Mercurial > hg > nginx-site
annotate xml/en/docs/dev/development_guide.xml @ 1920:de5251816480
Fixed a typo.
author | Vladimir Homutov <vl@nginx.com> |
---|---|
date | Thu, 02 Mar 2017 13:12:01 +0300 |
parents | dcfb4f3ac8a7 |
children | 2c14a16c61eb |
rev | line source |
---|---|
1899 | 1 <?xml version="1.0"?> |
2 | |
3 <!-- | |
4 Copyright (C) Nginx, Inc. | |
5 --> | |
6 | |
7 <!DOCTYPE article SYSTEM "../../../../dtd/article.dtd"> | |
8 | |
9 <article name="Development guide" | |
10 link="/en/docs/dev/development_guide.html" | |
11 lang="en" | |
1914
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
12 rev="2"> |
1899 | 13 |
14 <section name="Introduction" id="introduction"> | |
15 | |
16 | |
17 <section name="Code layout" id="code_layout"> | |
18 | |
19 <para> | |
20 <list type="bullet"> | |
21 <listitem> | |
22 <literal>auto</literal> - build scripts | |
23 </listitem> | |
24 | |
25 <listitem> | |
26 <literal>src</literal> | |
27 | |
28 <list type="bullet"> | |
29 | |
30 <listitem> | |
31 <literal>core</literal> - basic types and functions - string, array, log, | |
32 pool etc | |
33 </listitem> | |
34 | |
35 <listitem> | |
36 <literal>event</literal> - event core | |
37 | |
38 <list type="bullet"> | |
39 | |
40 <listitem> | |
41 <literal>modules</literal> - event notification modules: epoll, kqueue, | |
42 select etc | |
43 </listitem> | |
44 | |
45 </list> | |
46 | |
47 </listitem> | |
48 | |
49 <listitem> | |
50 <literal>http</literal> - core HTTP module and common code | |
51 | |
52 <list type="bullet"> | |
53 | |
54 <listitem> | |
55 <literal>modules</literal> - other HTTP modules | |
56 </listitem> | |
57 | |
58 <listitem> | |
59 <literal>v2</literal> - HTTPv2 | |
60 </listitem> | |
61 | |
62 </list> | |
63 | |
64 </listitem> | |
65 | |
66 <listitem> | |
67 <literal>mail</literal> - mail modules | |
68 </listitem> | |
69 | |
70 <listitem> | |
71 <literal>os</literal> - platform-specific code | |
72 | |
73 <list type="bullet"> | |
74 | |
75 <listitem> | |
76 <literal>unix</literal> | |
77 </listitem> | |
78 | |
79 <listitem> | |
80 <literal>win32</literal> | |
81 </listitem> | |
82 | |
83 </list> | |
84 | |
85 </listitem> | |
86 | |
87 <listitem> | |
88 <literal>stream</literal> - stream modules | |
89 </listitem> | |
90 | |
91 </list> | |
92 | |
93 </listitem> | |
94 | |
95 </list> | |
96 </para> | |
97 | |
98 </section> | |
99 | |
100 | |
101 <section name="Include files" id="include_files"> | |
102 | |
103 <para> | |
104 Each nginx file should start with including the following two files: | |
105 </para> | |
106 | |
107 | |
108 <programlisting> | |
109 #include <ngx_config.h> | |
110 #include <ngx_core.h> | |
111 </programlisting> | |
112 | |
113 <para> | |
114 In addition to that, HTTP code should include | |
115 </para> | |
116 | |
117 | |
118 <programlisting> | |
119 #include <ngx_http.h> | |
120 </programlisting> | |
121 | |
122 <para> | |
123 Mail code should include | |
124 </para> | |
125 | |
126 | |
127 <programlisting> | |
128 #include <ngx_mail.h> | |
129 </programlisting> | |
130 | |
131 <para> | |
132 Stream code should include | |
133 </para> | |
134 | |
135 | |
136 <programlisting> | |
137 #include <ngx_stream.h> | |
138 </programlisting> | |
139 | |
140 </section> | |
141 | |
142 | |
143 <section name="Integers" id="integers"> | |
144 | |
145 <para> | |
146 For general purpose, nginx code uses the following two integer types | |
147 <literal>ngx_int_t</literal> and <literal>ngx_uint_t</literal> which are | |
148 typedefs for <literal>intptr_t</literal> and <literal>uintptr_t</literal>. | |
149 </para> | |
150 | |
151 </section> | |
152 | |
153 | |
154 <section name="Common return codes" id="common_return_codes"> | |
155 | |
156 <para> | |
157 Most functions in nginx return the following codes: | |
158 </para> | |
159 | |
160 <para> | |
161 <list type="bullet"> | |
162 | |
163 <listitem> | |
164 <literal>NGX_OK</literal> - operation succeeded | |
165 </listitem> | |
166 | |
167 <listitem> | |
168 <literal>NGX_ERROR</literal> - operation failed | |
169 </listitem> | |
170 | |
171 <listitem> | |
172 <literal>NGX_AGAIN</literal> - operation incomplete, function should be called | |
173 again | |
174 </listitem> | |
175 | |
176 <listitem> | |
177 <literal>NGX_DECLINED</literal> - operation rejected, for example, if disabled | |
178 in configuration. This is never an error | |
179 </listitem> | |
180 | |
181 <listitem> | |
182 <literal>NGX_BUSY</literal> - resource is not available | |
183 </listitem> | |
184 | |
185 <listitem> | |
186 <literal>NGX_DONE</literal> - operation done or continued elsewhere. | |
187 Also used as an alternative success code | |
188 </listitem> | |
189 | |
190 <listitem> | |
191 <literal>NGX_ABORT</literal> - function was aborted. | |
192 Also used as an alternative error code | |
193 </listitem> | |
194 | |
195 </list> | |
196 </para> | |
197 | |
198 </section> | |
199 | |
200 | |
201 <section name="Error handling" id="error_handling"> | |
202 | |
203 <para> | |
204 For getting the last system error code, the <literal>ngx_errno</literal> macro | |
205 is available. | |
206 It's mapped to <literal>errno</literal> on POSIX platforms and to | |
207 <literal>GetLastError()</literal> call in Windows. | |
208 For getting the last socket error number, the | |
209 <literal>ngx_socket_errno</literal> macro is available. | |
210 It's mapped to <literal>errno</literal> on POSIX systems as well, | |
211 and to <literal>WSAGetLastError()</literal> call on Windows. | |
212 For performance reasons the values of <literal>ngx_errno</literal> or | |
213 <literal>ngx_socket_errno</literal> should not be accessed more than | |
214 once in a row. | |
215 The error value should be stored in a local variable of type | |
216 <literal>ngx_err_t</literal> for using multiple times, if required. | |
217 For setting errors, <literal>ngx_set_errno(errno)</literal> and | |
218 <literal>ngx_set_socket_errno(errno)</literal> macros are available. | |
219 </para> | |
220 | |
221 <para> | |
222 The values of <literal>ngx_errno</literal> or | |
223 <literal>ngx_socket_errno</literal> can be passed to logging functions | |
224 <literal>ngx_log_error()</literal> and <literal>ngx_log_debugX()</literal>, in | |
225 which case system error text is added to the log message. | |
226 </para> | |
227 | |
228 <para> | |
229 Example using <literal>ngx_errno</literal>: | |
230 </para> | |
231 | |
232 | |
233 <programlisting> | |
234 void | |
235 ngx_my_kill(ngx_pid_t pid, ngx_log_t *log, int signo) | |
236 { | |
237 ngx_err_t err; | |
238 | |
239 if (kill(pid, signo) == -1) { | |
240 err = ngx_errno; | |
241 | |
242 ngx_log_error(NGX_LOG_ALERT, log, err, "kill(%P, %d) failed", pid, signo); | |
243 | |
244 if (err == NGX_ESRCH) { | |
245 return 2; | |
246 } | |
247 | |
248 return 1; | |
249 } | |
250 | |
251 return 0; | |
252 } | |
253 </programlisting> | |
254 | |
255 </section> | |
256 | |
257 | |
258 </section> | |
259 | |
260 | |
261 <section name="Strings" id="strings"> | |
262 | |
263 | |
264 <section name="Overview" id="overview"> | |
265 | |
266 <para> | |
267 For C strings, nginx code uses unsigned character type pointer | |
268 <literal>u_char *</literal>. | |
269 </para> | |
270 | |
271 <para> | |
272 The nginx string type <literal>ngx_str_t</literal> is defined as follows: | |
273 </para> | |
274 | |
275 | |
276 <programlisting> | |
277 typedef struct { | |
278 size_t len; | |
279 u_char *data; | |
280 } ngx_str_t; | |
281 </programlisting> | |
282 | |
283 <para> | |
284 The <literal>len</literal> field holds the string length, | |
285 <literal>data</literal> holds the string data. | |
286 The string, held in <literal>ngx_str_t</literal>, may or may not be | |
287 null-terminated after the <literal>len</literal> bytes. | |
288 In most cases it’s not. | |
289 However, in certain parts of code (for example, when parsing configuration), | |
290 <literal>ngx_str_t</literal> objects are known to be null-terminated, and that | |
291 knowledge is used to simplify string comparison and makes it easier to pass | |
292 those strings to syscalls. | |
293 </para> | |
294 | |
295 <para> | |
296 A number of string operations are provided in nginx. | |
1915
8b7c3b0ef1a4
Semantically marked paths.
Vladimir Homutov <vl@nginx.com>
parents:
1914
diff
changeset
|
297 They are declared in <path>src/core/ngx_string.h</path>. |
1899 | 298 Some of them are wrappers around standard C functions: |
299 </para> | |
300 | |
301 <para> | |
302 <list type="bullet"> | |
303 | |
304 <listitem> | |
305 <literal>ngx_strcmp()</literal> | |
306 </listitem> | |
307 | |
308 <listitem> | |
309 <literal>ngx_strncmp()</literal> | |
310 </listitem> | |
311 | |
312 <listitem> | |
313 <literal>ngx_strstr()</literal> | |
314 </listitem> | |
315 | |
316 <listitem> | |
317 <literal>ngx_strlen()</literal> | |
318 </listitem> | |
319 | |
320 <listitem> | |
321 <literal>ngx_strchr()</literal> | |
322 </listitem> | |
323 | |
324 <listitem> | |
325 <literal>ngx_memcmp()</literal> | |
326 </listitem> | |
327 | |
328 <listitem> | |
329 <literal>ngx_memset()</literal> | |
330 </listitem> | |
331 | |
332 <listitem> | |
333 <literal>ngx_memcpy()</literal> | |
334 </listitem> | |
335 | |
336 <listitem> | |
337 <literal>ngx_memmove()</literal> | |
338 </listitem> | |
339 | |
340 </list> | |
341 | |
342 </para> | |
343 | |
344 <para> | |
345 Some nginx-specific string functions: | |
346 </para> | |
347 | |
348 <para> | |
349 <list type="bullet"> | |
350 | |
351 <listitem> | |
352 <literal>ngx_memzero()</literal> fills memory with zeroes | |
353 </listitem> | |
354 | |
355 <listitem> | |
356 <literal>ngx_cpymem()</literal> does the same as | |
357 <literal>ngx_memcpy()</literal>, but returns the final destination address | |
358 This one is handy for appending multiple strings in a row | |
359 </listitem> | |
360 | |
361 <listitem> | |
362 <literal>ngx_movemem()</literal> does the same as | |
363 <literal>ngx_memmove()</literal>, but returns the final destination address. | |
364 </listitem> | |
365 | |
366 <listitem> | |
367 <literal>ngx_strlchr()</literal> searches for a character in a string, | |
368 delimited by two pointers | |
369 </listitem> | |
370 </list> | |
371 </para> | |
372 | |
373 <para> | |
374 Some case conversion and comparison functions: | |
375 </para> | |
376 | |
377 <para> | |
378 <list type="bullet"> | |
379 | |
380 <listitem> | |
381 <literal>ngx_tolower()</literal> | |
382 </listitem> | |
383 | |
384 <listitem> | |
385 <literal>ngx_toupper()</literal> | |
386 </listitem> | |
387 | |
388 <listitem> | |
389 <literal>ngx_strlow()</literal> | |
390 </listitem> | |
391 | |
392 <listitem> | |
393 <literal>ngx_strcasecmp()</literal> | |
394 </listitem> | |
395 | |
396 <listitem> | |
397 <literal>ngx_strncasecmp()</literal> | |
398 </listitem> | |
399 | |
400 </list> | |
401 </para> | |
402 | |
403 </section> | |
404 | |
405 | |
406 <section name="Formatting" id="formatting"> | |
407 | |
408 <para> | |
409 A number of formatting functions are provided by nginx. These functions support nginx-specific types: | |
410 </para> | |
411 | |
412 | |
413 <para> | |
414 <list type="bullet"> | |
415 | |
416 <listitem> | |
417 <literal>ngx_sprintf(buf, fmt, ...)</literal> | |
418 </listitem> | |
419 | |
420 <listitem> | |
421 <literal>ngx_snprintf(buf, max, fmt, ...)</literal> | |
422 </listitem> | |
423 | |
424 <listitem> | |
425 <literal>ngx_slrintf(buf, last, fmt, ...)</literal> | |
426 </listitem> | |
427 | |
428 <listitem> | |
429 <literal>ngx_vslprint(buf, last, fmt, args)</literal> | |
430 </listitem> | |
431 | |
432 <listitem> | |
433 <literal>ngx_vsnprint(buf, max, fmt, args)</literal> | |
434 </listitem> | |
435 | |
436 </list> | |
437 </para> | |
438 | |
439 <para> | |
440 The full list of formatting options, supported by these functions, can be found | |
1915
8b7c3b0ef1a4
Semantically marked paths.
