Training courses

Kernel and Embedded Linux

Bootlin training courses

Embedded Linux, kernel,
Yocto Project, Buildroot, real-time,
graphics, boot time, debugging...

Bootlin logo

Elixir Cross Referencer

  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
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
/*	$NetBSD: tok822_parse.c,v 1.1.1.1 2009/06/23 10:08:48 tron Exp $	*/

/*++
/* NAME
/*	tok822_parse 3
/* SUMMARY
/*	RFC 822 address parser
/* SYNOPSIS
/*	#include <tok822.h>
/*
/*	TOK822 *tok822_scan_limit(str, tailp, limit)
/*	const char *str;
/*	TOK822	**tailp;
/*	int	limit;
/*
/*	TOK822 *tok822_scan(str, tailp)
/*	const char *str;
/*	TOK822	**tailp;
/*
/*	TOK822	*tok822_parse_limit(str, limit)
/*	const char *str;
/*	int	limit;
/*
/*	TOK822	*tok822_parse(str)
/*	const char *str;
/*
/*	TOK822	*tok822_scan_addr(str)
/*	const char *str;
/*
/*	VSTRING	*tok822_externalize(buffer, tree, flags)
/*	VSTRING	*buffer;
/*	TOK822	*tree;
/*	int	flags;
/*
/*	VSTRING	*tok822_internalize(buffer, tree, flags)
/*	VSTRING	*buffer;
/*	TOK822	*tree;
/*	int	flags;
/* DESCRIPTION
/*	This module converts address lists between string form and parse
/*	tree formats. The string form can appear in two different ways:
/*	external (or quoted) form, as used in message headers, and internal
/*	(unquoted) form, as used internally by the mail software.
/*	Although RFC 822 expects 7-bit data, these routines pay no
/*	special attention to 8-bit characters.
/*
/*	tok822_scan() converts the external-form string in \fIstr\fR
/*	to a linear token list. The \fItailp\fR argument is a null pointer
/*	or receives the pointer value of the last result list element.
/*
/*	tok822_scan_limit() implements tok822_scan(), which is a macro.
/*	The \fIlimit\fR argument is either zero or an upper bound on the
/*	number of tokens produced.
/*
/*	tok822_parse() converts the external-form address list in
/*	\fIstr\fR to the corresponding token tree. The parser is permissive
/*	and will not throw away information that it does not understand.
/*	The parser adds missing commas between addresses.
/*
/*	tok822_parse_limit() implements tok822_parse(), which is a macro.
/*	The \fIlimit\fR argument is either zero or an upper bound on the
/*	number of tokens produced.
/*
/*	tok822_scan_addr() converts the external-form string in
/*	\fIstr\fR to an address token tree. This is just string to
/*	token list conversion; no parsing is done. This routine is
/*	suitable for data that should contain just one address and no
/*	other information.
/*
/*	tok822_externalize() converts a token list to external form.
/*	Where appropriate, characters and strings are quoted and white
/*	space is inserted. The \fIflags\fR argument is the binary OR of
/*	zero or more of the following:
/* .IP TOK822_STR_WIPE
/*	Initially, truncate the result to zero length.
/* .IP TOK822_STR_TERM
/*	Append a null terminator to the result when done.
/* .IP TOK822_STR_LINE
/*	Append a line break after each comma token, instead of appending
/*	whitespace.  It is up to the caller to concatenate short lines to
/*	produce longer ones.
/* .IP TOK822_STR_TRNC
/*	Truncate non-address information to 250 characters per address, to
/*	protect Sendmail systems that are vulnerable to the problem in CERT
/*	advisory CA-2003-07.
/*	This flag has effect with tok822_externalize() only.
/* .PP
/*	The macro TOK_822_NONE expresses that none of the above features
/*	should be activated.
/*
/*	The macro TOK822_STR_DEFL combines the TOK822_STR_WIPE and
/*	TOK822_STR_TERM flags. This is useful for most token to string
/*	conversions.
/*
/*	The macro TOK822_STR_HEAD combines the TOK822_STR_TERM,
/*	TOK822_STR_LINE and TOK822_STR_TRNC flags. This is useful for
/*	the special case of token to mail header conversion.
/*
/*	tok822_internalize() converts a token list to string form,
/*	without quoting. White space is inserted where appropriate.
/*	The \fIflags\fR argument is as with tok822_externalize().
/* STANDARDS
/* .ad
/* .fi
/*	RFC 822 (ARPA Internet Text Messages). In addition to this standard
/*	this module implements additional operators such as % and !. These
/*	are needed because the real world is not all RFC 822. Also, the ':'
/*	operator is allowed to appear inside addresses, to accommodate DECnet.
/*	In addition, 8-bit data is not given special treatment.
/* LICENSE
/* .ad
/* .fi
/*	The Secure Mailer license must be distributed with this software.
/* AUTHOR(S)
/*	Wietse Venema
/*	IBM T.J. Watson Research
/*	P.O. Box 704
/*	Yorktown Heights, NY 10598, USA
/*--*/

