summaryrefslogtreecommitdiffstats
path: root/src/common/ConfUtils.cc
blob: 7c8f1cc78b7f02cdfea3dd26d49a31a9298216a8 (plain)
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
// -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-
// vim: ts=8 sw=2 smarttab
/*
 * Ceph - scalable distributed file system
 *
 * Copyright (C) 2011 New Dream Network
 *
 * This is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License version 2.1, as published by the Free Software
 * Foundation.  See file COPYING.
 *
 */

#include <algorithm>
#include <map>
#include <sstream>
#include <sys/stat.h>
#include <iostream>

#include "include/buffer.h"
#include "common/errno.h"
#include "common/utf8.h"
#include "common/ConfUtils.h"

using std::ostringstream;
using std::pair;
using std::string;

#define MAX_CONFIG_FILE_SZ 0x40000000

////////////////////////////// ConfLine //////////////////////////////
ConfLine::
ConfLine(const std::string &key_, const std::string &val_,
      const std::string &newsection_, const std::string &comment_, int line_no_)
  : key(key_), val(val_), newsection(newsection_)
{
  // If you want to implement writable ConfFile support, you'll need to save
  // the comment and line_no arguments here.
}

bool ConfLine::
operator<(const ConfLine &rhs) const
{
  // We only compare keys.
  // If you have more than one line with the same key in a given section, the
  // last one wins.
  if (key < rhs.key)
    return true;
  else
    return false;
}

std::ostream &operator<<(std::ostream& oss, const ConfLine &l)
{
  oss << "ConfLine(key = '" << l.key << "', val='"
      << l.val << "', newsection='" << l.newsection << "')";
  return oss;
}
///////////////////////// ConfFile //////////////////////////
ConfFile::
ConfFile()
{
}

ConfFile::
~ConfFile()
{
}

void ConfFile::
clear()
{
  sections.clear();
}

/* We load the whole file into memory and then parse it.  Although this is not
 * the optimal approach, it does mean that most of this code can be shared with
 * the bufferlist loading function. Since bufferlists are always in-memory, the
 * load_from_buffer interface works well for them.
 * In general, configuration files should be a few kilobytes at maximum, so
 * loading the whole configuration into memory shouldn't be a problem.
 */
int ConfFile::
parse_file(const std::string &fname, std::deque<std::string> *errors,
	   std::ostream *warnings)
{
  clear();

  int ret = 0;
  size_t sz;
  char *buf = NULL;
  FILE *fp = fopen(fname.c_str(), "r");
  if (!fp) {
    ostringstream oss;
    oss << __func__ << ": cannot open " << fname << ": " << cpp_strerror(errno);
    errors->push_back(oss.str());
    ret = -errno;
    return ret;
  }

  struct stat st_buf;
  if (fstat(fileno(fp), &st_buf)) {
    ret = -errno;
    ostringstream oss;
    oss << __func__ << ": failed to fstat '" << fname << "': " << cpp_strerror(ret);
    errors->push_back(oss.str());
    goto done;
  }

  if (st_buf.st_size > MAX_CONFIG_FILE_SZ) {
    ostringstream oss;
    oss << __func__ << ": config file '" << fname << "' is " << st_buf.st_size
	<< " bytes, but the maximum is " << MAX_CONFIG_FILE_SZ;
    errors->push_back(oss.str());
    ret = -EINVAL;
    goto done;
  }

  sz = (size_t)st_buf.st_size;
  buf = (char*)malloc(sz);
  if (!buf) {
    ret = -ENOMEM;
    goto done;
  }

  if (fread(buf, 1, sz, fp) != sz) {
    if (ferror(fp)) {
      ret = -errno;
      ostringstream oss;
      oss << __func__ << ": fread error while reading '" << fname << "': "
	  << cpp_strerror(ret);
      errors->push_back(oss.str());
      goto done;
    }
    else {
      ostringstream oss;
      oss << __func__ << ": unexpected EOF while reading '" << fname << "': "
	  << "possible concurrent modification?";
      errors->push_back(oss.str());
      ret = -EIO;
      goto done;
    }
  }

  load_from_buffer(buf, sz, errors, warnings);
  ret = 0;

