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         unsigned long long const ed_size = ed_cmds.Size();
 239         unsigned long long const in_size = in_file.Size();
 240         if (ed_size == 0 || in_size == 0)
 241                 return MMAP_FAILED;
 242
 243         EdCommand* commands = 0;
 244         size_t command_count = 0;
 245         size_t command_alloc = 0;
 246
 247         const char* begin = (char*) ed_cmds.Data();
 248         const char* end = begin;
 249         const char* ed_end = (char*) ed_cmds.Data() + ed_size;
 250
 251         const char* input = (char*) in_file.Data();
 252         const char* input_end = (char*) in_file.Data() + in_size;
 253
 254         size_t i;
 255
 256         /* 1. Parse entire script.  It is executed in reverse order, so we cather it
 257          *    in the `commands' buffer first
 258          */
 259
 260         for(;;) {
 261                 EdCommand cmd;
 262                 cmd.data_start = 0;
 263                 cmd.data_end = 0;
 264
 265                 while(begin != ed_end && *begin == '\n')
 266                         ++begin;
 267                 while(end != ed_end && *end != '\n')
 268                         ++end;
 269                 if(end == ed_end && begin == end)
 270                         break;
 271
 272                 /* Determine command range */
 273                 const char* tmp = begin;
 274
 275                 for(;;) {
 276                         /* atoll is safe despite lacking NUL-termination; we know there's an
 277                          * alphabetic character at end[-1]
 278                          */
 279                         if(tmp == end) {
 280                                 cmd.first_line = atol(begin);
 281                                 cmd.last_line = cmd.first_line;
 282                                 break;
 283                         }
 284                         if(*tmp == ',') {
 285                                 cmd.first_line = atol(begin);
 286                                 cmd.last_line = atol(tmp + 1);
 287                                 break;
 288                         }
 289                         ++tmp;
 290                 }
 291
 292                 // which command to execute on this line(s)?
 293                 switch (end[-1]) {
 294                         case MODE_CHANGED:
 295                                 if (Debug == true)
 296                                         std::clog << "Change from line " << cmd.first_line << " to " << cmd.last_line << std::endl;
 297                                 break;
 298                         case MODE_ADDED:
 299                                 if (Debug == true)
 300                                         std::clog << "Insert after line " << cmd.first_line << std::endl;
 301                                 break;
 302                         case MODE_DELETED:
 303                                 if (Debug == true)
 304                                         std::clog << "Delete from line " << cmd.first_line << " to " << cmd.last_line << std::endl;
 305                                 break;
 306                         default:
 307                                 _error->Error("rred: Unknown ed command '%c'. Abort.", end[-1]);
 308                                 free(commands);
 309                                 return ED_PARSER;
 310                 }
 311                 cmd.type = end[-1];
 312
 313                 /* Determine the size of the inserted text, so we don't have to scan this
 314                  * text again later.
 315                  */
 316                 begin = end + 1;
 317                 end = begin;
 318                 cmd.data_lines = 0;
 319
 320                 if(cmd.type == MODE_ADDED || cmd.type == MODE_CHANGED) {
 321                         cmd.data_start = begin - (char*) ed_cmds.Data();
 322                         while(end != ed_end) {
 323                                 if(*end == '\n') {
 324                                         if(end[-1] == '.' && end[-2] == '\n')
 325                                                 break;
 326                                         ++cmd.data_lines;
 327                                 }
 328                                 ++end;
 329                         }
 330                         cmd.data_end = end - (char*) ed_cmds.Data() - 1;
 331                         begin = end + 1;
 332                         end = begin;
 333                 }
 334                 if(command_count == command_alloc) {
 335                         command_alloc = (command_alloc + 64) * 3 / 2;
 336                         commands = (EdCommand*) realloc(commands, command_alloc * sizeof(EdCommand));
 337                 }
 338                 commands[command_count++] = cmd;
 339         }
 340
 341         struct iovec* iov = new struct iovec[IOV_COUNT];
 342         size_t iov_size = 0;
 343
 344         size_t amount, remaining;
 345         size_t line = 1;
 346         EdCommand* cmd;
 347
 348         /* 2. Execute script.  We gather writes in a `struct iov' array, and flush
 349          *    using writev to minimize the number of system calls.  Data is read
 350          *    directly from the memory mappings of the input file and the script.