/* System library. */

#include <sys_defs.h>
#include <ctype.h>
#include <string.h>

/* Utility library. */

#include <vstring.h>
#include <msg.h>
#include <stringops.h>

/* Global library. */

#include "lex_822.h"
#include "quote_822_local.h"
#include "tok822.h"

 /*
  * I suppose this is my favorite macro. Used heavily for tokenizing.
  */
#define COLLECT(t,s,c,cond) { \
	while ((c = *(unsigned char *) s) != 0) { \
	    if (c == '\\') { \
		if ((c = *(unsigned char *)++s) == 0) \
		    break; \
	    } else if (!(cond)) { \
		break; \
	    } \
	    VSTRING_ADDCH(t->vstr, IS_SPACE_TAB_CR_LF(c) ? ' ' : c); \
	    s++; \
	} \
	VSTRING_TERMINATE(t->vstr); \
    }

#define COLLECT_SKIP_LAST(t,s,c,cond) { COLLECT(t,s,c,cond); if (*s) s++; }

 /*
  * Not quite as complex. The parser depends heavily on it.
  */
#define SKIP(tp, cond) { \
	while (tp->type && (cond)) \
	    tp = tp->prev; \
    }

#define MOVE_COMMENT_AND_CONTINUE(tp, right) { \
	TOK822 *prev = tok822_unlink(tp); \
	right = tok822_prepend(right, tp); \
	tp = prev; \
	continue; \
    }

#define SKIP_MOVE_COMMENT(tp, cond, right) { \
	while (tp->type && (cond)) { \
	    if (tp->type == TOK822_COMMENT) \
		MOVE_COMMENT_AND_CONTINUE(tp, right); \
	    tp = tp->prev; \
	} \
    }

 /*
  * Single-character operators. We include the % and ! operators because not
  * all the world is RFC822. XXX Make this operator list configurable when we
  * have a real rewriting language. Include | for aliases file parsing.
  */
static char tok822_opchar[] = "|%!" LEX_822_SPECIALS;
static void tok822_quote_atom(TOK822 *);
static const char *tok822_comment(TOK822 *, const char *);
static TOK822 *tok822_group(int, TOK822 *, TOK822 *, int);
static void tok822_copy_quoted(VSTRING *, char *, char *);
static int tok822_append_space(TOK822 *);

#define DO_WORD		(1<<0)		/* finding a word is ok here */
#define DO_GROUP	(1<<1)		/* doing an address group */