Vladimir Homutov <vl@nginx.com>
parents:
1914
diff
changeset
|
441 in <path>src/core/ngx_string.c</path>. Some of them are: |
1899 | 442 </para> |
443 | |
444 | |
445 <programlisting> | |
446 %O - off_t | |
447 %T - time_t | |
448 %z - size_t | |
449 %i - ngx_int_t | |
450 %p - void * | |
451 %V - ngx_str_t * | |
452 %s - u_char * (null-terminated) | |
453 %*s - size_t + u_char * | |
454 </programlisting> | |
455 | |
456 <para> | |
457 The ‘u’ modifier makes most types unsigned, ‘X’/‘x’ convert output to hex. | |
458 </para> | |
459 | |
460 <para> | |
461 Example: | |
462 | |
463 <programlisting> | |
464 u_char buf[NGX_INT_T_LEN]; | |
465 size_t len; | |
466 ngx_int_t n; | |
467 | |
468 /* set n here */ | |
469 | |
470 len = ngx_sprintf(buf, "%ui", n) - buf; | |
471 </programlisting> | |
472 | |
473 </para> | |
474 | |
475 </section> | |
476 | |
477 | |
478 <section name="Numeric conversion" id="numeric_conversion"> | |
479 | |
480 <para> | |
481 Several functions for numeric conversion are implemented in nginx: | |
482 </para> | |
483 | |
484 <para> | |
485 <list type="bullet"> | |
486 | |
487 <listitem> | |
488 <literal>ngx_atoi(line, n)</literal> - converts a string of given length to a | |
489 positive integer of type <literal>ngx_int_t</literal>. | |
490 Returns <literal>NGX_ERROR</literal> on error | |
491 </listitem> | |
492 | |
493 <listitem> | |
494 <literal>ngx_atosz(line, n)</literal> - same for <literal>ssize_t</literal> | |
495 type | |
496 </listitem> | |
497 | |
498 <listitem> | |
499 <literal>ngx_atoof(line, n)</literal> - same for <literal>off_t</literal> | |
500 type | |
501 </listitem> | |
502 | |
503 <listitem> | |
504 <literal>ngx_atotm(line, n)</literal> - same for <literal>time_t</literal> | |
505 type | |
506 </listitem> | |
507 | |
508 <listitem> | |
509 <literal>ngx_atofp(line, n, point)</literal> - converts a fixed point floating | |
510 number of given length to a positive integer of type | |
511 <literal>ngx_int_t</literal>. | |
512 The result is shifted left by <literal>points</literal> decimal | |
513 positions. The string representation of the number is expected to have no more | |
514 than <literal>points</literal> fractional digits. | |
515 Returns <literal>NGX_ERROR</literal> on error. For example, | |
516 <literal>ngx_atofp("10.5", 4, 2)</literal> returns <literal>1050</literal> | |
517 </listitem> | |
518 | |
519 <listitem> | |
520 <literal>ngx_hextoi(line, n)</literal> - converts hexadecimal representation of | |
521 a positive integer to <literal>ngx_int_t</literal>. Returns | |
522 <literal>NGX_ERROR</literal> on error | |
523 </listitem> | |
524 | |
525 </list> | |
526 </para> | |
527 | |
528 </section> | |
529 | |
1919
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
530 <section name="Regular expressions" id="regex"> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
531 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
532 <para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
533 The regular expressions interface in nginx is a wrapper around |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
534 the <link url="http://www.pcre.org">PCRE</link> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
535 library. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
536 The corresponding header file is <path>src/core/ngx_regex.h</path>. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
537 </para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
538 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
539 <para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
540 To use a regular expression for string matching, first, it needs to be |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
541 compiled, this is usually done at configuration phase. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
542 Note that since PCRE support is optional, all code using the interface must |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
543 be protected by the surrounding <literal>NGX_PCRE</literal> macro: |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
544 <programlisting> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
545 #if (NGX_PCRE) |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
546 ngx_regex_t *re; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
547 ngx_regex_compile_t rc; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
548 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
549 u_char errstr[NGX_MAX_CONF_ERRSTR]; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
550 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
551 ngx_str_t value = ngx_string("message (\\d\\d\\d).*Codeword is '(?<cw>\\w+)'"); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
552 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
553 ngx_memzero(&rc, sizeof(ngx_regex_compile_t)); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
554 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
555 rc.pattern = value; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
556 rc.pool = cf->pool; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
557 rc.err.len = NGX_MAX_CONF_ERRSTR; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
558 rc.err.data = errstr; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
559 /* rc.options are passed as is to pcre_compile() */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
560 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
561 if (ngx_regex_compile(&rc) != NGX_OK) { |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
562 ngx_conf_log_error(NGX_LOG_EMERG, cf, 0, "%V", &rc.err); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
563 return NGX_CONF_ERROR; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
564 } |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
565 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
566 re = rc.regex; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
567 #endif |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
568 </programlisting> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
569 After successful compilation, <literal>ngx_regex_compile_t</literal> structure |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
570 fields <literal>captures</literal> and <literal>named_captures</literal> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
571 are filled with count of all and named captures respectively found in the |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
572 regular expression. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
573 </para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
574 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
575 <para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
576 Later, the compiled regular expression may be used to match strings against it: |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
577 <programlisting> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
578 ngx_int_t n; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
579 int captures[(1 + rc.captures) * 3]; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
580 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
581 ngx_str_t input = ngx_string("This is message 123. Codeword is 'foobar'."); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
582 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
583 n = ngx_regex_exec(re, &input, captures, (1 + rc.captures) * 3); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
584 if (n >= 0) { |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
585 /* string matches expression */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
586 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
587 } else if (n == NGX_REGEX_NO_MATCHED) { |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
588 /* no match was found */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
589 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
590 } else { |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
591 /* some error */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
592 ngx_log_error(NGX_LOG_ALERT, log, 0, ngx_regex_exec_n " failed: %i", n); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
593 } |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
594 </programlisting> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
595 The arguments of <literal>ngx_regex_exec()</literal> are: the compiled regular |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
596 expression <literal>re</literal>, string to match <literal>s</literal>, |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
597 optional array of integers to hold found <literal>captures</literal> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
598 and its <literal>size</literal>. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
599 The <literal>captures</literal> array size must be a multiple of three, |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
600 per requirements of the |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
601 <link url="http://www.pcre.org/original/doc/html/pcreapi.html">PCRE API</link>. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
602 In the example, its size is calculated from a total number of captures plus |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
603 one for the matched string itself. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
604 </para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
605 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
606 <para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
607 Now, if there are matches, captures may be accessed: |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
608 <programlisting> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
609 u_char *p; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
610 size_t size; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
611 ngx_str_t name, value; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
612 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
613 /* all captures */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
614 for (i = 0; i < n * 2; i += 2) { |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
615 value.data = input.data + captures[i]; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
616 value.len = captures[i + 1] - captures[i]; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
617 } |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
618 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
619 /* accessing named captures */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
620 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
621 size = rc.name_size; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
622 p = rc.names; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
623 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
624 for (i = 0; i < rc.named_captures; i++, p += size) { |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
625 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
626 /* capture name */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
627 name.data = &p[2]; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
628 name.len = ngx_strlen(name.data); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
629 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
630 n = 2 * ((p[0] << 8) + p[1]); |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
631 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
632 /* captured value */ |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
633 value.data = &input.data[captures[n]]; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
634 value.len = captures[n + 1] - captures[n]; |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
635 } |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
636 </programlisting> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
637 </para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
638 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
639 <para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
640 The <literal>ngx_regex_exec_array()</literal> function accepts the array of |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
641 <literal>ngx_regex_elt_t</literal> elements (which are just compiled regular |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
642 expressions with associated names), a string to match and a log. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
643 The function will apply expressions from the array to the string until |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
644 the match is found or no more expressions are left. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
645 The return value is <literal>NGX_OK</literal> in case of match and |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
646 <literal>NGX_DECLINED</literal> otherwise, or <literal>NGX_ERROR</literal> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
647 in case of error. |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
648 </para> |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
649 |
dcfb4f3ac8a7
Added the "Regular expressions" section to the development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1915
diff
changeset
|
650 </section> |
1899 | 651 |
652 </section> | |
653 | |
654 | |
655 <section name="Containers" id="containers"> | |
656 | |
657 | |
658 <section name="Array" id="array"> | |
659 | |
660 <para> | |