done:
  free(buf);
  fclose(fp);
  return ret;
}

int ConfFile::
parse_bufferlist(ceph::bufferlist *bl, std::deque<std::string> *errors,
		 std::ostream *warnings)
{
  clear();

  load_from_buffer(bl->c_str(), bl->length(), errors, warnings);
  return 0;
}

int ConfFile::
read(const std::string &section, const std::string &key, std::string &val) const
{
  string k(normalize_key_name(key));

  const_section_iter_t s = sections.find(section);
  if (s == sections.end())
    return -ENOENT;
  ConfLine exemplar(k, "", "", "", 0);
  ConfSection::const_line_iter_t l = s->second.lines.find(exemplar);
  if (l == s->second.lines.end())
    return -ENOENT;
  val = l->val;
  return 0;
}

ConfFile::const_section_iter_t ConfFile::
sections_begin() const
{
  return sections.begin();
}

ConfFile::const_section_iter_t ConfFile::
sections_end() const
{
  return sections.end();
}

void ConfFile::
trim_whitespace(std::string &str, bool strip_internal)
{
  // strip preceding
  const char *in = str.c_str();
  while (true) {
    char c = *in;
    if ((!c) || (!isspace(c)))
      break;
    ++in;
  }
  char output[strlen(in) + 1];
  strcpy(output, in);

  // strip trailing
  char *o = output + strlen(output);
  while (true) {
    if (o == output)
      break;
    --o;
    if (!isspace(*o)) {
      ++o;
      *o = '\0';
      break;
    }
  }

  if (!strip_internal) {
    str.assign(output);
    return;
  }

  // strip internal
  char output2[strlen(output) + 1];
  char *out2 = output2;
  bool prev_was_space = false;
  for (char *u = output; *u; ++u) {
    char c = *u;
    if (isspace(c)) {
      if (!prev_was_space)
	*out2++ = c;
      prev_was_space = true;
    }
    else {
      *out2++ = c;
      prev_was_space = false;
    }
  }
  *out2++ = '\0';
  str.assign(output2);
}

/* Normalize a key name.
 *
 * Normalized key names have no leading or trailing whitespace, and all
 * whitespace is stored as underscores.  The main reason for selecting this
 * normal form is so that in common/config.cc, we can use a macro to stringify
 * the field names of md_config_t and get a key in normal form.
 */
std::string ConfFile::
normalize_key_name(const std::string &key)
{
  if (key.find_first_of(" \t\r\n\f\v\xa0") == string::npos) {
    return key;
  }

  string k(key);
  ConfFile::trim_whitespace(k, true);
  std::replace(k.begin(), k.end(), ' ', '_');
  return k;
}

std::ostream &operator<<(std::ostream &oss, const ConfFile &cf)
{
  for (ConfFile::const_section_iter_t s = cf.sections_begin();
       s != cf.sections_end(); ++s) {
    oss << "[" << s->first << "]\n";
    for (ConfSection::const_line_iter_t l = s->second.lines.begin();
	 l != s->second.lines.end(); ++l) {
      if (!l->key.empty()) {
	oss << "\t" << l->key << " = \"" << l->val << "\"\n";
      }
    }
  }
  return oss;
}

void ConfFile::
load_from_buffer(const char *buf, size_t sz, std::deque<std::string> *errors,
		 std::ostream *warnings)
{
  errors->clear();

  section_iter_t::value_type vt("global", ConfSection());
  pair < section_iter_t, bool > vr(sections.insert(vt));
  ceph_assert(vr.second);
  section_iter_t cur_section = vr.first;
  std::string acc;

  const char *b = buf;
  int line_no = 0;
  size_t line_len = -1;
  size_t rem = sz;
  while (1) {
    b += line_len + 1;
    if ((line_len + 1) > rem)
      break;
    rem -= line_len + 1;
    if (rem == 0)
      break;
    line_no++;

    // look for the next newline
    const char *end = (const char*)memchr(b, '\n', rem);
    if (!end) {
      ostringstream oss;
      oss << "read_conf: ignoring line " << line_no << " because it doesn't "
	  << "end with a newline! Please end the config file with a newline.";
      errors->push_back(oss.str());
      break;
    }