#define ADD_COMMA	','		/* resynchronize */
#define NO_MISSING_COMMA 0

/* tok822_internalize - token tree to string, internal form */

VSTRING *tok822_internalize(VSTRING *vp, TOK822 *tree, int flags)
{
    TOK822 *tp;

    if (flags & TOK822_STR_WIPE)
	VSTRING_RESET(vp);

    for (tp = tree; tp; tp = tp->next) {
	switch (tp->type) {
	case ',':
	    VSTRING_ADDCH(vp, tp->type);
	    if (flags & TOK822_STR_LINE) {
		VSTRING_ADDCH(vp, '\n');
		continue;
	    }
	    break;
	case TOK822_ADDR:
	    tok822_internalize(vp, tp->head, TOK822_STR_NONE);
	    break;
	case TOK822_COMMENT:
	case TOK822_ATOM:
	case TOK822_QSTRING:
	    vstring_strcat(vp, vstring_str(tp->vstr));
	    break;
	case TOK822_DOMLIT:
	    VSTRING_ADDCH(vp, '[');
	    vstring_strcat(vp, vstring_str(tp->vstr));
	    VSTRING_ADDCH(vp, ']');
	    break;
	case TOK822_STARTGRP:
	    VSTRING_ADDCH(vp, ':');
	    break;
	default:
	    if (tp->type >= TOK822_MINTOK)
		msg_panic("tok822_internalize: unknown operator %d", tp->type);
	    VSTRING_ADDCH(vp, tp->type);
	}
	if (tok822_append_space(tp))
	    VSTRING_ADDCH(vp, ' ');
    }
    if (flags & TOK822_STR_TERM)
	VSTRING_TERMINATE(vp);
    return (vp);
}

/* strip_address - strip non-address text from address expression */

static void strip_address(VSTRING *vp, ssize_t start, TOK822 *addr)
{
    VSTRING *tmp;

    /*
     * Emit plain <address>. Discard any comments or phrases.
     */
    VSTRING_TERMINATE(vp);
    msg_warn("stripping too many comments from address: %.100s...",
	     printable(vstring_str(vp) + start, '?'));
    vstring_truncate(vp, start);
    VSTRING_ADDCH(vp, '<');
    if (addr) {
	tmp = vstring_alloc(100);
	tok822_internalize(tmp, addr, TOK822_STR_TERM);
	quote_822_local_flags(vp, vstring_str(tmp),
			      QUOTE_FLAG_8BITCLEAN | QUOTE_FLAG_APPEND);
	vstring_free(tmp);
    }
    VSTRING_ADDCH(vp, '>');
}

/* tok822_externalize - token tree to string, external form */

VSTRING *tok822_externalize(VSTRING *vp, TOK822 *tree, int flags)
{
    VSTRING *tmp;
    TOK822 *tp;
    ssize_t start;
    TOK822 *addr;
    ssize_t addr_len;

    /*
     * Guard against a Sendmail buffer overflow (CERT advisory CA-2003-07).
     * The problem was that Sendmail could store too much non-address text
     * (comments, phrases, etc.) into a static 256-byte buffer.
     * 
     * When the buffer fills up, fixed Sendmail versions remove comments etc.
     * and reduce the information to just <$g>, which expands to <address>.
     * No change is made when an address expression (text separated by
     * commas) contains no address. This fix reportedly also protects
     * Sendmail systems that are still vulnerable to this problem.
     * 
     * Postfix takes the same approach, grudgingly. To avoid unnecessary damage,
     * Postfix removes comments etc. only when the amount of non-address text
     * in an address expression (text separated by commas) exceeds 250 bytes.
     * 
     * With Sendmail, the address part of an address expression is the
     * right-most <> instance in that expression. If an address expression
     * contains no <>, then Postfix guarantees that it contains at most one
     * non-comment string; that string is the address part of the address
     * expression, so there is no ambiguity.
     * 
     * Finally, we note that stress testing shows that other code in Sendmail
     * 8.12.8 bluntly truncates ``text <address>'' to 256 bytes even when
     * this means chopping the <address> somewhere in the middle. This is a
     * loss of control that we're not entirely comfortable with. However,
     * unbalanced quotes and dangling backslash do not seem to influence the
     * way that Sendmail parses headers, so this is not an urgent problem.
     */
#define MAX_NONADDR_LENGTH 250

