methods/rred.cc

   1 // Includes                                                                     /*{{{*/
   2 #include <config.h>
   3
   4 #include <apt-pkg/fileutl.h>
   5 #include <apt-pkg/mmap.h>
   6 #include <apt-pkg/error.h>
   7 #include <apt-pkg/acquire-method.h>
   8 #include <apt-pkg/strutl.h>
   9 #include <apt-pkg/hashes.h>
  10 #include <apt-pkg/configuration.h>
  11
  12 #include <sys/stat.h>
  13 #include <sys/uio.h>
  14 #include <unistd.h>
  15 #include <utime.h>
  16 #include <stdio.h>
  17 #include <errno.h>
  18 #include <zlib.h>
  19 #include <apti18n.h>
  20                                                                                 /*}}}*/
  21 /** \brief RredMethod - ed-style incremential patch method                      {{{
  22  *
  23  *  This method implements a patch functionality similar to "patch --ed" that is
  24  *  used by the "tiffany" incremental packages download stuff. It differs from
  25  *  "ed" insofar that it is way more restricted (and therefore secure).
  26  *  The currently supported ed commands are "<em>c</em>hange", "<em>a</em>dd" and
  27  *  "<em>d</em>elete" (diff doesn't output any other).
  28  *  Additionally the records must be reverse sorted by line number and
  29  *  may not overlap (diff *seems* to produce this kind of output).
  30  * */
  31 class RredMethod : public pkgAcqMethod {
  32         bool Debug;
  33         // the size of this doesn't really matter (except for performance)
  34         const static int BUF_SIZE = 1024;
  35         // the supported ed commands
  36         enum Mode {MODE_CHANGED='c', MODE_DELETED='d', MODE_ADDED='a'};
  37         // return values
  38         enum State {ED_OK, ED_ORDERING, ED_PARSER, ED_FAILURE, MMAP_FAILED};
  39
  40         State applyFile(FileFd &ed_cmds, FileFd &in_file, FileFd &out_file,
  41                      unsigned long &line, char *buffer, Hashes *hash) const;
  42         void ignoreLineInFile(FileFd &fin, char *buffer) const;
  43         void copyLinesFromFileToFile(FileFd &fin, FileFd &fout, unsigned int lines,
  44                                     Hashes *hash, char *buffer) const;
  45
  46         State patchFile(FileFd &Patch, FileFd &From, FileFd &out_file, Hashes *hash) const;
  47         State patchMMap(FileFd &Patch, FileFd &From, FileFd &out_file, Hashes *hash) const;
  48
  49 protected:
  50         // the methods main method
  51         virtual bool Fetch(FetchItem *Itm);
  52
  53 public:
  54         RredMethod() : pkgAcqMethod("1.1",SingleInstance | SendConfig), Debug(false) {};
  55 };
  56                                                                                 /*}}}*/
  57 /** \brief applyFile - in reverse order with a tail recursion                   {{{
  58  *
  59  *  As it is expected that the commands are in reversed order in the patch file
  60  *  we check in the first half if the command is valid, but doesn't execute it
  61  *  and move a step deeper. After reaching the end of the file we apply the
  62  *  patches in the correct order: last found command first.
  63  *
  64  *  \param ed_cmds patch file to apply
  65  *  \param in_file base file we want to patch
  66  *  \param out_file file to write the patched result to
  67  *  \param line of command operation
  68  *  \param buffer internal used read/write buffer
  69  *  \param hash the created file for correctness
  70  *  \return the success State of the ed command executor
  71  */
  72 RredMethod::State RredMethod::applyFile(FileFd &ed_cmds, FileFd &in_file, FileFd &out_file,
  73                         unsigned long &line, char *buffer, Hashes *hash) const {
  74         // get the current command and parse it
  75         if (ed_cmds.ReadLine(buffer, BUF_SIZE) == NULL) {
  76                 if (Debug == true)
  77                         std::clog << "rred: encounter end of file - we can start patching now." << std::endl;
  78                 line = 0;
  79                 return ED_OK;
  80         }
  81
  82         // parse in the effected linenumbers
  83         char* idx;
  84         errno=0;
  85         unsigned long const startline = strtol(buffer, &idx, 10);
  86         if (errno == ERANGE || errno == EINVAL) {
  87                 _error->Errno("rred", "startline is an invalid number");
  88                 return ED_PARSER;
  89         }
  90         if (startline > line) {
  91                 _error->Error("rred: The start line (%lu) of the next command is higher than the last line (%lu). This is not allowed.", startline, line);
  92                 return ED_ORDERING;
  93         }
  94         unsigned long stopline;
  95         if (*idx == ',') {
  96                 idx++;
  97                 errno=0;
  98                 stopline = strtol(idx, &idx, 10);
  99                 if (errno == ERANGE || errno == EINVAL) {
 100                         _error->Errno("rred", "stopline is an invalid number");
 101                         return ED_PARSER;
 102                 }
 103         }
 104         else {
 105                 stopline = startline;
 106         }
 107         line = startline;
 108
 109         // which command to execute on this line(s)?
