1 /*
2 * SysDB - src/frontend/scanner.l
3 * Copyright (C) 2013 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"
35 #include "frontend/connection.h"
36 #include "frontend/parser.h"
37 #include "frontend/grammar.h"
38 #include "utils/error.h"
40 #include <assert.h>
41 #include <errno.h>
43 #include <string.h>
44 #include <stdlib.h>
46 #include <time.h>
48 #define YY_EXTRA_TYPE sdb_fe_yyextra_t *
50 static struct {
51 const char *name;
52 int id;
53 } reserved_words[] = {
54 { "AND", AND },
55 { "END", END },
56 { "FETCH", FETCH },
57 { "FILTER", FILTER },
58 { "IN", IN },
59 { "IS", IS },
60 { "LIST", LIST },
61 { "LOOKUP", LOOKUP },
62 { "MATCHING", MATCHING },
63 { "NOT", NOT },
64 { "NULL", NULL_T },
65 { "OR", OR },
66 { "START", START },
67 { "TIMESERIES", TIMESERIES },
68 };
70 void
71 sdb_fe_yyerror(YYLTYPE *lval, sdb_fe_yyscan_t scanner, const char *msg);
73 %}
75 %option never-interactive
76 %option reentrant
77 %option bison-bridge
78 %option bison-locations
79 %option 8bit
80 %option yylineno
81 %option nodefault
82 %option noinput
83 %option nounput
84 %option noyywrap
85 %option verbose
86 %option warn
87 %option prefix="sdb_fe_yy" outfile="lex.yy.c"
89 %x CSC
91 whitespace ([ \t\n\r\f]+)
92 simple_comment ("--"[^\n\r]*)
94 /*
95 * C style comments
96 */
97 csc_start \/\*
98 csc_inside ([^*/]+|[^*]\/|\*[^/])
99 csc_end \*\/
101 /*
102 * Strings and identifiers.
103 */
104 identifier ([A-Za-z_][A-Za-z_0-9$]*)
105 /* TODO: fully support SQL strings */
106 string ('([^']|'')*')
108 /*
109 * Numeric constants.
110 */
111 dec ([\+\-]?[0-9]+)
112 exp ([\+\-]?[0-9]+[Ee]\+?[0-9]+)
113 integer ({dec}|{exp})
114 float1 ([\+\-]?[0-9]+\.[0-9]*([Ee][\+\-]?[0-9]+)?)
115 float2 ([\+\-]?[0-9]*\.[0-9]+([Ee][\+\-]?[0-9]+)?)
116 float3 ([\+\-]?[0-9]+[Ee]\-[0-9]+)
117 float4 ([\+\-]?[Ii][Nn][Ff]([Ii][Nn][Ii][Tt][Yy])?)
118 float5 ([Nn][Aa][Nn])
119 float ({float1}|{float2}|{float3}|{float4}|{float5})
121 /*
122 * Time constants.
123 */
124 date ([0-9]{4}-[0-9]{2}-[0-9]{2})
125 time ([0-9]{1,2}:[0-9]{1,2}(:[0-9]{1,2}(\.[0-9]{1,9})?)?)
