ automaton); |
162 |
/* MR1 */ if (interactive && !charfull) { |
163 |
/* MR1 */ /* MR23 */ printMessage(stdout, " char=empty"); |
164 |
/* MR1 */ } else { |
165 |
/* MR1 */ if (ch=='\n') { |
166 |
/* MR1 */ /* MR23 */ printMessage(stdout, " char=newline"); |
167 |
/* MR1 */ } else { |
168 |
/* MR1 */ /* MR23 */ printMessage(stdout, " char=(%c)",ch); |
169 |
/* MR1 */ }; |
170 |
/* MR1 */ }; |
171 |
/* MR1 */ /* MR23 */ printMessage(stdout, " %s\n", |
172 |
/* MR1 */ (add_erase==1 ? "skip()" : |
173 |
/* MR1 */ add_erase==2 ? "more()" : |
174 |
/* MR1 */ "")); |
175 |
/* MR1 */ }; |
176 |
|
177 |
#endif |
178 |
|
179 |
switch (add_erase) { |
180 |
case 1: goto skip; |
181 |
case 2: goto more; |
182 |
} |
183 |
return tk; |
184 |
} |
185 |
|
186 |
void DLGLexer:: |
187 |
advance() |
188 |
{ |
189 |
if ( input==NULL ) err_in(); |
190 |
ZZGETC; charfull = 1; ZZINC; |
191 |
} |