661 The nginx array type <literal>ngx_array_t</literal> is defined as follows | |
662 </para> | |
663 | |
664 | |
665 <programlisting> | |
666 typedef struct { | |
667 void *elts; | |
668 ngx_uint_t nelts; | |
669 size_t size; | |
670 ngx_uint_t nalloc; | |
671 ngx_pool_t *pool; | |
672 } ngx_array_t; | |
673 </programlisting> | |
674 | |
675 <para> | |
676 The elements of array are available through the <literal>elts</literal> field. | |
677 The number of elements is held in the <literal>nelts</literal> field. | |
678 The <literal>size</literal> field holds the size of a single element and is set | |
679 when initializing the array. | |
680 </para> | |
681 | |
682 <para> | |
683 An array can be created in a pool with the | |
684 <literal>ngx_array_create(pool, n, size)</literal> call. | |
685 An already allocated array object can be initialized with the | |
686 <literal>ngx_array_init(array, pool, n, size)</literal> call. | |
687 </para> | |
688 | |
689 | |
690 <programlisting> | |
691 ngx_array_t *a, b; | |
692 | |
693 /* create an array of strings with preallocated memory for 10 elements */ | |
694 a = ngx_array_create(pool, 10, sizeof(ngx_str_t)); | |
695 | |
696 /* initialize string array for 10 elements */ | |
697 ngx_array_init(&b, pool, 10, sizeof(ngx_str_t)); | |
698 </programlisting> | |
699 | |
700 <para> | |
701 Adding elements to array are done with the following functions: | |
702 </para> | |
703 | |
704 <para> | |
705 <list type="bullet"> | |
706 | |
707 <listitem> | |
708 <literal>ngx_array_push(a)</literal> adds one tail element and returns pointer | |
709 to it | |
710 </listitem> | |
711 | |
712 <listitem> | |
713 <literal>ngx_array_push_n(a, n)</literal> adds <literal>n</literal> tail elements | |
714 and returns pointer to the first one | |
715 </listitem> | |
716 | |
717 </list> | |
718 </para> | |
719 | |
720 <para> | |
721 If currently allocated memory is not enough for new elements, a new memory for | |
722 elements is allocated and existing elements are copied to that memory. | |
723 The new memory block is normally twice as large, as the existing one. | |
724 </para> | |
725 | |
726 | |
727 <programlisting> | |
728 s = ngx_array_push(a); | |
729 ss = ngx_array_push_n(&b, 3); | |
730 </programlisting> | |
731 | |
732 </section> | |
733 | |
734 | |
735 <section name="List" id="list"> | |
736 | |
737 <para> | |
738 List in nginx is a sequence of arrays, optimized for inserting a potentially | |
739 large number of items. The list type is defined as follows: | |
740 </para> | |
741 | |
742 | |
743 <programlisting> | |
744 typedef struct { | |
745 ngx_list_part_t *last; | |
746 ngx_list_part_t part; | |
747 size_t size; | |
748 ngx_uint_t nalloc; | |
749 ngx_pool_t *pool; | |
750 } ngx_list_t; | |
751 </programlisting> | |
752 | |
753 <para> | |
754 The actual items are store in list parts, defined as follows: | |
755 </para> | |
756 | |
757 | |
758 <programlisting> | |
759 typedef struct ngx_list_part_s ngx_list_part_t; | |
760 | |
761 struct ngx_list_part_s { | |
762 void *elts; | |
763 ngx_uint_t nelts; | |
764 ngx_list_part_t *next; | |
765 }; | |
766 </programlisting> | |
767 | |
768 <para> | |
769 Initially, a list must be initialized by calling | |
770 <literal>ngx_list_init(list, pool, n, size)</literal> or created by calling | |
771 <literal>ngx_list_create(pool, n, size)</literal>. | |
772 Both functions receive the size of a single item and a number of items per list | |
773 part. | |
774 The <literal>ngx_list_push(list)</literal> function is used to add an item to the | |
775 list. Iterating over the items is done by direct accessing the list fields, as | |
776 seen in the example: | |
777 </para> | |
778 | |
779 | |
780 <programlisting> | |
781 ngx_str_t *v; | |
782 ngx_uint_t i; | |
783 ngx_list_t *list; | |
784 ngx_list_part_t *part; | |
785 | |
786 list = ngx_list_create(pool, 100, sizeof(ngx_str_t)); | |
787 if (list == NULL) { /* error */ } | |
788 | |
789 /* add items to the list */ | |
790 | |
791 v = ngx_list_push(list); | |
792 if (v == NULL) { /* error */ } | |
793 ngx_str_set(v, “foo”); | |
794 | |
795 v = ngx_list_push(list); | |
796 if (v == NULL) { /* error */ } | |
797 ngx_str_set(v, “bar”); | |
798 | |
799 /* iterate over the list */ | |
800 | |
801 part = &list->part; | |
802 v = part->elts; | |
803 | |
804 for (i = 0; /* void */; i++) { | |
805 | |
806 if (i >= part->nelts) { | |
807 if (part->next == NULL) { | |
808 break; | |
809 } | |
810 | |
811 part = part->next; | |
812 v = part->elts; | |
813 i = 0; | |
814 } | |
815 | |
816 ngx_do_smth(&v[i]); | |
817 } | |
818 </programlisting> | |
819 | |
820 <para> | |
821 The primary use for the list in nginx is HTTP input and output headers. | |
822 </para> | |
823 | |
824 <para> | |
825 The list does not support item removal. | |
826 However, when needed, items can internally be marked as missing without actual | |
827 removing from the list. | |
828 For example, HTTP output headers which are stored as | |
829 <literal>ngx_table_elt_t</literal> objects, are marked as missing by setting | |
830 the <literal>hash</literal> field of <literal>ngx_table_elt_t</literal> to | |
831 zero. Such items are explicitly skipped, when iterating over the headers. | |
832 </para> | |
833 | |
834 </section> | |
835 | |
836 | |
837 <section name="Queue" id="queue"> | |
838 | |
839 <para> | |
840 Queue in nginx is an intrusive doubly linked list, with each node defined as | |
841 follows: | |
842 </para> | |
843 | |
844 | |
845 <programlisting> | |
846 typedef struct ngx_queue_s ngx_queue_t; | |
847 | |
848 struct ngx_queue_s { | |
849 ngx_queue_t *prev; | |
850 ngx_queue_t *next; | |
851 }; | |
852 </programlisting> | |
853 | |
854 <para> | |
855 The head queue node is not linked with any data. Before using, the list head | |
856 should be initialized with <literal>ngx_queue_init(q)</literal> call. | |
857 Queues support the following operations: | |
858 </para> | |
859 | |
860 <para> | |
861 <list type="bullet"> | |
862 | |
863 <listitem> | |
864 <literal>ngx_queue_insert_head(h, x)</literal>, | |
865 <literal>ngx_queue_insert_tail(h, x)</literal> - insert a new node | |
866 </listitem> | |
867 | |
868 <listitem> | |
869 <literal>ngx_queue_remove(x)</literal> - remove a queue node | |
870 </listitem> | |
871 | |
872 <listitem> | |
873 <literal>ngx_queue_split(h, q, n)</literal> - split a queue at a node, | |
874 queue tail is returned in a separate queue | |
875 </listitem> | |
876 | |
877 <listitem> | |
878 <literal>ngx_queue_add(h, n)</literal> - add second queue to the first queue | |
879 </listitem> | |
880 | |
881 <listitem> | |
882 <literal>ngx_queue_head(h)</literal>, | |
883 <literal>ngx_queue_last(h)</literal> - get first or last queue node | |
884 </listitem> | |
885 | |
886 <listitem> | |
887 <literal>ngx_queue_sentinel(h)</literal> | |
888 - get a queue sentinel object to end iteration at | |
889 </listitem> | |
890 | |
891 <listitem> | |
892 <literal>ngx_queue_data(q, type, link)</literal> - get reference to the beginning of a | |
893 queue node data structure, considering the queue field offset in it | |
894 </listitem> | |
895 | |
896 </list> | |
897 </para> | |
898 | |
899 <para> | |
900 Example: | |
901 </para> | |
902 | |
903 | |
904 <programlisting> | |
905 typedef struct { | |
906 ngx_str_t value; | |
907 ngx_queue_t queue; | |
908 } ngx_foo_t; | |
909 | |
910 ngx_foo_t *f; | |
911 ngx_queue_t values; | |
912 | |
913 ngx_queue_init(&values); | |
914 | |
915 f = ngx_palloc(pool, sizeof(ngx_foo_t)); | |
916 if (f == NULL) { /* error */ } | |
917 ngx_str_set(&f->value, “foo”); | |
918 | |
919 ngx_queue_insert_tail(&values, f); | |
920 | |
921 /* insert more nodes here */ | |
922 | |
923 for (q = ngx_queue_head(&values); | |
924 q != ngx_queue_sentinel(&values); | |
925 q = ngx_queue_next(q)) | |
926 { | |
927 f = ngx_queue_data(q, ngx_foo_t, queue); | |
928 | |
929 ngx_do_smth(&f->value); | |
930 } | |
931 </programlisting> | |
932 | |
933 </section> | |
934 | |
935 | |
936 <section name="Red-Black tree" id="red_black_tree"> | |
937 | |
938 <para> | |
1915
8b7c3b0ef1a4
Semantically marked paths.
Vladimir Homutov <vl@nginx.com>
parents:
1914
diff
changeset
|
939 The <path>src/core/ngx_rbtree.h</path> header file provides access to the |
1899 | 940 effective implementation of red-black trees. |
941 </para> | |
942 | |
943 | |
944 <programlisting> | |
945 typedef struct { | |
946 ngx_rbtree_t rbtree; | |
947 ngx_rbtree_node_t sentinel; | |
948 | |
949 /* custom per-tree data here */ | |
950 } my_tree_t; | |
951 | |
952 typedef struct { | |
953 ngx_rbtree_node_t rbnode; | |
954 | |
955 /* custom per-node data */ | |
956 foo_t val; | |
957 } my_node_t; | |
958 </programlisting> | |
959 | |
960 <para> | |
961 To deal with a tree as a whole, you need two nodes: root and sentinel. | |
962 Typically, they are added to some custom structure, thus allowing to | |
963 organize your data into a tree which leaves contain a link to or embed | |
964 your data. | |
965 </para> | |
966 | |
967 <para> | |
968 To initialize a tree: | |
969 </para> | |
970 | |
971 | |
972 <programlisting> | |
973 my_tree_t root; | |
974 | |
975 ngx_rbtree_init(&root.rbtree, &root.sentinel, insert_value_function); | |
976 </programlisting> | |
977 | |
978 <para> | |
979 The <literal>insert_value_function</literal> is a function that is | |
980 responsible for traversing the tree and inserting new values into correct | |
981 place. | |
982 For example, the <literal>ngx_str_rbtree_insert_value</literal> functions is | |
983 designed to deal with <literal>ngx_str_t</literal> type. | |
984 </para> | |
985 | |
986 | |
987 <programlisting> | |
988 void ngx_str_rbtree_insert_value(ngx_rbtree_node_t *temp, | |
989 ngx_rbtree_node_t *node, | |
990 ngx_rbtree_node_t *sentinel) | |
991 </programlisting> | |
992 | |
993 <para> | |
994 Its arguments are pointers to a root node of an insertion, newly created node | |
995 to be added, and a tree sentinel. | |
996 </para> | |
997 | |
998 <para> | |
999 The traversal is pretty straightforward and can be demonstrated with the | |
1000 following lookup function pattern: | |
1001 </para> | |
1002 | |
1003 | |
1004 <programlisting> | |
1005 my_node_t * | |
1006 my_rbtree_lookup(ngx_rbtree_t *rbtree, foo_t *val, uint32_t hash) | |
1007 { | |
1008 ngx_int_t rc; | |
1009 my_node_t *n; | |
1010 ngx_rbtree_node_t *node, *sentinel; | |
1011 | |
1012 node = rbtree->root; | |
1013 sentinel = rbtree->sentinel; | |
1014 | |
1015 while (node != sentinel) { | |
1016 | |
1017 n = (my_node_t *) node; | |
1018 | |
1019 if (hash != node->key) { | |
1020 node = (hash < node->key) ? node->left : node->right; | |
1021 continue; | |
1022 } | |
1023 | |
1024 rc = compare(val, node->val); | |
1025 | |
1026 if (rc < 0) { | |
1027 node = node->left; | |
1028 continue; | |
1029 } | |
1030 | |
1031 if (rc > 0) { | |
1032 node = node->right; | |
1033 continue; | |
1034 } | |
1035 | |
1036 return n; | |
1037 } | |
1038 | |
1039 return NULL; | |
1040 } | |
1041 </programlisting> | |
1042 | |
1043 <para> | |
1044 The <literal>compare()</literal> is a classic comparator function returning | |
1045 value less, equal or greater than zero. To speed up lookups and avoid comparing | |
1046 user objects that can be big, integer hash field is used. | |
1047 </para> | |
1048 | |
1049 <para> | |
1050 To add a node to a tree, allocate a new node, initialize it and call | |
1051 <literal>ngx_rbtree_insert()</literal>: | |
1052 </para> | |
1053 | |
1054 | |
1055 <programlisting> | |
1056 my_node_t *my_node; | |
1057 ngx_rbtree_node_t *node; | |
1058 | |
1059 my_node = ngx_palloc(...); | |
1060 init_custom_data(&my_node->val); | |
1061 | |
1062 node = &my_node->rbnode; | |
1063 node->key = create_key(my_node->val); | |
1064 | |
1065 ngx_rbtree_insert(&root->rbtree, node); | |
1066 </programlisting> | |
1067 | |
1068 <para> | |
1069 to remove a node: | |
1070 </para> | |
1071 | |
1072 | |
1073 <programlisting> | |
1074 ngx_rbtree_delete(&root->rbtree, node); | |
1075 </programlisting> | |
1076 | |
1077 </section> | |
1078 | |
1914
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1079 <section name="Hash" id="hash"> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1080 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1081 <para> |
1920 | 1082 Hash table functions are declared in <path>src/core/ngx_hash.h</path>. |
1914
