]> git.saurik.com Git - apt.git/blame - apt-pkg/contrib/strutl.cc
fix "Mismatched free() / delete / delete []" in simple_buffer
[apt.git] / apt-pkg / contrib / strutl.cc
CommitLineData
6c139d6e
AL
1// -*- mode: cpp; mode: fold -*-
2// Description /*{{{*/
d48c6a7d 3// $Id: strutl.cc,v 1.48 2003/07/18 14:15:11 mdz Exp $
6c139d6e
AL
4/* ######################################################################
5
b2e465d6 6 String Util - Some useful string functions.
6c139d6e 7
b2e465d6
AL
8 These have been collected from here and there to do all sorts of useful
9 things to strings. They are useful in file parsers, URI handlers and
24231681 10 especially in APT methods.
6c139d6e
AL
11
12 This source is placed in the Public Domain, do with it what you will
24231681 13 It was originally written by Jason Gunthorpe <jgg@gpu.srv.ualberta.ca>
6c139d6e
AL
14
15 ##################################################################### */
16 /*}}}*/
17// Includes /*{{{*/
ea542140
DK
18#include <config.h>
19
cdcc6d34 20#include <apt-pkg/strutl.h>
7049d16d 21#include <apt-pkg/fileutl.h>
b2e465d6 22#include <apt-pkg/error.h>
0a8a80e5 23
453b82a3
DK
24#include <stddef.h>
25#include <stdlib.h>
26#include <time.h>
27#include <string>
28#include <vector>
6c139d6e
AL
29#include <ctype.h>
30#include <string.h>
5076b3c2 31#include <sstream>
6c139d6e 32#include <stdio.h>
152ab79e 33#include <algorithm>
2b154e53 34#include <unistd.h>
b2e465d6 35#include <regex.h>
b0db36b1 36#include <errno.h>
b2e465d6 37#include <stdarg.h>
a52f938b 38#include <iconv.h>
0db4a45b 39
ea542140 40#include <apti18n.h>
6c139d6e 41 /*}}}*/
453b82a3
DK
42using namespace std;
43
65dbd5a1
MV
44// Strip - Remove white space from the front and back of a string /*{{{*/
45// ---------------------------------------------------------------------
46namespace APT {
47 namespace String {
b5787388 48std::string Strip(const std::string &str)
65dbd5a1 49{
b5787388
DK
50 // ensure we have at least one character
51 if (str.empty() == true)
52 return str;
53
54 char const * const s = str.c_str();
55 size_t start = 0;
56 for (; isspace(s[start]) != 0; ++start)
57 ; // find the first not-space
58
59 // string contains only whitespaces
60 if (s[start] == '\0')
65dbd5a1 61 return "";
b5787388
DK
62
63 size_t end = str.length() - 1;
64 for (; isspace(s[end]) != 0; --end)
65 ; // find the last not-space
66
67 return str.substr(start, end - start + 1);
65dbd5a1 68}
cf993341
MV
69
70bool Endswith(const std::string &s, const std::string &end)
71{
72 if (end.size() > s.size())
73 return false;
74 return (s.substr(s.size() - end.size(), s.size()) == end);
75}
76
a5bb5e1e
MV
77bool Startswith(const std::string &s, const std::string &start)
78{
79 if (start.size() > s.size())
80 return false;
81 return (s.substr(0, start.size()) == start);
82}
83
65dbd5a1
MV
84}
85}
86 /*}}}*/
a52f938b
OS
87// UTF8ToCodeset - Convert some UTF-8 string for some codeset /*{{{*/
88// ---------------------------------------------------------------------
89/* This is handy to use before display some information for enduser */
90bool UTF8ToCodeset(const char *codeset, const string &orig, string *dest)
91{
92 iconv_t cd;
93 const char *inbuf;
b39c1859
MV
94 char *inptr, *outbuf;
95 size_t insize, bufsize;
96 dest->clear();
97
a52f938b
OS
98 cd = iconv_open(codeset, "UTF-8");
99 if (cd == (iconv_t)(-1)) {
100 // Something went wrong
101 if (errno == EINVAL)
102 _error->Error("conversion from 'UTF-8' to '%s' not available",
103 codeset);
104 else
105 perror("iconv_open");
106
a52f938b
OS
107 return false;
108 }
109
b39c1859 110 insize = bufsize = orig.size();
a52f938b
OS
111 inbuf = orig.data();
112 inptr = (char *)inbuf;
b39c1859
MV
113 outbuf = new char[bufsize];
114 size_t lastError = -1;
a52f938b 115
1f99b6d3
DK
116 while (insize != 0)
117 {
b39c1859
MV
118 char *outptr = outbuf;
119 size_t outsize = bufsize;
1f99b6d3 120 size_t const err = iconv(cd, &inptr, &insize, &outptr, &outsize);
b39c1859 121 dest->append(outbuf, outptr - outbuf);
1f99b6d3
DK
122 if (err == (size_t)(-1))
123 {
b39c1859
MV
124 switch (errno)
125 {
126 case EILSEQ:
127 insize--;
128 inptr++;
129 // replace a series of unknown multibytes with a single "?"
130 if (lastError != insize) {
131 lastError = insize - 1;
132 dest->append("?");
133 }
134 break;
135 case EINVAL:
136 insize = 0;
137 break;
138 case E2BIG:
139 if (outptr == outbuf)
140 {
141 bufsize *= 2;
142 delete[] outbuf;
143 outbuf = new char[bufsize];
144 }
145 break;
146 }
1f99b6d3
DK
147 }
148 }
a52f938b 149
a52f938b
OS
150 delete[] outbuf;
151
152 iconv_close(cd);
153
154 return true;
155}
156 /*}}}*/
6c139d6e
AL
157// strstrip - Remove white space from the front and back of a string /*{{{*/
158// ---------------------------------------------------------------------
159/* This is handy to use when parsing a file. It also removes \n's left
160 over from fgets and company */
161char *_strstrip(char *String)
162{
163 for (;*String != 0 && (*String == ' ' || *String == '\t'); String++);
164
165 if (*String == 0)
166 return String;
4fb400a6
MV
167 return _strrstrip(String);
168}
169 /*}}}*/
170// strrstrip - Remove white space from the back of a string /*{{{*/
171// ---------------------------------------------------------------------
172char *_strrstrip(char *String)
173{
6c139d6e
AL
174 char *End = String + strlen(String) - 1;
175 for (;End != String - 1 && (*End == ' ' || *End == '\t' || *End == '\n' ||
176 *End == '\r'); End--);
177 End++;
178 *End = 0;
179 return String;
d3e8fbb3 180}
6c139d6e
AL
181 /*}}}*/
182// strtabexpand - Converts tabs into 8 spaces /*{{{*/
183// ---------------------------------------------------------------------
184/* */
185char *_strtabexpand(char *String,size_t Len)
186{
187 for (char *I = String; I != I + Len && *I != 0; I++)
188 {
189 if (*I != '\t')
190 continue;
191 if (I + 8 > String + Len)
192 {
193 *I = 0;
194 return String;
195 }
196
197 /* Assume the start of the string is 0 and find the next 8 char
198 division */
199 int Len;
200 if (String == I)
201 Len = 1;
202 else
203 Len = 8 - ((String - I) % 8);
204 Len -= 2;
205 if (Len <= 0)
206 {
207 *I = ' ';
208 continue;
209 }
210
211 memmove(I + Len,I + 1,strlen(I) + 1);
212 for (char *J = I; J + Len != I; *I = ' ', I++);
213 }
214 return String;
215}
216 /*}}}*/
217// ParseQuoteWord - Parse a single word out of a string /*{{{*/
218// ---------------------------------------------------------------------
219/* This grabs a single word, converts any % escaped characters to their
220 proper values and advances the pointer. Double quotes are understood
7834cb57
AL
221 and striped out as well. This is for URI/URL parsing. It also can
222 understand [] brackets.*/
6c139d6e
AL
223bool ParseQuoteWord(const char *&String,string &Res)
224{
225 // Skip leading whitespace
226 const char *C = String;
227 for (;*C != 0 && *C == ' '; C++);
228 if (*C == 0)
229 return false;
230
231 // Jump to the next word
36f610f1 232 for (;*C != 0 && isspace(*C) == 0; C++)
6c139d6e
AL
233 {
234 if (*C == '"')
235 {
404528bd
DK
236 C = strchr(C + 1, '"');
237 if (C == NULL)
7834cb57
AL
238 return false;
239 }
240 if (*C == '[')
241 {
404528bd
DK
242 C = strchr(C + 1, ']');
243 if (C == NULL)
6c139d6e
AL
244 return false;
245 }
246 }
247
248 // Now de-quote characters
249 char Buffer[1024];
250 char Tmp[3];
251 const char *Start = String;
252 char *I;
253 for (I = Buffer; I < Buffer + sizeof(Buffer) && Start != C; I++)
254 {
436d7eab
DK
255 if (*Start == '%' && Start + 2 < C &&
256 isxdigit(Start[1]) && isxdigit(Start[2]))
6c139d6e
AL
257 {
258 Tmp[0] = Start[1];
259 Tmp[1] = Start[2];
1bc849af 260 Tmp[2] = 0;
6c139d6e
AL
261 *I = (char)strtol(Tmp,0,16);
262 Start += 3;
263 continue;
264 }
265 if (*Start != '"')
266 *I = *Start;
267 else
268 I--;
269 Start++;
270 }
271 *I = 0;
272 Res = Buffer;
273
274 // Skip ending white space
36f610f1 275 for (;*C != 0 && isspace(*C) != 0; C++);
6c139d6e
AL
276 String = C;
277 return true;
278}
279 /*}}}*/
08e8f724
AL
280// ParseCWord - Parses a string like a C "" expression /*{{{*/
281// ---------------------------------------------------------------------
b2e465d6 282/* This expects a series of space separated strings enclosed in ""'s.