#define RESET_NONADDR_LENGTH { \
	start = VSTRING_LEN(vp); \
	addr = 0; \
	addr_len = 0; \
    }

#define ENFORCE_NONADDR_LENGTH do { \
	if (addr && VSTRING_LEN(vp) - addr_len > start + MAX_NONADDR_LENGTH) \
	    strip_address(vp, start, addr->head); \
    } while(0)

    if (flags & TOK822_STR_WIPE)
	VSTRING_RESET(vp);

    if (flags & TOK822_STR_TRNC)
	RESET_NONADDR_LENGTH;

    for (tp = tree; tp; tp = tp->next) {
	switch (tp->type) {
	case ',':
	    if (flags & TOK822_STR_TRNC)
		ENFORCE_NONADDR_LENGTH;
	    VSTRING_ADDCH(vp, tp->type);
	    VSTRING_ADDCH(vp, (flags & TOK822_STR_LINE) ? '\n' : ' ');
	    if (flags & TOK822_STR_TRNC)
		RESET_NONADDR_LENGTH;
	    continue;

	    /*
	     * XXX In order to correctly externalize an address, it is not
	     * sufficient to quote individual atoms. There are higher-level
	     * rules that say when an address localpart needs to be quoted.
	     * We wing it with the quote_822_local() routine, which ignores
	     * the issue of atoms in the domain part that would need quoting.
	     */
	case TOK822_ADDR:
	    addr = tp;
	    tmp = vstring_alloc(100);
	    tok822_internalize(tmp, tp->head, TOK822_STR_TERM);
	    addr_len = VSTRING_LEN(vp);
	    quote_822_local_flags(vp, vstring_str(tmp),
				  QUOTE_FLAG_8BITCLEAN | QUOTE_FLAG_APPEND);
	    addr_len = VSTRING_LEN(vp) - addr_len;
	    vstring_free(tmp);
	    break;
	case TOK822_ATOM:
	case TOK822_COMMENT:
	    vstring_strcat(vp, vstring_str(tp->vstr));
	    break;
	case TOK822_QSTRING:
	    VSTRING_ADDCH(vp, '"');
	    tok822_copy_quoted(vp, vstring_str(tp->vstr), "\"\\\r\n");
	    VSTRING_ADDCH(vp, '"');
	    break;
	case TOK822_DOMLIT:
	    VSTRING_ADDCH(vp, '[');
	    tok822_copy_quoted(vp, vstring_str(tp->vstr), "\\\r\n");
	    VSTRING_ADDCH(vp, ']');
	    break;
	case TOK822_STARTGRP:
	    VSTRING_ADDCH(vp, ':');
	    break;
	case '<':
	    if (tp->next && tp->next->type == '>') {
		addr = tp;
		addr_len = 0;
	    }
	    VSTRING_ADDCH(vp, '<');
	    break;
	default:
	    if (tp->type >= TOK822_MINTOK)
		msg_panic("tok822_externalize: unknown operator %d", tp->type);
	    VSTRING_ADDCH(vp, tp->type);
	}
	if (tok822_append_space(tp))
	    VSTRING_ADDCH(vp, ' ');
    }
    if (flags & TOK822_STR_TRNC)
	ENFORCE_NONADDR_LENGTH;

    if (flags & TOK822_STR_TERM)
	VSTRING_TERMINATE(vp);
    return (vp);
}

/* tok822_copy_quoted - copy a string while quoting */

static void tok822_copy_quoted(VSTRING *vp, char *str, char *quote_set)
{
    int     ch;

    while ((ch = *(unsigned char *) str++) != 0) {
	if (strchr(quote_set, ch))
	    VSTRING_ADDCH(vp, '\\');
	VSTRING_ADDCH(vp, ch);
    }
}

/* tok822_append_space - see if space is needed after this token */

static int tok822_append_space(TOK822 *tp)
{
    TOK822 *next;

    if (tp == 0 || (next = tp->next) == 0 || tp->owner != 0)
	return (0);
    if (tp->type == ',' || tp->type == TOK822_STARTGRP || next->type == '<')
	return (1);