 110         switch (*idx) {
 111                 case MODE_CHANGED:
 112                         if (Debug == true)
 113                                 std::clog << "Change from line " << startline << " to " << stopline << std::endl;
 114                         break;
 115                 case MODE_ADDED:
 116                         if (Debug == true)
 117                                 std::clog << "Insert after line " << startline << std::endl;
 118                         break;
 119                 case MODE_DELETED:
 120                         if (Debug == true)
 121                                 std::clog << "Delete from line " << startline << " to " << stopline << std::endl;
 122                         break;
 123                 default:
 124                         _error->Error("rred: Unknown ed command '%c'. Abort.", *idx);
 125                         return ED_PARSER;
 126         }
 127         unsigned char mode = *idx;
 128
 129         // save the current position
 130         unsigned const long long pos = ed_cmds.Tell();
 131
 132         // if this is add or change then go to the next full stop
 133         unsigned int data_length = 0;
 134         if (mode == MODE_CHANGED || mode == MODE_ADDED) {
 135                 do {
 136                         ignoreLineInFile(ed_cmds, buffer);
 137                         data_length++;
 138                 }
 139                 while (strncmp(buffer, ".", 1) != 0);
 140                 data_length--; // the dot should not be copied
 141         }
 142
 143         // do the recursive call - the last command is the one we need to execute at first
 144         const State child = applyFile(ed_cmds, in_file, out_file, line, buffer, hash);
 145         if (child != ED_OK) {
 146                 return child;
 147         }
 148
 149         // change and delete are working on "line" - add is done after "line"
 150         if (mode != MODE_ADDED)
 151                 line++;
 152
 153         // first wind to the current position and copy over all unchanged lines
 154         if (line < startline) {
 155                 copyLinesFromFileToFile(in_file, out_file, (startline - line), hash, buffer);
 156                 line = startline;
 157         }
 158
 159         if (mode != MODE_ADDED)
 160                 line--;
 161
 162         // include data from ed script
 163         if (mode == MODE_CHANGED || mode == MODE_ADDED) {
 164                 ed_cmds.Seek(pos);
 165                 copyLinesFromFileToFile(ed_cmds, out_file, data_length, hash, buffer);
 166         }
 167
 168         // ignore the corresponding number of lines from input
 169         if (mode == MODE_CHANGED || mode == MODE_DELETED) {
 170                 while (line < stopline) {
 171                         ignoreLineInFile(in_file, buffer);
 172                         line++;
 173                 }
 174         }
 175         return ED_OK;
 176 }
 177                                                                                 /*}}}*/
 178 void RredMethod::copyLinesFromFileToFile(FileFd &fin, FileFd &fout, unsigned int lines,/*{{{*/
 179                                         Hashes *hash, char *buffer) const {
 180         while (0 < lines--) {
 181                 do {
 182                         fin.ReadLine(buffer, BUF_SIZE);
 183                         unsigned long long const towrite = strlen(buffer);
 184                         fout.Write(buffer, towrite);
 185                         hash->Add((unsigned char*)buffer, towrite);
 186                 } while (strlen(buffer) == (BUF_SIZE - 1) &&
 187                        buffer[BUF_SIZE - 2] != '\n');
 188         }
 189 }
 190                                                                                 /*}}}*/
 191 void RredMethod::ignoreLineInFile(FileFd &fin, char *buffer) const {            /*{{{*/
 192         fin.ReadLine(buffer, BUF_SIZE);
 193         while (strlen(buffer) == (BUF_SIZE - 1) &&
 194                buffer[BUF_SIZE - 2] != '\n') {
 195                 fin.ReadLine(buffer, BUF_SIZE);
 196                 buffer[0] = ' ';
 197         }
 198 }
 199                                                                                 /*}}}*/
 200 RredMethod::State RredMethod::patchFile(FileFd &Patch, FileFd &From,            /*{{{*/
 201                                         FileFd &out_file, Hashes *hash) const {
 202    char buffer[BUF_SIZE];
 203
 204    /* we do a tail recursion to read the commands in the right order */
 205    unsigned long line = -1; // assign highest possible value
 206    State const result = applyFile(Patch, From, out_file, line, buffer, hash);
 207
