[lnkForumImage]
TotalShareware - Download Free Software
Confronta i prezzi di migliaia di prodotti.
Asp Forum
 Home | Login | Register | Search 


 

Forums >

comp.lang.ruby

Getting files out of a .tar.gz archive

Chris Duncan

3/10/2008 2:09:00 PM

Hi,

I have a .tar.gz file containing some xml files. I need to locate
particular files in the archive and process them. I'm looking for a
pure Ruby way to do this without resorting to external system
commands.

I found Archive::Tar::Minitar that allows me to process my files once
I have expanded the .tar.gz file to a .tar file. So I can do this :-

open(@tarfile, "rb") do |f|
Archive::Tar::Minitar::Reader.open(f).each do |entry|
fpl = StringIO.new( entry.read) if entry.name =~ /
#{@date}#{channel}_pl/
fpi = StringIO.new( entry.read) if entry.name =~ /
#{@date}#{channel}_pi/
end
end

However, in order to get a tar file, I have to call gunzip to expand
my .tar.gz file. Does anybody know of a way for me to replace the
gunzip call with a Ruby library call of some sort? Or does anyone have
suggestions for an alternative way to do this?

Cheers,

Chris
http:/...
10 Answers

Todd Benson

3/10/2008 3:20:00 PM

0

On Mon, Mar 10, 2008 at 9:09 AM, celldee <celldee@gmail.com> wrote:
> Hi,
>
> I have a .tar.gz file containing some xml files. I need to locate
> particular files in the archive and process them. I'm looking for a
> pure Ruby way to do this without resorting to external system
> commands.
>
> I found Archive::Tar::Minitar that allows me to process my files once
> I have expanded the .tar.gz file to a .tar file. So I can do this :-
>
> open(@tarfile, "rb") do |f|
> Archive::Tar::Minitar::Reader.open(f).each do |entry|
> fpl = StringIO.new( entry.read) if entry.name =~ /
> #{@date}#{channel}_pl/
> fpi = StringIO.new( entry.read) if entry.name =~ /
> #{@date}#{channel}_pi/
> end
> end
>
> However, in order to get a tar file, I have to call gunzip to expand
> my .tar.gz file. Does anybody know of a way for me to replace the
> gunzip call with a Ruby library call of some sort? Or does anyone have
> suggestions for an alternative way to do this?
>
> Cheers,
>
> Chris

There's Zlib::Gzip* classes. I've never used them, though.

Todd

Chris Duncan

3/10/2008 4:46:00 PM

0

Hi Todd,

I've tried to use Zlib::GzipReader, but that just gives me a
continuous stream of text or a series of strings that do not resemble
the actual file structure, unless I've missed something.

Thanks,

Chris
http:/...

On Mar 10, 3:19=A0pm, "Todd Benson" <caduce...@gmail.com> wrote:
> On Mon, Mar 10, 2008 at 9:09 AM, celldee <cell...@gmail.com> wrote:
> > Hi,
>
> > =A0I have a .tar.gz file containing some xml files. I need to locate
> > =A0particular files in the archive and process them. I'm looking for a
> > =A0pure Ruby way to do this without resorting to external system
> > =A0commands.
>
> > =A0I found Archive::Tar::Minitar that allows me to process my files once=

> > =A0I have expanded the .tar.gz file to a .tar file. So I can do this :-
>
> > =A0open(@tarfile, "rb") do |f|
> > =A0 Archive::Tar::Minitar::Reader.open(f).each do |entry|
> > =A0 =A0 fpl =3D StringIO.new( entry.read) if entry.name =3D~ /
> > =A0#{@date}#{channel}_pl/
> > =A0 =A0 fpi =3D StringIO.new( entry.read) if entry.name =3D~ /
> > =A0#{@date}#{channel}_pi/
> > =A0 end
> > =A0end
>
> > =A0However, in order to get a tar file, I have to call gunzip to expand
> > =A0my .tar.gz file. Does anybody know of a way for me to replace the
> > =A0gunzip call with a Ruby library call of some sort? Or does anyone hav=
e
> > =A0suggestions for an alternative way to do this?
>
> > =A0Cheers,
>
> > =A0Chris
>
> There's Zlib::Gzip* classes. =A0I've never used them, though.
>
> Todd

Daniel Brumbaugh Keeney

3/10/2008 5:04:00 PM

0

On Mon, Mar 10, 2008 at 9:09 AM, celldee <celldee@gmail.com> wrote:
> Hi,
>
> I have a .tar.gz file containing some xml files. I need to locate
> particular files in the archive and process them. I'm looking for a
> pure Ruby way to do this without resorting to external system
> commands.
>
> I found Archive::Tar::Minitar that allows me to process my files once
> I have expanded the .tar.gz file to a .tar file. So I can do this :-
>
> open(@tarfile, "rb") do |f|
> Archive::Tar::Minitar::Reader.open(f).each do |entry|
> fpl = StringIO.new( entry.read) if entry.name =~ /
> #{@date}#{channel}_pl/
> fpi = StringIO.new( entry.read) if entry.name =~ /
> #{@date}#{channel}_pi/
> end
> end
>
> However, in order to get a tar file, I have to call gunzip to expand
> my .tar.gz file. Does anybody know of a way for me to replace the
> gunzip call with a Ruby library call of some sort? Or does anyone have
> suggestions for an alternative way to do this?
>
> Cheers,
>
> Chris
> http:/...
>
>

Use the docs.