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1083 Exact and wildcard matching is supported. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1084 The latter requires extra setup and is described in a separate section below. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1085 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1086 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1087 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1088 To initialize a hash, one needs to know the number of elements in advance, |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1089 so that nginx can build the hash optimally. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1090 Two parameters that need to be configured are <literal>max_size</literal> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1091 and <literal>bucket_size</literal>. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1092 The details of setting up these are provided in a separate |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1093 <link doc="../hash.xml">document</link>. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1094 Usually, these two parameters are configurable by user. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1095 Hash initialization settings are stored as the |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1096 <literal>ngx_hash_init_t</literal> type, |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1097 and the hash itself is <literal>ngx_hash_t</literal>: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1098 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1099 ngx_hash_t foo_hash; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1100 ngx_hash_init_t hash; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1101 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1102 hash.hash = &foo_hash; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1103 hash.key = ngx_hash_key; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1104 hash.max_size = 512; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1105 hash.bucket_size = ngx_align(64, ngx_cacheline_size); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1106 hash.name = "foo_hash"; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1107 hash.pool = cf->pool; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1108 hash.temp_pool = cf->temp_pool; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1109 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1110 The <literal>key</literal> is a pointer to a function that creates hash integer |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1111 key from a string. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1112 Two generic functions are provided: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1113 <literal>ngx_hash_key(data, len)</literal> and |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1114 <literal>ngx_hash_key_lc(data, len)</literal>. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1115 The latter converts a string to lowercase and thus requires the passed string to |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1116 be writable. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1117 If this is not true, <literal>NGX_HASH_READONLY_KEY</literal> flag |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1118 may be passed to the function, initializing array keys (see below). |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1119 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1120 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1121 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1122 The hash keys are stored in <literal>ngx_hash_keys_arrays_t</literal> and |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1123 are initialized with <literal>ngx_hash_keys_array_init(arr, type)</literal>: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1124 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1125 ngx_hash_keys_arrays_t foo_keys; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1126 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1127 foo_keys.pool = cf->pool; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1128 foo_keys.temp_pool = cf->temp_pool; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1129 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1130 ngx_hash_keys_array_init(&foo_keys, NGX_HASH_SMALL); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1131 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1132 The second parameter can be either <literal>NGX_HASH_SMALL</literal> or |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1133 <literal>NGX_HASH_LARGE</literal> and controls the amount of preallocated |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1134 resources for the hash. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1135 If you expect the hash to contain thousands elements, |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1136 use <literal>NGX_HASH_LARGE</literal>. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1137 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1138 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1139 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1140 The <literal>ngx_hash_add_key(keys_array, key, value, flags)</literal> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1141 function is used to insert keys into hash keys array; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1142 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1143 ngx_str_t k1 = ngx_string("key1"); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1144 ngx_str_t k2 = ngx_string("key2"); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1145 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1146 ngx_hash_add_key(&foo_keys, &k1, &my_data_ptr_1, NGX_HASH_READONLY_KEY); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1147 ngx_hash_add_key(&foo_keys, &k2, &my_data_ptr_2, NGX_HASH_READONLY_KEY); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1148 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1149 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1150 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1151 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1152 Now, the hash table may be built using the call to |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1153 <literal>ngx_hash_init(hinit, key_names, nelts)</literal>: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1154 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1155 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1156 ngx_hash_init(&hash, foo_keys.keys.elts, foo_keys.keys.nelts); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1157 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1158 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1159 This may fail, if <literal>max_size</literal> or <literal>bucket_size</literal> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1160 parameters are not big enough. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1161 When the hash is built, <literal>ngx_hash_find(hash, key, name, len)</literal> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1162 function may be used to look up elements: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1163 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1164 my_data_t *data; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1165 ngx_uint_t key; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1166 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1167 key = ngx_hash_key(k1.data, k1.len); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1168 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1169 data = ngx_hash_find(&foo_hash, key, k1.data, k1.len); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1170 if (data == NULL) { |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1171 /* key not found */ |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1172 } |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1173 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1174 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1175 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1176 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1177 <section name="Wildcard matching" id="wildcard_matching"> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1178 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1179 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1180 To create a hash that works with wildcards, |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1181 <literal>ngx_hash_combined_t</literal> type is used. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1182 It includes the hash type described above and has two additional keys arrays: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1183 <literal>dns_wc_head</literal> and <literal>dns_wc_tail</literal>. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1184 The initialization of basic properties is done similarly to a usual hash: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1185 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1186 ngx_hash_init_t hash |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1187 ngx_hash_combined_t foo_hash; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1188 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1189 hash.hash = &foo_hash.hash; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1190 hash.key = ...; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1191 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1192 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1193 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1194 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1195 It is possible to add wildcard keys using the |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1196 <literal>NGX_HASH_WILDCARD_KEY</literal> flag: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1197 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1198 /* k1 = ".example.org"; */ |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1199 /* k2 = "foo.*"; */ |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1200 ngx_hash_add_key(&foo_keys, &k1, &data1, NGX_HASH_WILDCARD_KEY); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1201 ngx_hash_add_key(&foo_keys, &k2, &data2, NGX_HASH_WILDCARD_KEY); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1202 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1203 The function recognizes wildcards and adds keys into corresponding arrays. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1204 Please refer to the |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1205 <link doc="../http/ngx_http_map_module.xml" id="map"/> module |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1206 documentation for the description of the wildcard syntax and |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1207 matching algorithm. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1208 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1209 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1210 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1211 Depending on the contents of added keys, you may need to initialize up to three |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1212 keys arrays: one for exact matching (described above), and two for matching |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1213 starting from head or tail of a string: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1214 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1215 if (foo_keys.dns_wc_head.nelts) { |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1216 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1217 ngx_qsort(foo_keys.dns_wc_head.elts, |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1218 (size_t) foo_keys.dns_wc_head.nelts, |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1219 sizeof(ngx_hash_key_t), |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1220 cmp_dns_wildcards); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1221 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1222 hash.hash = NULL; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1223 hash.temp_pool = pool; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1224 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1225 if (ngx_hash_wildcard_init(&hash, foo_keys.dns_wc_head.elts, |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1226 foo_keys.dns_wc_head.nelts) |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1227 != NGX_OK) |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1228 { |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1229 return NGX_ERROR; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1230 } |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1231 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1232 foo_hash.wc_head = (ngx_hash_wildcard_t *) hash.hash; |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1233 } |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1234 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1235 The keys array needs to be sorted, and initialization results must be added |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1236 to the combined hash. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1237 The initialization of <literal>dns_wc_tail</literal> array is done similarly. |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1238 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1239 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1240 <para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1241 The lookup in a combined hash is handled by the |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1242 <literal>ngx_hash_find_combined(chash, key, name, len)</literal>: |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1243 <programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1244 /* key = "bar.example.org"; - will match ".example.org" */ |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1245 /* key = "foo.example.com"; - will match "foo.*" */ |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1246 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1247 hkey = ngx_hash_key(key.data, key.len); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1248 res = ngx_hash_find_combined(&foo_hash, hkey, key.data, key.len); |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1249 </programlisting> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1250 </para> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1251 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1252 </section> |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1253 |