 208    /* read the rest from infile */
 209    if (result == ED_OK) {
 210       while (From.ReadLine(buffer, BUF_SIZE) != NULL) {
 211          unsigned long long const towrite = strlen(buffer);
 212          out_file.Write(buffer, towrite);
 213          hash->Add((unsigned char*)buffer, towrite);
 214       }
 215    }
 216    return result;
 217 }
 218                                                                                 /*}}}*/
 219 /* struct EdCommand                                                             {{{*/
 220 #ifdef _POSIX_MAPPED_FILES
 221 struct EdCommand {
 222   size_t data_start;
 223   size_t data_end;
 224   size_t data_lines;
 225   size_t first_line;
 226   size_t last_line;
 227   char type;
 228 };
 229 #define IOV_COUNT 1024 /* Don't really want IOV_MAX since it can be arbitrarily large */
 230 #endif
 231                                                                                 /*}}}*/
 232 RredMethod::State RredMethod::patchMMap(FileFd &Patch, FileFd &From,            /*{{{*/
 233                                         FileFd &out_file, Hashes *hash) const {
 234 #ifdef _POSIX_MAPPED_FILES
 235         MMap ed_cmds(Patch, MMap::ReadOnly);
 236         MMap in_file(From, MMap::ReadOnly);
 237
 238         if (ed_cmds.Size() == 0 || in_file.Size() == 0)
 239                 return MMAP_FAILED;
 240
 241         EdCommand* commands = 0;
 242         size_t command_count = 0;
 243         size_t command_alloc = 0;
 244
 245         const char* begin = (char*) ed_cmds.Data();
 246         const char* end = begin;
 247         const char* ed_end = (char*) ed_cmds.Data() + ed_cmds.Size();
 248
 249         const char* input = (char*) in_file.Data();
 250         const char* input_end = (char*) in_file.Data() + in_file.Size();
 251
 252         size_t i;
 253
 254         /* 1. Parse entire script.  It is executed in reverse order, so we cather it
 255          *    in the `commands' buffer first
 256          */
 257
 258         for(;;) {
 259                 EdCommand cmd;
 260                 cmd.data_start = 0;
 261                 cmd.data_end = 0;
 262
 263                 while(begin != ed_end && *begin == '\n')
 264                         ++begin;
 265                 while(end != ed_end && *end != '\n')
 266                         ++end;
 267                 if(end == ed_end && begin == end)
 268                         break;
 269
 270                 /* Determine command range */
 271                 const char* tmp = begin;
 272
 273                 for(;;) {
 274                         /* atoll is safe despite lacking NUL-termination; we know there's an
 275                          * alphabetic character at end[-1]
 276                          */
 277                         if(tmp == end) {
 278                                 cmd.first_line = atol(begin);
 279                                 cmd.last_line = cmd.first_line;
 280                                 break;
 281                         }
 282                         if(*tmp == ',') {
 283                                 cmd.first_line = atol(begin);
 284                                 cmd.last_line = atol(tmp + 1);
 285                                 break;
 286                         }
 287                         ++tmp;
 288                 }
 289
 290                 // which command to execute on this line(s)?
 291                 switch (end[-1]) {
 292                         case MODE_CHANGED:
 293                                 if (Debug == true)
 294                                         std::clog << "Change from line " << cmd.first_line << " to " << cmd.last_line << std::endl;
 295                                 break;
 296                         case MODE_ADDED:
 297                                 if (Debug == true)
 298                                         std::clog << "Insert after line " << cmd.first_line << std::endl;
 299                                 break;
 300                         case MODE_DELETED:
 301                                 if (Debug == true)
 302                                         std::clog << "Delete from line " << cmd.first_line << " to " << cmd.last_line << std::endl;
 303                                 break;
 304                         default:
 305                                 _error->Error("rred: Unknown ed command '%c'. Abort.", end[-1]);
 306                                 free(commands);
 307                                 return ED_PARSER;
 308                 }
 309                 cmd.type = end[-1];
 310
 311                 /* Determine the size of the inserted text, so we don't have to scan this
 312                  * text again later.