#define NON_OPERATOR(x) \
    (x->type == TOK822_ATOM || x->type == TOK822_QSTRING \
     || x->type == TOK822_COMMENT || x->type == TOK822_DOMLIT \
     || x->type == TOK822_ADDR)

    return (NON_OPERATOR(tp) && NON_OPERATOR(next));
}

/* tok822_scan_limit - tokenize string */

TOK822 *tok822_scan_limit(const char *str, TOK822 **tailp, int tok_count_limit)
{
    TOK822 *head = 0;
    TOK822 *tail = 0;
    TOK822 *tp;
    int     ch;
    int     tok_count = 0;

    /*
     * XXX 2822 new feature: Section 4.1 allows "." to appear in a phrase (to
     * allow for forms such as: Johnny B. Goode <johhny@domain.org>. I cannot
     * handle that at the tokenizer level - it is not context sensitive. And
     * to fix this at the parser level requires radical changes to preserve
     * white space as part of the token stream. Thanks a lot, people.
     */
    while ((ch = *(unsigned char *) str++) != 0) {
	if (IS_SPACE_TAB_CR_LF(ch))
	    continue;
	if (ch == '(') {
	    tp = tok822_alloc(TOK822_COMMENT, (char *) 0);
	    str = tok822_comment(tp, str);
	} else if (ch == '[') {
	    tp = tok822_alloc(TOK822_DOMLIT, (char *) 0);
	    COLLECT_SKIP_LAST(tp, str, ch, ch != ']');
	} else if (ch == '"') {
	    tp = tok822_alloc(TOK822_QSTRING, (char *) 0);
	    COLLECT_SKIP_LAST(tp, str, ch, ch != '"');
	} else if (ch != '\\' && strchr(tok822_opchar, ch)) {
	    tp = tok822_alloc(ch, (char *) 0);
	} else {
	    tp = tok822_alloc(TOK822_ATOM, (char *) 0);
	    str -= 1;				/* \ may be first */
	    COLLECT(tp, str, ch, !IS_SPACE_TAB_CR_LF(ch) && !strchr(tok822_opchar, ch));
	    tok822_quote_atom(tp);
	}
	if (head == 0) {
	    head = tail = tp;
	    while (tail->next)
		tail = tail->next;
	} else {
	    tail = tok822_append(tail, tp);
	}
	if (tok_count_limit > 0 && ++tok_count >= tok_count_limit)
	    break;
    }
    if (tailp)
	*tailp = tail;
    return (head);
}

/* tok822_parse_limit - translate external string to token tree */

TOK822 *tok822_parse_limit(const char *str, int tok_count_limit)
{
    TOK822 *head;
    TOK822 *tail;
    TOK822 *right;
    TOK822 *first_token;
    TOK822 *last_token;
    TOK822 *tp;
    int     state;

    /*
     * First, tokenize the string, from left to right. We are not allowed to
     * throw away any information that we do not understand. With a flat
     * token list that contains all tokens, we can always convert back to
     * string form.
     */
    if ((first_token = tok822_scan_limit(str, &last_token, tok_count_limit)) == 0)
	return (0);

    /*
     * For convenience, sandwich the token list between two sentinel tokens.
     */
#define GLUE(left,rite) { left->next = rite; rite->prev = left; }

    head = tok822_alloc(0, (char *) 0);
    GLUE(head, first_token);
    tail = tok822_alloc(0, (char *) 0);
    GLUE(last_token, tail);

