1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
|
.TH MAGI 3 2021-01-07 v0.0.1 "Magi Library Manual"
.SH NAME
.B magi
\- Common Gateway Interface (CGI) library
.SH SYNOPSIS
.B #include <magi.h>
.SH DESCRIPTION
The
.B magi
library gives the user a way to interact with the web server
through CGI as described in
.IR "RFC 3875" .
The library is free and open source software.
It is written in ANSI C without any dependencies except C standard library.
.SH GETTING STARTED
Once you have obtained the sources (from
.I http://git.veresov.xyz/aversey/magi
or other way) run
.P
.RS
.nf
make
.fi
.RE
.P
in order to build the library. The results will be in the
.I build
directory.
.P
To build the examples run:
.P
.RS
.nf
make examples
.fi
.RE
.P
Add any single-file C program using
.B magi
into the
.I examples
directory in order to build it with the same command.
.P
To give you a
.B magi
try here is the minimal program:
.P
.RS
.nf
#include <magi.h>
#include <stdio.h>
int main()
{
struct magi_request request;
magi_request_init(&request);
if (magi_parse(&request)) {
magi_response_default();
puts("<html><head><title>Hi!</title></head><body>Hi!</body></html>");
} else {
magi_error_response(request.error);
}
magi_request_free(&request);
return 0;
}
.fi
.RE
.P
You can install
.B magi
system-wide once you have got root privileges by running:
.P
.RS
.nf
make install
.fi
.RE
.P
Once you have installed the library system-wide all you need to do to compile
you programs is to specify
.I -lmagi
flag. So to compile
.I some.c
using this library you can run:
.P
.RS
.nf
gcc some.c -lmagi
.fi
.RE
.P
Default installation locations are
.I /usr/include
for the headers,
.I /usr/lib
for library binaries and
.I /usr/share/man
for this manual.
.SH OVERVIEW
Main task of the
.B magi
library is to analyze CGI request and represent it to the user in
.B magi_request
data structure.
CGI request is passed in environment variables and standard input.
You can access environment directly even then using magi,
however reading from standard input will cause a mess.
.P
The library also provides response functionality with
.B magi_response
and its methods.
Since body can be everything and only headers have specific CGI-related
structure,
.B magi
provides response functionality only for them.
Body can be outputted directly into standard output with any method you want
after forming and sending CGI headers via
.BR magi_response .
.SH REQUEST
Request structure initialization is done with
.BR magi_request_init ().
Don't forget to free everything in it then the work is done via
.BR magi_request_free ().
.P
.RS
.nf
struct magi_request request;
magi_request_init(&request);
/* do your work with request */
magi_request_free(&request);
.fi
.RE
.P
Obtaining of request is done in two parts, i.e. head via
.BR magi_parse_head ()
and body via
.BR magi_parse_body ().
Analysis of head will fill everything in
.B magi_request
structure, except
.I body
and
.I files
fields.
Such separation has place since body processing can depend on data from head,
e.g. in case of loading files, if file size is defined by user group.
If analysis succeeded they will return 1, otherwise 0 as error mark.
More distinct error code will be placed in
.IR request.error .
.P
.RS
.nf
if (!magi_parse_head(&request)) {
/* handle error and exit */
}
/* configure body processing using head data */
if (magi_parse_body(&request)) {
/* send response using request data */
} else {
/* handle error */
}
.fi
.RE
.P
If your case is not as advanced, you can use shortcut
.BR magi_parse ()
to analyze both head and body.
.BR magi_parse (&request)
is literally
.BR magi_parse_head (&request)
&&
.BR magi_parse_body (&request).
.P
.RS
.nf
if (magi_parse(&request)) {
/* send response using request data */
} else {
/* handle error */
}
.fi
.RE
.SS General information in request
Request method is case-sensitive ("GET", "POST", etc.) and stored in
.I method
field.
.P
Server's document root (e.g. /var/www/htdocs) is in
.I document_root
field. It is useful in case of processing documents on server, e.g.
if you are building CMS.
.P
If request was done over HTTPS
.I is_secure
field will be set as 1, otherwise as zero (actually it checks port to be 443,
which is standard HTTPS port).
.P
You can restore URL by following formula:
.P
.RS
.nf
http[s]://{host}:{port}{script}{path}[?...]
.fi
.RE
.P
there
.I host
(e.g. "example.com"),
.I port
(usually 80 or 443),
.I script
(e.g. "/cgi-bin/script") and
.I path
(e.g. "/login") are fields. All that values are lowercased, since they are
case-insensitive.
.SS Form fields values in request
Parameters passed in URL after question mark (e.g. "?q=example&color=red")
are stored in linked list of key-values
.RB ( magi_params )
field
.IR head ,
while HTML form fields values passed in body (except for files,
described lower) are stored in
.IR body .