 313                  */
 314                 begin = end + 1;
 315                 end = begin;
 316                 cmd.data_lines = 0;
 317
 318                 if(cmd.type == MODE_ADDED || cmd.type == MODE_CHANGED) {
 319                         cmd.data_start = begin - (char*) ed_cmds.Data();
 320                         while(end != ed_end) {
 321                                 if(*end == '\n') {
 322                                         if(end[-1] == '.' && end[-2] == '\n')
 323                                                 break;
 324                                         ++cmd.data_lines;
 325                                 }
 326                                 ++end;
 327                         }
 328                         cmd.data_end = end - (char*) ed_cmds.Data() - 1;
 329                         begin = end + 1;
 330                         end = begin;
 331                 }
 332                 if(command_count == command_alloc) {
 333                         command_alloc = (command_alloc + 64) * 3 / 2;
 334                         commands = (EdCommand*) realloc(commands, command_alloc * sizeof(EdCommand));
 335                 }
 336                 commands[command_count++] = cmd;
 337         }
 338
 339         struct iovec* iov = new struct iovec[IOV_COUNT];
 340         size_t iov_size = 0;
 341
 342         size_t amount, remaining;
 343         size_t line = 1;
 344         EdCommand* cmd;
 345
 346         /* 2. Execute script.  We gather writes in a `struct iov' array, and flush
 347          *    using writev to minimize the number of system calls.  Data is read
 348          *    directly from the memory mappings of the input file and the script.
 349          */
 350
 351         for(i = command_count; i-- > 0; ) {
 352                 cmd = &commands[i];
 353                 if(cmd->type == MODE_ADDED)
 354                         amount = cmd->first_line + 1;
 355                 else
 356                         amount = cmd->first_line;
 357
 358                 if(line < amount) {
 359                         begin = input;
 360                         while(line != amount) {
 361                                 input = (const char*) memchr(input, '\n', input_end - input);
 362                                 if(!input)
 363                                         break;
 364                                 ++line;
 365                                 ++input;
 366                         }
 367
 368                         iov[iov_size].iov_base = (void*) begin;
 369                         iov[iov_size].iov_len = input - begin;
 370                         hash->Add((const unsigned char*) begin, input - begin);
 371
 372                         if(++iov_size == IOV_COUNT) {
 373                                 writev(out_file.Fd(), iov, IOV_COUNT);
 374                                 iov_size = 0;
 375                         }
 376                 }
 377
 378                 if(cmd->type == MODE_DELETED || cmd->type == MODE_CHANGED) {
 379                         remaining = (cmd->last_line - cmd->first_line) + 1;
 380                         line += remaining;
 381                         while(remaining) {
 382                                 input = (const char*) memchr(input, '\n', input_end - input);
 383                                 if(!input)
 384                                         break;
 385                                 --remaining;
 386                                 ++input;
 387                         }
 388                 }
 389
 390                 if(cmd->type == MODE_CHANGED || cmd->type == MODE_ADDED) {
 391                         if(cmd->data_end != cmd->data_start) {
 392                                 iov[iov_size].iov_base = (void*) ((char*)ed_cmds.Data() + cmd->data_start);
 393                                 iov[iov_size].iov_len = cmd->data_end - cmd->data_start;
 394                                 hash->Add((const unsigned char*) ((char*)ed_cmds.Data() + cmd->data_start),
 395                                 iov[iov_size].iov_len);
 396
 397                                 if(++iov_size == IOV_COUNT) {
 398                                         writev(out_file.Fd(), iov, IOV_COUNT);
 399                                         iov_size = 0;
 400                                 }
 401                         }
 402                 }
 403         }
 404
 405         if(input != input_end) {
 406                 iov[iov_size].iov_base = (void*) input;
 407                 iov[iov_size].iov_len = input_end - input;
 408                 hash->Add((const unsigned char*) input, input_end - input);
 409                 ++iov_size;
 410         }
 411
 412         if(iov_size) {
 413                 writev(out_file.Fd(), iov, iov_size);
 414                 iov_size = 0;
 415         }
 416
 417         for(i = 0; i < iov_size; i += IOV_COUNT) {
 418                 if(iov_size - i < IOV_COUNT)
 419                         writev(out_file.Fd(), iov + i, iov_size - i);
 420                 else
 421                         writev(out_file.Fd(), iov + i, IOV_COUNT);
 422         }
 423
 424         delete [] iov;
 425         free(commands);
 426
 427         return ED_OK;
 428 #else
 429         return MMAP_FAILED;
 430 #endif
 431 }