f8659301a260
Added hash API description.
Vladimir Homutov <vl@nginx.com>
parents:
1907
diff
changeset
|
1254 </section> |
1899 | 1255 |
1256 </section> | |
1257 | |
1258 | |
1259 <section name="Memory management" id="memory_management"> | |
1260 | |
1261 | |
1262 <section name="Heap" id="heap"> | |
1263 | |
1264 <para> | |
1265 To allocate memory from system heap, the following functions are provided by | |
1266 nginx: | |
1267 </para> | |
1268 | |
1269 <para> | |
1270 <list type="bullet"> | |
1271 | |
1272 <listitem> | |
1273 <literal>ngx_alloc(size, log)</literal> - allocate memory from system heap. | |
1274 This is a wrapper around <literal>malloc()</literal> with logging support. | |
1275 Allocation error and debugging information is logged to <literal>log</literal> | |
1276 </listitem> | |
1277 | |
1278 <listitem> | |
1279 <literal>ngx_calloc(size, log)</literal> - same as | |
1280 <literal>ngx_alloc()</literal>, but memory is filled with zeroes after | |
1281 allocation | |
1282 </listitem> | |
1283 | |
1284 <listitem> | |
1285 <literal>ngx_memalign(alignment, size, log)</literal> - allocate aligned memory | |
1286 from system heap. This is a wrapper around <literal>posix_memalign()</literal> | |
1287 on those platforms which provide it. | |
1288 Otherwise implementation falls back to <literal>ngx_alloc()</literal> which | |
1289 provides maximum alignment | |
1290 </listitem> | |
1291 | |
1292 <listitem> | |
1293 <literal>ngx_free(p)</literal> - free allocated memory. | |
1294 This is a wrapper around <literal>free()</literal> | |
1295 </listitem> | |
1296 | |
1297 </list> | |
1298 </para> | |
1299 | |
1300 </section> | |
1301 | |
1302 | |
1303 <section name="Pool" id="pool"> | |
1304 | |
1305 <para> | |
1306 Most nginx allocations are done in pools. Memory allocated in an nginx pool is | |
1307 freed automatically when the pool in destroyed. This provides good | |
1308 allocation performance and makes memory control easy. | |
1309 </para> | |
1310 | |
1311 <para> | |
1312 A pool internally allocates objects in continuous blocks of memory. Once a | |
1313 block is full, a new one is allocated and added to the pool memory | |
1314 block list. When a large allocation is requested which does not fit into | |
1315 a block, such allocation is forwarded to the system allocator and the | |
1316 returned pointer is stored in the pool for further deallocation. | |
1317 </para> | |
1318 | |
1319 <para> | |
1320 Nginx pool has the type <literal>ngx_pool_t</literal>. | |
1321 The following operations are supported: | |
1322 </para> | |
1323 | |
1324 <para> | |
1325 <list type="bullet"> | |
1326 | |
1327 <listitem> | |
1328 <literal>ngx_create_pool(size, log)</literal> - create a pool with given | |
1329 block size. The pool object returned is allocated in the pool as well. | |
1330 </listitem> | |
1331 | |
1332 <listitem> | |
1333 <literal>ngx_destroy_pool(pool)</literal> - free all pool memory, including | |
1334 the pool object itself. | |
1335 </listitem> | |
1336 | |
1337 <listitem> | |
1338 <literal>ngx_palloc(pool, size)</literal> - allocate aligned memory from pool | |
1339 </listitem> | |
1340 | |
1341 <listitem> | |
1342 <literal>ngx_pcalloc(pool, size)</literal> - allocated aligned memory | |
1343 from pool and fill it with zeroes | |
1344 </listitem> | |
1345 | |
1346 <listitem> | |
1347 <literal>ngx_pnalloc(pool, size)</literal> - allocate unaligned memory from pool. | |
1348 Mostly used for allocating strings | |
1349 </listitem> | |
1350 | |
1351 <listitem> | |
1352 <literal>ngx_pfree(pool, p)</literal> - free memory, previously allocated | |
1353 in the pool. | |
1354 Only allocations, forwarded to the system allocator, can be freed. | |
1355 </listitem> | |
1356 | |
1357 </list> | |
1358 </para> | |
1359 | |
1360 <programlisting> | |
1361 u_char *p; | |
1362 ngx_str_t *s; | |
1363 ngx_pool_t *pool; | |
1364 | |
1365 pool = ngx_create_pool(1024, log); | |
1366 if (pool == NULL) { /* error */ } | |
1367 | |
1368 s = ngx_palloc(pool, sizeof(ngx_str_t)); | |
1369 if (s == NULL) { /* error */ } | |
1370 ngx_str_set(s, “foo”); | |
1371 | |
1372 p = ngx_pnalloc(pool, 3); | |
1373 if (p == NULL) { /* error */ } | |
1374 ngx_memcpy(p, “foo”, 3); | |
1375 </programlisting> | |
1376 | |
1377 <para> | |
1378 Since chain links <literal>ngx_chain_t</literal> are actively used in nginx, | |
1379 nginx pool provides a way to reuse them. | |
1380 The <literal>chain</literal> field of <literal>ngx_pool_t</literal> keeps a | |
1381 list of previously allocated links ready for reuse. For efficient allocation of | |
1382 a chain link in a pool, the function | |
1383 <literal>ngx_alloc_chain_link(pool)</literal> should be used. | |
1384 This function looks up a free chain link in the pool list and only if it's | |
1385 empty allocates a new one. To free a link <literal>ngx_free_chain(pool, cl)</literal> | |
1386 should be called. | |
1387 </para> | |
1388 | |
1389 <para> | |
1390 Cleanup handlers can be registered in a pool. | |
1391 Cleanup handler is a callback with an argument which is called when pool is | |
1392 destroyed. | |
1393 Pool is usually tied with a specific nginx object (like HTTP request) and | |
1394 destroyed in the end of that object’s lifetime, releasing the object itself. | |
1902 | 1395 Registering a pool cleanup is a convenient way to release resources, close file |
1899 | 1396 descriptors or make final adjustments to shared data, associated with the main |
1397 object. | |
1398 </para> | |
1399 | |
1400 <para> | |
1401 A pool cleanup is registered by calling <literal>ngx_pool_cleanup_add(pool, | |
1402 size)</literal> which returns <literal>ngx_pool_cleanup_t</literal> pointer to | |
1403 be filled by the caller. The <literal>size</literal> argument allows allocating | |
1404 context for the cleanup handler. | |
1405 </para> | |
1406 | |
1407 | |
1408 <programlisting> | |
1409 ngx_pool_cleanup_t *cln; | |
1410 | |
1411 cln = ngx_pool_cleanup_add(pool, 0); | |
1412 if (cln == NULL) { /* error */ } | |
1413 | |
1414 cln->handler = ngx_my_cleanup; | |
1415 cln->data = “foo”; | |
1416 | |
1417 ... | |
1418 | |
1419 static void | |
1420 ngx_my_cleanup(void *data) | |
1421 { | |
1422 u_char *msg = data; | |
1423 | |
1424 ngx_do_smth(msg); | |
1425 } | |
1426 </programlisting> | |
1427 | |
1428 </section> | |
1429 | |
1430 | |
1431 <section name="Shared memory" id="shared_memory"> | |
1432 | |
1433 <para> | |
1434 Shared memory is used by nginx to share common data between processes. | |
1435 Function <literal>ngx_shared_memory_add(cf, name, size, tag)</literal> adds a | |
1436 new shared memory entry <literal>ngx_shm_zone_t</literal> to the cycle. The | |
1437 function receives <literal>name</literal> and <literal>size</literal> of the | |
1438 zone. | |
1439 Each shared zone must have a unique name. | |
1440 If a shared zone entry with the provided name exists, the old zone entry is | |
1441 reused, if its tag value matches too. | |
1442 Mismatched tag is considered an error. | |
1443 Usually, the address of the module structure is passed as tag, making it | |
1444 possible to reuse shared zones by name within one nginx module. | |
1445 </para> | |
1446 | |
1447 <para> | |
1448 The shared memory entry structure <literal>ngx_shm_zone_t</literal> has the | |
1449 following fields: | |
1450 </para> | |
1451 | |
1452 <para> | |
1453 <list type="bullet"> | |
1454 | |
1455 <listitem> | |
1456 <literal>init</literal> - initialization callback, called after shared zone is | |
1457 mapped to actual memory | |
1458 </listitem> | |
1459 | |
1460 <listitem> | |
1461 <literal>data</literal> - data context, used to pass arbitrary data to the | |
1462 <literal>init</literal> callback | |
1463 </listitem> | |
1464 | |
1465 <listitem> | |
1466 <literal>noreuse</literal> - flag, disabling shared zone reuse from the | |
1467 old cycle | |
1468 </listitem> | |
1469 | |
1470 <listitem> | |
1471 <literal>tag</literal> - shared zone tag | |
1472 </listitem> | |
1473 | |
1474 <listitem> | |
1475 <literal>shm</literal> - platform-specific object of type | |
1476 <literal>ngx_shm_t</literal>, having at least the following fields: | |
1477 <list type="bullet"> | |
1478 | |
1479 <listitem> | |
1480 <literal>addr</literal> - mapped shared memory address, initially NULL | |
1481 </listitem> | |
1482 | |
1483 <listitem> | |
1484 <literal>size</literal> - shared memory size | |
1485 </listitem> | |
1486 | |
1487 <listitem> | |
1488 <literal>name</literal> - shared memory name | |
1489 </listitem> | |
1490 | |
1491 <listitem> | |
1492 <literal>log</literal> - shared memory log | |
1493 </listitem> | |
1494 | |
1495 <listitem> | |
1496 <literal>exists</literal> - flag, showing that shared memory was inherited | |
1497 from the master process (Windows-specific) | |
1498 </listitem> | |
1499 | |
1500 </list> | |
1501 </listitem> | |
1502 | |
1503 </list> | |
1504 </para> | |
1505 | |
1506 <para> | |
1507 Shared zone entries are mapped to actual memory in | |
1508 <literal>ngx_init_cycle()</literal> after configuration is parsed. | |
1509 On POSIX systems, <literal>mmap()</literal> syscall is used to create shared | |
1510 anonymous mapping. | |
1511 On Windows, <literal>CreateFileMapping()/MapViewOfFileEx()</literal> pair is | |
1512 used. | |
1513 </para> | |
1514 | |
1515 <para> | |
1516 For allocating in shared memory, nginx provides slab pool | |
1517 <literal>ngx_slab_pool_t</literal>. | |
1518 In each nginx shared zone, a slab pool is automatically created for allocating | |
1519 memory in that zone. | |
1520 The pool is located in the beginning of the shared zone and can be accessed by | |
1521 the expression <literal>(ngx_slab_pool_t *) shm_zone->shm.addr</literal>. | |
1522 Allocation in shared zone is done by calling one of the functions | |
1523 <literal>ngx_slab_alloc(pool, size)/ngx_slab_calloc(pool, size)</literal>. | |
1524 Memory is freed by calling <literal>ngx_slab_free(pool, p)</literal>. | |
1525 </para> | |
1526 | |
1527 <para> | |
1528 Slab pool divides all shared zone into pages. | |
1529 Each page is used for allocating objects of the same size. | |
1530 Only the sizes which are powers of 2, and not less than 8, are considered. | |
1531 Other sizes are rounded up to one of these values. | |
1532 For each page, a bitmask is kept, showing which blocks within that page are in | |
1533 use and which are free for allocation. | |
1534 For sizes greater than half-page (usually, 2048 bytes), allocation is done by | |
1535 entire pages. | |
1536 </para> | |
1537 | |
1538 <para> | |
1539 To protect data in shared memory from concurrent access, mutex is available in | |
1540 the <literal>mutex</literal> field of <literal>ngx_slab_pool_t</literal>. | |
1541 The mutex is used by the slab pool while allocating and freeing memory. | |
1542 However, it can be used to protect any other user data structures, | |
1543 allocated in the shared zone. | |
1544 Locking is done by calling | |
1545 <literal>ngx_shmtx_lock(&shpool->mutex)</literal>, unlocking is done by | |
1546 calling <literal>ngx_shmtx_unlock(&shpool->mutex)</literal>. | |
1547 </para> | |
1548 | |
1549 | |
1550 <programlisting> | |
1551 ngx_str_t name; | |
1552 ngx_foo_ctx_t *ctx; | |
1553 ngx_shm_zone_t *shm_zone; | |
1554 | |
1555 ngx_str_set(&name, "foo"); | |
1556 | |
1557 /* allocate shared zone context */ | |
1558 ctx = ngx_pcalloc(cf->pool, sizeof(ngx_foo_ctx_t)); | |
1559 if (ctx == NULL) { | |
1560 /* error */ | |
1561 } | |
1562 | |
1563 /* add an entry for 65k shared zone */ | |
1564 shm_zone = ngx_shared_memory_add(cf, &name, 65536, &ngx_foo_module); | |
1565 if (shm_zone == NULL) { | |
1566 /* error */ | |
1567 } | |
1568 | |
1569 /* register init callback and context */ | |
1570 shm_zone->init = ngx_foo_init_zone; | |
1571 shm_zone->data = ctx; | |
1572 | |
1573 | |
1574 ... | |
1575 | |
1576 | |
1577 static ngx_int_t | |
1578 ngx_foo_init_zone(ngx_shm_zone_t *shm_zone, void *data) | |
1579 { | |
1580 ngx_foo_ctx_t *octx = data; | |
1581 | |
1582 size_t len; | |
1583 ngx_foo_ctx_t *ctx; | |
1584 ngx_slab_pool_t *shpool; | |
1585 | |
1586 value = shm_zone->data; | |
1587 | |
1588 if (octx) { | |
1589 /* reusing a shared zone from old cycle */ | |
1590 ctx->value = octx->value; | |
1591 return NGX_OK; | |
1592 } | |
1593 | |
1594 shpool = (ngx_slab_pool_t *) shm_zone->shm.addr; | |
1595 | |
1596 if (shm_zone->shm.exists) { | |
1597 /* initialize shared zone context in Windows nginx worker */ | |
1598 ctx->value = shpool->data; | |
1599 return NGX_OK; | |
1600 } | |
1601 | |
1602 /* initialize shared zone */ | |
1603 | |
1604 ctx->value = ngx_slab_alloc(shpool, sizeof(ngx_uint_t)); | |
1605 if (ctx->value == NULL) { | |
1606 return NGX_ERROR; | |
1607 } | |
1608 | |
1609 shpool->data = ctx->value; | |
1610 | |
1611 return NGX_OK; | |
1612 } | |
1613 </programlisting> | |
1614 | |
1615 </section> | |
1616 | |
1617 | |
1618 </section> | |
1619 | |
1620 | |
1621 <section name="Logging" id="logging"> | |
1622 | |
1623 <para> | |
1624 For logging nginx code uses <literal>ngx_log_t</literal> objects. | |
1625 Nginx logger provides support for several types of output: | |
1626 | |
1627 <list type="bullet"> | |
1628 | |
1629 <listitem> | |