From Minitar's readme:
tgz = Zlib::GzipReader.new(File.open('test.tgz', 'rb'))
# Warning: tgz and the file will be closed.
Minitar.unpack(tgz, 'x')

For GZip and the rest of the standard library,
http://www.ruby-doc.o...

Daniel Brumbaugh Keeney

Chris Duncan

3/10/2008 5:44:00 PM

0

Hi Daniel,

I saw that, but I don't want to expand the .tar.gz any more than I
have to. The code that I put up earlier is getting what I want out of
the .tar file using Minitar::Reader which I quite like, I'm just
looking to eliminate the gunzip step. Minitar.unpack expands
the .tar.gz and writes the files to disk, which means that I'll have
to mess around in the filesystem more than I need to.

Thanks for your reply,

Chris
http:/...

On Mar 10, 5:04=A0pm, "Daniel Brumbaugh Keeney"
<devi.webmas...@gmail.com> wrote:
> On Mon, Mar 10, 2008 at 9:09 AM, celldee <cell...@gmail.com> wrote:
> > Hi,
>
> > =A0I have a .tar.gz file containing some xml files. I need to locate
> > =A0particular files in the archive and process them. I'm looking for a
> > =A0pure Ruby way to do this without resorting to external system
> > =A0commands.
>
> > =A0I found Archive::Tar::Minitar that allows me to process my files once=

> > =A0I have expanded the .tar.gz file to a .tar file. So I can do this :-
>
> > =A0open(@tarfile, "rb") do |f|
> > =A0 Archive::Tar::Minitar::Reader.open(f).each do |entry|
> > =A0 =A0 fpl =3D StringIO.new( entry.read) if entry.name =3D~ /
> > =A0#{@date}#{channel}_pl/
> > =A0 =A0 fpi =3D StringIO.new( entry.read) if entry.name =3D~ /
> > =A0#{@date}#{channel}_pi/
> > =A0 end
> > =A0end
>
> > =A0However, in order to get a tar file, I have to call gunzip to expand
> > =A0my .tar.gz file. Does anybody know of a way for me to replace the
> > =A0gunzip call with a Ruby library call of some sort? Or does anyone hav=
e
> > =A0suggestions for an alternative way to do this?
>
> > =A0Cheers,
>
> > =A0Chris
> > =A0http:/...
>
> Use the docs.
>
> From Minitar's readme:
> tgz =3D Zlib::GzipReader.new(File.open('test.tgz', 'rb'))
> # Warning: tgz and the file will be closed.
> Minitar.unpack(tgz, 'x')
>
> For GZip and the rest of the standard library,http://www.ruby-doc...
ib/
>
> Daniel Brumbaugh Keeney

Joachim Glauche

3/10/2008 5:56:00 PM

0

celldee wrote:
> I saw that, but I don't want to expand the .tar.gz any more than I
> have to. The code that I put up earlier is getting what I want out of
> the .tar file using Minitar::Reader which I quite like, I'm just
> looking to eliminate the gunzip step.

If packed with Gzip, you always have to unpack it in order to read the
content.

Tar is the container and Gzip does the compression. So the files in a
tar.gz file are first put into a container then packed. So in order you
read anything inside the tar you have to unpack the file to have access
the tarball to read your file inside.
--
Posted via http://www.ruby-....

Paul Mckibbin

3/10/2008 6:53:00 PM

0

Chris,

Persevere with Zlib::GzipReader, that'll give you what you want.

Mac
--
Posted via http://www.ruby-....

