Re: [Bug-wget] Check file length before download

[Micah Cowan]

On 06/16/2012 08:31 AM, Ángel González wrote:
> On 16/06/12 12:07, jjDaNiMoTh wrote:
>> Hi list,
>>
>> It's not a bug, but I don't find any other ml for this awesome project.
> Don't worry. It's the appropiate list.
> 
>> I want to download files from a Web Server which hasn't the Range
>> support, so basically -c doesn't work.
>> What I need is the ability to automatically remove the incomplete file
>> (if remote and local file size aren't the same) and restart
>> downloading from scratch that file: could I achieve this with wget? If
>> not, there is some way or some other tool that allow such behavior?
> -c option will do what you want even if the server doesn't support range
> requests.
> 

It would be nice if it did, (or does it now? If so, it's changed
behavior), but what it actually does (last I checked) is _skip_ over the
data, up to the current file length. That's probably a bit dangerous,
potentially - particularly when the file is dynamically generated (a
frequent case for resources that don't support ranged requests).

I suppose part of the theory was that, for crappy connections, you
didn't want to wind up with a smaller file in one run than you do on
another...

I believe at some point in the past (somewhere pre-1.10?), it did
actually start the download over again. Perhaps we should reconsider this.

-mjc