 351          */
 352
 353         for(i = command_count; i-- > 0; ) {
 354                 cmd = &commands[i];
 355                 if(cmd->type == MODE_ADDED)
 356                         amount = cmd->first_line + 1;
 357                 else
 358                         amount = cmd->first_line;
 359
 360                 if(line < amount) {
 361                         begin = input;
 362                         while(line != amount) {
 363                                 input = (const char*) memchr(input, '\n', input_end - input);
 364                                 if(!input)
 365                                         break;
 366                                 ++line;
 367                                 ++input;
 368                         }
 369
 370                         iov[iov_size].iov_base = (void*) begin;
 371                         iov[iov_size].iov_len = input - begin;
 372                         hash->Add((const unsigned char*) begin, input - begin);
 373
 374                         if(++iov_size == IOV_COUNT) {
 375                                 writev(out_file.Fd(), iov, IOV_COUNT);
 376                                 iov_size = 0;
 377                         }
 378                 }
 379
 380                 if(cmd->type == MODE_DELETED || cmd->type == MODE_CHANGED) {
 381                         remaining = (cmd->last_line - cmd->first_line) + 1;
 382                         line += remaining;
 383                         while(remaining) {
 384                                 input = (const char*) memchr(input, '\n', input_end - input);
 385                                 if(!input)
 386                                         break;
 387                                 --remaining;
 388                                 ++input;
 389                         }
 390                 }
 391
 392                 if(cmd->type == MODE_CHANGED || cmd->type == MODE_ADDED) {
 393                         if(cmd->data_end != cmd->data_start) {
 394                                 iov[iov_size].iov_base = (void*) ((char*)ed_cmds.Data() + cmd->data_start);
 395                                 iov[iov_size].iov_len = cmd->data_end - cmd->data_start;
 396                                 hash->Add((const unsigned char*) ((char*)ed_cmds.Data() + cmd->data_start),
 397                                 iov[iov_size].iov_len);
 398
 399                                 if(++iov_size == IOV_COUNT) {
 400                                         writev(out_file.Fd(), iov, IOV_COUNT);
 401                                         iov_size = 0;
 402                                 }
 403                         }
 404                 }
 405         }
 406
 407         if(input != input_end) {
 408                 iov[iov_size].iov_base = (void*) input;
 409                 iov[iov_size].iov_len = input_end - input;
 410                 hash->Add((const unsigned char*) input, input_end - input);
 411                 ++iov_size;
 412         }
 413
 414         if(iov_size) {
 415                 writev(out_file.Fd(), iov, iov_size);
 416                 iov_size = 0;
 417         }
 418
 419         for(i = 0; i < iov_size; i += IOV_COUNT) {
 420                 if(iov_size - i < IOV_COUNT)
 421                         writev(out_file.Fd(), iov + i, iov_size - i);
 422                 else
 423                         writev(out_file.Fd(), iov + i, IOV_COUNT);
 424         }
 425
 426         delete [] iov;
 427         free(commands);
 428
 429         return ED_OK;
 430 #else
 431         return MMAP_FAILED;
 432 #endif
 433 }
 434                                                                                 /*}}}*/
 435 bool RredMethod::Fetch(FetchItem *Itm)                                          /*{{{*/
 436 {
 437    Debug = _config->FindB("Debug::pkgAcquire::RRed", false);
 438    URI Get = Itm->Uri;
 439    std::string Path = Get.Host + Get.Path; // To account for relative paths
 440
 441    FetchResult Res;
 442    Res.Filename = Itm->DestFile;
 443    if (Itm->Uri.empty() == true) {
 444       Path = Itm->DestFile;
 445       Itm->DestFile.append(".result");
 446    } else
 447       URIStart(Res);
 448
 449    if (Debug == true)
 450       std::clog << "Patching " << Path << " with " << Path
 451          << ".ed and putting result into " << Itm->DestFile << std::endl;
 452    // Open the source and destination files (the d'tor of FileFd will do