To get value of parameter by its key use
.BR magi_request_param ().
This function priorities body params, so if you need specifically
parameter from URL use
.B magi_request_urlparam ().
If no parameter was passed with given name, these functions will return null.
.P
For example, if we build search engine and want to access query passed in
form field with name
.IR q ,
we must call:
.P
.RS
.nf
const char *query = magi_request_param(&request, "q");
if (query) {
/* calculate and show result */
} else {
/* show welcoming page */
}
.fi
.RE
.P
Note what values returned by
.BR magi_request_urlparam ()
are decoded from URL encoding, so you don't need to deal with all that
percents and pluses. Result of
.BR magi_request_param ()
is also raw data.
.SS Cookies in request
Same situation with cookies, stored in
.I cookies
field, and accessible via
.BR magi_request_cookie ().
.P
For example, we can check session ID of user, stored in cookie "sid":
.P
.RS
.nf
struct awesome_session session;
const char *sid = magi_request_cookie(&request, "sid");
if (!sid) {
/* render response for unauthorized user */
} else if (session_exists(sid)) {
load_session(&session, sid);
/* render response for this user */
} else {
/* show error page of invalid session */
}
.fi
.RE
.P
While for cookies as described in
.I RFC 6265
it's all you need, your server can use Cookie2, as described in
.IR "RFC 2965" .
In both cases
.BR magi_request_cookie ()
will return the most accurate cookie for this request, however Cookie2
allows more complex behavior, setting and receiving different host and path
for cookies. You able to receive all information about cookie using
.BR magi_request_cookie_complex (),
which returns pointer to
.B magi_cookie
structure. If no cookie is found, result is null. It has fields
.IR name ,
.IR data ,
.I path
(without '/' at the end),
.I domain
(with dot at the beginning) and
.I max_age
(the last one is used only in response).
.SS Files in request
While ordinary parameters are not very big and don't need special processing,
files can be huge and need special processing.
Due to their size, files shouldn't be loaded into memory in general.
The
.B magi
library allows you to setup your own callback for file loading,
or use predefined one,
.BR magi_loadfiles .
.P
You can load file from field
.I to_load
into
.I ./uploaded
with limits on its size of 1MB as following:
.P
.RS
.nf
struct magi_loadfiles cb;
magi_loadfiles_init(&cb);
magi_loadfiles_add(&cb, "to_load", "./uploaded", 1024 * 1024);
magi_loadfiles_set(&request, &cb);
magi_parse_body(&request);
magi_loadfiles_free(&rules);
/* Use file ./uploaded */
.fi
.RE
.P
There
.I request
is initialized
.B magi_request
with parsed head and not parsed body.
.P
To access information about file, use
.B magi_request_file ().
It will get your
.B magi_file
structure for file loaded from field with passed name.
This structure have file name on user host in
.I filename
field.
Other parameters (as Content-Type) are in
.I params
field, accessible with
.BR magi_file_param ().
.P
For example, we can get user feedback, returning Content-Type of loaded file:
.P
.RS
.nf
struct magi_file *loaded = magi_request_file(&request, "to_load");
if (loaded) {
char *type = magi_file_param(loaded, "Content-Type");
if (type_to_response) {
/* Report the user that file of 'type' was loaded. */
} else {
/* Report the user that file was loaded without
specified Content-Type. */
}
} else {
/* Report the user that file wasn't loaded. */
}
.SS File processing callback
In some cases
.B magi_loadfiles
can be not enough.
Then you can specify your own
.BR magi_file_callback .
The
.I act
field contains the callback function itself.
The
.I userdata
field has type
.I "void *"
allowing you to exchange state across different calls of callback.
The
.I addon_max
field specify how much bytes can be passed to your callback with one call.
.P
Callback function has type
.BR magi_file_callback_act ,
which is function returning void, with
.IR userdata ,
.I newfile
flag,
.I file
.B magi_file
structure,
.I addon
and
.IR addon_len .
.P
Files are passed sequentially addon by addon.
At the file end callback will be called with
.I addon
and
.I addon_len
as nulls.
If current
.I addon
is first in current
.I file
.I newfile
flag will be set up.
.P
You can load file from field 'file' into current directory with name,
as specified by
.I filename
field as:
.P
.RS
.nf
static void cb(void *userdata,
int newfile,
const struct magi_file *file,
const char *addon,
int addon_len)
{
static FILE *f = 0;
if (!strcmp("file", file->field)) {
if (newfile) {
f = fopen(file->filename, "wb");
}
fwrite(addon, 1, addon_len, f);
if (!addon) {
fclose(f);
}
}
}
.fi
.RE
.P
Set it as callback for processing files for initialized
.I request
with parsed head and not parsed body, and then parse the body:
.P
.RS
.nf
request.callback.act = cb;
magi_parse_body(&request);
/* Now file is loaded into filename */
.fi
.RE
.SH RESPONSE
Response headers are formed with
.B magi_response
structure.
