methods/rred.cc

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