- Edwin - wrote:

On Mon, 20 Sep 2004 07:11:37 -0500 Jim Sabatke wrote:

...

How do I access a link with wget written in the following form?

http://prdownloads.sourceforge.net/abiword/abiword-2.0.11.tar.gz?use_mirror=...

Just type it after wget and it'll download the file.

$ wget <the-long-url-here>

I've tried that, a whole bunch of times. It creates a subdirectory prdownloads.sourceforge.net with the following: robot.txt abiword/ and abiword/ contains abiword-2.0.11.tar.gz?mirror=mesh and it's not the right file. It's very short and am HTML file that lynx won't decode. I've tried every option in man and in the online manual. -- Jim Sabatke Hire Me!! - See my resume at http://my.execpc.com/~jsabatke Do not meddle in the affairs of Dragons, for you are crunchy and good with ketchup. NOTE: Please do not email me any attachments with Microsoft extensions. They are deleted on my ISP's server before I ever see them, and no bounce message is sent.

Felix Miata wrote:

Jim Sabatke wrote:

...

...

...

How do I access a link with wget written in the following form?

...

...

...

http://prdownloads.sourceforge.net/abiword/abiword-2.0.11.tar.gz?use_mirror=...

...

I've tried every option in man and in the online manual.

I typically find it necessary to put quotes around a URL that contains a "?".

You need to escape the ? using a \ because it's an special character, like this: http://prdownloads.sourceforge.net/abiword/abiword-2.0.11.tar.gz\?use_mirror=unc Or try the double-quotes proposed by Felix. HTH, Martin -- Martin Mielke Senior UNIX SysAdmin martin.mielke@thales-is.com THALES Information Systems http://www.thales-is.com/ Tel.: (+34) 91 556 92 62 TimeZone: GMT+1 :-)

Randall R Schulz wrote:

Jim,

Wget is very powerful. There are many options, but they're not particularly difficult to understand, so you'd be well advised to familiarize yourself with them. The "--help" output is over 100 lines long.

As I posted, I spent hours reading man, then online manual, and searching google for the answer. This situation isn't covered in any of them that I could find. I think the wget community would be well served by having this type of usage covered by an example.

Randall R Schulz wrote:

Jim,

On Monday 20 September 2004 11:59, Jim Sabatke wrote:

...

Randall R Schulz wrote:

...

Jim,

Wget is very powerful. There are many options, but they're not particularly difficult to understand, so you'd be well advised to familiarize yourself with them. The "--help" output is over 100 lines long.

As I posted, I spent hours reading man, then online manual, and searching google for the answer. This situation isn't covered in any of them that I could find. I think the wget community would be well served by having this type of usage covered by an example.

Here's an excerpt from the manual page:

-==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==-

--cut-dirs=number

Ignore number directory components. This is useful for getting a fine-grained control over the directory where recursive retrieval will be saved.

Take, for example, the directory at ftp://ftp.xemacs.org/pub/xemacs/. If you retrieve it with -r, it will be saved locally under ftp.xemacs.org/pub/xemacs/. While the -nH option can remove the ftp.xemacs.org/ part, you are still stuck with pub/xemacs. This is where --cut-dirs comes in handy; it makes Wget not ‘‘see'' number remote directory components. Here are several examples of how --cut-dirs option works.

No options -> ftp.xemacs.org/pub/xemacs/ -nH -> pub/xemacs/ -nH --cut-dirs=1 -> xemacs/ -nH --cut-dirs=2 -> .

--cut-dirs=1 -> ftp.xemacs.org/xemacs/ ...

If you just want to get rid of the directory structure, this option is similar to a combination of -nd and -P. However, unlike -nd, --cut-dirs does not lose with subdirectories---for instance, with -nH --cut-dirs=1, a beta/ subdirectory will be placed to xemacs/beta, as one would expect.

-==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==-

To me, wget ain't rocket science.

And if you want to get hired for computer work, you're surely going to have to be able to figure this sort of thing out under your own intellectual power, it seems to me.

Randall Schulz

That's an idiotic reply. Who would know that --cut-dirs would have anything to do with the URL mentioned? I read through everything you posted and still don't get the reference. What on earth does that have to do with www.abc.com?something=nothing type structures? I still have no idea why I want to get rid of a directory structure. As I originally said, I didn't understand the URL structure for the get. I couldn't find doc's on it. I still have no idea what the '?' is for. Sometimes it's very easy when you have working knowledge of an area that is helpful, and competely obsucre if you don't. I once made a very good living from solvng difficult problems, although my area is really project management. Some things just require assistance. Perhaps your bullying should convince me to slit my wrists and get it over with. The reason I'm not working is because of severe depression. I'm desperately (yes, I am desperate) trying to recover from that by doing things to spur me in a positive direction. In the past couple days I've compiled, installed and configured gnome 2.8 and the latest abiword. That isn't easy for most people, and I did it without asking anyone a thing.

Jim Sabatke wrote:

Randall R Schulz wrote:

...

