Hi again,

I found something what can be called a bug.

The command line and the output (shortened):

$ wget -k www.seznam.cz
--14:14:28--  http://www.seznam.cz/
           => `index.html'
Resolving www.seznam.cz... done.
Connecting to www.seznam.cz[212.80.76.18]:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: unspecified [text/html]

    [ <=>                                 ] 19,975         3.17M/s

14:14:28 (3.17 MB/s) - `index.html' saved [19975]

Converting index.html... 5-123
Converted 1 files in 0.01 seconds.

---
That is, newly created file is really link-converted.

Now I run:

$ wget -k -O myfile www.seznam.cz
--14:16:07--  http://www.seznam.cz/
           => `myfile'
Resolving www.seznam.cz... done.
Connecting to www.seznam.cz[212.80.76.3]:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: unspecified [text/html]

    [ <=>                                 ] 19,980         3.18M/s

14:16:07 (3.18 MB/s) - `myfile' saved [19980]

index.html.1: No such file or directory
Converting index.html.1... nothing to do.
Converted 1 files in 0.00 seconds.

---
Now myfile is created and then wget tries to convert index.html.1, i.e. the 
file it normally *would* create if there was no -O option... 

When I wish the content to be sent to stdout (-O -), this postponed 
converting function is run again on index.html.1. Which is totally wrong, all 
content has been sent out to stdout already.

Not only my content is not link-converted. Is not here a possibility, that
wget can inadvertently garble files on disk it has nothing to do with?

Vlada Macek


Reply via email to