1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
|
/* Lex file for rsyslog config format v2 (RainerScript).
* Please note: this file introduces the new config format, but maintains
* backward compatibility. In order to do so, the grammar is not 100% clean,
* but IMHO still sufficiently easy both to understand for programmers
* maitaining the code as well as users writing the config file. Users are,
* of course, encouraged to use new constructs only. But it needs to be noted
* that some of the legacy constructs (specifically the in-front-of-action
* PRI filter) are very hard to beat in ease of use, at least for simpler
* cases. So while we hope that cfsysline support can be dropped some time in
* the future, we will probably keep these useful constructs.
*
* Copyright 2011-2014 Rainer Gerhards and Adiscon GmbH.
*
* This file is part of the rsyslog runtime library.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
* -or-
* see COPYING.ASL20 in the source distribution
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
%{
#include "config.h"
#include "parserif.h"
extern char *strdup(char*); /* somehow we do not get this from string.h... */
%}
%option noyywrap nodefault case-insensitive yylineno
/*%option noyywrap nodefault case-insensitive */
/* avoid compiler warning: `yyunput' defined but not used */
%option nounput noinput
%x INOBJ
/* INOBJ is selected if we are inside an object (name/value pairs!) */
%x COMMENT
/* COMMENT is "the usual trick" to handle C-style comments */
%x INCL
/* INCL is in $IncludeConfig processing (skip to include file) */
%x LINENO
/* LINENO: support for setting the linenumber */
%x INCALL
/* INCALL: support for the call statement */
%x EXPR
/* EXPR is a bit ugly, but we need it to support pre v6-syntax. The problem
* is that cfsysline statement start with $..., the same like variables in
* an expression. However, cfsysline statements can never appear inside an
* expression. So we create a specific expr mode, which is turned on after
* we lexed a keyword that needs to be followed by an expression (using
* knowledge from the upper layer...). In expr mode, we strictly do
* expression-based parsing. Expr mode is stopped when we reach a token
* that can not be part of an expression (currently only "then"). As I
* wrote this ugly, but the price needed to pay in order to remain
* compatible to the previous format.
*/
%{
#include <ctype.h>
#include <stdio.h>
#include <stdlib.h>
#include <strings.h>
#include <libestr.h>
#include "rainerscript.h"
#include "parserif.h"
#include "grammar.h"
static int preCommentState; /* save for lex state before a comment */
struct bufstack {
struct bufstack *prev;
YY_BUFFER_STATE bs;
int lineno;
char *fn;
es_str_t *estr;
} *currbs = NULL;
char *cnfcurrfn; /* name of currently processed file */
int popfile(void);
int cnfSetLexFile(char *fname);
extern int yydebug;
/* somehow, I need these prototype even though the headers are
* included. I guess that's some autotools magic I don't understand...
*/
#if !defined(__FreeBSD__)
int fileno(FILE *stream);
#endif
%}
%%
/* keywords */
"if" { BEGIN EXPR; return IF; }
<EXPR>"then" { BEGIN INITIAL; return THEN; }
<EXPR>";" { BEGIN INITIAL; return ';'; }
<EXPR>"or" { return OR; }
<EXPR>"and" { return AND; }
<EXPR>"not" { return NOT; }
<EXPR>"=" |
<EXPR>"," |
<EXPR>"*" |
<EXPR>"/" |
<EXPR>"%" |
<EXPR>"+" |
<EXPR>"&" |
<EXPR>"-" |
<EXPR>"[" |
<EXPR>"]" |
<EXPR>"(" |
<EXPR>")" { return yytext[0]; }
<EXPR>"==" { return CMP_EQ; }
<EXPR>"<=" { return CMP_LE; }
<EXPR>">=" { return CMP_GE; }
<EXPR>"!=" |
<EXPR>"<>" { return CMP_NE; }
<EXPR>"<" { return CMP_LT; }
<EXPR>">" { return CMP_GT; }
<EXPR>"contains" { return CMP_CONTAINS; }
<EXPR>"contains_i" { return CMP_CONTAINSI; }
<EXPR>"startswith" { return CMP_STARTSWITH; }
<EXPR>"startswith_i" { return CMP_STARTSWITHI; }
<EXPR>0[0-7]+ | /* octal number */
<EXPR>0x[0-7a-f] | /* hex number, following rule is dec; strtoll handles all! */
<EXPR>([1-9][0-9]*|0) { yylval.n = strtoll(yytext, NULL, 0); return NUMBER; }
<EXPR>\$[$!./]{0,1}[@a-z][!@a-z0-9\-_\.]* { yylval.s = strdup(yytext+1); return VAR; }
<EXPR>\'([^'\\]|\\['"\\$bntr]|\\x[0-9a-f][0-9a-f]|\\[0-7][0-7][0-7])*\' {
yytext[yyleng-1] = '\0';
unescapeStr((uchar*)yytext+1, yyleng-2);
yylval.estr = es_newStrFromBuf(yytext+1, strlen(yytext)-1);
return STRING; }
<EXPR>\"([^"\\$]|\\["'\\$bntr]|\\x[0-9a-f][0-9a-f]|\\[0-7][0-7][0-7])*\" {
yytext[yyleng-1] = '\0';
unescapeStr((uchar*)yytext+1, yyleng-2);
yylval.estr = es_newStrFromBuf(yytext+1, strlen(yytext)-1);
return STRING; }
<EXPR>[ \t\n]
<EXPR>[a-z][a-z0-9_]* { yylval.estr = es_newStrFromCStr(yytext, yyleng);
return FUNC; }
<EXPR>. { parser_errmsg("invalid character '%s' in expression "
"- is there an invalid escape sequence somewhere?",
yytext); }
<INCALL>[ \t\n]
<INCALL>. { parser_errmsg("invalid character '%s' in 'call' statement"
"- is there an invalid escape sequence somewhere?",
yytext); }
<INCALL>[a-zA-Z][a-zA-Z0-9_\.]* { yylval.estr = es_newStrFromCStr(yytext, yyleng);
BEGIN INITIAL;
return NAME; }
"&" { return '&'; }
"{" { return '{'; }
"}" { return '}'; }
"stop" { return STOP; }
"else" { return ELSE; }
"call" { BEGIN INCALL; return CALL; }
"set" { BEGIN EXPR; return SET; }
"unset" { BEGIN EXPR; return UNSET; }
"continue" { return CONTINUE; }
/* line number support because the "preprocessor" combines lines and so needs
* to tell us the real source line.