08e8f724 283 It concatenates the ""'s into a single string. */
b2e465d6 284bool ParseCWord(const char *&String,string &Res)
08e8f724
AL
285{
286 // Skip leading whitespace
287 const char *C = String;
288 for (;*C != 0 && *C == ' '; C++);
289 if (*C == 0)
290 return false;
291
292 char Buffer[1024];
293 char *Buf = Buffer;
294 if (strlen(String) >= sizeof(Buffer))
295 return false;
296
297 for (; *C != 0; C++)
298 {
299 if (*C == '"')
300 {
301 for (C++; *C != 0 && *C != '"'; C++)
302 *Buf++ = *C;
303
304 if (*C == 0)
305 return false;
306
307 continue;
308 }
309
310 if (C != String && isspace(*C) != 0 && isspace(C[-1]) != 0)
311 continue;
312 if (isspace(*C) == 0)
313 return false;
314 *Buf++ = ' ';
b2e465d6 315 }
08e8f724
AL
316 *Buf = 0;
317 Res = Buffer;
b2e465d6 318 String = C;
08e8f724
AL
319 return true;
320}
321 /*}}}*/
6d5dd02a 322// QuoteString - Convert a string into quoted from /*{{{*/
1bc849af 323// ---------------------------------------------------------------------
6d5dd02a 324/* */
171c75f1 325string QuoteString(const string &Str, const char *Bad)
1bc849af 326{
b8eba208 327 std::stringstream Res;
f7f0d6c7 328 for (string::const_iterator I = Str.begin(); I != Str.end(); ++I)
1bc849af 329 {
b8eba208 330 if (strchr(Bad,*I) != 0 || isprint(*I) == 0 ||
436d7eab
DK
331 *I == 0x25 || // percent '%' char
332 *I <= 0x20 || *I >= 0x7F) // control chars
1bc849af 333 {
4453cfdc 334 ioprintf(Res, "%%%02hhx", *I);
1bc849af
AL
335 }
336 else
b8eba208 337 Res << *I;
1bc849af 338 }
b8eba208 339 return Res.str();
1bc849af
AL
340}
341 /*}}}*/
6d5dd02a 342// DeQuoteString - Convert a string from quoted from /*{{{*/
6c139d6e 343// ---------------------------------------------------------------------
6d5dd02a 344/* This undoes QuoteString */
171c75f1 345string DeQuoteString(const string &Str)
436d7eab
DK
346{
347 return DeQuoteString(Str.begin(),Str.end());
348}
349string DeQuoteString(string::const_iterator const &begin,
350 string::const_iterator const &end)
6c139d6e
AL
351{
352 string Res;
f7f0d6c7 353 for (string::const_iterator I = begin; I != end; ++I)
6c139d6e 354 {
436d7eab
DK
355 if (*I == '%' && I + 2 < end &&
356 isxdigit(I[1]) && isxdigit(I[2]))
6c139d6e 357 {
6d5dd02a
AL
358 char Tmp[3];
359 Tmp[0] = I[1];
360 Tmp[1] = I[2];
361 Tmp[2] = 0;
362 Res += (char)strtol(Tmp,0,16);
363 I += 2;
364 continue;
6c139d6e
AL
365 }
366 else
367 Res += *I;
368 }
6d5dd02a 369 return Res;
6c139d6e 370}
6d5dd02a
AL
371
372 /*}}}*/
6c139d6e
AL
373// SizeToStr - Convert a long into a human readable size /*{{{*/
374// ---------------------------------------------------------------------
24231681
AL
375/* A max of 4 digits are shown before conversion to the next highest unit.
376 The max length of the string will be 5 chars unless the size is > 10
6c139d6e
AL
377 YottaBytes (E24) */
378string SizeToStr(double Size)
379{
6c139d6e
AL
380 double ASize;
381 if (Size >= 0)
382 ASize = Size;
383 else
384 ASize = -1*Size;
b8eba208 385
6c139d6e
AL
386 /* bytes, KiloBytes, MegaBytes, GigaBytes, TeraBytes, PetaBytes,
387 ExaBytes, ZettaBytes, YottaBytes */
7f25bdff 388 char Ext[] = {'\0','k','M','G','T','P','E','Z','Y'};
6c139d6e
AL
389 int I = 0;
390 while (I <= 8)
391 {
392 if (ASize < 100 && I != 0)
393 {
b8eba208
DK
394 std::string S;
395 strprintf(S, "%'.1f %c", ASize, Ext[I]);
396 return S;
6c139d6e 397 }
b8eba208 398
6c139d6e
AL
399 if (ASize < 10000)
400 {
b8eba208
DK
401 std::string S;
402 strprintf(S, "%'.0f %c", ASize, Ext[I]);
403 return S;
6c139d6e
AL
404 }
405 ASize /= 1000.0;
406 I++;
407 }
b8eba208 408 return "";
6c139d6e
AL
409}
410 /*}}}*/
411// TimeToStr - Convert the time into a string /*{{{*/
412// ---------------------------------------------------------------------
413/* Converts a number of seconds to a hms format */
414string TimeToStr(unsigned long Sec)
415{
b8eba208
DK
416 std::string S;
417 if (Sec > 60*60*24)
6c139d6e 418 {
b8eba208
DK
419 //TRANSLATOR: d means days, h means hours, min means minutes, s means seconds
420 strprintf(S,_("%lid %lih %limin %lis"),Sec/60/60/24,(Sec/60/60) % 24,(Sec/60) % 60,Sec % 60);
421 }
422 else if (Sec > 60*60)
423 {
424 //TRANSLATOR: h means hours, min means minutes, s means seconds
425 strprintf(S,_("%lih %limin %lis"),Sec/60/60,(Sec/60) % 60,Sec % 60);
426 }
427 else if (Sec > 60)
428 {
429 //TRANSLATOR: min means minutes, s means seconds
430 strprintf(S,_("%limin %lis"),Sec/60,Sec % 60);
431 }
432 else
433 {
434 //TRANSLATOR: s means seconds
435 strprintf(S,_("%lis"),Sec);
6c139d6e 436 }
6c139d6e
AL
437 return S;
438}
439 /*}}}*/
440// SubstVar - Substitute a string for another string /*{{{*/
441// ---------------------------------------------------------------------
1e3f4083 442/* This replaces all occurrences of Subst with Contents in Str. */
171c75f1 443string SubstVar(const string &Str,const string &Subst,const string &Contents)
6c139d6e 444{
224dc038
DK
445 if (Subst.empty() == true)
446 return Str;
447
8efa2a3b 448 string::size_type Pos = 0;
6c139d6e
AL
449 string::size_type OldPos = 0;
450 string Temp;
224dc038
DK
451
452 while (OldPos < Str.length() &&
6c139d6e
AL
453 (Pos = Str.find(Subst,OldPos)) != string::npos)
454 {
224dc038
DK
455 if (OldPos != Pos)
456 Temp.append(Str, OldPos, Pos - OldPos);
457 if (Contents.empty() == false)
458 Temp.append(Contents);
459 OldPos = Pos + Subst.length();
6c139d6e 460 }
224dc038 461
6c139d6e
AL
462 if (OldPos == 0)
463 return Str;
224dc038
DK
464
465 if (OldPos >= Str.length())
466 return Temp;
6c139d6e
AL
467 return Temp + string(Str,OldPos);
468}
b2e465d6
AL
469string SubstVar(string Str,const struct SubstVar *Vars)
470{
471 for (; Vars->Subst != 0; Vars++)
472 Str = SubstVar(Str,Vars->Subst,*Vars->Contents);
473 return Str;
474}
6c139d6e 475 /*}}}*/
fa3b0945
MV
476// OutputInDepth - return a string with separator multiplied with depth /*{{{*/
477// ---------------------------------------------------------------------
478/* Returns a string with the supplied separator depth + 1 times in it */
479std::string OutputInDepth(const unsigned long Depth, const char* Separator)
480{
481 std::string output = "";
482 for(unsigned long d=Depth+1; d > 0; d--)
483 output.append(Separator);
484 return output;
485}
486 /*}}}*/
ad00ae81
AL
487// URItoFileName - Convert the uri into a unique file name /*{{{*/
488// ---------------------------------------------------------------------
489/* This converts a URI into a safe filename. It quotes all unsafe characters
490 and converts / to _ and removes the scheme identifier. The resulting
491 file name should be unique and never occur again for a different file */
171c75f1 492string URItoFileName(const string &URI)
ad00ae81 493{
54cf15cb
AL
494 // Nuke 'sensitive' items
495 ::URI U(URI);
171c75f1
MV
496 U.User.clear();
497 U.Password.clear();
498 U.Access.clear();
54cf15cb 499
ad00ae81 500 // "\x00-\x20{}|\\\\^\\[\\]<>\"\x7F-\xFF";
171c75f1
MV
501 string NewURI = QuoteString(U,"\\|{}[]<>\"^~_=!@#$%^&*");
502 replace(NewURI.begin(),NewURI.end(),'/','_');
503 return NewURI;
ad00ae81
AL
504}
505 /*}}}*/
6c139d6e
AL
506// Base64Encode - Base64 Encoding routine for short strings /*{{{*/
507// ---------------------------------------------------------------------
508/* This routine performs a base64 transformation on a string. It was ripped
509 from wget and then patched and bug fixed.