1630 stderr - logging to standard error output | |
1631 </listitem> | |
1632 | |
1633 <listitem> | |
1634 file - logging to file | |
1635 </listitem> | |
1636 | |
1637 <listitem> | |
1638 syslog - logging to syslog | |
1639 </listitem> | |
1640 | |
1641 <listitem> | |
1642 memory - logging to internal memory storage for development purposes. | |
1643 The memory could be accessed later with debugger | |
1644 </listitem> | |
1645 | |
1646 </list> | |
1647 </para> | |
1648 | |
1649 <para> | |
1650 A logger instance may actually be a chain of loggers, linked to each other with | |
1651 the <literal>next</literal> field. | |
1652 Each message is written to all loggers in chain. | |
1653 </para> | |
1654 | |
1655 <para> | |
1656 Each logger has an error level which limits the messages written to that log. | |
1657 The following error levels are supported by nginx: | |
1658 </para> | |
1659 | |
1660 <para> | |
1661 <list type="bullet"> | |
1662 | |
1663 <listitem> | |
1664 <literal>NGX_LOG_EMERG</literal> | |
1665 </listitem> | |
1666 | |
1667 <listitem> | |
1668 <literal>NGX_LOG_ALERT</literal> | |
1669 </listitem> | |
1670 | |
1671 <listitem> | |
1672 <literal>NGX_LOG_CRIT</literal> | |
1673 </listitem> | |
1674 | |
1675 <listitem> | |
1676 <literal>NGX_LOG_ERR</literal> | |
1677 </listitem> | |
1678 | |
1679 <listitem> | |
1680 <literal>NGX_LOG_WARN</literal> | |
1681 </listitem> | |
1682 | |
1683 <listitem> | |
1684 <literal>NGX_LOG_NOTICE</literal> | |
1685 </listitem> | |
1686 | |
1687 <listitem> | |
1688 <literal>NGX_LOG_INFO</literal> | |
1689 </listitem> | |
1690 | |
1691 <listitem> | |
1692 <literal>NGX_LOG_DEBUG</literal> | |
1693 </listitem> | |
1694 | |
1695 </list> | |
1696 </para> | |
1697 | |
1698 <para> | |
1699 For debug logging, debug mask is checked as well. The following debug masks | |
1700 exist: | |
1701 </para> | |
1702 | |
1703 <para> | |
1704 <list type="bullet"> | |
1705 | |
1706 <listitem> | |
1707 <literal>NGX_LOG_DEBUG_CORE</literal> | |
1708 </listitem> | |
1709 | |
1710 <listitem> | |
1711 <literal>NGX_LOG_DEBUG_ALLOC</literal> | |
1712 </listitem> | |
1713 | |
1714 <listitem> | |
1715 <literal>NGX_LOG_DEBUG_MUTEX</literal> | |
1716 </listitem> | |
1717 | |
1718 <listitem> | |
1719 <literal>NGX_LOG_DEBUG_EVENT</literal> | |
1720 </listitem> | |
1721 | |
1722 <listitem> | |
1723 <literal>NGX_LOG_DEBUG_HTTP</literal> | |
1724 </listitem> | |
1725 | |
1726 <listitem> | |
1727 <literal>NGX_LOG_DEBUG_MAIL</literal> | |
1728 </listitem> | |
1729 | |
1730 <listitem> | |
1731 <literal>NGX_LOG_DEBUG_STREAM</literal> | |
1732 </listitem> | |
1733 | |
1734 </list> | |
1735 </para> | |
1736 | |
1737 <para> | |
1738 Normally, loggers are created by existing nginx code from | |
1739 <literal>error_log</literal> directives and are available at nearly every stage | |
1740 of processing in cycle, configuration, client connection and other objects. | |
1741 </para> | |
1742 | |
1743 <para> | |
1744 Nginx provides the following logging macros: | |
1745 </para> | |
1746 | |
1747 <para> | |
1748 <list type="bullet"> | |
1749 | |
1750 <listitem> | |
1751 <literal>ngx_log_error(level, log, err, fmt, ...)</literal> - error logging | |
1752 </listitem> | |
1753 | |
1754 <listitem> | |
1755 <literal>ngx_log_debug0(level, log, err, fmt)</literal>, | |
1756 <literal>ngx_log_debug1(level, log, err, fmt, arg1)</literal> etc - debug | |
1757 logging, up to 8 formatting arguments are supported | |
1758 </listitem> | |
1759 | |
1760 </list> | |
1761 </para> | |
1762 | |
1763 <para> | |
1764 A log message is formatted in a buffer of size | |
1765 <literal>NGX_MAX_ERROR_STR</literal> (currently, 2048 bytes) on stack. | |
1766 The message is prepended with error level, process PID, connection id (stored | |
1767 in <literal>log->connection</literal>) and system error text. | |
1768 For non-debug messages <literal>log->handler</literal> is called as well to | |
1769 prepend more specific information to the log message. | |
1770 HTTP module sets <literal>ngx_http_log_error()</literal> function as log | |
1771 handler to log client and server addresses, current action (stored in | |
1772 <literal>log->action</literal>), client request line, server name etc. | |
1773 </para> | |
1774 | |
1775 <para> | |
1776 Example: | |
1777 </para> | |
1778 | |
1779 | |
1780 <programlisting> | |
1781 /* specify what is currently done */ | |
1782 log->action = "sending mp4 to client”; | |
1783 | |
1784 /* error and debug log */ | |
1785 ngx_log_error(NGX_LOG_INFO, c->log, 0, "client prematurely | |
1786 closed connection”); | |
1787 | |
1788 ngx_log_debug2(NGX_LOG_DEBUG_HTTP, mp4->file.log, 0, | |
1789 "mp4 start:%ui, length:%ui”, mp4->start, mp4->length); | |
1790 </programlisting> | |
1791 | |
1792 <para> | |
1793 Logging result: | |
1794 </para> | |
1795 | |
1796 | |
1797 <programlisting> | |
1798 2016/09/16 22:08:52 [info] 17445#0: *1 client prematurely closed connection while | |
1799 sending mp4 to client, client: 127.0.0.1, server: , request: "GET /file.mp4 HTTP/1.1” | |
1800 2016/09/16 23:28:33 [debug] 22140#0: *1 mp4 start:0, length:10000 | |
1801 </programlisting> | |
1802 | |
1803 </section> | |
1804 | |
1805 | |
1806 <section name="Cycle" id="cycle"> | |
1807 | |
1808 <para> | |
1809 Cycle object keeps nginx runtime context, created from a specific | |
1810 configuration. | |
1811 The type of the cycle is <literal>ngx_cycle_t</literal>. | |
1812 Upon configuration reload a new cycle is created from the new version of nginx | |
1813 configuration. | |
1814 The old cycle is usually deleted after a new one is successfully created. | |
1815 Currently active cycle is held in the <literal>ngx_cycle</literal> global | |
1816 variable and is inherited by newly started nginx workers. | |
1817 </para> | |
1818 | |
1819 <para> | |
1820 A cycle is created by the function <literal>ngx_init_cycle()</literal>. | |
1821 The function receives the old cycle as the argument. | |
1822 It's used to locate the configuration file and inherit as much resources as | |
1823 possible from the old cycle to keep nginx running smoothly. | |
1824 When nginx starts, a fake cycle called "init cycle" is created and is then | |
1825 replaced by a normal cycle, built from configuration. | |
1826 </para> | |
1827 | |
1828 <para> | |
1829 Some members of the cycle: | |
1830 </para> | |
1831 | |
1832 <para> | |
1833 <list type="bullet"> | |
1834 | |
1835 <listitem> | |
1836 <literal>pool</literal> - cycle pool. Created for each new cycle | |
1837 </listitem> | |
1838 | |
1839 <listitem> | |
1840 <literal>log</literal> - cycle log. Initially, this log is inherited from the | |
1841 old cycle. | |
1842 After reading configuration, this member is set to point to | |
1843 <literal>new_log</literal> | |
1844 </listitem> | |
1845 | |
1846 <listitem> | |
1847 <literal>new_log</literal> - cycle log, created by the configuration. | |
1848 It's affected by the root scope <literal>error_log</literal> directive | |
1849 </listitem> | |
1850 | |
1851 <listitem> | |
1852 <literal>connections</literal>, <literal>connections_n</literal> - per-worker | |
1853 array of connections of type <literal>ngx_connection_t</literal>, created by | |
1854 the event module while initializing each nginx worker. | |
1855 The number of connections is set by the <literal>worker_connections</literal> | |
1856 directive | |
1857 </listitem> | |
1858 | |
1859 <listitem> | |
1860 <literal>free_connections</literal>, | |
1861 <literal>free_connections_n</literal> - the and number of currently available | |
1862 connections. | |
1863 If no connections are available, nginx worker refuses to accept new clients | |
1864 </listitem> | |
1865 | |
1866 <listitem> | |
1867 <literal>files</literal>, <literal>files_n</literal> - array for mapping file | |
1868 descriptors to nginx connections. | |
1869 This mapping is used by the event modules, having the | |
1870 <literal>NGX_USE_FD_EVENT</literal> flag (currently, it's poll and devpoll) | |
1871 </listitem> | |
1872 | |
1873 <listitem> | |
1874 <literal>conf_ctx</literal> - array of core module configurations. | |
1875 The configurations are created and filled while reading nginx configuration | |
1876 files | |
1877 </listitem> | |
1878 | |
1879 <listitem> | |
1880 <literal>modules</literal>, <literal>modules_n</literal> - array of modules | |
1881 <literal>ngx_module_t</literal>, both static and dynamic, loaded by current | |
1882 configuration | |
1883 </listitem> | |
1884 | |
1885 <listitem> | |
1886 <literal>listening</literal> - array of listening objects | |
1887 <literal>ngx_listening_t</literal>. | |
1888 Listening objects are normally added by the the <literal>listen</literal> | |
1889 directive of different modules which call the | |
1890 <literal>ngx_create_listening()</literal> function. | |
1891 Based on listening objects, listen sockets are created by nginx | |
1892 </listitem> | |
1893 | |
1894 <listitem> | |
1895 <literal>paths</literal> - array of paths <literal>ngx_path_t</literal>. | |
1896 Paths are added by calling the function <literal>ngx_add_path()</literal> from | |
1897 modules which are going to operate on certain directories. | |
1898 These directories are created by nginx after reading configuration, if missing. | |
1899 Moreover, two handlers can be added for each path: | |
1900 | |
1901 <list type="bullet"> | |
1902 | |
1903 <listitem> | |
1904 path loader - executed only once in 60 seconds after starting or reloading | |
1905 nginx. Normally, reads the directory and stores data in nginx shared | |
1906 memory. The handler is called from a dedicated nginx process "nginx | |
1907 cache loader" | |
1908 </listitem> | |
1909 | |
1910 <listitem> | |
1911 path manager - executed periodically. Normally, removes old files from the | |
1912 directory and reflects these changes in nginx memory. The handler is | |
1913 called from a dedicated nginx process "nginx cache manager" | |
1914 </listitem> | |
1915 | |
1916 </list> | |
1917 </listitem> | |
1918 | |
1919 <listitem> | |
1920 <literal>open_files</literal> - list of <literal>ngx_open_file_t</literal> | |
1921 objects. | |
1922 An open file object is created by calling the function | |
1923 <literal>ngx_conf_open_file()</literal>. | |
1924 After reading configuration nginx opens all files from the | |
1925 <literal>open_files</literal> list and stores file descriptors in the | |
1926 <literal>fd</literal> field of each open file object. | |
1927 The files are opened in append mode and created if missing. | |
1928 The files from this list are reopened by nginx workers upon receiving the | |
1929 reopen signal (usually it's <literal>USR1</literal>). | |
1930 In this case the <literal>fd</literal> fields are changed to new descriptors. | |
1931 The open files are currently used for logging | |
1932 </listitem> | |
1933 | |
1934 <listitem> | |
1935 <literal>shared_memory</literal> - list of shared memory zones, each added by | |
1936 calling the <literal>ngx_shared_memory_add()</literal> function. | |
1937 Shared zones are mapped to the same address range in all nginx processes and | |
1938 are used to share common data, for example HTTP cache in-memory tree | |
1939 </listitem> | |
1940 | |
1941 </list> | |
1942 </para> | |
1943 | |
1944 </section> | |
1945 | |
1946 <section name="Buffer" id="buffer"> | |
1947 | |
1948 <para> | |
1949 For input/output operations, nginx provides the buffer type | |
1950 <literal>ngx_buf_t</literal>. | |
1951 Normally, it's used to hold data to be written to a destination or read from a | |
1952 source. | |
1953 Buffer can reference data in memory and in file. | |
1954 Technically it's possible that a buffer references both at the same time. | |
1955 Memory for the buffer is allocated separately and is not related to the buffer | |
1956 structure <literal>ngx_buf_t</literal>. | |
1957 </para> | |
1958 | |
1959 <para> | |
1960 The structure <literal>ngx_buf_t</literal> has the following fields: | |
1961 </para> | |
1962 | |
1963 <para> | |
1964 <list type="bullet"> | |
1965 | |
1966 <listitem> | |
1967 <literal>start</literal>, <literal>end</literal> - the boundaries of memory | |
1968 block, allocated for the buffer | |
1969 </listitem> | |
1970 | |
1971 <listitem> | |
1972 <literal>pos</literal>, <literal>last</literal> - memory buffer boundaries, | |
1973 normally a subrange of <literal>start</literal> .. <literal>end</literal> | |
1974 </listitem> | |
1975 | |
1976 <listitem> | |
1977 <literal>file_pos</literal>, <literal>file_last</literal> - file buffer | |
1978 boundaries, these are offsets from the beginning of the file | |
1979 </listitem> | |
1980 | |
1981 <listitem> | |
1982 <literal>tag</literal> - unique value, used to distinguish buffers, created by | |
1983 different nginx module, usually, for the purpose of buffer reuse | |
1984 </listitem> | |
1985 | |
1986 <listitem> | |
1987 <literal>file</literal> - file object | |
1988 </listitem> | |
1989 | |
1990 <listitem> | |
1991 <literal>temporary</literal> - flag, meaning that the buffer references | |
1992 writable memory | |
1993 </listitem> | |
1994 | |
1995 <listitem> | |
1996 <literal>memory</literal> - flag, meaning that the buffer references read-only | |
1997 memory | |
1998 </listitem> | |
1999 | |
2000 <listitem> | |
2001 <literal>in_file</literal> - flag, meaning that current buffer references data | |
2002 in a file | |
2003 </listitem> | |
2004 | |
2005 <listitem> | |
2006 <literal>flush</literal> - flag, meaning that all data prior to this buffer | |