Jim,

Wget is very powerful. There are many options, but they're not particularly difficult to understand, so you'd be well advised to familiarize yourself with them. The "--help" output is over 100 lines long.

As I posted, I spent hours reading man, then online manual, and searching google for the answer. This situation isn't covered in any of them that I could find. I think the wget community would be well served by having this type of usage covered by an example.

May I suggest that you subscribe to the wget list? Perhaps they can solve your problem/confirm the bug has been reported etc. Send a blank mail to: wget-subscribe@sunsite.dk -- The Little Helper ======================================================================== Hylton Conacher - Linux user # 229959 at http://counter.li.org Currently using SuSE 9.0 Professional with KDE 3.1 Licenced Windows user ========================================================================

Jim, On Monday 20 September 2004 17:39, Jim Sabatke wrote:

Mike McMullin wrote: ...

...

wget http://prdownloads.sourceforge.net/abiword/abiword-2.0.11.tar.gz

Will work for you. I have just tried this in a shell, and the file transfered.

Mike

Actually, that was one of the first things I tried. It downloads a file called abiword-2.0.11.tar.gz, but it is way too small (supposed to be almost 25MB) and it isn't a tar.gz file. It looked like an HTML file, but lynx couldn't decode it.

Thanks for trying!

Sourceforge has mirror selection built in to their main Web site design and it is not the most "wget-friendly," if you will. The link you copied is misleading in its appearance. It actually leads to a Web page from which you select a mirror (buried amid the boilerplate and advertising is the text "You are requesting file: /abiword/abiword-2.0.11.tar.gz Please select a mirror"). The links on that page are the ones that include the query parameter ("?...") that selects the mirror. Without that parameter, you just won't get the actual file you're trying to download. Thus there's no way around either using the "-O" option to wget or manually renaming the file with the mirror selection query parameter appended. Randall Schulz

Jim, On Monday 20 September 2004 18:46, Jim Sabatke wrote:

...

I think the path I followed was this:

From freshmeat's listing I browsed to:

http://freshmeat.net/redir/abiword/56/url_tgz/abiword-2.0.10.tar.gz

Which took me to the mirror selection page. From there I copied the link for the download from the mirror and used that for input to wget:

http://prdownloads.sourceforge.net/abiword/abiword-2.0.10.tar.gz?use_mi rror=mesh

That was the URL that consumed all the time trying to get the resumed download. I thought quoting it had worked, but I tried it again and it downloads a small file, like the previous attempts.

Quoting isn't going to make a difference for this URL unless in the directory in which you issue the wget command there happens to be a directory named "http:" which contains a directory named "prdownloads.sourceforge.net" which contains a directory named "abiword" which contains a file or directory named "abiword-2.0.10.tar.gz?use_mirror=mesh" with any character in the position of the '?'. Otherwise, that argument, which _is_ a shell glob pattern, is passed unchanged. (And even that detail depends on which shell you're using and / or the options you've selected.)

I didn't want to click on the link, because I didn't want the browser to truncate the 20MB+ that I had already retrieved. Had I moved the file and clicked on the link, I would have seen that the next screen shows the direct link in case the download doesn't start. I used that link, provided by someone else in an earlier email, with wget -c and the download resumed to completion.

From now on, I'm just going to start the download, kill it, and then use wget with the backup link.

That isn't going to work any better than using the browser's "copy link address" command (typically accessed via a context menu from the link itself) and then using the resulting URL with wget (including any necessary directory trimming options, the "-O" to get the right output name and the "-c" to get "wget" to retry if an error occurs).

BTW, a few weeks ago, when I was asking for alternative download programs, several people mentioned that wget has no problem with following this kind of link. I did ask in advance. Then I couldn't figure out how to make it work and focused too hard on making wget do something it apparenlty can't do. Telling someone they may not be competent to get a job because they couldn't follow the simple man pages was really annoying after doing that much research into the problem.

You were not misinformed. Wget handles those URLs just fine. If you want query strings stripped (or other options such as the directory adjustments and continue options, etc) you can write a simple shell script to save you the tedium of handling those details manually. Randall Schulz

Leen, On Tuesday 21 September 2004 03:32, Leendert Meyer wrote:

...

wget 'http://prdownloads.sourceforge.net/abiword/abiword-2.0.11.tar.gz?use_m irror=unc' yields another HTML-page, so clearly this is not the right URL to use with wget. Put quotes around the URL because of the '?'.

Let's not encourage magical thinking w.r.t. shell glob characters. The question mark (or an asterisk or a sequence of characters enclosed in square brackets) will only be changed by the shell's glob processing when the full string in which it appears can be replaced by the name of an existing entity in the file system (and one that can be referred to relative to the current directory, whenever the argument string does not start with a slash). If the shell is set to reject glob patterns that don't match an existing file system entity, then you'll get an explicit error from the shell. If not, the unaltered pattern will be passed on as an argument to the program. So in cases like this, the likelihood that such an argument will be altered silently by shell glob processing is exceedingly remote.

..

Cheers,

Leen

Randall Schulz