*/
"preprocfilelinenumber(" { BEGIN LINENO; }
<LINENO>[0-9]+ { yylineno = atoi(yytext) - 1; }
<LINENO>")" { BEGIN INITIAL; }
<LINENO>.|\n
/* $IncludeConfig must be detected as part of CFSYSLINE, because this is
* always the longest match :-(
*/
<INCL>.|\n
<INCL>[^ \t\n]+ { if(cnfDoInclude(yytext) != 0)
yyterminate();
BEGIN INITIAL; }
"main_queue"[ \n\t]*"(" { yylval.objType = CNFOBJ_MAINQ;
BEGIN INOBJ; return BEGINOBJ; }
"timezone"[ \n\t]*"(" { yylval.objType = CNFOBJ_TIMEZONE;
BEGIN INOBJ; return BEGINOBJ; }
"parser"[ \n\t]*"(" { yylval.objType = CNFOBJ_PARSER;
BEGIN INOBJ; return BEGINOBJ; }
"global"[ \n\t]*"(" { yylval.objType = CNFOBJ_GLOBAL;
BEGIN INOBJ; return BEGINOBJ; }
"template"[ \n\t]*"(" { yylval.objType = CNFOBJ_TPL;
BEGIN INOBJ; return BEGIN_TPL; }
"ruleset"[ \n\t]*"(" { yylval.objType = CNFOBJ_RULESET;
BEGIN INOBJ; return BEGIN_RULESET; }
"property"[ \n\t]*"(" { yylval.objType = CNFOBJ_PROPERTY;
BEGIN INOBJ; return BEGIN_PROPERTY; }
"constant"[ \n\t]*"(" { yylval.objType = CNFOBJ_CONSTANT;
BEGIN INOBJ; return BEGIN_CONSTANT; }
"input"[ \n\t]*"(" { yylval.objType = CNFOBJ_INPUT;
BEGIN INOBJ; return BEGINOBJ; }
"module"[ \n\t]*"(" { yylval.objType = CNFOBJ_MODULE;
BEGIN INOBJ; return BEGINOBJ; }
"lookup_table"[ \n\t]*"(" { yylval.objType = CNFOBJ_LOOKUP_TABLE;
BEGIN INOBJ; return BEGINOBJ; }
"action"[ \n\t]*"(" { BEGIN INOBJ; return BEGIN_ACTION; }
^[ \t]*:\$?[a-z\-]+[ ]*,[ ]*!?[a-z]+[ ]*,[ ]*\"(\\\"|[^\"])*\" {
yylval.s = strdup(rmLeadingSpace(yytext));
dbgprintf("lexer: propfilt is '%s'\n", yylval.s);
return PROPFILT;
}
^[ \t]*[\*a-z][\*a-z]*[0-7]*[\.,][,!=;\.\*a-z0-7]+ { yylval.s = strdup(rmLeadingSpace(yytext)); return PRIFILT; }
"~" |
"*" |
\-\/[^*][^\n]* |
\/[^*][^\n]* |
:[a-z0-9]+:[^\n]* |
[\|\.\-\@\^?~>][^\n]+ |
[a-z0-9_][a-z0-9_\-\+,;]* { yylval.s = yytext; return LEGACY_ACTION; }
<INOBJ>")" { BEGIN INITIAL; return ENDOBJ; }
<INOBJ>[a-z][a-z0-9_\.]* { yylval.estr = es_newStrFromCStr(yytext, yyleng);
return NAME; }
<INOBJ>"," |
<INOBJ>"[" |
<INOBJ>"]" |
<INOBJ>"=" { return(yytext[0]); }
<INOBJ>\"([^"\\]|\\['"?\\abfnrtv]|\\[0-7]{1,3})*\" {
yytext[yyleng-1] = '\0';
unescapeStr((uchar*)yytext+1, yyleng-2);
yylval.estr = es_newStrFromBuf(yytext+1, strlen(yytext)-1);
return STRING; }
/*yylval.estr = es_newStrFromBuf(yytext+1, yyleng-2);
return VALUE; }*/
"/*" { preCommentState = YY_START; BEGIN COMMENT; }
<INOBJ>"/*" { preCommentState = YY_START; BEGIN COMMENT; }
<EXPR>"/*" { preCommentState = YY_START; BEGIN COMMENT; }
<COMMENT>"*/" { BEGIN preCommentState; }
<COMMENT>([^*]|\n)+|.