    // find length of line, and search for NULLs
    line_len = 0;
    bool found_null = false;
    for (const char *tmp = b; tmp != end; ++tmp) {
      line_len++;
      if (*tmp == '\0') {
	found_null = true;
      }
    }

    if (found_null) {
      ostringstream oss;
      oss << "read_conf: ignoring line " << line_no << " because it has "
	  << "an embedded null.";
      errors->push_back(oss.str());
      acc.clear();
      continue;
    }

    if (check_utf8(b, line_len)) {
      ostringstream oss;
      oss << "read_conf: ignoring line " << line_no << " because it is not "
	  << "valid UTF8.";
      errors->push_back(oss.str());
      acc.clear();
      continue;
    }

    if ((line_len >= 1) && (b[line_len-1] == '\\')) {
      // A backslash at the end of a line serves as a line continuation marker.
      // Combine the next line with this one.
      // Remove the backslash itself from the text.
      acc.append(b, line_len - 1);
      continue;
    }

    acc.append(b, line_len);

    //cerr << "acc = '" << acc << "'" << std::endl;
    ConfLine *cline = process_line(line_no, acc.c_str(), errors);
    acc.clear();
    if (!cline)
      continue;
    const std::string &csection(cline->newsection);
    if (!csection.empty()) {
      std::map <std::string, ConfSection>::value_type nt(csection, ConfSection());
      pair < section_iter_t, bool > nr(sections.insert(nt));
      cur_section = nr.first;
    }
    else {
      if (cur_section->second.lines.count(*cline)) {
	// replace an existing key/line in this section, so that
	//  [mysection]
	//    foo = 1
	//    foo = 2
	// will result in foo = 2.
	cur_section->second.lines.erase(*cline);
	if (cline->key.length() && warnings)
	  *warnings << "warning: line " << line_no << ": '" << cline->key << "' in section '"
		    << cur_section->first << "' redefined " << std::endl;
      }
      // add line to current section
      //std::cerr << "cur_section = " << cur_section->first << ", " << *cline << std::endl;
      cur_section->second.lines.insert(*cline);
    }
    delete cline;
  }

  if (!acc.empty()) {
    ostringstream oss;
    oss << "read_conf: don't end with lines that end in backslashes!";
    errors->push_back(oss.str());
  }
}

