1 /*
2 * SysDB - src/parser/scanner.l
3 * Copyright (C) 2013-2015 Sebastian 'tokkee' Harl <sh@tokkee.org>
4 * All rights reserved.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
8 * are met:
9 * 1. Redistributions of source code must retain the above copyright
10 * notice, this list of conditions and the following disclaimer.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
14 *
15 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
16 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
17 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
18 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR
19 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
20 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
21 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
22 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
23 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
24 * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
25 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26 */
28 %{
30 #if HAVE_CONFIG_H
31 # include "config.h"
32 #endif /* HAVE_CONFIG_H */
34 #include "core/data.h"
36 #include "parser/parser.h"
37 #include "parser/grammar.h"
39 #include "utils/error.h"
41 #include <assert.h>
42 #include <errno.h>
44 #include <string.h>
45 #include <stdlib.h>
47 #include <time.h>
49 #define YY_EXTRA_TYPE sdb_parser_yyextra_t *
51 static struct {
52 const char *name;
53 int id;
54 } reserved_words[] = {
55 { "ALL", ALL },
56 { "AND", AND },
57 { "ANY", ANY },
58 { "END", END },
59 { "FALSE", FALSE },
60 { "FETCH", FETCH },
61 { "FILTER", FILTER },
62 { "IN", IN },
63 { "IS", IS },
64 { "LAST", LAST },
65 { "LIST", LIST },
66 { "LOOKUP", LOOKUP },
67 { "MATCHING", MATCHING },
68 { "NOT", NOT },
69 { "NULL", NULL_T },
70 { "OR", OR },
71 { "START", START },
72 { "STORE", STORE },
73 { "TIMESERIES", TIMESERIES },
74 { "TRUE", TRUE },
75 { "UPDATE", UPDATE },
77 /* object types */
78 { "host", HOST_T },
79 { "hosts", HOSTS_T },
80 { "service", SERVICE_T },
81 { "services", SERVICES_T },
82 { "metric", METRIC_T },
83 { "metrics", METRICS_T },
84 { "attribute", ATTRIBUTE_T },
85 { "attributes", ATTRIBUTES_T },
86 /* queryable fields */
87 { "name", NAME_T },
88 { "last_update", LAST_UPDATE_T },
89 { "age", AGE_T },
90 { "interval", INTERVAL_T },
91 { "backend", BACKEND_T },
92 { "value", VALUE_T },
93 };
95 void
96 sdb_parser_yyerror(YYLTYPE *lval, sdb_parser_yyscan_t scanner, const char *msg);
98 %}
100 %option never-interactive
101 %option reentrant
102 %option bison-bridge
103 %option bison-locations
104 %option 8bit
105 %option yylineno
106 %option nodefault
107 %option noinput
108 %option nounput
109 %option noyywrap
110 %option verbose
111 %option warn
112 %option prefix="sdb_parser_yy" outfile="lex.yy.c"
114 %x CSC
116 whitespace ([ \t\n\r\f]+)
117 simple_comment ("--"[^\n\r]*)
119 /*
120 * C style comments
121 */
122 csc_start \/\*
123 csc_inside ([^*/]+|[^*]\/|\*[^/])
124 csc_end \*\/
126 /*
127 * Strings and identifiers.
128 */
129 identifier ([A-Za-z_][A-Za-z_0-9$]*)
130 /* TODO: fully support SQL strings */
131 string ('([^']|'')*')
133 /*
134 * Numeric constants.
135 */
136 dec ([\+\-]?[0-9]+)
137 exp ([\+\-]?[0-9]+[Ee]\+?[0-9]+)
138 integer ({dec}|{exp})
139 float1 ([\+\-]?[0-9]+\.[0-9]*([Ee][\+\-]?[0-9]+)?)
140 float2 ([\+\-]?[0-9]*\.[0-9]+([Ee][\+\-]?[0-9]+)?)
141 float3 ([\+\-]?[0-9]+[Ee]\-[0-9]+)
142 float4 ([\+\-]?[Ii][Nn][Ff]([Ii][Nn][Ii][Tt][Yy])?)
143 float5 ([Nn][Aa][Nn])
144 float ({float1}|{float2}|{float3}|{float4}|{float5})
146 /*
147 * Time constants.
148 */
149 date ([0-9]{4}-[0-9]{2}-[0-9]{2})
150 time ([0-9]{1,2}:[0-9]{1,2}(:[0-9]{1,2}(\.[0-9]{1,9})?)?)