Daniel Brumbaugh Keeney

3/11/2008 12:01:00 AM

0

On Mon, Mar 10, 2008 at 12:43 PM, celldee <celldee@gmail.com> wrote:
> Hi Daniel,
>
> I saw that, but I don't want to expand the .tar.gz any more than I
> have to. The code that I put up earlier is getting what I want out of
> the .tar file using Minitar::Reader which I quite like, I'm just
> looking to eliminate the gunzip step. Minitar.unpack expands
> the .tar.gz and writes the files to disk, which means that I'll have
> to mess around in the filesystem more than I need to.
>
> Thanks for your reply,
>
> Chris


My apologies for failing to understand the issue. GZip and Minitar
both provide incremental readers, although I have not used them. I
believe the correct combination for what you're asking is this:

tgz = Zlib::GzipReader.new(File.open('test.tgz', 'rb'))
# Warning: tgz and the file will be closed.
reader = Minitar::Reader.new(tgz)
reader.each_entry do |file|
#do something with each file, and break if you like
end
reader.close # does this do anything?
tgz.close

Daniel Brumbaugh Keeney

Daniel Brumbaugh Keeney

3/11/2008 12:04:00 AM

0

On Mon, Mar 10, 2008 at 7:01 PM, Daniel Brumbaugh Keeney
<devi.webmaster@gmail.com> wrote:
> tgz = Zlib::GzipReader.new(File.open('test.tgz', 'rb'))
> # Warning: tgz and the file will be closed.

I seem to have failed to remove the former comment. It looks like
Minitar::Reader at no point closes tgz or the file, and therefore tgz
needs to be closed manually afterward, as I my code correctly
demonstrated, though the comment did not.

Daniel Brumbaugh Keeney

Paul Mckibbin

3/11/2008 12:45:00 AM

0

Daniel Brumbaugh Keeney wrote:
> On Mon, Mar 10, 2008 at 7:01 PM, Daniel Brumbaugh Keeney
> <devi.webmaster@gmail.com> wrote:
>> tgz = Zlib::GzipReader.new(File.open('test.tgz', 'rb'))
>> # Warning: tgz and the file will be closed.
>
> I seem to have failed to remove the former comment. It looks like
> Minitar::Reader at no point closes tgz or the file, and therefore tgz
> needs to be closed manually afterward, as I my code correctly
> demonstrated, though the comment did not.
>
> Daniel Brumbaugh Keeney

You can use Zlib::GzipReader.open with a block. (Since I set Chris the
original problem, and this isn't the hard part, I'll show the code that
is used in the app, with a couple of minor mods to make it look like
Chris's example.)

Zlib::GzipReader.open(@tarfile) { |tgz|
Archive::Tar::Minitar::Reader.open(tgz).each do |entry|

# Chris's code
# fpl = StringIO.new( entry.read) if entry.name =~
/#{@date}#{channel}_pl/
# fpi = StringIO.new( entry.read) if entry.name =~
/#{@date}#{channel}_pi/

# Or test the verification by using the XML document directly
# fpl=REXML::Document.new(StringIO.new(entry.read)) if entry.name
=~/#{date}#{channel}_pl/
# fpi=REXML::Document.new(StringIO.new(entry.read)) if entry.name
=~/#{date}#{channel}_pi/

end
}

Mac
--
Posted via http://www.ruby-....

Chris Duncan

3/11/2008 10:03:00 AM

0

Thanks Mac, works like a charm. Thanks for your efforts everybody
else.

On Mar 11, 12:44=A0am, Paul Mckibbin <pmckib...@gmail.com> wrote:
> Daniel Brumbaugh Keeney wrote:
> > On Mon, Mar 10, 2008 at 7:01 PM, Daniel Brumbaugh Keeney
> > <devi.webmas...@gmail.com> wrote:
> >> =A0tgz =3D Zlib::GzipReader.new(File.open('test.tgz', 'rb'))
> >> =A0# Warning: tgz and the file will be closed.
>
> > I seem to have failed to remove the former comment. It looks like
> > Minitar::Reader at no point closes tgz or the file, and therefore tgz
> > needs to be closed manually afterward, as I my code correctly
> > demonstrated, though the comment did not.
>
> > Daniel Brumbaugh Keeney
>
> You can use Zlib::GzipReader.open with a block. (Since I set Chris the
> original problem, and this isn't the hard part, I'll show the code that
> is used in the app, with a couple of minor mods to make it look like
> Chris's example.)
>
> Zlib::GzipReader.open(@tarfile) { |tgz|
> =A0 Archive::Tar::Minitar::Reader.open(tgz).each do |entry|
>
> # Chris's code
> # =A0 =A0fpl =3D StringIO.new( entry.read) if entry.name =3D~
> /#{@date}#{channel}_pl/
> # =A0 =A0fpi =3D StringIO.new( entry.read) if entry.name =3D~
> /#{@date}#{channel}_pi/
>
> # Or test the verification by using the XML document directly
> # =A0 fpl=3DREXML::Document.new(StringIO.new(entry.read)) if entry.name
> =3D~/#{date}#{channel}_pl/
> # =A0 fpi=3DREXML::Document.new(StringIO.new(entry.read)) if entry.name
> =3D~/#{date}#{channel}_pi/
>
> =A0 end
>
> }
>
> Mac
> --
> Posted viahttp://www.ruby-....