510
511 This spec can be found in rfc2045 */
171c75f1 512string Base64Encode(const string &S)
6c139d6e
AL
513{
514 // Conversion table.
515 static char tbl[64] = {'A','B','C','D','E','F','G','H',
516 'I','J','K','L','M','N','O','P',
517 'Q','R','S','T','U','V','W','X',
518 'Y','Z','a','b','c','d','e','f',
519 'g','h','i','j','k','l','m','n',
520 'o','p','q','r','s','t','u','v',
521 'w','x','y','z','0','1','2','3',
522 '4','5','6','7','8','9','+','/'};
523
524 // Pre-allocate some space
525 string Final;
526 Final.reserve((4*S.length() + 2)/3 + 2);
527
528 /* Transform the 3x8 bits to 4x6 bits, as required by
529 base64. */
5933aab2 530 for (string::const_iterator I = S.begin(); I < S.end(); I += 3)
6c139d6e
AL
531 {
532 char Bits[3] = {0,0,0};
533 Bits[0] = I[0];
5933aab2 534 if (I + 1 < S.end())
6c139d6e 535 Bits[1] = I[1];
5933aab2 536 if (I + 2 < S.end())
6c139d6e
AL
537 Bits[2] = I[2];
538
539 Final += tbl[Bits[0] >> 2];
540 Final += tbl[((Bits[0] & 3) << 4) + (Bits[1] >> 4)];
541
5933aab2 542 if (I + 1 >= S.end())
6c139d6e
AL
543 break;
544
545 Final += tbl[((Bits[1] & 0xf) << 2) + (Bits[2] >> 6)];
546
5933aab2 547 if (I + 2 >= S.end())
6c139d6e
AL
548 break;
549
550 Final += tbl[Bits[2] & 0x3f];
551 }
552
553 /* Apply the padding elements, this tells how many bytes the remote
554 end should discard */
555 if (S.length() % 3 == 2)
556 Final += '=';
557 if (S.length() % 3 == 1)
558 Final += "==";
559
560 return Final;
561}
562 /*}}}*/
0da8987a 563// stringcmp - Arbitrary string compare /*{{{*/
6c139d6e 564// ---------------------------------------------------------------------
7365ff46 565/* This safely compares two non-null terminated strings of arbitrary
6c139d6e
AL
566 length */
567int stringcmp(const char *A,const char *AEnd,const char *B,const char *BEnd)
47db8997
AL
568{
569 for (; A != AEnd && B != BEnd; A++, B++)
570 if (*A != *B)
571 break;
572
573 if (A == AEnd && B == BEnd)
574 return 0;
575 if (A == AEnd)
576 return 1;
577 if (B == BEnd)
578 return -1;
579 if (*A < *B)
580 return -1;
581 return 1;
582}
ae0b19f5
AL
583
584#if __GNUC__ >= 3
47db8997
AL
585int stringcmp(string::const_iterator A,string::const_iterator AEnd,
586 const char *B,const char *BEnd)
587{
588 for (; A != AEnd && B != BEnd; A++, B++)
589 if (*A != *B)
590 break;
591
592 if (A == AEnd && B == BEnd)
593 return 0;
594 if (A == AEnd)
595 return 1;
596 if (B == BEnd)
597 return -1;
598 if (*A < *B)
599 return -1;
600 return 1;
601}
602int stringcmp(string::const_iterator A,string::const_iterator AEnd,
603 string::const_iterator B,string::const_iterator BEnd)
6c139d6e
AL
604{
605 for (; A != AEnd && B != BEnd; A++, B++)
606 if (*A != *B)
607 break;
608
609 if (A == AEnd && B == BEnd)
610 return 0;
611 if (A == AEnd)
612 return 1;
613 if (B == BEnd)
614 return -1;
615 if (*A < *B)
616 return -1;
617 return 1;
618}
ae0b19f5 619#endif
6c139d6e 620 /*}}}*/
0da8987a 621// stringcasecmp - Arbitrary case insensitive string compare /*{{{*/
6c139d6e
AL
622// ---------------------------------------------------------------------
623/* */
624int stringcasecmp(const char *A,const char *AEnd,const char *B,const char *BEnd)
47db8997
AL
625{
626 for (; A != AEnd && B != BEnd; A++, B++)
6dc60370 627 if (tolower_ascii(*A) != tolower_ascii(*B))
47db8997
AL
628 break;
629
630 if (A == AEnd && B == BEnd)
631 return 0;
632 if (A == AEnd)
633 return 1;
634 if (B == BEnd)
635 return -1;
6dc60370 636 if (tolower_ascii(*A) < tolower_ascii(*B))
47db8997
AL
637 return -1;
638 return 1;
639}
ae0b19f5 640#if __GNUC__ >= 3
47db8997
AL
641int stringcasecmp(string::const_iterator A,string::const_iterator AEnd,
642 const char *B,const char *BEnd)
643{
644 for (; A != AEnd && B != BEnd; A++, B++)
6dc60370 645 if (tolower_ascii(*A) != tolower_ascii(*B))
47db8997
AL
646 break;
647
648 if (A == AEnd && B == BEnd)
649 return 0;
650 if (A == AEnd)
651 return 1;
652 if (B == BEnd)
653 return -1;
6dc60370 654 if (tolower_ascii(*A) < tolower_ascii(*B))
47db8997
AL
655 return -1;
656 return 1;
657}
658int stringcasecmp(string::const_iterator A,string::const_iterator AEnd,
659 string::const_iterator B,string::const_iterator BEnd)
6c139d6e
AL
660{
661 for (; A != AEnd && B != BEnd; A++, B++)
6dc60370 662 if (tolower_ascii(*A) != tolower_ascii(*B))
6c139d6e 663 break;
3b5421b4 664
6c139d6e
AL
665 if (A == AEnd && B == BEnd)
666 return 0;
667 if (A == AEnd)
668 return 1;
669 if (B == BEnd)
670 return -1;
6dc60370 671 if (tolower_ascii(*A) < tolower_ascii(*B))
6c139d6e
AL
672 return -1;
673 return 1;
674}
ae0b19f5 675#endif
6c139d6e 676 /*}}}*/
3b5421b4
AL
677// LookupTag - Lookup the value of a tag in a taged string /*{{{*/
678// ---------------------------------------------------------------------
679/* The format is like those used in package files and the method
680 communication system */
171c75f1 681string LookupTag(const string &Message,const char *Tag,const char *Default)
3b5421b4
AL
682{
683 // Look for a matching tag.
