Re: [Bug-wget] [PATCH] Bug 40426 follow-up

[Ander Juaristi]

On 03/13/2015 12:03 PM, Tim Ruehsen wrote:
Hi Tim,

> wget -O - http://foo > file
> is not using -r, so strictly the doc would be still correct when prohibiting
> the combination of -r and -O -.

True. Although I didn't notice that '-r' was not explicitly mentioned, I had (and still have) an idea in 
mind of how Wget should behave. If "wget -O - http://foo > file" dumps the contents in file, 
I, as a final user, would expect "wget -r -O - http://foo > file" to do the same but 
recursively.

> One approach would be to download everything 'as normal', keeping a list of
> all downloaded files in download order. And when ready and all downloading /
> conversion has been done, go through the list and copy all the file data to
> stdout (+ removing files / cleaning up).
I assume that when you say "download everything as normal" you mean the same as 
"wget -r http://foo";, without '-O -'. Right?
Let Wget download all the files, place them all together in a directory named 
"foo", and then in the end dump them all (honoring the order in which they were 
downloaded) and remove them.

> What do you think ?

Yep, it looks better to me.

I'll try that with HTTP first, and let you know.

--
Regards,
- AJ