    /*
     * Next step is to transform the token list into a parse tree. This is
     * done most conveniently from right to left. If there is something that
     * we do not understand, just leave it alone, don't throw it away. The
     * address information that we're looking for sits in-between the current
     * node (tp) and the one called right. Add missing commas on the fly.
     */
    state = DO_WORD;
    right = tail;
    tp = tail->prev;
    while (tp->type) {
	if (tp->type == TOK822_COMMENT) {	/* move comment to the side */
	    MOVE_COMMENT_AND_CONTINUE(tp, right);
	} else if (tp->type == ';') {		/* rh side of named group */
	    right = tok822_group(TOK822_ADDR, tp, right, ADD_COMMA);
	    state = DO_GROUP | DO_WORD;
	} else if (tp->type == ':' && (state & DO_GROUP) != 0) {
	    tp->type = TOK822_STARTGRP;
	    (void) tok822_group(TOK822_ADDR, tp, right, NO_MISSING_COMMA);
	    SKIP(tp, tp->type != ',');
	    right = tp;
	    continue;
	} else if (tp->type == '>') {		/* rh side of <route> */
	    right = tok822_group(TOK822_ADDR, tp, right, ADD_COMMA);
	    SKIP_MOVE_COMMENT(tp, tp->type != '<', right);
	    (void) tok822_group(TOK822_ADDR, tp, right, NO_MISSING_COMMA);
	    SKIP(tp, tp->type > 0xff || strchr(">;,:", tp->type) == 0);
	    right = tp;
	    state |= DO_WORD;
	    continue;
	} else if (tp->type == TOK822_ATOM || tp->type == TOK822_QSTRING
		   || tp->type == TOK822_DOMLIT) {
	    if ((state & DO_WORD) == 0)
		right = tok822_group(TOK822_ADDR, tp, right, ADD_COMMA)->next;
	    state &= ~DO_WORD;
	} else if (tp->type == ',') {
	    right = tok822_group(TOK822_ADDR, tp, right, NO_MISSING_COMMA);
	    state |= DO_WORD;
	} else {
	    state |= DO_WORD;
	}
	tp = tp->prev;
    }
    (void) tok822_group(TOK822_ADDR, tp, right, NO_MISSING_COMMA);

    /*
     * Discard the sentinel tokens on the left and right extremes. Properly
     * terminate the resulting list.
     */
    tp = (head->next != tail ? head->next : 0);
    tok822_cut_before(head->next);
    tok822_free(head);
    tok822_cut_before(tail);
    tok822_free(tail);
    return (tp);
}

/* tok822_quote_atom - see if an atom needs quoting when externalized */

static void tok822_quote_atom(TOK822 *tp)
{
    char   *cp;
    int     ch;

    /*
     * RFC 822 expects 7-bit data. Rather than quoting every 8-bit character
     * (and still passing it on as 8-bit data) we leave 8-bit data alone.
     */
    for (cp = vstring_str(tp->vstr); (ch = *(unsigned char *) cp) != 0; cp++) {
	if ( /* !ISASCII(ch) || */ ch == ' '
	    || ISCNTRL(ch) || strchr(tok822_opchar, ch)) {
	    tp->type = TOK822_QSTRING;
	    break;
	}
    }
}

/* tok822_comment - tokenize comment */

static const char *tok822_comment(TOK822 *tp, const char *str)
{
    int     level = 1;
    int     ch;

    /*
     * XXX We cheat by storing comments in their external form. Otherwise it
     * would be a royal pain to preserve \ before (. That would require a
     * recursive parser; the easy to implement stack-based recursion would be
     * too expensive.
     */
    VSTRING_ADDCH(tp->vstr, '(');

    while ((ch = *(unsigned char *) str) != 0) {
	VSTRING_ADDCH(tp->vstr, ch);
	str++;
	if (ch == '(') {			/* comments can nest! */
	    level++;
	} else if (ch == ')') {
	    if (--level == 0)
		break;
	} else if (ch == '\\') {
	    if ((ch = *(unsigned char *) str) == 0)
		break;
	    VSTRING_ADDCH(tp->vstr, ch);
	    str++;
	}
    }
    VSTRING_TERMINATE(tp->vstr);
    return (str);
}