684 int Length = strlen(Tag);
f7f0d6c7 685 for (string::const_iterator I = Message.begin(); I + Length < Message.end(); ++I)
3b5421b4
AL
686 {
687 // Found the tag
688 if (I[Length] == ':' && stringcasecmp(I,I+Length,Tag) == 0)
689 {
690 // Find the end of line and strip the leading/trailing spaces
171c75f1 691 string::const_iterator J;
3b5421b4 692 I += Length + 1;
74dedb4a 693 for (; isspace_ascii(*I) != 0 && I < Message.end(); ++I);
f7f0d6c7 694 for (J = I; *J != '\n' && J < Message.end(); ++J);
74dedb4a 695 for (; J > I && isspace_ascii(J[-1]) != 0; --J);
3b5421b4 696
0db4a45b 697 return string(I,J);
3b5421b4
AL
698 }
699
f7f0d6c7 700 for (; *I != '\n' && I < Message.end(); ++I);
3b5421b4
AL
701 }
702
703 // Failed to find a match
704 if (Default == 0)
705 return string();
706 return Default;
707}
708 /*}}}*/
709// StringToBool - Converts a string into a boolean /*{{{*/
710// ---------------------------------------------------------------------
711/* This inspects the string to see if it is true or if it is false and
712 then returns the result. Several varients on true/false are checked. */
171c75f1 713int StringToBool(const string &Text,int Default)
3b5421b4 714{
08be0ca3
MV
715 char *ParseEnd;
716 int Res = strtol(Text.c_str(),&ParseEnd,0);
717 // ensure that the entire string was converted by strtol to avoid
718 // failures on "apt-cache show -a 0ad" where the "0" is converted
719 const char *TextEnd = Text.c_str()+Text.size();
720 if (ParseEnd == TextEnd && Res >= 0 && Res <= 1)
3b5421b4
AL
721 return Res;
722
723 // Check for positives
724 if (strcasecmp(Text.c_str(),"no") == 0 ||
725 strcasecmp(Text.c_str(),"false") == 0 ||
726 strcasecmp(Text.c_str(),"without") == 0 ||
7f25bdff 727 strcasecmp(Text.c_str(),"off") == 0 ||
3b5421b4
AL
728 strcasecmp(Text.c_str(),"disable") == 0)
729 return 0;
730
731 // Check for negatives
732 if (strcasecmp(Text.c_str(),"yes") == 0 ||
733 strcasecmp(Text.c_str(),"true") == 0 ||
734 strcasecmp(Text.c_str(),"with") == 0 ||
7f25bdff 735 strcasecmp(Text.c_str(),"on") == 0 ||
3b5421b4
AL
736 strcasecmp(Text.c_str(),"enable") == 0)
737 return 1;
738
739 return Default;
740}
741 /*}}}*/
0a8a80e5
AL
742// TimeRFC1123 - Convert a time_t into RFC1123 format /*{{{*/
743// ---------------------------------------------------------------------
744/* This converts a time_t into a string time representation that is
745 year 2000 complient and timezone neutral */
746string TimeRFC1123(time_t Date)
747{
410327e1
DK
748 struct tm Conv;
749 if (gmtime_r(&Date, &Conv) == NULL)
750 return "";
0a8a80e5 751
410327e1 752 char Buf[300];
0a8a80e5
AL
753 const char *Day[] = {"Sun","Mon","Tue","Wed","Thu","Fri","Sat"};
754 const char *Month[] = {"Jan","Feb","Mar","Apr","May","Jun","Jul",
755 "Aug","Sep","Oct","Nov","Dec"};
756
8ff84cf3 757 snprintf(Buf, sizeof(Buf), "%s, %02i %s %i %02i:%02i:%02i GMT",Day[Conv.tm_wday],
0a8a80e5
AL
758 Conv.tm_mday,Month[Conv.tm_mon],Conv.tm_year+1900,Conv.tm_hour,
759 Conv.tm_min,Conv.tm_sec);
760 return Buf;
761}
762 /*}}}*/
763// ReadMessages - Read messages from the FD /*{{{*/
764// ---------------------------------------------------------------------
765/* This pulls full messages from the input FD into the message buffer.
766 It assumes that messages will not pause during transit so no
ffc36991
DB
767 fancy buffering is used.
768
769 In particular: this reads blocks from the input until it believes
770 that it's run out of input text. Each block is terminated by a
d8c71b3b 771 double newline ('\n' followed by '\n').
ffc36991 772 */
0a8a80e5
AL
773bool ReadMessages(int Fd, vector<string> &List)
774{
aee70518 775 char Buffer[64000];
ffc36991
DB
776 // Represents any left-over from the previous iteration of the
777 // parse loop. (i.e., if a message is split across the end
778 // of the buffer, it goes here)
779 string PartialMessage;
d8c71b3b
DK
780
781 do {
782 int const Res = read(Fd, Buffer, sizeof(Buffer));
b0db36b1
AL
783 if (Res < 0 && errno == EINTR)
784 continue;
d8c71b3b
DK
785
786 // process we read from has died
0a8a80e5
AL
787 if (Res == 0)
788 return false;
d8c71b3b 789
0a8a80e5 790 // No data
3b0e76ec 791#if EAGAIN != EWOULDBLOCK
d8c71b3b 792 if (Res < 0 && (errno == EAGAIN || errno == EWOULDBLOCK))
3b0e76ec
DK
793#else
794 if (Res < 0 && errno == EAGAIN)
795#endif
0a8a80e5 796 return true;
b2e465d6
AL
797 if (Res < 0)
798 return false;
d8c71b3b
DK
799
800 // extract the message(s) from the buffer
801 char const *Start = Buffer;
802 char const * const End = Buffer + Res;
803
804 char const * NL = (char const *) memchr(Start, '\n', End - Start);
805 if (NL == NULL)
0a8a80e5 806 {
d8c71b3b
DK
807 // end of buffer: store what we have so far and read new data in
808 PartialMessage.append(Start, End - Start);
809 Start = End;
0a8a80e5 810 }
ffc36991 811 else
d8c71b3b
DK
812 ++NL;
813
814 if (PartialMessage.empty() == false && Start < End)
815 {
816 // if we start with a new line, see if the partial message we have ended with one
817 // so that we properly detect records ending between two read() runs
818 // cases are: \n|\n , \r\n|\r\n and \r\n\r|\n
819 // the case \r|\n\r\n is handled by the usual double-newline handling
820 if ((NL - Start) == 1 || ((NL - Start) == 2 && *Start == '\r'))
821 {
822 if (APT::String::Endswith(PartialMessage, "\n") || APT::String::Endswith(PartialMessage, "\r\n\r"))
823 {
824 PartialMessage.erase(PartialMessage.find_last_not_of("\r\n") + 1);
825 List.push_back(PartialMessage);
826 PartialMessage.clear();
827 while (NL < End && (*NL == '\n' || *NL == '\r')) ++NL;
828 Start = NL;
829 }
830 }
831 }
832
833 while (Start < End) {
834 char const * NL2 = (char const *) memchr(NL, '\n', End - NL);
835 if (NL2 == NULL)
836 {
837 // end of buffer: store what we have so far and read new data in
838 PartialMessage.append(Start, End - Start);
839 break;
840 }
841 ++NL2;
842
843 // did we find a double newline?
844 if ((NL2 - NL) == 1 || ((NL2 - NL) == 2 && *NL == '\r'))
845 {
846 PartialMessage.append(Start, NL2 - Start);
847 PartialMessage.erase(PartialMessage.find_last_not_of("\r\n") + 1);
848 List.push_back(PartialMessage);
849 PartialMessage.clear();
850 while (NL2 < End && (*NL2 == '\n' || *NL2 == '\r')) ++NL2;
851 Start = NL2;
852 }
853 NL = NL2;
854 }
855
856 // we have read at least one complete message and nothing left
857 if (PartialMessage.empty() == true)
858 return true;
0a8a80e5
AL
859
860 if (WaitFd(Fd) == false)
861 return false;
d8c71b3b 862 } while (true);
0a8a80e5
AL
863}
864 /*}}}*/
24231681
AL
865// MonthConv - Converts a month string into a number /*{{{*/
866// ---------------------------------------------------------------------
867/* This was lifted from the boa webserver which lifted it from 'wn-v1.07'
6dc60370 868 Made it a bit more robust with a few tolower_ascii though. */
24231681
AL
869static int MonthConv(char *Month)
870{
6dc60370 871 switch (tolower_ascii(*Month))
24231681 872 {
6dc60370
DK
873 case 'a':
874 return tolower_ascii(Month[1]) == 'p'?3:7;
875 case 'd':
24231681 876 return 11;
6dc60370 877 case 'f':
24231681 878 return 1;
6dc60370
DK
879 case 'j':
880 if (tolower_ascii(Month[1]) == 'a')
24231681 881 return 0;
6dc60370
DK
882 return tolower_ascii(Month[2]) == 'n'?5:6;
883 case 'm':
884 return tolower_ascii(Month[2]) == 'r'?2:4;
885 case 'n':
24231681 886 return 10;
6dc60370 887 case 'o':
24231681 888 return 9;
6dc60370 889 case 's':
24231681
AL
890 return 8;
891
892 // Pretend it is January..