152 %%
154 {whitespace} |
155 {simple_comment} { /* ignore */ }
157 {csc_start} { BEGIN(CSC); }
158 <CSC>{csc_inside} { /* ignore */ }
159 <CSC>{csc_end} { BEGIN(INITIAL); }
160 <CSC><<EOF>> {
161 sdb_parser_yyerror(yylloc, yyscanner, "unterminated C-style comment");
162 return SCANNER_ERROR;
163 }
165 {identifier} {
166 size_t i;
167 for (i = 0; i < SDB_STATIC_ARRAY_LEN(reserved_words); ++i)
168 if (! strcasecmp(reserved_words[i].name, yytext))
169 return reserved_words[i].id;
171 yylval->str = strdup(yytext);
172 return IDENTIFIER;
173 }
174 {string} {
175 char *quot;
176 size_t len;
178 /* remove the leading and trailing quote */
179 yytext[yyleng - 1] = '\0';
180 yylval->str = strdup(yytext + 1);
182 quot = yylval->str;
183 len = yyleng - 2;
184 while ((quot = strstr(quot, "''")) != NULL) {
185 memmove(quot, quot + 1, len - (quot - yylval->str) - 1);
186 yylval->str[len - 1] = '\0';
187 --len;
188 ++quot;
189 }
190 return STRING;
191 }
192 {integer} {
193 yylval->data.data.integer = (int64_t)strtoll(yytext, NULL, 10);
194 yylval->data.type = SDB_TYPE_INTEGER;
195 return INTEGER;
196 }
197 {float} {
198 yylval->data.data.decimal = strtod(yytext, NULL);
199 yylval->data.type = SDB_TYPE_DECIMAL;
200 return FLOAT;
201 }
203 {date} {
204 struct tm tm;
205 memset(&tm, 0, sizeof(tm));
206 if (! strptime(yytext, "%Y-%m-%d", &tm)) {
207 char errmsg[1024];
208 snprintf(errmsg, sizeof(errmsg),
209 "Failed to parse '%s' as date", yytext);
210 sdb_parser_yyerror(yylloc, yyscanner, errmsg);
211 return SCANNER_ERROR;
212 }
213 yylval->datetime = SECS_TO_SDB_TIME(mktime(&tm));
214 return DATE;
215 }
216 {time} {
217 struct tm tm;
218 char time[9], ns[10];
219 char *tmp;
220 int t;
222 memset(&tm, 0, sizeof(tm));
223 memset(time, '\0', sizeof(time));
224 memset(ns, '0', sizeof(ns));
225 ns[sizeof(ns) - 1] = '\0';
227 tmp = strchr(yytext, '.');
228 if (tmp) {
229 size_t i;
230 *tmp = '\0';
231 ++tmp;
232 strncpy(ns, tmp, sizeof(ns));
233 for (i = strlen(ns); i < 9; ++i)
234 ns[i] = '0';
235 }
236 strncpy(time, yytext, sizeof(time));
237 if (tmp) {
238 /* reset for better error messages */
239 --tmp;
240 *tmp = '.';
241 }
243 tmp = strchr(time, ':');
244 assert(tmp);
245 tmp = strchr(tmp + 1, ':');
246 if (! tmp)
247 strncat(time, ":00", sizeof(time) - strlen(time) - 1);
249 if (! strptime(time, "%H:%M:%S", &tm)) {
250 char errmsg[1024];
251 snprintf(errmsg, sizeof(errmsg),
252 "Failed to parse '%s' as time", yytext);
253 sdb_parser_yyerror(yylloc, yyscanner, errmsg);
254 return SCANNER_ERROR;
255 }
257 t = tm.tm_sec + 60 * tm.tm_min + 3600 * tm.tm_hour;
258 yylval->datetime = SECS_TO_SDB_TIME(t);
259 yylval->datetime += (sdb_time_t)strtoll(ns, NULL, 10);
260 return TIME;
261 }
263 = { return CMP_EQUAL; }
264 != { return CMP_NEQUAL; }
265 =~ { return CMP_REGEX; }
266 !~ { return CMP_NREGEX; }
267 \< { return CMP_LT; }
268 \<= { return CMP_LE; }
269 \>= { return CMP_GE; }
270 \> { return CMP_GT; }
271 \|\| { return CONCAT; }
273 . { /* XXX: */ return yytext[0]; }
275 %%
277 sdb_parser_yyscan_t
278 sdb_parser_scanner_init(const char *str, int len, sdb_parser_yyextra_t *yyext)
279 {
280 yyscan_t scanner;
282 if (! str)
283 return NULL;
285 if (sdb_parser_yylex_init(&scanner)) {
286 char errbuf[1024];
287 sdb_strbuf_sprintf(yyext->errbuf, "yylex_init_failed: %s",
288 sdb_strerror(errno, errbuf, sizeof(errbuf)));
289 return NULL;
290 }
292 sdb_parser_yyset_extra(yyext, scanner);
294 if (len < 0)
295 len = strlen(str);
297 /* the newly allocated buffer state (YY_BUFFER_STATE) is stored inside the
298 * scanner and, thus, will be freed by yylex_destroy */
299 sdb_parser_yy_scan_bytes(str, len, scanner);
300 return scanner;
301 } /* sdb_parser_scanner_init */
303 void
304 sdb_parser_scanner_destroy(sdb_parser_yyscan_t scanner)
305 {
306 sdb_parser_yylex_destroy(scanner);
307 } /* sdb_parser_scanner_destroy */
309 /* vim: set tw=78 sw=4 ts=4 noexpandtab : */