It is initiated with
.BR magi_response_init (),
sent with
.BR magi_response_send (),
and freed with
.BR magi_response_free ().
The only defaults are
.I text/html
as Content-Type and
.I 200 Ok
as status.
You can send them with
.BR magi_response_default ().
.P
.BR magi_response_content_type (),
.BR magi_reponse_content_length ()
and
.BR magi_response_status ()
are used to change corresponding headers.
Any other header can be only added, not changed with
.BR magi_response_header ().
Don't use it in cases above.
.P
For cookies use
.BR magi_response_cookie ()
to set cookies and
.BR magi_response_cookie_discard ()
to discard them.
In case of Cookie2 use
.BR magi_response_cookie_complex ().
.P
Lets send headers for XHTML body, setting cookie 'monster' as 'cookie':
.P
.RS
.nf
struct magi_response response;
magi_response_init(&response);
magi_response_content_type(&response, "application/xhtml+xml");
magi_response_cookie(&response, "monster", "cookie");
magi_response_send(&response);
magi_response_free(&response);
.fi
.RE
.SS URL encoding
It is described in
.IR "RFC 3986" .
Briefly it is replacement of every space into plus sign and every not
alpha-numerical or not one of "~-._" character into percent sign
followed by hexadecimal representation of given character byte.
.P
The
.B magi
library provides functions to form url-encoded strings, which is very useful
in forming response. Use
.BR magi_urlenc_size ()
to find what the size of code will be and then call
.BR magi_urlenc ()
for encoding itself.
.P
For example, lets form URL to search in DuckDuckGo for provided char
.RI * query
in char
.RI * url .
.P
.RS
.nf
const char *prefix = "http://duckduckgo.com/?q=";
const int prelen = strlen(prefix);
const int urlencsize = magi_urlenc_size(query);
url = malloc(prelen + urlencsize + 1);
strcpy(url, prefix);
magi_urlenc(query, url + prelen);
url[prelen + urlencsize] = 0;
.fi
.RE
.P
Note that second argument of
.BR magi_urlenc ()
which is encoding destination should be at least size of
.BR magi_urlenc_size ()
of plain data, and that
.BR magi_urlenc ()
doesn't write zero to form null-terminated string in its output.
.SH ERRORS
If function is returning pointer, error is only in case of null returned.
If function is returning
.I int
as success mark it will be null only in case of error, and one otherwise.
Exact
.B magi_error
code is in
.I error
field of
.B magi_request
structure. For other modules error codes seem to be overkill.
.P
You can access default error message with
.BR magi_error_message ()
or send default error page with
.BR magi_error_response ().
.SH DEBUGGING
To debug your CGI scripts with
.I gdb
you can include signal.h:
.P
.RS
.nf
#include <signal.h>
.fi
.RE
.P
and stop your script in the beginning with:
.P
.RS
.nf
raise(SIGSTOP);
.fi
.RE
.P
Then compile your script and place it in directory which is handled by your
CGI server.
Check timeout settings for CGI scripts on your server and make it big enough,
since the server will kill your script by timeout.
Now make request to your server and run the following in the shell as root:
.P
.RS
.nf
gdb path_to_your_executable `pgrep name_of_your_executable`
.fi
.RE
.P
This will connect
.I gdb
to your running CGI script. It will be paused by SIGSTOP, so you will be able
to setup your breakpoints. In order to continue run:
.P
.RS
.nf
continue
.fi
.RE
.SH COMPATIBILTY
Since
.B magi
only uses C standard library and the C language itself it should be able
to run on every environment with them, i.e. Linux, OpenBSD, FreeBSD, macOS,
Windows, etc.
.P
The
.B magi
library should work well with any server supporting CGI, since it is
compatible with
.IR "RFC 3875" ,
i.e. Apache, nginx, Caddy, etc.
.P
The library is compatible with C++, since it avoid using typedefs for structs.
However it doesn't have 'extern "C"' in the headers, so you need to wrap your
includes with it manually or use
.B #include <magi.hpp>
which is a shortcut for include of
.I magi.h
wrapped in 'extern "C"' construct.
.SH AUTHORS AND LICENSE
Copyright 2019-2021
.B Aleksey Veresov
.RI < aleksey@veresov.pro >
.P
This software is provided 'as-is', without any express or implied warranty.
In no event will the authors be held liable for any damages arising from
the use of this software.
.P
Permission to use, copy, modify, and distribute this software for any
purpose with or without fee is hereby granted, provided that the above
copyright notice and this permission notice appear in all copies.
|