2007 should be flushed | |
2008 </listitem> | |
2009 | |
2010 <listitem> | |
2011 <literal>recycled</literal> - flag, meaning that the buffer can be reused and | |
2012 should be consumed as soon as possible | |
2013 </listitem> | |
2014 | |
2015 <listitem> | |
2016 <literal>sync</literal> - flag, meaning that the buffer carries no data or | |
2017 special signal like <literal>flush</literal> or <literal>last_buf</literal>. | |
2018 Normally, such buffers are considered an error by nginx. This flags allows | |
2019 skipping the error checks | |
2020 </listitem> | |
2021 | |
2022 <listitem> | |
2023 <literal>last_buf</literal> - flag, meaning that current buffer is the last in | |
2024 output | |
2025 </listitem> | |
2026 | |
2027 <listitem> | |
2028 <literal>last_in_chain</literal> - flag, meaning that there's no more data | |
2029 buffers in a (sub)request | |
2030 </listitem> | |
2031 | |
2032 <listitem> | |
2033 <literal>shadow</literal> - reference to another buffer, related to the current | |
2034 buffer. Usually current buffer uses data from the shadow buffer. Once current | |
2035 buffer is consumed, the shadow buffer should normally also be marked as | |
2036 consumed | |
2037 </listitem> | |
2038 | |
2039 <listitem> | |
2040 <literal>last_shadow</literal> - flag, meaning that current buffer is the last | |
2041 buffer, referencing a particular shadow buffer | |
2042 </listitem> | |
2043 | |
2044 <listitem> | |
2045 <literal>temp_file</literal> - flag, meaning that the buffer is in a temporary | |
2046 file | |
2047 </listitem> | |
2048 | |
2049 </list> | |
2050 </para> | |
2051 | |
2052 <para> | |
2053 For input and output buffers are linked in chains. | |
2054 Chain is a sequence of chain links <literal>ngx_chain_t</literal>, defined as | |
2055 follows: | |
2056 </para> | |
2057 | |
2058 | |
2059 <programlisting> | |
2060 typedef struct ngx_chain_s ngx_chain_t; | |
2061 | |
2062 struct ngx_chain_s { | |
2063 ngx_buf_t *buf; | |
2064 ngx_chain_t *next; | |
2065 }; | |
2066 </programlisting> | |
2067 | |
2068 <para> | |
2069 Each chain link keeps a reference to its buffer and a reference to the next | |
2070 chain link. | |
2071 </para> | |
2072 | |
2073 <para> | |
2074 Example of using buffers and chains: | |
2075 </para> | |
2076 | |
2077 | |
2078 <programlisting> | |
2079 ngx_chain_t * | |
2080 ngx_get_my_chain(ngx_pool_t *pool) | |
2081 { | |
2082 ngx_buf_t *b; | |
2083 ngx_chain_t *out, *cl, **ll; | |
2084 | |
2085 /* first buf */ | |
2086 cl = ngx_alloc_chain_link(pool); | |
2087 if (cl == NULL) { /* error */ } | |
2088 | |
2089 b = ngx_calloc_buf(pool); | |
2090 if (b == NULL) { /* error */ } | |
2091 | |
2092 b->start = (u_char *) "foo"; | |
2093 b->pos = b->start; | |
2094 b->end = b->start + 3; | |
2095 b->last = b->end; | |
2096 b->memory = 1; /* read-only memory */ | |
2097 | |
2098 cl->buf = b; | |
2099 out = cl; | |
2100 ll = &cl->next; | |
2101 | |
2102 /* second buf */ | |
2103 cl = ngx_alloc_chain_link(pool); | |
2104 if (cl == NULL) { /* error */ } | |
2105 | |
2106 b = ngx_create_temp_buf(pool, 3); | |
2107 if (b == NULL) { /* error */ } | |
2108 | |
2109 b->last = ngx_cpymem(b->last, "foo", 3); | |
2110 | |
2111 cl->buf = b; | |
2112 cl->next = NULL; | |
2113 *ll = cl; | |
2114 | |
2115 return out; | |
2116 } | |
2117 </programlisting> | |
2118 | |
2119 </section> | |
2120 | |
2121 | |
2122 <section name="Networking" id="networking"> | |
2123 | |
2124 | |
2125 <!-- | |
2126 <section name="Network data types" id="network_data_types"> | |
2127 | |
2128 <para> | |
2129 TBD: ngx_addr_t, ngx_url_t, ngx_socket_t, ngx_sockaddr_t, parse url, parse | |
2130 address... | |
2131 </para> | |
2132 | |
2133 </section> | |
2134 --> | |
2135 | |
2136 <section name="Connection" id="connection"> | |
2137 | |
2138 <para> | |
2139 Connection type <literal>ngx_connection_t</literal> is a wrapper around a | |
2140 socket descriptor. Some of the structure fields are: | |
2141 </para> | |
2142 | |
2143 <para> | |
2144 <list type="bullet"> | |
2145 | |
2146 <listitem> | |
2147 <literal>fd</literal> - socket descriptor | |
2148 </listitem> | |
2149 | |
2150 <listitem> | |
2151 <literal>data</literal> - arbitrary connection context. | |
2152 Normally, a pointer to a higher level object, built on top of the connection, | |
2153 like HTTP request or Stream session | |
2154 </listitem> | |
2155 | |
2156 <listitem> | |
2157 <literal>read</literal>, <literal>write</literal> - read and write events for | |
2158 the connection | |
2159 </listitem> | |
2160 | |
2161 <listitem> | |
2162 <literal>recv</literal>, <literal>send</literal>, | |
2163 <literal>recv_chain</literal>, <literal>send_chain</literal> - I/O operations | |
2164 for the connection | |
2165 </listitem> | |
2166 | |
2167 <listitem> | |
2168 <literal>pool</literal> - connection pool | |
2169 </listitem> | |
2170 | |
2171 <listitem> | |
2172 <literal>log</literal> - connection log | |
2173 </listitem> | |
2174 | |
2175 <listitem> | |
2176 <literal>sockaddr</literal>, <literal>socklen</literal>, | |
2177 <literal>addr_text</literal> - remote socket address in binary and text forms | |
2178 </listitem> | |
2179 | |
2180 <listitem> | |
2181 <literal>local_sockaddr</literal>, <literal>local_socklen</literal> - local | |
2182 socket address in binary form. | |
2183 Initially, these fields are empty. | |
2184 Function <literal>ngx_connection_local_sockaddr()</literal> should be used to | |
2185 get socket local address | |
2186 </listitem> | |
2187 | |
2188 <listitem> | |
2189 <literal>proxy_protocol_addr</literal>, <literal>proxy_protocol_port</literal> | |
2190 - PROXY protocol client address and port, if PROXY protocol is enabled for the | |
2191 connection | |
2192 </listitem> | |
2193 | |
2194 <listitem> | |
2195 <literal>ssl</literal> - nginx connection SSL context | |
2196 </listitem> | |
2197 | |
2198 <listitem> | |
2199 <literal>reusable</literal> - flag, meaning, that the connection is at the | |
2200 state, when it can be reused | |
2201 </listitem> | |
2202 | |
2203 <listitem> | |
2204 <literal>close</literal> - flag, meaning, that the connection is being reused | |
2205 and should be closed | |
2206 </listitem> | |
2207 | |
2208 </list> | |
2209 </para> | |
2210 | |
2211 <para> | |
2212 An nginx connection can transparently encapsulate SSL layer. | |
2213 In this case the connection <literal>ssl</literal> field holds a pointer to an | |
2214 <literal>ngx_ssl_connection_t</literal> structure, keeping all SSL-related data | |
2215 for the connection, including <literal>SSL_CTX</literal> and | |
2216 <literal>SSL</literal>. | |
2217 The handlers <literal>recv</literal>, <literal>send</literal>, | |
2218 <literal>recv_chain</literal>, <literal>send_chain</literal> are set as well to | |
2219 SSL functions. | |
2220 </para> | |
2221 | |
2222 <para> | |
2223 The number of connections per nginx worker is limited by the | |
2224 <literal>worker_connections</literal> value. | |
2225 All connection structures are pre-created when a worker starts and stored in | |
2226 the <literal>connections</literal> field of the cycle object. | |
2227 To reach out for a connection structure, <literal>ngx_get_connection(s, | |
2228 log)</literal> function is used. | |
2229 The function receives a socket descriptor <literal>s</literal> which needs to | |
2230 be wrapped in a connection structure. | |
2231 </para> | |
2232 | |
2233 <para> | |
2234 Since the number of connections per worker is limited, nginx provides a | |
2235 way to grab connections which are currently in use. | |
2236 To enable or disable reuse of a connection, function | |
2237 <literal>ngx_reusable_connection(c, reusable)</literal> is called. | |
2238 Calling <literal>ngx_reusable_connection(c, 1)</literal> sets the | |
2239 <literal>reuse</literal> flag of the connection structure and inserts the | |
2240 connection in the <literal>reusable_connections_queue</literal> of the cycle. | |
2241 Whenever <literal>ngx_get_connection()</literal> finds out there are no | |
2242 available connections in the <literal>free_connections</literal> list of the | |
2243 cycle, it calls <literal>ngx_drain_connections()</literal> to release a | |
2244 specific number of reusable connections. | |
2245 For each such connection, the <literal>close</literal> flag is set and its read | |
2246 handler is called which is supposed to free the connection by calling | |
2247 <literal>ngx_close_connection(c)</literal> and make it available for reuse. | |
2248 To exit the state when a connection can be reused | |
2249 <literal>ngx_reusable_connection(c, 0)</literal> is called. | |
2250 An example of reusable connections in nginx is HTTP client connections which | |
2251 are marked as reusable until some data is received from the client. | |
2252 </para> | |
2253 | |
2254 </section> | |
2255 | |
2256 | |
2257 </section> | |
2258 | |
2259 | |
2260 <section name="Events" id="events"> | |
2261 | |
2262 | |
2263 <section name="Event" id="event"> | |
2264 | |
2265 <para> | |
2266 Event object <literal>ngx_event_t</literal> in nginx provides a way to be | |
2267 notified of a specific event happening. | |
2268 </para> | |
2269 | |
2270 <para> | |
2271 Some of the fields of the <literal>ngx_event_t</literal> are: | |
2272 </para> | |
2273 | |
2274 <para> | |
2275 <list type="bullet"> | |
2276 | |
2277 <listitem> | |
2278 <literal>data</literal> - arbitrary event context, used in event handler, | |
2279 usually, a pointer to a connection, tied with the event | |
2280 </listitem> | |
2281 | |
2282 <listitem> | |
2283 <literal>handler</literal> - callback function to be invoked when the event | |
2284 happens | |
2285 </listitem> | |
2286 | |
2287 <listitem> | |
2288 <literal>write</literal> - flag, meaning that this is the write event. | |
2289 Used to distinguish between read and write events | |
2290 </listitem> | |
2291 | |
2292 <listitem> | |
2293 <literal>active</literal> - flag, meaning that the event is registered for | |
2294 receiving I/O notifications, normally from notification mechanisms like epoll, | |
2295 kqueue, poll | |
2296 </listitem> | |
2297 | |
2298 <listitem> | |
2299 <literal>ready</literal> - flag, meaning that the event has received an | |
2300 I/O notification | |
2301 </listitem> | |
2302 | |
2303 <listitem> | |
2304 <literal>delayed</literal> - flag, meaning that I/O is delayed due to rate | |
2305 limiting | |
2306 </listitem> | |
2307 | |
2308 <listitem> | |
2309 <literal>timer</literal> - Red-Black tree node for inserting the event into | |
2310 the timer tree | |
2311 </listitem> | |
2312 | |
2313 <listitem> | |
2314 <literal>timer_set</literal> - flag, meaning that the event timer is set, | |
2315 but not yet expired | |
2316 </listitem> | |
2317 | |
2318 <listitem> | |
2319 <literal>timedout</literal> - flag, meaning that the event timer has expired | |
2320 </listitem> | |
2321 | |
2322 <listitem> | |
2323 <literal>eof</literal> - read event flag, meaning that the eof has happened | |
2324 while reading data | |
2325 </listitem> | |
2326 | |
2327 <listitem> | |
2328 <literal>pending_eof</literal> - flag, meaning that the eof is pending on the | |
2329 socket, even though there may be some data available before it. | |
2330 The flag is delivered via <literal>EPOLLRDHUP</literal> epoll event or | |
2331 <literal>EV_EOF</literal> kqueue flag | |
2332 </listitem> | |
2333 | |
2334 <listitem> | |
2335 <literal>error</literal> - flag, meaning that an error has happened while | |
2336 reading (for read event) or writing (for write event) | |
2337 </listitem> | |
2338 | |
2339 <listitem> | |
2340 <literal>cancelable</literal> - timer event flag, meaning that the event | |
2341 handler should be called while performing nginx worker graceful shutdown, event | |
2342 though event timeout has not yet expired. The flag provides a way to finalize | |
2343 certain activities, for example, flush log files | |
2344 </listitem> | |
2345 | |
2346 <listitem> | |
2347 <literal>posted</literal> - flag, meaning that the event is posted to queue | |
2348 </listitem> | |
2349 | |
2350 <listitem> | |
2351 <literal>queue</literal> - queue node for posting the event to a queue | |
2352 </listitem> | |
2353 | |
2354 </list> | |
2355 </para> | |
2356 | |
2357 </section> | |
2358 | |
2359 | |
2360 <section name="I/O events" id="i_o_events"> | |
2361 | |
2362 <para> | |
2363 Each connection, received with the | |
2364 <literal>ngx_get_connection()</literal> call, has two events attached to it: | |
2365 <literal>c->read</literal> and <literal>c->write</literal>. | |
2366 These events are used to receive notifications about the socket being ready for | |
2367 reading or writing. | |
2368 All such events operate in Edge-Triggered mode, meaning that they only trigger | |
2369 notifications when the state of the socket changes. | |
2370 For example, doing a partial read on a socket will not make nginx deliver a | |
2371 repeated read notification until more data arrive in the socket. | |
2372 Even when the underlying I/O notification mechanism is essentially | |
2373 Level-Triggered (poll, select etc), nginx will turn the notifications into | |
2374 Edge-Triggered. | |
2375 To make nginx event notifications consistent across all notifications systems | |
2376 on different platforms, it's required, that the functions | |
2377 <literal>ngx_handle_read_event(rev, flags)</literal> and | |
1907
42ed974b83a5
Fixed the duplicated reference in development guide.