893 default:
894 return 0;
895 }
896}
897 /*}}}*/
55089145 898// timegm - Internal timegm if the gnu version is not available /*{{{*/
6d5dd02a 899// ---------------------------------------------------------------------
55089145 900/* Converts struct tm to time_t, assuming the data in tm is UTC rather
6d5dd02a 901 than local timezone (mktime assumes the latter).
41b6caf4 902
55089145
DK
903 This function is a nonstandard GNU extension that is also present on
904 the BSDs and maybe other systems. For others we follow the advice of
905 the manpage of timegm and use his portable replacement. */
906#ifndef HAVE_TIMEGM
6d5dd02a
AL
907static time_t timegm(struct tm *t)
908{
55089145
DK
909 char *tz = getenv("TZ");
910 setenv("TZ", "", 1);
911 tzset();
912 time_t ret = mktime(t);
913 if (tz)
914 setenv("TZ", tz, 1);
915 else
916 unsetenv("TZ");
917 tzset();
918 return ret;
6d5dd02a
AL
919}
920#endif
921 /*}}}*/
cd8cf88f
DK
922// FullDateToTime - Converts a HTTP1.1 full date strings into a time_t /*{{{*/
923// ---------------------------------------------------------------------
924/* tries to parses a full date as specified in RFC2616 Section 3.3.1
925 with one exception: All timezones (%Z) are accepted but the protocol
926 says that it MUST be GMT, but this one is equal to UTC which we will
927 encounter from time to time (e.g. in Release files) so we accept all
928 here and just assume it is GMT (or UTC) later on */
929bool RFC1123StrToTime(const char* const str,time_t &time)
930{
931 struct tm Tm;
24d7b626
DK
932 setlocale (LC_ALL,"C");
933 bool const invalid =
cd8cf88f 934 // Sun, 06 Nov 1994 08:49:37 GMT ; RFC 822, updated by RFC 1123
24d7b626 935 (strptime(str, "%a, %d %b %Y %H:%M:%S %Z", &Tm) == NULL &&
cd8cf88f
DK
936 // Sunday, 06-Nov-94 08:49:37 GMT ; RFC 850, obsoleted by RFC 1036
937 strptime(str, "%A, %d-%b-%y %H:%M:%S %Z", &Tm) == NULL &&
938 // Sun Nov 6 08:49:37 1994 ; ANSI C's asctime() format
24d7b626
DK
939 strptime(str, "%a %b %d %H:%M:%S %Y", &Tm) == NULL);
940 setlocale (LC_ALL,"");
941 if (invalid == true)
cd8cf88f
DK
942 return false;
943
944 time = timegm(&Tm);
945 return true;
946}
947 /*}}}*/
948// FTPMDTMStrToTime - Converts a ftp modification date into a time_t /*{{{*/
949// ---------------------------------------------------------------------
950/* */
951bool FTPMDTMStrToTime(const char* const str,time_t &time)
952{
953 struct tm Tm;
954 // MDTM includes no whitespaces but recommend and ignored by strptime
955 if (strptime(str, "%Y %m %d %H %M %S", &Tm) == NULL)
956 return false;
957
958 time = timegm(&Tm);
959 return true;
960}
961 /*}}}*/
24231681
AL
962// StrToTime - Converts a string into a time_t /*{{{*/
963// ---------------------------------------------------------------------
1e3f4083 964/* This handles all 3 popular time formats including RFC 1123, RFC 1036
24231681
AL
965 and the C library asctime format. It requires the GNU library function
966 'timegm' to convert a struct tm in UTC to a time_t. For some bizzar
f58a97d3
AL
967 reason the C library does not provide any such function :< This also
968 handles the weird, but unambiguous FTP time format*/
171c75f1 969bool StrToTime(const string &Val,time_t &Result)
24231681
AL
970{
971 struct tm Tm;
972 char Month[10];
404528bd 973
24231681 974 // Skip the day of the week
404528bd
DK
975 const char *I = strchr(Val.c_str(), ' ');
976
24231681 977 // Handle RFC 1123 time
f58a97d3 978 Month[0] = 0;
324cbd56 979 if (sscanf(I," %2d %3s %4d %2d:%2d:%2d GMT",&Tm.tm_mday,Month,&Tm.tm_year,
24231681
AL
980 &Tm.tm_hour,&Tm.tm_min,&Tm.tm_sec) != 6)
981 {
982 // Handle RFC 1036 time
324cbd56 983 if (sscanf(I," %2d-%3s-%3d %2d:%2d:%2d GMT",&Tm.tm_mday,Month,
24231681
AL
984 &Tm.tm_year,&Tm.tm_hour,&Tm.tm_min,&Tm.tm_sec) == 6)
985 Tm.tm_year += 1900;
986 else
987 {
988 // asctime format
324cbd56 989 if (sscanf(I," %3s %2d %2d:%2d:%2d %4d",Month,&Tm.tm_mday,
24231681 990 &Tm.tm_hour,&Tm.tm_min,&Tm.tm_sec,&Tm.tm_year) != 6)
f58a97d3
AL
991 {
992 // 'ftp' time
7ef72446 993 if (sscanf(Val.c_str(),"%4d%2d%2d%2d%2d%2d",&Tm.tm_year,&Tm.tm_mon,
f58a97d3
AL
994 &Tm.tm_mday,&Tm.tm_hour,&Tm.tm_min,&Tm.tm_sec) != 6)
995 return false;
996 Tm.tm_mon--;
997 }
24231681
AL
998 }
999 }
1000
1001 Tm.tm_isdst = 0;
f58a97d3
AL
1002 if (Month[0] != 0)
1003 Tm.tm_mon = MonthConv(Month);
70e0c168
MV
1004 else
1005 Tm.tm_mon = 0; // we don't have a month, so pick something
24231681
AL
1006 Tm.tm_year -= 1900;
1007
1008 // Convert to local time and then to GMT
1009 Result = timegm(&Tm);
1010 return true;
1011}
1012 /*}}}*/
ddc1d8d0
AL
1013// StrToNum - Convert a fixed length string to a number /*{{{*/
1014// ---------------------------------------------------------------------
1015/* This is used in decoding the crazy fixed length string headers in
1016 tar and ar files. */
1017bool StrToNum(const char *Str,unsigned long &Res,unsigned Len,unsigned Base)
1018{
1019 char S[30];
1020 if (Len >= sizeof(S))
1021 return false;
1022 memcpy(S,Str,Len);
1023 S[Len] = 0;
1024
1025 // All spaces is a zero
1026 Res = 0;
1027 unsigned I;
1028 for (I = 0; S[I] == ' '; I++);
1029 if (S[I] == 0)
1030 return true;
1031
1032 char *End;
1033 Res = strtoul(S,&End,Base);
1034 if (End == S)
1035 return false;
1036
1037 return true;
1038}
1039 /*}}}*/
650faab0
DK
1040// StrToNum - Convert a fixed length string to a number /*{{{*/
1041// ---------------------------------------------------------------------
1042/* This is used in decoding the crazy fixed length string headers in
1043 tar and ar files. */
1044bool StrToNum(const char *Str,unsigned long long &Res,unsigned Len,unsigned Base)
1045{
1046 char S[30];
1047 if (Len >= sizeof(S))
1048 return false;
1049 memcpy(S,Str,Len);
1050 S[Len] = 0;
1051
1052 // All spaces is a zero
1053 Res = 0;
1054 unsigned I;
1055 for (I = 0; S[I] == ' '; I++);
1056 if (S[I] == 0)
1057 return true;
1058
1059 char *End;
1060 Res = strtoull(S,&End,Base);
1061 if (End == S)
1062 return false;
1063
1064 return true;
1065}
1066 /*}}}*/
1067
54f2f0a3
NH
1068// Base256ToNum - Convert a fixed length binary to a number /*{{{*/
1069// ---------------------------------------------------------------------
1070/* This is used in decoding the 256bit encoded fixed length fields in
1071 tar files */
3c09d634 1072bool Base256ToNum(const char *Str,unsigned long long &Res,unsigned int Len)
54f2f0a3 1073{
54f2f0a3
NH
1074 if ((Str[0] & 0x80) == 0)
1075 return false;
1076 else
1077 {
1078 Res = Str[0] & 0x7F;
f688d1d3 1079 for(unsigned int i = 1; i < Len; ++i)
54f2f0a3
NH
1080 Res = (Res<<8) + Str[i];
1081 return true;
1082 }
1083}
1084 /*}}}*/
3c09d634
GJ
1085// Base256ToNum - Convert a fixed length binary to a number /*{{{*/
1086// ---------------------------------------------------------------------
1087/* This is used in decoding the 256bit encoded fixed length fields in
1088 tar files */
1089bool Base256ToNum(const char *Str,unsigned long &Res,unsigned int Len)
1090{
1091 unsigned long long Num;
1092 bool rc;
1093
1094 rc = Base256ToNum(Str, Num, Len);
1095 Res = Num;