/*
 * A simple state-machine based parser.
 * This probably could/should be rewritten with something like boost::spirit
 * or yacc if the grammar ever gets more complex.
 */
ConfLine* ConfFile::
process_line(int line_no, const char *line, std::deque<std::string> *errors)
{
  enum acceptor_state_t {
    ACCEPT_INIT,
    ACCEPT_SECTION_NAME,
    ACCEPT_KEY,
    ACCEPT_VAL_START,
    ACCEPT_UNQUOTED_VAL,
    ACCEPT_QUOTED_VAL,
    ACCEPT_COMMENT_START,
    ACCEPT_COMMENT_TEXT,
  };
  const char *l = line;
  acceptor_state_t state = ACCEPT_INIT;
  string key, val, newsection, comment;
  bool escaping = false;
  while (true) {
    char c = *l++;
    switch (state) {
      case ACCEPT_INIT:
	if (c == '\0')
	  return NULL; // blank line. Not an error, but not interesting either.
	else if (c == '[')
	  state = ACCEPT_SECTION_NAME;
	else if ((c == '#') || (c == ';'))
	  state = ACCEPT_COMMENT_TEXT;
	else if (c == ']') {
	  ostringstream oss;
	  oss << "unexpected right bracket at char " << (l - line)
	      << ", line " << line_no;
	  errors->push_back(oss.str());
	  return NULL;
	}
	else if (isspace(c)) {
	  // ignore whitespace here
	}
	else {
	  // try to accept this character as a key
	  state = ACCEPT_KEY;
	  --l;
	}
	break;
      case ACCEPT_SECTION_NAME:
	if (c == '\0') {
	  ostringstream oss;
	  oss << "error parsing new section name: expected right bracket "
	      << "at char " << (l - line) << ", line " << line_no;
	  errors->push_back(oss.str());
	  return NULL;
	}
	else if ((c == ']') && (!escaping)) {
	  trim_whitespace(newsection, true);
	  if (newsection.empty()) {
	    ostringstream oss;
	    oss << "error parsing new section name: no section name found? "
	        << "at char " << (l - line) << ", line " << line_no;
	    errors->push_back(oss.str());
	    return NULL;
	  }
	  state = ACCEPT_COMMENT_START;
	}
	else if (((c == '#') || (c == ';')) && (!escaping)) {
	  ostringstream oss;
	  oss << "unexpected comment marker while parsing new section name, at "
	      << "char " << (l - line) << ", line " << line_no;
	  errors->push_back(oss.str());
	  return NULL;
	}
	else if ((c == '\\') && (!escaping)) {
	  escaping = true;
	}
	else {
	  escaping = false;
	  newsection += c;
	}
	break;
      case ACCEPT_KEY:
	if ((((c == '#') || (c == ';')) && (!escaping)) || (c == '\0')) {
	  ostringstream oss;
	  if (c == '\0') {
	    oss << "end of key=val line " << line_no
	        << " reached, no \"=val\" found...missing =?";
	  } else {
	    oss << "unexpected character while parsing putative key value, "
		<< "at char " << (l - line) << ", line " << line_no;
	  }
	  errors->push_back(oss.str());
	  return NULL;
	}
	else if ((c == '=') && (!escaping)) {
	  key = normalize_key_name(key);
	  if (key.empty()) {
	    ostringstream oss;
	    oss << "error parsing key name: no key name found? "
	        << "at char " << (l - line) << ", line " << line_no;
	    errors->push_back(oss.str());
	    return NULL;
	  }
	  state = ACCEPT_VAL_START;
	}
	else if ((c == '\\') && (!escaping)) {
	  escaping = true;
	}
	else {
	  escaping = false;
	  key += c;
	}
	break;
      case ACCEPT_VAL_START:
	if (c == '\0')
	  return new ConfLine(key, val, newsection, comment, line_no);
	else if ((c == '#') || (c == ';'))
	  state = ACCEPT_COMMENT_TEXT;
	else if (c == '"')
	  state = ACCEPT_QUOTED_VAL;
	else if (isspace(c)) {
	  // ignore whitespace
	}
	else {
	  // try to accept character as a val
	  state = ACCEPT_UNQUOTED_VAL;
	  --l;
	}
	break;
      case ACCEPT_UNQUOTED_VAL:
	if (c == '\0') {
	  if (escaping) {
	    ostringstream oss;
	    oss << "error parsing value name: unterminated escape sequence "
	        << "at char " << (l - line) << ", line " << line_no;
	    errors->push_back(oss.str());
	    return NULL;
	  }
	  trim_whitespace(val, false);
	  return new ConfLine(key, val, newsection, comment, line_no);
	}
	else if (((c == '#') || (c == ';')) && (!escaping)) {
	  trim_whitespace(val, false);
	  state = ACCEPT_COMMENT_TEXT;
	}
	else if ((c == '\\') && (!escaping)) {
	  escaping = true;
	}
	else {
	  escaping = false;
	  val += c;
	}
	break;
      case ACCEPT_QUOTED_VAL:
	if (c == '\0') {
	  ostringstream oss;
	  oss << "found opening quote for value, but not the closing quote. "
	      << "line " << line_no;
	  errors->push_back(oss.str());
	  return NULL;
	}
	else if ((c == '"') && (!escaping)) {
	  state = ACCEPT_COMMENT_START;
	}
	else if ((c == '\\') && (!escaping)) {
	  escaping = true;
	}
	else {
	  escaping = false;
	  // Add anything, including whitespace.
	  val += c;
	}
	break;
      case ACCEPT_COMMENT_START:
	if (c == '\0') {
	  return new ConfLine(key, val, newsection, comment, line_no);
	}
	else if ((c == '#') || (c == ';')) {
	  state = ACCEPT_COMMENT_TEXT;
	}
	else if (isspace(c)) {
	  // ignore whitespace
	}
	else {
	  ostringstream oss;
	  oss << "unexpected character at char " << (l - line) << " of line "
	      << line_no;
	  errors->push_back(oss.str());
	  return NULL;
	}
	break;
      case ACCEPT_COMMENT_TEXT:
	if (c == '\0')
	  return new ConfLine(key, val, newsection, comment, line_no);
	else
	  comment += c;
	break;
      default:
	ceph_abort();
	break;
    }
    ceph_assert(c != '\0'); // We better not go past the end of the input string.
  }
}