 432                                                                                 /*}}}*/
 433 bool RredMethod::Fetch(FetchItem *Itm)                                          /*{{{*/
 434 {
 435    Debug = _config->FindB("Debug::pkgAcquire::RRed", false);
 436    URI Get = Itm->Uri;
 437    std::string Path = Get.Host + Get.Path; // To account for relative paths
 438
 439    FetchResult Res;
 440    Res.Filename = Itm->DestFile;
 441    if (Itm->Uri.empty() == true) {
 442       Path = Itm->DestFile;
 443       Itm->DestFile.append(".result");
 444    } else
 445       URIStart(Res);
 446
 447    if (Debug == true)
 448       std::clog << "Patching " << Path << " with " << Path
 449          << ".ed and putting result into " << Itm->DestFile << std::endl;
 450    // Open the source and destination files (the d'tor of FileFd will do
 451    // the cleanup/closing of the fds)
 452    FileFd From(Path,FileFd::ReadOnly);
 453    FileFd Patch(Path+".ed",FileFd::ReadOnly, FileFd::Gzip);
 454    FileFd To(Itm->DestFile,FileFd::WriteAtomic);
 455    To.EraseOnFailure();
 456    if (_error->PendingError() == true)
 457       return false;
 458
 459    Hashes Hash;
 460    // now do the actual patching
 461    State const result = patchMMap(Patch, From, To, &Hash);
 462    if (result == MMAP_FAILED) {
 463       // retry with patchFile
 464       Patch.Seek(0);
 465       From.Seek(0);
 466       To.Open(Itm->DestFile,FileFd::WriteAtomic);
 467       if (_error->PendingError() == true)
 468          return false;
 469       if (patchFile(Patch, From, To, &Hash) != ED_OK) {
 470          return _error->WarningE("rred", _("Could not patch %s with mmap and with file operation usage - the patch seems to be corrupt."), Path.c_str());
 471       } else if (Debug == true) {
 472          std::clog << "rred: finished file patching of " << Path  << " after mmap failed." << std::endl;
 473       }
 474    } else if (result != ED_OK) {
 475       return _error->Errno("rred", _("Could not patch %s with mmap (but no mmap specific fail) - the patch seems to be corrupt."), Path.c_str());
 476    } else if (Debug == true) {
 477       std::clog << "rred: finished mmap patching of " << Path << std::endl;
 478    }
 479
 480    // write out the result
 481    From.Close();
 482    Patch.Close();
 483    To.Close();
 484
 485    /* Transfer the modification times from the patch file
 486       to be able to see in which state the file should be
 487       and use the access time from the "old" file */
 488    struct stat BufBase, BufPatch;
 489    if (stat(Path.c_str(),&BufBase) != 0 ||
 490        stat(std::string(Path+".ed").c_str(),&BufPatch) != 0)
 491       return _error->Errno("stat",_("Failed to stat"));
 492
 493    struct utimbuf TimeBuf;
 494    TimeBuf.actime = BufBase.st_atime;
 495    TimeBuf.modtime = BufPatch.st_mtime;
 496    if (utime(Itm->DestFile.c_str(),&TimeBuf) != 0)
 497       return _error->Errno("utime",_("Failed to set modification time"));
 498
 499    if (stat(Itm->DestFile.c_str(),&BufBase) != 0)
 500       return _error->Errno("stat",_("Failed to stat"));
 501
 502    // return done
 503    Res.LastModified = BufBase.st_mtime;
 504    Res.Size = BufBase.st_size;
 505    Res.TakeHashes(Hash);
 506    URIDone(Res);
 507
 508    return true;
 509 }
 510                                                                                 /*}}}*/
 511 /** \brief Wrapper class for testing rred */                                    /*{{{*/
 512 class TestRredMethod : public RredMethod {
 513 public:
 514         /** \brief Run rred in debug test mode
 515          *
 516          *  This method can be used to run the rred method outside
 517          *  of the "normal" acquire environment for easier testing.
 518          *
 519          *  \param base basename of all files involved in this rred test
 520          */
 521         bool Run(char const *base) {
 522                 _config->CndSet("Debug::pkgAcquire::RRed", "true");
 523                 FetchItem *test = new FetchItem;
 524                 test->DestFile = base;
 525                 return Fetch(test);
 526         }
 527 };
 528                                                                                 /*}}}*/
 529 /** \brief Starter for the rred method (or its test method)                     {{{
 530  *
 531  *  Used without parameters is the normal behavior for methods for
 532  *  the APT acquire system. While this works great for the acquire system
 533  *  it is very hard to test the method and therefore the method also
 534  *  accepts one parameter which will switch it directly to debug test mode:
 535  *  The test mode expects that if "Testfile" is given as parameter
 536  *  the file "Testfile" should be ed-style patched with "Testfile.ed"
 537  *  and will write the result to "Testfile.result".
 538  */
 539 int main(int argc, char *argv[]) {
 540         if (argc <= 1) {
 541                 RredMethod Mth;
 542                 return Mth.Run();
 543         } else {
 544                 TestRredMethod Mth;
 545                 bool result = Mth.Run(argv[1]);
 546                 _error->DumpErrors();
 547                 return result;
 548         }
 549 }
 550                                                                                 /*}}}*/