1096 if (Res != Num)
1097 return false;
1098
1099 return rc;
1100}
1101 /*}}}*/
6e52073f
AL
1102// HexDigit - Convert a hex character into an integer /*{{{*/
1103// ---------------------------------------------------------------------
1104/* Helper for Hex2Num */
1105static int HexDigit(int c)
1106{
1107 if (c >= '0' && c <= '9')
1108 return c - '0';
1109 if (c >= 'a' && c <= 'f')
1110 return c - 'a' + 10;
1111 if (c >= 'A' && c <= 'F')
1112 return c - 'A' + 10;
fc8f1c22 1113 return -1;
6e52073f
AL
1114}
1115 /*}}}*/
1116// Hex2Num - Convert a long hex number into a buffer /*{{{*/
1117// ---------------------------------------------------------------------
1118/* The length of the buffer must be exactly 1/2 the length of the string. */
171c75f1 1119bool Hex2Num(const string &Str,unsigned char *Num,unsigned int Length)
eff0c22e
JAK
1120{
1121 return Hex2Num(APT::StringView(Str), Num, Length);
1122}
1123
1124bool Hex2Num(const APT::StringView Str,unsigned char *Num,unsigned int Length)
6e52073f 1125{
0db4a45b 1126 if (Str.length() != Length*2)
6e52073f
AL
1127 return false;
1128
1129 // Convert each digit. We store it in the same order as the string
1130 int J = 0;
eff0c22e 1131 for (auto I = Str.begin(); I != Str.end();J++, I += 2)
6e52073f 1132 {
fc8f1c22
NT
1133 int first_half = HexDigit(I[0]);
1134 int second_half;
1135 if (first_half < 0)
6e52073f
AL
1136 return false;
1137
fc8f1c22
NT
1138 second_half = HexDigit(I[1]);
1139 if (second_half < 0)
1140 return false;
1141 Num[J] = first_half << 4;
1142 Num[J] += second_half;
6e52073f
AL
1143 }
1144
1145 return true;
1146}
1147 /*}}}*/
b2e465d6
AL
1148// TokSplitString - Split a string up by a given token /*{{{*/
1149// ---------------------------------------------------------------------
1150/* This is intended to be a faster splitter, it does not use dynamic
1151 memories. Input is changed to insert nulls at each token location. */
1152bool TokSplitString(char Tok,char *Input,char **List,
1153 unsigned long ListMax)
1154{
1155 // Strip any leading spaces
1156 char *Start = Input;
1157 char *Stop = Start + strlen(Start);
1158 for (; *Start != 0 && isspace(*Start) != 0; Start++);
1159
1160 unsigned long Count = 0;
1161 char *Pos = Start;
1162 while (Pos != Stop)
1163 {
1164 // Skip to the next Token
1165 for (; Pos != Stop && *Pos != Tok; Pos++);
1166
1167 // Back remove spaces
1168 char *End = Pos;
1169 for (; End > Start && (End[-1] == Tok || isspace(End[-1]) != 0); End--);
1170 *End = 0;
1171
1172 List[Count++] = Start;
1173 if (Count >= ListMax)
1174 {
1175 List[Count-1] = 0;
1176 return false;
1177 }
1178
1179 // Advance pos
1180 for (; Pos != Stop && (*Pos == Tok || isspace(*Pos) != 0 || *Pos == 0); Pos++);
1181 Start = Pos;
1182 }
1183
1184 List[Count] = 0;
1185 return true;
1186}
1187 /*}}}*/
3f42500d 1188// VectorizeString - Split a string up into a vector of strings /*{{{*/
d7cf5923
DK
1189// ---------------------------------------------------------------------
1190/* This can be used to split a given string up into a vector, so the
1191 propose is the same as in the method above and this one is a bit slower
3f42500d
DK
1192 also, but the advantage is that we have an iteratable vector */
1193vector<string> VectorizeString(string const &haystack, char const &split)
d7cf5923 1194{
a5414e56
DK
1195 vector<string> exploded;
1196 if (haystack.empty() == true)
1197 return exploded;
d7cf5923
DK
1198 string::const_iterator start = haystack.begin();
1199 string::const_iterator end = start;
d7cf5923
DK
1200 do {
1201 for (; end != haystack.end() && *end != split; ++end);
1202 exploded.push_back(string(start, end));
1203 start = end + 1;
1204 } while (end != haystack.end() && (++end) != haystack.end());
1205 return exploded;
1206}
1207 /*}}}*/
9572a54b 1208// StringSplit - split a string into a string vector by token /*{{{*/
00f4d9ff 1209// ---------------------------------------------------------------------
41053d72 1210/* See header for details.
00f4d9ff 1211 */
41053d72 1212vector<string> StringSplit(std::string const &s, std::string const &sep,
85bf0019 1213 unsigned int maxsplit)
00f4d9ff
MV
1214{
1215 vector<string> split;
1216 size_t start, pos;
85bf0019 1217
9572a54b 1218 // no seperator given, this is bogus
00f4d9ff
MV
1219 if(sep.size() == 0)
1220 return split;
85bf0019
MV
1221
1222 start = pos = 0;
9572a54b
MV
1223 while (pos != string::npos)
1224 {
00f4d9ff
MV
1225 pos = s.find(sep, start);
1226 split.push_back(s.substr(start, pos-start));
85bf0019 1227
9572a54b 1228 // if maxsplit is reached, the remaining string is the last item
2ddab3fb 1229 if(split.size() >= maxsplit)
85bf0019
MV
1230 {
1231 split[split.size()-1] = s.substr(start);
1232 break;
1233 }
1234 start = pos+sep.size();
9572a54b 1235 }
00f4d9ff
MV
1236 return split;
1237}
1238 /*}}}*/
b2e465d6
AL
1239// RegexChoice - Simple regex list/list matcher /*{{{*/
1240// ---------------------------------------------------------------------
1241/* */
1242unsigned long RegexChoice(RxChoiceList *Rxs,const char **ListBegin,
1243 const char **ListEnd)
1244{
1245 for (RxChoiceList *R = Rxs; R->Str != 0; R++)
1246 R->Hit = false;
1247
1248 unsigned long Hits = 0;
ef74268b 1249 for (; ListBegin < ListEnd; ++ListBegin)
b2e465d6
AL
1250 {
1251 // Check if the name is a regex
1252 const char *I;
1253 bool Regex = true;
1254 for (I = *ListBegin; *I != 0; I++)
1255 if (*I == '.' || *I == '?' || *I == '*' || *I == '|')
1256 break;
1257 if (*I == 0)
1258 Regex = false;
1259
1260 // Compile the regex pattern
1261 regex_t Pattern;
1262 if (Regex == true)
1263 if (regcomp(&Pattern,*ListBegin,REG_EXTENDED | REG_ICASE |
1264 REG_NOSUB) != 0)
1265 Regex = false;
1266
1267 // Search the list
1268 bool Done = false;
1269 for (RxChoiceList *R = Rxs; R->Str != 0; R++)
1270 {
1271 if (R->Str[0] == 0)
1272 continue;
1273
1274 if (strcasecmp(R->Str,*ListBegin) != 0)
1275 {
1276 if (Regex == false)
1277 continue;
1278 if (regexec(&Pattern,R->Str,0,0,0) != 0)
1279 continue;
1280 }
1281 Done = true;
1282
1283 if (R->Hit == false)
1284 Hits++;
1285
1286 R->Hit = true;
1287 }
1288
1289 if (Regex == true)
1290 regfree(&Pattern);
1291
1292 if (Done == false)
1293 _error->Warning(_("Selection %s not found"),*ListBegin);
1294 }
1295
1296 return Hits;
1297}
1298 /*}}}*/
5076b3c2 1299// {str,io}printf - C format string outputter to C++ strings/iostreams /*{{{*/
b2e465d6 1300// ---------------------------------------------------------------------
1168596f
AL
1301/* This is used to make the internationalization strings easier to translate
1302 and to allow reordering of parameters */
5076b3c2
DK
1303static bool iovprintf(ostream &out, const char *format,
1304 va_list &args, ssize_t &size) {
1305 char *S = (char*)malloc(size);
1306 ssize_t const n = vsnprintf(S, size, format, args);
1307 if (n > -1 && n < size) {
1308 out << S;
1309 free(S);
1310 return true;
1311 } else {
1312 if (n > -1)
1313 size = n + 1;
1314 else
1315 size *= 2;
1316 }
1317 free(S);
1318 return false;
1319}
1320void ioprintf(ostream &out,const char *format,...)