Vladimir Homutov <vl@nginx.com>
parents:
1902
diff
changeset
|
2378 <literal>ngx_handle_write_event(wev, lowat)</literal> are called after handling |
1899 | 2379 an I/O socket notification or calling any I/O functions on that socket. |
2380 Normally, these functions are called once in the end of each read or write | |
2381 event handler. | |
2382 </para> | |
2383 | |
2384 </section> | |
2385 | |
2386 | |
2387 <section name="Timer events" id="timer_events"> | |
2388 | |
2389 <para> | |
2390 An event can be set to notify a timeout expiration. | |
2391 The function <literal>ngx_add_timer(ev, timer)</literal> sets a timeout for an | |
2392 event, <literal>ngx_del_timer(ev)</literal> deletes a previously set timeout. | |
2393 Timeouts currently set for all existing events, are kept in a global timeout | |
2394 Red-Black tree <literal>ngx_event_timer_rbtree</literal>. | |
2395 The key in that tree has the type <literal>ngx_msec_t</literal> and is the time | |
2396 in milliseconds since the beginning of January 1, 1970 (modulus | |
2397 <literal>ngx_msec_t</literal> max value) at which the event should expire. | |
2398 The tree structure provides fast inserting and deleting operations, as well as | |
2399 accessing the nearest timeouts. | |
2400 The latter is used by nginx to find out for how long to wait for I/O events | |
2401 and for expiring timeout events afterwards. | |
2402 </para> | |
2403 | |
2404 </section> | |
2405 | |
2406 | |
2407 <section name="Posted events" id="posted_events"> | |
2408 | |
2409 <para> | |
2410 An event can be posted which means that its handler will be called at some | |
2411 point later within the current event loop iteration. | |
2412 Posting events is a good practice for simplifying code and escaping stack | |
2413 overflows. | |
2414 Posted events are held in a post queue. | |
2415 The macro <literal>ngx_post_event(ev, q)</literal> posts the event | |
2416 <literal>ev</literal> to the post queue <literal>q</literal>. | |
2417 Macro <literal>ngx_delete_posted_event(ev)</literal> deletes the event | |
2418 <literal>ev</literal> from whatever queue it's currently posted. | |
2419 Normally, events are posted to the <literal>ngx_posted_events</literal> queue. | |
2420 This queue is processed late in the event loop - after all I/O and timer | |
2421 events are already handled. | |
2422 The function <literal>ngx_event_process_posted()</literal> is called to process | |
2423 an event queue. | |
2424 This function calls event handlers until the queue is not empty. This means | |
2425 that a posted event handler can post more events to be processed within the | |
2426 current event loop iteration. | |
2427 </para> | |
2428 | |
2429 <para> | |
2430 Example: | |
2431 </para> | |
2432 | |
2433 | |
2434 <programlisting> | |
2435 void | |
2436 ngx_my_connection_read(ngx_connection_t *c) | |
2437 { | |
2438 ngx_event_t *rev; | |
2439 | |
2440 rev = c->read; | |
2441 | |
2442 ngx_add_timer(rev, 1000); | |
2443 | |
2444 rev->handler = ngx_my_read_handler; | |
2445 | |
2446 ngx_my_read(rev); | |
2447 } | |
2448 | |
2449 | |
2450 void | |
2451 ngx_my_read_handler(ngx_event_t *rev) | |
2452 { | |
2453 ssize_t n; | |
2454 ngx_connection_t *c; | |
2455 u_char buf[256]; | |
2456 | |
2457 if (rev->timedout) { /* timeout expired */ } | |
2458 | |
2459 c = rev->data; | |
2460 | |
2461 while (rev->ready) { | |
2462 n = c->recv(c, buf, sizeof(buf)); | |
2463 | |
2464 if (n == NGX_AGAIN) { | |
2465 break; | |
2466 } | |
2467 | |
2468 if (n == NGX_ERROR) { /* error */ } | |
2469 | |
2470 /* process buf */ | |
2471 } | |
2472 | |
2473 if (ngx_handle_read_event(rev, 0) != NGX_OK) { /* error */ } | |
2474 } | |
2475 </programlisting> | |
2476 | |
2477 </section> | |
2478 | |
2479 | |
2480 <section name="Event loop" id="event_loop"> | |
2481 | |
2482 <para> | |
2483 All nginx processes which do I/O, have an event loop. | |
2484 The only type of process which does not have I/O, is nginx master process which | |
2485 spends most of its time in <literal>sigsuspend()</literal> call waiting for | |
2486 signals to arrive. | |
2487 Event loop is implemented in <literal>ngx_process_events_and_timers()</literal> | |
2488 function. | |
2489 This function is called repeatedly until the process exits. | |
2490 It has the following stages: | |
2491 </para> | |
2492 | |
2493 <para> | |
2494 <list type="bullet"> | |
2495 | |
2496 <listitem> | |
2497 find nearest timeout by calling <literal>ngx_event_find_timer()</literal>. | |
2498 This function finds the leftmost timer tree node and returns the number of | |
2499 milliseconds until that node expires | |
2500 </listitem> | |
2501 | |
2502 <listitem> | |
2503 process I/O events by calling a handler, specific to event notification | |
2504 mechanism, chosen by nginx configuration. | |
2505 This handler waits for at least one I/O event to happen, but no longer, than | |
2506 the nearest timeout. | |
2507 For each read or write event which has happened, the <literal>ready</literal> | |
2508 flag is set and its handler is called. | |
2509 For Linux, normally, the <literal>ngx_epoll_process_events()</literal> handler | |
2510 is used which calls <literal>epoll_wait()</literal> to wait for I/O events | |
2511 </listitem> | |
2512 | |
2513 <listitem> | |
2514 expire timers by calling <literal>ngx_event_expire_timers()</literal>. | |
2515 The timer tree is iterated from the leftmost element to the right until a not | |
2516 yet expired timeout is found. | |
2517 For each expired node the <literal>timedout</literal> event flag is set, | |
2518 <literal>timer_set</literal> flag is reset, and the event handler is called | |
2519 </listitem> | |
2520 | |
2521 <listitem> | |
2522 process posted events by calling <literal>ngx_event_process_posted()</literal>. | |
2523 The function repeatedly removes the first element from the posted events | |
2524 queue and calls its handler until the queue gets empty | |
2525 </listitem> | |
2526 | |
2527 </list> | |
2528 </para> | |
2529 | |
2530 <para> | |
2531 All nginx processes handle signals as well. | |
2532 Signal handlers only set global variables which are checked after the | |
2533 <literal>ngx_process_events_and_timers()</literal> call. | |
2534 </para> | |
2535 | |
2536 </section> | |
2537 | |
2538 | |
2539 </section> | |
2540 | |
2541 | |
2542 <section name="Processes" id="processes"> | |
2543 | |
2544 <para> | |
2545 There are several types of processes in nginx. | |
2546 The type of current process is kept in the <literal>ngx_process</literal> | |
2547 global variable: | |
2548 </para> | |
2549 | |
2550 <list type="bullet"> | |
2551 | |
2552 <listitem> | |
2553 | |
2554 <para> | |
2555 <literal>NGX_PROCESS_MASTER</literal> - the master process runs the | |
2556 <literal>ngx_master_process_cycle()</literal> function. | |
2557 Master process does not have any I/O and responds only to signals. | |
2558 It reads configuration, creates cycles, starts and controls child processes | |
2559 </para> | |
2560 | |
2561 | |
2562 </listitem> | |
2563 | |
2564 <listitem> | |
2565 | |
2566 <para> | |
2567 <literal>NGX_PROCESS_WORKER</literal> - the worker process runs the | |
2568 <literal>ngx_worker_process_cycle()</literal> function. | |
2569 Worker processes are started by master and handle client connections. | |
2570 They also respond to signals and channel commands, sent from master | |
2571 </para> | |
2572 | |
2573 | |
2574 </listitem> | |
2575 | |
2576 <listitem> | |
2577 | |
2578 <para> | |
2579 <literal>NGX_PROCESS_SINGLE</literal> - single process is the only type | |
2580 of processes which exist in the <literal>master_process off</literal> mode. | |
2581 The cycle function for this process is | |
2582 <literal>ngx_single_process_cycle()</literal>. | |
2583 This process creates cycles and handles client connections | |
2584 </para> | |
2585 | |
2586 | |
2587 </listitem> | |
2588 | |
2589 <listitem> | |
2590 | |
2591 <para> | |
2592 <literal>NGX_PROCESS_HELPER</literal> - currently, there are two types of | |
2593 helper processes: cache manager and cache loader. | |
2594 Both of them share the same cycle function | |
2595 <literal>ngx_cache_manager_process_cycle()</literal>. | |
2596 </para> | |
2597 | |
2598 | |
2599 </listitem> | |
2600 | |
2601 </list> | |
2602 | |
2603 <para> | |
2604 All nginx processes handle the following signals: | |
2605 </para> | |
2606 | |
2607 <list type="bullet"> | |
2608 | |
2609 <listitem> | |
2610 | |
2611 <para> | |
2612 <literal>NGX_SHUTDOWN_SIGNAL</literal> (<literal>SIGQUIT</literal>) - graceful | |
2613 shutdown. | |
2614 Upon receiving this signal master process sends shutdown signal to all child | |
2615 processes. | |
2616 When no child processes are left, master destroys cycle pool and exits. | |
2617 A worker process which received this signal, closes all listening sockets and | |
2618 waits until timeout tree becomes empty, then destroys cycle pool and exits. | |
2619 A cache manager process exits right after receiving this signal. | |
2620 The variable <literal>ngx_quit</literal> is set to one after receiving this | |
2621 signal and immediately reset after being processed. | |
2622 The variable <literal>ngx_exiting</literal> is set to one when worker process | |
2623 is in shutdown state | |
2624 </para> | |
2625 | |
2626 | |
2627 </listitem> | |
2628 | |
2629 <listitem> | |
2630 | |
2631 <para> | |
2632 <literal>NGX_TERMINATE_SIGNAL</literal> (<literal>SIGTERM</literal>) - | |
2633 terminate. | |
2634 Upon receiving this signal master process sends terminate signal to all child | |
2635 processes. | |
2636 If child processes do not exit in 1 second, they are killed with the | |
2637 <literal>SIGKILL</literal> signal. | |
2638 When no child processes are left, master process destroys cycle pool and exits. | |
2639 A worker or cache manager process which received this signal destroys cycle | |
2640 pool and exits. | |
2641 The variable <literal>ngx_terminate</literal> is set to one after receiving | |
2642 this signal | |
2643 </para> | |
2644 | |
2645 | |
2646 </listitem> | |
2647 | |
2648 <listitem> | |
2649 | |
2650 <para> | |
2651 <literal>NGX_NOACCEPT_SIGNAL</literal> (<literal>SIGWINCH</literal>) | |
2652 - gracefully shut down worker processes | |
2653 </para> | |
2654 | |
2655 | |
2656 </listitem> | |
2657 | |
2658 <listitem> | |
2659 | |
2660 <para> | |
2661 <literal>NGX_RECONFIGURE_SIGNAL</literal> (<literal>SIGHUP</literal>) - | |
2662 reconfigure. | |
2663 Upon receiving this signal master process creates a new cycle from | |
2664 configuration file. | |
2665 If the new cycle was created successfully, the old cycle is deleted and new | |
2666 child processes are started. | |
2667 Meanwhile, the old processes receive the shutdown signal. | |
2668 In single-process mode, nginx creates a new cycle as well, but keeps the old | |
2669 one until all clients, tied to the old cycle, are gone. | |
2670 Worker and helper processes ignore this signal | |
2671 </para> | |
2672 | |
2673 | |
2674 </listitem> | |
2675 | |
2676 <listitem> | |
2677 | |
2678 <para> | |
2679 <literal>NGX_REOPEN_SIGNAL</literal> (<literal>SIGUSR1</literal>) - reopen | |
2680 files. | |
2681 Master process passes this signal to workers. | |
2682 Worker processes reopen all <literal>open_files</literal> from the cycle | |
2683 </para> | |
2684 | |
2685 | |
2686 </listitem> | |
2687 | |
2688 <listitem> | |
2689 | |
2690 <para> | |
2691 <literal>NGX_CHANGEBIN_SIGNAL</literal> (<literal>SIGUSR2</literal>) - change | |
2692 nginx binary. | |
2693 Master process starts a new nginx binary and passes there a list of all listen | |
2694 sockets. | |
2695 The list is passed in the environment variable <literal>"NGINX"</literal> in | |
2696 text format, where descriptor numbers separated with semicolons. | |
2697 A new nginx instance reads that variable and adds the sockets to its init | |
2698 cycle. | |
2699 Other processes ignore this signal | |
2700 </para> | |
2701 | |
2702 | |
2703 </listitem> | |
2704 | |
2705 </list> | |
2706 | |
2707 <para> | |
2708 While all nginx worker processes are able to receive and properly handle POSIX | |
2709 signals, master process normally does not pass any signals to workers and | |
2710 helpers with the standard <literal>kill()</literal> syscall. | |
2711 Instead, nginx uses inter-process channels which allow sending messages between | |
2712 all nginx processes. | |
2713 Currently, however, messages are only sent from master to its children. | |
2714 Those messages carry the same signals. | |
2715 The channels are socketpairs with their ends in different processes. | |
2716 </para> | |
2717 | |
2718 <para> | |
2719 When running nginx binary, several values can be specified next to | |
2720 <literal>-s</literal> parameter. | |
2721 Those values are <literal>stop</literal>, <literal>quit</literal>, | |
2722 <literal>reopen</literal>, <literal>reload</literal>. | |
2723 They are converted to signals <literal>NGX_TERMINATE_SIGNAL</literal>, | |
2724 <literal>NGX_SHUTDOWN_SIGNAL</literal>, <literal>NGX_REOPEN_SIGNAL</literal> | |
2725 and <literal>NGX_RECONFIGURE_SIGNAL</literal> and sent to the nginx master | |
2726 process, whose pid is read from nginx pid file. | |
2727 </para> | |
2728 | |
2729 </section> | |
2730 | |
2731 </article> |