127 %%
129 {whitespace} |
130 {simple_comment} { /* ignore */ }
132 {csc_start} { BEGIN(CSC); }
133 <CSC>{csc_inside} { /* ignore */ }
134 <CSC>{csc_end} { BEGIN(INITIAL); }
135 <CSC><<EOF>> {
136 sdb_fe_yyerror(yylloc, yyscanner, "unterminated C-style comment");
137 return SCANNER_ERROR;
138 }
140 {identifier} {
141 size_t i;
142 for (i = 0; i < SDB_STATIC_ARRAY_LEN(reserved_words); ++i)
143 if (! strcasecmp(reserved_words[i].name, yytext))
144 return reserved_words[i].id;
146 yylval->str = strdup(yytext);
147 return IDENTIFIER;
148 }
149 {string} {
150 char *quot;
151 size_t len;
153 /* remove the leading and trailing quote */
154 yytext[yyleng - 1] = '\0';
155 yylval->str = strdup(yytext + 1);
157 quot = yylval->str;
158 len = yyleng - 2;
159 while ((quot = strstr(quot, "''")) != NULL) {
160 memmove(quot, quot + 1, len - (quot - yylval->str) - 1);
161 yylval->str[len - 1] = '\0';
162 --len;
163 ++quot;
164 }
165 return STRING;
166 }
167 {integer} {
168 yylval->data.data.integer = (int64_t)strtoll(yytext, NULL, 10);
169 yylval->data.type = SDB_TYPE_INTEGER;
170 return INTEGER;
171 }
172 {float} {
173 yylval->data.data.decimal = strtod(yytext, NULL);
174 yylval->data.type = SDB_TYPE_DECIMAL;
175 return FLOAT;
176 }
178 {date} {
179 struct tm tm;
180 memset(&tm, 0, sizeof(tm));
181 if (! strptime(yytext, "%Y-%m-%d", &tm)) {
182 char errmsg[1024];
183 snprintf(errmsg, sizeof(errmsg),
184 "Failed to parse '%s' as date", yytext);
185 sdb_fe_yyerror(yylloc, yyscanner, errmsg);
186 return SCANNER_ERROR;
187 }
188 yylval->datetime = SECS_TO_SDB_TIME(mktime(&tm));
189 return DATE;
190 }
191 {time} {
192 struct tm tm;
193 char time[9], ns[10];
194 char *tmp;
196 memset(&tm, 0, sizeof(tm));
197 memset(time, '\0', sizeof(time));
198 memset(ns, '0', sizeof(ns));
199 ns[sizeof(ns) - 1] = '\0';
201 tmp = strchr(yytext, '.');
202 if (tmp) {
203 size_t i;
204 *tmp = '\0';
205 ++tmp;
206 strncpy(ns, tmp, sizeof(ns));
207 for (i = strlen(ns); i < 9; ++i)
208 ns[i] = '0';
209 }
210 strncpy(time, yytext, sizeof(time));
211 if (tmp) {
212 /* reset for better error messages */
213 --tmp;
214 *tmp = '.';
215 }
217 tmp = strchr(time, ':');
218 assert(tmp);
219 tmp = strchr(tmp + 1, ':');
220 if (! tmp)
221 strncat(time, ":00", sizeof(time));
223 if (! strptime(time, "%H:%M:%S", &tm)) {
224 char errmsg[1024];
225 snprintf(errmsg, sizeof(errmsg),
226 "Failed to parse '%s' as time", yytext);
227 sdb_fe_yyerror(yylloc, yyscanner, errmsg);
228 return SCANNER_ERROR;
229 }
231 yylval->datetime = SECS_TO_SDB_TIME(mktime(&tm));
232 yylval->datetime += (sdb_time_t)strtoll(ns, NULL, 10);
233 return TIME;
234 }
236 = { return CMP_EQUAL; }
237 != { return CMP_NEQUAL; }
238 =~ { return CMP_REGEX; }
239 !~ { return CMP_NREGEX; }
240 \< { return CMP_LT; }
241 \<= { return CMP_LE; }
242 \>= { return CMP_GE; }
243 \> { return CMP_GT; }
244 \|\| { return CONCAT; }
246 . { /* XXX: */ return yytext[0]; }
248 %%
250 sdb_fe_yyscan_t
251 sdb_fe_scanner_init(const char *str, int len, sdb_fe_yyextra_t *yyext)
252 {
253 yyscan_t scanner;
255 if (! str)
256 return NULL;
258 if (sdb_fe_yylex_init(&scanner)) {
259 char errbuf[1024];
260 sdb_log(SDB_LOG_ERR, "frontend: yylex_init failed: %s",
261 sdb_strerror(errno, errbuf, sizeof(errbuf)));
262 return NULL;
263 }
265 sdb_fe_yyset_extra(yyext, scanner);
267 if (len < 0)
268 len = strlen(str);
270 /* the newly allocated buffer state (YY_BUFFER_STATE) is stored inside the
271 * scanner and, thus, will be freed by yylex_destroy */
272 sdb_fe_yy_scan_bytes(str, len, scanner);
273 return scanner;
274 } /* sdb_fe_scanner_init */
276 void
277 sdb_fe_scanner_destroy(sdb_fe_yyscan_t scanner)
278 {
279 sdb_fe_yylex_destroy(scanner);
280 } /* sdb_fe_scanner_destroy */
282 /* vim: set tw=78 sw=4 ts=4 noexpandtab : */