b2e465d6
AL
1321{
1322 va_list args;
5076b3c2
DK
1323 ssize_t size = 400;
1324 while (true) {
e8afd168 1325 bool ret;
5076b3c2 1326 va_start(args,format);
ce105e87 1327 ret = iovprintf(out, format, args, size);
5076b3c2 1328 va_end(args);
ce105e87
DK
1329 if (ret == true)
1330 return;
5076b3c2 1331 }
1168596f 1332}
5076b3c2 1333void strprintf(string &out,const char *format,...)
d4cd303e
MV
1334{
1335 va_list args;
5076b3c2
DK
1336 ssize_t size = 400;
1337 std::ostringstream outstr;
1338 while (true) {
e8afd168 1339 bool ret;
5076b3c2 1340 va_start(args,format);
ce105e87 1341 ret = iovprintf(outstr, format, args, size);
5076b3c2 1342 va_end(args);
ce105e87
DK
1343 if (ret == true)
1344 break;
5076b3c2
DK
1345 }
1346 out = outstr.str();
d4cd303e
MV
1347}
1348 /*}}}*/
1168596f
AL
1349// safe_snprintf - Safer snprintf /*{{{*/
1350// ---------------------------------------------------------------------
1351/* This is a snprintf that will never (ever) go past 'End' and returns a
1352 pointer to the end of the new string. The returned string is always null
1353 terminated unless Buffer == end. This is a better alterantive to using
1354 consecutive snprintfs. */
1355char *safe_snprintf(char *Buffer,char *End,const char *Format,...)
1356{
1357 va_list args;
ea6db08d 1358 int Did;
1168596f 1359
1168596f
AL
1360 if (End <= Buffer)
1361 return End;
163dc55b 1362 va_start(args,Format);
1168596f 1363 Did = vsnprintf(Buffer,End - Buffer,Format,args);
163dc55b
MV
1364 va_end(args);
1365
1168596f
AL
1366 if (Did < 0 || Buffer + Did > End)
1367 return End;
1368 return Buffer + Did;
b2e465d6
AL
1369}
1370 /*}}}*/
cdb9307c
MV
1371// StripEpoch - Remove the version "epoch" from a version string /*{{{*/
1372// ---------------------------------------------------------------------
1373string StripEpoch(const string &VerStr)
1374{
1375 size_t i = VerStr.find(":");
1376 if (i == string::npos)
1377 return VerStr;
1378 return VerStr.substr(i+1);
1379}
69c2ecbd 1380 /*}}}*/
98eb4e9e 1381
4e86942a
MV
1382// tolower_ascii - tolower() function that ignores the locale /*{{{*/
1383// ---------------------------------------------------------------------
6dc60370 1384/* This little function is the most called method we have and tries
1e3f4083 1385 therefore to do the absolut minimum - and is notable faster than
6dc60370
DK
1386 standard tolower/toupper and as a bonus avoids problems with different
1387 locales - we only operate on ascii chars anyway. */
98eb4e9e 1388#undef tolower_ascii
390344f9 1389int tolower_ascii(int const c) APT_CONST APT_COLD;
6dc60370 1390int tolower_ascii(int const c)
4e86942a 1391{
98eb4e9e 1392 return tolower_ascii_inline(c);
4e86942a
MV
1393}
1394 /*}}}*/
1395
98b06343
JAK
1396// isspace_ascii - isspace() function that ignores the locale /*{{{*/
1397// ---------------------------------------------------------------------
1398/* This little function is one of the most called methods we have and tries
1399 therefore to do the absolut minimum - and is notable faster than
1400 standard isspace() and as a bonus avoids problems with different
1401 locales - we only operate on ascii chars anyway. */
98eb4e9e 1402#undef isspace_ascii
390344f9 1403int isspace_ascii(int const c) APT_CONST APT_COLD;
98b06343
JAK
1404int isspace_ascii(int const c)
1405{
98eb4e9e 1406 return isspace_ascii_inline(c);
98b06343
JAK
1407}
1408 /*}}}*/
1409
1e3f4083 1410// CheckDomainList - See if Host is in a , separate list /*{{{*/
f8081133 1411// ---------------------------------------------------------------------
1e3f4083 1412/* The domain list is a comma separate list of domains that are suffix
f8081133 1413 matched against the argument */
171c75f1 1414bool CheckDomainList(const string &Host,const string &List)
f8081133 1415{
47db8997 1416 string::const_iterator Start = List.begin();
f7f0d6c7 1417 for (string::const_iterator Cur = List.begin(); Cur <= List.end(); ++Cur)
f8081133 1418 {
47db8997 1419 if (Cur < List.end() && *Cur != ',')
f8081133
AL
1420 continue;
1421
1422 // Match the end of the string..
e2c7e6b5 1423 if ((Host.size() >= (unsigned)(Cur - Start)) &&
f8081133 1424 Cur - Start != 0 &&
47db8997 1425 stringcasecmp(Host.end() - (Cur - Start),Host.end(),Start,Cur) == 0)
f8081133
AL
1426 return true;
1427
1428 Start = Cur + 1;
1429 }
1430 return false;
1431}
1432 /*}}}*/
b9179170
MV
1433// strv_length - Return the length of a NULL-terminated string array /*{{{*/
1434// ---------------------------------------------------------------------
1435/* */
1436size_t strv_length(const char **str_array)
1437{
1438 size_t i;
1439 for (i=0; str_array[i] != NULL; i++)
1440 /* nothing */
1441 ;
1442 return i;
1443}
b8eba208 1444 /*}}}*/
69c2ecbd 1445// DeEscapeString - unescape (\0XX and \xXX) from a string /*{{{*/
a513ace2 1446// ---------------------------------------------------------------------
cca2efe6
MV
1447/* */
1448string DeEscapeString(const string &input)
a513ace2 1449{
b9dc4706 1450 char tmp[3];
69c2ecbd
DK
1451 string::const_iterator it;
1452 string output;
f7f0d6c7 1453 for (it = input.begin(); it != input.end(); ++it)
a513ace2
MV
1454 {
1455 // just copy non-escape chars
1456 if (*it != '\\')
1457 {
1458 output += *it;
1459 continue;
1460 }
f8081133 1461
a513ace2
MV
1462 // deal with double escape
1463 if (*it == '\\' &&
1464 (it + 1 < input.end()) && it[1] == '\\')
1465 {
1466 // copy
1467 output += *it;
1468 // advance iterator one step further
f7f0d6c7 1469 ++it;
a513ace2
MV
1470 continue;
1471 }
1472
1473 // ensure we have a char to read
1474 if (it + 1 == input.end())
1475 continue;
f8081133 1476
a513ace2 1477 // read it
f7f0d6c7 1478 ++it;
a513ace2
MV
1479 switch (*it)
1480 {
1481 case '0':
b9dc4706 1482 if (it + 2 <= input.end()) {
a513ace2
MV
1483 tmp[0] = it[1];
1484 tmp[1] = it[2];
b9dc4706 1485 tmp[2] = 0;
a513ace2
MV
1486 output += (char)strtol(tmp, 0, 8);
1487 it += 2;
1488 }
1489 break;
1490 case 'x':
1491 if (it + 2 <= input.end()) {
1492 tmp[0] = it[1];
1493 tmp[1] = it[2];
1494 tmp[2] = 0;
1495 output += (char)strtol(tmp, 0, 16);
1496 it += 2;
1497 }
1498 break;
1499 default:
1500 // FIXME: raise exception here?
