Extract a Tar - Tar Extractor
Some performance measurements showed that zlib performed quite poorly
- in comparision to a forked gzip process. This tar extractor makes use
+ in comparison to a forked gzip process. This tar extractor makes use
of the fact that dup'd file descriptors have the same seek pointer
and that gzip will not read past the end of a compressed stream,
even if there is more data. We use the dup property to track extraction
##################################################################### */
/*}}}*/
// Include Files /*{{{*/
-#include <apt-pkg/extracttar.h>
+#include<config.h>
+#include <apt-pkg/dirstream.h>
+#include <apt-pkg/extracttar.h>
#include <apt-pkg/error.h>
#include <apt-pkg/strutl.h>
#include <apt-pkg/configuration.h>
-#include <apt-pkg/macros.h>
+#include <apt-pkg/fileutl.h>
-#include <stdlib.h>
+#include <string.h>
+#include <algorithm>
+#include <string>
#include <unistd.h>
#include <signal.h>
#include <fcntl.h>
#include <iostream>
+
#include <apti18n.h>
/*}}}*/
// ExtractTar::ExtractTar - Constructor /*{{{*/
// ---------------------------------------------------------------------
/* */
-ExtractTar::ExtractTar(FileFd &Fd,unsigned long Max,string DecompressionProgram) : File(Fd),
- MaxInSize(Max), DecompressProg(DecompressionProgram)
-
+ExtractTar::ExtractTar(FileFd &Fd,unsigned long long Max,string DecompressionProgram)
+ : File(Fd), MaxInSize(Max), DecompressProg(DecompressionProgram)
{
GZPid = -1;
- InFd = -1;
Eof = false;
}
/*}}}*/
ExtractTar::~ExtractTar()
{
// Error close
- Done(true);
+ Done();
}
/*}}}*/
// ExtractTar::Done - Reap the gzip sub process /*{{{*/
-// ---------------------------------------------------------------------
-/* If the force flag is given then error messages are suppressed - this
- means we hit the end of the tar file but there was still gzip data. */
-bool ExtractTar::Done(bool Force)
+bool ExtractTar::Done(bool)
{
- InFd.Close();
- if (GZPid <= 0)
- return true;
-
- /* If there is a pending error then we are cleaning up gzip and are
- not interested in it's failures */
- if (_error->PendingError() == true)
- Force = true;
-
- // Make sure we clean it up!
- kill(GZPid,SIGINT);
- string confvar = string("dir::bin::") + DecompressProg;
- if (ExecWait(GZPid,_config->Find(confvar.c_str(),DecompressProg.c_str()).c_str(),
- Force) == false)
- {
- GZPid = -1;
- return Force;
- }
-
- GZPid = -1;
- return true;
+ return Done();
+}
+bool ExtractTar::Done()
+{
+ return InFd.Close();
}
/*}}}*/
// ExtractTar::StartGzip - Startup gzip /*{{{*/
gzip will efficiently ignore the extra bits. */
bool ExtractTar::StartGzip()
{
- int Pipes[2];
- if (pipe(Pipes) != 0)
- return _error->Errno("pipe",_("Failed to create pipes"));
-
- // Fork off the process
- GZPid = ExecFork();
-
- // Spawn the subprocess
- if (GZPid == 0)
+ if (DecompressProg.empty())
{
- // Setup the FDs
- dup2(Pipes[1],STDOUT_FILENO);
- dup2(File.Fd(),STDIN_FILENO);
- int Fd = open("/dev/null",O_RDWR);
- if (Fd == -1)
- _exit(101);
- dup2(Fd,STDERR_FILENO);
- close(Fd);
- SetCloseExec(STDOUT_FILENO,false);
- SetCloseExec(STDIN_FILENO,false);
- SetCloseExec(STDERR_FILENO,false);
-
- const char *Args[3];
- string confvar = string("dir::bin::") + DecompressProg;
- string argv0 = _config->Find(confvar.c_str(),DecompressProg.c_str());
- Args[0] = argv0.c_str();
- Args[1] = "-d";
- Args[2] = 0;
- execvp(Args[0],(char **)Args);
- cerr << _("Failed to exec gzip ") << Args[0] << endl;
- _exit(100);
+ InFd.OpenDescriptor(File.Fd(), FileFd::ReadOnly, FileFd::None, false);
+ return true;
+ }
+
+ std::vector<APT::Configuration::Compressor> const compressors = APT::Configuration::getCompressors();
+ std::vector<APT::Configuration::Compressor>::const_iterator compressor = compressors.begin();