 453    // the cleanup/closing of the fds)
 454    FileFd From(Path,FileFd::ReadOnly);
 455    FileFd Patch(Path+".ed",FileFd::ReadOnly, FileFd::Gzip);
 456    FileFd To(Itm->DestFile,FileFd::WriteAtomic);
 457    To.EraseOnFailure();
 458    if (_error->PendingError() == true)
 459       return false;
 460
 461    Hashes Hash;
 462    // now do the actual patching
 463    State const result = patchMMap(Patch, From, To, &Hash);
 464    if (result == MMAP_FAILED) {
 465       // retry with patchFile
 466       Patch.Seek(0);
 467       From.Seek(0);
 468       To.Open(Itm->DestFile,FileFd::WriteAtomic);
 469       if (_error->PendingError() == true)
 470          return false;
 471       if (patchFile(Patch, From, To, &Hash) != ED_OK) {
 472          return _error->WarningE("rred", _("Could not patch %s with mmap and with file operation usage - the patch seems to be corrupt."), Path.c_str());
 473       } else if (Debug == true) {
 474          std::clog << "rred: finished file patching of " << Path  << " after mmap failed." << std::endl;
 475       }
 476    } else if (result != ED_OK) {
 477       return _error->Errno("rred", _("Could not patch %s with mmap (but no mmap specific fail) - the patch seems to be corrupt."), Path.c_str());
 478    } else if (Debug == true) {
 479       std::clog << "rred: finished mmap patching of " << Path << std::endl;
 480    }
 481
 482    // write out the result
 483    From.Close();
 484    Patch.Close();
 485    To.Close();
 486
 487    /* Transfer the modification times from the patch file
 488       to be able to see in which state the file should be
 489       and use the access time from the "old" file */
 490    struct stat BufBase, BufPatch;
 491    if (stat(Path.c_str(),&BufBase) != 0 ||
 492        stat(std::string(Path+".ed").c_str(),&BufPatch) != 0)
 493       return _error->Errno("stat",_("Failed to stat"));
 494
 495    struct utimbuf TimeBuf;
 496    TimeBuf.actime = BufBase.st_atime;
 497    TimeBuf.modtime = BufPatch.st_mtime;
 498    if (utime(Itm->DestFile.c_str(),&TimeBuf) != 0)
 499       return _error->Errno("utime",_("Failed to set modification time"));
 500
 501    if (stat(Itm->DestFile.c_str(),&BufBase) != 0)
 502       return _error->Errno("stat",_("Failed to stat"));
 503
 504    // return done
 505    Res.LastModified = BufBase.st_mtime;
 506    Res.Size = BufBase.st_size;
 507    Res.TakeHashes(Hash);
 508    URIDone(Res);
 509
 510    return true;
 511 }
 512                                                                                 /*}}}*/
 513 /** \brief Wrapper class for testing rred */                                    /*{{{*/
 514 class TestRredMethod : public RredMethod {
 515 public:
 516         /** \brief Run rred in debug test mode
 517          *
 518          *  This method can be used to run the rred method outside
 519          *  of the "normal" acquire environment for easier testing.
 520          *
 521          *  \param base basename of all files involved in this rred test
 522          */
 523         bool Run(char const *base) {
 524                 _config->CndSet("Debug::pkgAcquire::RRed", "true");
 525                 FetchItem *test = new FetchItem;
 526                 test->DestFile = base;
 527                 return Fetch(test);
 528         }
 529 };
 530                                                                                 /*}}}*/
 531 /** \brief Starter for the rred method (or its test method)                     {{{
 532  *
 533  *  Used without parameters is the normal behavior for methods for
 534  *  the APT acquire system. While this works great for the acquire system
 535  *  it is very hard to test the method and therefore the method also
 536  *  accepts one parameter which will switch it directly to debug test mode:
 537  *  The test mode expects that if "Testfile" is given as parameter
 538  *  the file "Testfile" should be ed-style patched with "Testfile.ed"
 539  *  and will write the result to "Testfile.result".
 540  */
 541 int main(int argc, char *argv[]) {
 542         if (argc <= 1) {
 543                 RredMethod Mth;
 544                 return Mth.Run();
 545         } else {
 546                 TestRredMethod Mth;
 547                 bool result = Mth.Run(argv[1]);
 548                 _error->DumpErrors();
 549                 return result;
 550         }
 551 }
 552                                                                                 /*}}}*/