a513ace2
MV
1501 break;
1502 }
1503 }
1504 return output;
1505}
1506 /*}}}*/
be4401bf 1507// URI::CopyFrom - Copy from an object /*{{{*/
93bf083d
AL
1508// ---------------------------------------------------------------------
1509/* This parses the URI into all of its components */
171c75f1 1510void URI::CopyFrom(const string &U)
93bf083d 1511{
5933aab2 1512 string::const_iterator I = U.begin();
93bf083d 1513
b2e465d6 1514 // Locate the first colon, this separates the scheme
f7f0d6c7 1515 for (; I < U.end() && *I != ':' ; ++I);
5933aab2 1516 string::const_iterator FirstColon = I;
93bf083d 1517
bfd22fc0
AL
1518 /* Determine if this is a host type URI with a leading double //
1519 and then search for the first single / */
5933aab2
AL
1520 string::const_iterator SingleSlash = I;
1521 if (I + 3 < U.end() && I[1] == '/' && I[2] == '/')
bfd22fc0 1522 SingleSlash += 3;
67ff87bf
AL
1523
1524 /* Find the / indicating the end of the hostname, ignoring /'s in the
1525 square brackets */
1526 bool InBracket = false;
f7f0d6c7 1527 for (; SingleSlash < U.end() && (*SingleSlash != '/' || InBracket == true); ++SingleSlash)
67ff87bf
AL
1528 {
1529 if (*SingleSlash == '[')
1530 InBracket = true;
1531 if (InBracket == true && *SingleSlash == ']')
1532 InBracket = false;
1533 }
1534
5933aab2
AL
1535 if (SingleSlash > U.end())
1536 SingleSlash = U.end();
93bf083d
AL
1537
1538 // We can now write the access and path specifiers
171c75f1 1539 Access.assign(U.begin(),FirstColon);
5933aab2 1540 if (SingleSlash != U.end())
171c75f1 1541 Path.assign(SingleSlash,U.end());
92e889c8
AL
1542 if (Path.empty() == true)
1543 Path = "/";
1544
93bf083d 1545 // Now we attempt to locate a user:pass@host fragment
d48c6a7d 1546 if (FirstColon + 2 <= U.end() && FirstColon[1] == '/' && FirstColon[2] == '/')
f46e7681
AL
1547 FirstColon += 3;
1548 else
1549 FirstColon += 1;
5933aab2 1550 if (FirstColon >= U.end())
93bf083d
AL
1551 return;
1552
1553 if (FirstColon > SingleSlash)
1554 FirstColon = SingleSlash;
1555
3856756b
AL
1556 // Find the colon...
1557 I = FirstColon + 1;
1d38d0e9
AL
1558 if (I > SingleSlash)
1559 I = SingleSlash;
f7f0d6c7 1560 for (; I < SingleSlash && *I != ':'; ++I);
5933aab2 1561 string::const_iterator SecondColon = I;
3856756b
AL
1562
1563 // Search for the @ after the colon
f7f0d6c7 1564 for (; I < SingleSlash && *I != '@'; ++I);
5933aab2 1565 string::const_iterator At = I;
93bf083d 1566
93bf083d
AL
1567 // Now write the host and user/pass
1568 if (At == SingleSlash)
1569 {
1570 if (FirstColon < SingleSlash)
171c75f1 1571 Host.assign(FirstColon,SingleSlash);
93bf083d
AL
1572 }
1573 else
1574 {
171c75f1 1575 Host.assign(At+1,SingleSlash);
436d7eab
DK
1576 // username and password must be encoded (RFC 3986)
1577 User.assign(DeQuoteString(FirstColon,SecondColon));
93bf083d 1578 if (SecondColon < At)
436d7eab 1579 Password.assign(DeQuoteString(SecondColon+1,At));
93bf083d
AL
1580 }
1581
67ff87bf
AL
1582 // Now we parse the RFC 2732 [] hostnames.
1583 unsigned long PortEnd = 0;
1584 InBracket = false;
1585 for (unsigned I = 0; I != Host.length();)
1586 {
1587 if (Host[I] == '[')
1588 {
1589 InBracket = true;
1590 Host.erase(I,1);
1591 continue;
1592 }
1593
1594 if (InBracket == true && Host[I] == ']')
1595 {
1596 InBracket = false;
1597 Host.erase(I,1);
1598 PortEnd = I;
1599 continue;
1600 }
1601 I++;
1602 }
1603
1604 // Tsk, weird.
1605 if (InBracket == true)
1606 {
171c75f1 1607 Host.clear();
67ff87bf
AL
1608 return;
1609 }
1610
1d38d0e9 1611 // Now we parse off a port number from the hostname
93bf083d
AL
1612 Port = 0;
1613 string::size_type Pos = Host.rfind(':');
67ff87bf 1614 if (Pos == string::npos || Pos < PortEnd)
93bf083d
AL
1615 return;
1616
1617 Port = atoi(string(Host,Pos+1).c_str());
171c75f1 1618 Host.assign(Host,0,Pos);
93bf083d
AL
1619}
1620 /*}}}*/
1621// URI::operator string - Convert the URI to a string /*{{{*/
1622// ---------------------------------------------------------------------
1623/* */
1624URI::operator string()
1625{
b8eba208
DK
1626 std::stringstream Res;
1627
54cf15cb 1628 if (Access.empty() == false)
b8eba208
DK
1629 Res << Access << ':';
1630
93bf083d 1631 if (Host.empty() == false)
b8eba208 1632 {
54cf15cb 1633 if (Access.empty() == false)
b8eba208
DK
1634 Res << "//";
1635
93bf083d
AL
1636 if (User.empty() == false)
1637 {
5b63d2a9
MV
1638 // FIXME: Technically userinfo is permitted even less
1639 // characters than these, but this is not conveniently
1640 // expressed with a blacklist.
b8eba208 1641 Res << QuoteString(User, ":/?#[]@");
93bf083d 1642 if (Password.empty() == false)
b8eba208
DK
1643 Res << ":" << QuoteString(Password, ":/?#[]@");
1644 Res << "@";
93bf083d 1645 }
b8eba208 1646
7834cb57 1647 // Add RFC 2732 escaping characters
b8eba208
DK
1648 if (Access.empty() == false && Host.find_first_of("/:") != string::npos)
1649 Res << '[' << Host << ']';
7834cb57 1650 else
b8eba208
DK
1651 Res << Host;
1652
492f957a 1653 if (Port != 0)
b8eba208 1654 Res << ':' << Port;
93bf083d 1655 }
b8eba208 1656
93bf083d 1657 if (Path.empty() == false)
492f957a
AL
1658 {
1659 if (Path[0] != '/')
b8eba208 1660 Res << "/" << Path;
492f957a 1661 else
b8eba208 1662 Res << Path;
492f957a 1663 }
b8eba208
DK
1664
1665 return Res.str();
93bf083d
AL
1666}
1667 /*}}}*/
b2e465d6 1668// URI::SiteOnly - Return the schema and site for the URI /*{{{*/
171c75f1 1669string URI::SiteOnly(const string &URI)
b2e465d6
AL
1670{
1671 ::URI U(URI);
171c75f1
MV
1672 U.User.clear();
1673 U.Password.clear();
1674 U.Path.clear();
b2e465d6
AL
1675 return U;
1676}
1677 /*}}}*/
1da3b7b8
DK
1678// URI::ArchiveOnly - Return the schema, site and cleaned path for the URI /*{{{*/
1679string URI::ArchiveOnly(const string &URI)
1680{
1681 ::URI U(URI);
1682 U.User.clear();
1683 U.Password.clear();
1684 if (U.Path.empty() == false && U.Path[U.Path.length() - 1] == '/')
1685 U.Path.erase(U.Path.length() - 1);
1686 return U;
1687}
1688 /*}}}*/
5e02df82 1689// URI::NoUserPassword - Return the schema, site and path for the URI /*{{{*/
5e02df82
MV
1690string URI::NoUserPassword(const string &URI)
1691{
1692 ::URI U(URI);
1693 U.User.clear();
1694 U.Password.clear();
5e02df82
MV
1695 return U;
1696}
1697 /*}}}*/