<INOBJ>#.*$ /* skip comments in input */
<INOBJ>[ \n\t]
<INOBJ>. { parser_errmsg("invalid character '%s' in object definition "
"- is there an invalid escape sequence somewhere?",
yytext); }
\$[a-z]+.*$ { /* see comment on $IncludeConfig above */
if(!strncasecmp(yytext, "$includeconfig ", 14)) {
yyless(14);
BEGIN INCL;
} else if(!strncasecmp(yytext, "$ruleset ", 9)) {
yylval.s = strdup(yytext);
return LEGACY_RULESET;
} else {
cnfDoCfsysline(strdup(yytext));
}
}
![^ \t\n]+[ \t]*$ { yylval.s = strdup(yytext); return BSD_TAG_SELECTOR; }
[+-]\*[ \t\n]*#.*$ { yylval.s = strdup(yytext); return BSD_HOST_SELECTOR; }
[+-]\*[ \t\n]*$ { yylval.s = strdup(yytext); return BSD_HOST_SELECTOR; }
^[ \t]*[+-][a-z0-9.:-]+[ \t]*$ { yylval.s = strdup(yytext); return BSD_HOST_SELECTOR; }
\#.*\n /* skip comments in input */
[\n\t ] /* drop whitespace */
. { parser_errmsg("invalid character '%s' "
"- is there an invalid escape sequence somewhere?",
yytext); }
<<EOF>> { if(popfile() != 0) yyterminate(); }
%%
int
cnfParseBuffer(char *buf, unsigned lenBuf)
{
struct bufstack *bs;
int r = 0;
yydebug = 1;
BEGIN INITIAL;
/* maintain stack */
if((bs = malloc(sizeof(struct bufstack))) == NULL) {
r = 1;
goto done;
}
if(currbs != NULL)
currbs->lineno = yylineno;
bs->prev = currbs;
bs->fn = strdup("*buffer*");
bs->bs = yy_scan_buffer(buf, lenBuf);
bs->estr = NULL;
currbs = bs;
cnfcurrfn = bs->fn;
yylineno = 1;
done: return r;
}
/* set a new buffers. Returns 0 on success, something else otherwise. */
int
cnfSetLexFile(char *fname)
{
es_str_t *str = NULL;
FILE *fp;
int r = 0;
struct bufstack *bs;
if(fname == NULL) {
fp = stdin;
} else {
if((fp = fopen(fname, "r")) == NULL) {
r = 1;
goto done;
}
}
readConfFile(fp, &str);
if(fp != stdin)
fclose(fp);
/* maintain stack */
if((bs = malloc(sizeof(struct bufstack))) == NULL) {
r = 1;
goto done;
}
if(currbs != NULL)
currbs->lineno = yylineno;
bs->prev = currbs;
bs->fn = strdup(fname == NULL ? "stdin" : fname);
bs->bs = yy_scan_buffer((char*)es_getBufAddr(str), es_strlen(str));
bs->estr = str; /* needed so we can free it later */
currbs = bs;
cnfcurrfn = bs->fn;
yylineno = 1;
dbgprintf("config parser: pushed file %s on top of stack\n", fname);
done:
if(r != 0) {
if(str != NULL)
es_deleteStr(str);
}
return r;
}
/* returns 0 on success, something else otherwise */
int
popfile(void)
{
struct bufstack *bs = currbs;
if(bs == NULL)
return 1;
/* delete current entry. But we must not free the file name if
* this is the top-level file, because then it may still be used
* in error messages for other processing steps.
* TODO: change this to another method which stores the file
* name inside the config objects. In the longer term, this is
* necessary, as otherwise we may provide wrong file name information
* at the end of include files as well. -- rgerhards, 2011-07-22
*/
dbgprintf("config parser: reached end of file %s\n", bs->fn);
yy_delete_buffer(bs->bs);
if(bs->prev != NULL)
free(bs->fn);
free(bs->estr);
/* switch back to previous */
currbs = bs->prev;
free(bs);
if(currbs == NULL) {
dbgprintf("config parser: parsing completed\n");
return 1; /* all processed */
}
yy_switch_to_buffer(currbs->bs);
yylineno = currbs->lineno;
cnfcurrfn = currbs->fn;
dbgprintf("config parser: resume parsing of file %s at line %d\n",
cnfcurrfn, yylineno);
return 0;
}
void
tellLexEndParsing(void)
{
free(cnfcurrfn);
cnfcurrfn= NULL;
}
|