+ for (; compressor != compressors.end(); ++compressor) {
+ if (compressor->Name == DecompressProg) {
+ return InFd.OpenDescriptor(File.Fd(), FileFd::ReadOnly, *compressor, false);
+ }
}
- // Fix up our FDs
- InFd.Fd(Pipes[0]);
- close(Pipes[1]);
- return true;
+ return _error->Error(_("Cannot find a configured compressor for '%s'"),
+ DecompressProg.c_str());
+
}
/*}}}*/
// ExtractTar::Go - Perform extraction /*{{{*/
return false;
// Loop over all blocks
- string LastLongLink;
- string LastLongName;
+ string LastLongLink, ItemLink;
+ string LastLongName, ItemName;
while (1)
{
bool BadRecord = false;
/* Check for a block of nulls - in this case we kill gzip, GNU tar
does this.. */
if (NewSum == ' '*sizeof(Tar->Checksum))
- return Done(true);
+ return Done();
if (NewSum != CheckSum)
return _error->Error(_("Tar checksum failed, archive corrupted"));
// Decode all of the fields
pkgDirStream::Item Itm;
if (StrToNum(Tar->Mode,Itm.Mode,sizeof(Tar->Mode),8) == false ||
- StrToNum(Tar->UserID,Itm.UID,sizeof(Tar->UserID),8) == false ||
- StrToNum(Tar->GroupID,Itm.GID,sizeof(Tar->GroupID),8) == false ||
- StrToNum(Tar->Size,Itm.Size,sizeof(Tar->Size),8) == false ||
- StrToNum(Tar->MTime,Itm.MTime,sizeof(Tar->MTime),8) == false ||
+ (Base256ToNum(Tar->UserID,Itm.UID,8) == false &&
+ StrToNum(Tar->UserID,Itm.UID,sizeof(Tar->UserID),8) == false) ||
+ (Base256ToNum(Tar->GroupID,Itm.GID,8) == false &&
+ StrToNum(Tar->GroupID,Itm.GID,sizeof(Tar->GroupID),8) == false) ||
+ (Base256ToNum(Tar->Size,Itm.Size,12) == false &&
+ StrToNum(Tar->Size,Itm.Size,sizeof(Tar->Size),8) == false) ||
+ (Base256ToNum(Tar->MTime,Itm.MTime,12) == false &&
+ StrToNum(Tar->MTime,Itm.MTime,sizeof(Tar->MTime),8) == false) ||
StrToNum(Tar->Major,Itm.Major,sizeof(Tar->Major),8) == false ||
StrToNum(Tar->Minor,Itm.Minor,sizeof(Tar->Minor),8) == false)
return _error->Error(_("Corrupted archive"));
-
- // Grab the filename
+
+ // Grab the filename and link target: use last long name if one was
+ // set, otherwise use the header value as-is, but remember that it may
+ // fill the entire 100-byte block and needs to be zero-terminated.
+ // See Debian Bug #689582.
if (LastLongName.empty() == false)
Itm.Name = (char *)LastLongName.c_str();
else
- {
- Tar->Name[sizeof(Tar->Name)-1] = 0;
- Itm.Name = Tar->Name;
- }
+ Itm.Name = (char *)ItemName.assign(Tar->Name, sizeof(Tar->Name)).c_str();
if (Itm.Name[0] == '.' && Itm.Name[1] == '/' && Itm.Name[2] != 0)
Itm.Name += 2;
-
- // Grab the link target
- Tar->Name[sizeof(Tar->LinkName)-1] = 0;
- Itm.LinkTarget = Tar->LinkName;
if (LastLongLink.empty() == false)
Itm.LinkTarget = (char *)LastLongLink.c_str();
-
+ else
+ Itm.LinkTarget = (char *)ItemLink.assign(Tar->LinkName, sizeof(Tar->LinkName)).c_str();
+
// Convert the type over
switch (Tar->LinkFlag)
{
case GNU_LongLink:
{
- unsigned long Length = Itm.Size;
+ unsigned long long Length = Itm.Size;
unsigned char Block[512];
while (Length > 0)
{
case GNU_LongName:
{
- unsigned long Length = Itm.Size;
+ unsigned long long Length = Itm.Size;
unsigned char Block[512];
while (Length > 0)
{
return false;
// Copy the file over the FD
- unsigned long Size = Itm.Size;
+ unsigned long long Size = Itm.Size;
while (Size != 0)
{
unsigned char Junk[32*1024];
- unsigned long Read = min(Size,(unsigned long)sizeof(Junk));
+ unsigned long Read = min(Size, (unsigned long long)sizeof(Junk));
if (InFd.Read(Junk,((Read+511)/512)*512) == false)
return false;
}
// And finish up
- if (Itm.Size >= 0 && BadRecord == false)
+ if (BadRecord == false)
if (Stream.FinishedFile(Itm,Fd) == false)
return false;
LastLongLink.erase();
}
- return Done(false);
+ return Done();
}
/*}}}*/
projects / apt.git / blobdiff