/* tok822_group - cluster a group of tokens */

static TOK822 *tok822_group(int group_type, TOK822 *left, TOK822 *right, int sync_type)
{
    TOK822 *group;
    TOK822 *sync;
    TOK822 *first;

    /*
     * Cluster the tokens between left and right under their own parse tree
     * node. Optionally insert a resync token.
     */
    if (left != right && (first = left->next) != right) {
	tok822_cut_before(right);
	tok822_cut_before(first);
	group = tok822_alloc(group_type, (char *) 0);
	tok822_sub_append(group, first);
	tok822_append(left, group);
	tok822_append(group, right);
	if (sync_type) {
	    sync = tok822_alloc(sync_type, (char *) 0);
	    tok822_append(left, sync);
	}
    }
    return (left);
}

/* tok822_scan_addr - convert external address string to address token */

TOK822 *tok822_scan_addr(const char *addr)
{
    TOK822 *tree = tok822_alloc(TOK822_ADDR, (char *) 0);

    tree->head = tok822_scan(addr, &tree->tail);
    return (tree);
}

#ifdef TEST

#include <unistd.h>
#include <vstream.h>
#include <readlline.h>

/* tok822_print - display token */

static void tok822_print(TOK822 *list, int indent)
{
    TOK822 *tp;

    for (tp = list; tp; tp = tp->next) {
	if (tp->type < TOK822_MINTOK) {
	    vstream_printf("%*s %s \"%c\"\n", indent, "", "OP", tp->type);
	} else if (tp->type == TOK822_ADDR) {
	    vstream_printf("%*s %s\n", indent, "", "address");
	    tok822_print(tp->head, indent + 2);
	} else if (tp->type == TOK822_STARTGRP) {
	    vstream_printf("%*s %s\n", indent, "", "group \":\"");
	} else {
	    vstream_printf("%*s %s \"%s\"\n", indent, "",
			   tp->type == TOK822_COMMENT ? "comment" :
			   tp->type == TOK822_ATOM ? "atom" :
			   tp->type == TOK822_QSTRING ? "quoted string" :
			   tp->type == TOK822_DOMLIT ? "domain literal" :
			   tp->type == TOK822_ADDR ? "address" :
			   "unknown\n", vstring_str(tp->vstr));
	}
    }
}

int     main(int unused_argc, char **unused_argv)
{
    VSTRING *vp = vstring_alloc(100);
    TOK822 *list;
    VSTRING *buf = vstring_alloc(100);

#define TEST_TOKEN_LIMIT 20

    while (readlline(buf, VSTREAM_IN, (int *) 0)) {
	while (VSTRING_LEN(buf) > 0 && vstring_end(buf)[-1] == '\n') {
	    vstring_end(buf)[-1] = 0;
	    vstring_truncate(buf, VSTRING_LEN(buf) - 1);
	}
	if (!isatty(vstream_fileno(VSTREAM_IN)))
	    vstream_printf(">>>%s<<<\n\n", vstring_str(buf));
	list = tok822_parse_limit(vstring_str(buf), TEST_TOKEN_LIMIT);
	vstream_printf("Parse tree:\n");
	tok822_print(list, 0);
	vstream_printf("\n");

	vstream_printf("Internalized:\n%s\n\n",
		vstring_str(tok822_internalize(vp, list, TOK822_STR_DEFL)));
	vstream_fflush(VSTREAM_OUT);
	vstream_printf("Externalized, no newlines inserted:\n%s\n\n",
		       vstring_str(tok822_externalize(vp, list,
				       TOK822_STR_DEFL | TOK822_STR_TRNC)));
	vstream_fflush(VSTREAM_OUT);
	vstream_printf("Externalized, newlines inserted:\n%s\n\n",
		       vstring_str(tok822_externalize(vp, list,
		     TOK822_STR_DEFL | TOK822_STR_LINE | TOK822_STR_TRNC)));
	vstream_fflush(VSTREAM_OUT);
	tok822_free_tree(list);
    }
    vstring_free(vp);
    vstring_free(buf);
    return (0);
}

#endif