Shell Programming and Scripting

Hi

I am downloading some files from the Internet. They are from different sources and they are all called

Code:

-.txt

that is: a dash "-" followed by a dot "." and then the file extension.
So i wrote a small script that hoped will downloads the first file and then rename them once they are completely downloaded. Here is the script I wrote:

Code:

#!/bin/sh
for i in `seq 1 1 30`
do
wget url
mv ./-.txt $i.txt
done

but what I get is that the script (above) downloads the first file "-.txt" and does not rename it to 1.txt as I wanted, and so keeps giving me the message that file "-.txt" has already been downloaded.

What is wrong in my script? (I have little experience with scripting in bash)

Thank you

Try either quoting or escapeing the filename.

As in either one of:

Code:

mv ./"-.txt" $i.txt
mv ./\-.txt $i.txt

Also about the urls, id try with:

Code:

urls="url1 url2 ..."
for url in $urls
do
   wget -o $(basename $url).txt $url 
done

So you avoid to rename the files, alternativly you could insert at the proper places:

Code:

# Below: urls=
i=0

# Replace on wget line:
$i-$(basename $url).txt

# Below wget
i=$(($i + 1))

Hope this helps & happy holidays

EDIT:
Understand that it saved the file as -.txt because - was the default input as you did not define an output file.

Hi

Thank you for your quick response.
Unfortunately, the files I download are contained in a python script (they are numerous) so I am left with the first answer of yours. I used the wget example to illustrate my problem. But it is fine since I just replace it with the python script.
I tried both escaping the dash and quoting it, but they didn't work. I still download the first file as "-.txt" and it never gets renamed to "1.txt".
Numbering these files is important, and I do use the first 3000 files of the collection I am trying to download (they are scientific txt files with lots of numbers)

Thank you again

When you invoke a "command line" in python, you can use the equivalent of exec or system. In either case, the child process (commands like mv and wget) are being executed by /bin/sh, which may or may not be the shell you think you are using.

There are two ways to approach this:
1. Play 20 questions and hope we get you where you need to be
2. OR make a suggestion for a quick workaround.

I vote for number two:
Use a separate script to download and rename files. Do not mess with doing that in python unless what I said about /bin/sh is very clear to you.

So your python code gets a few lines commented out - the ones doing downloads and renames. You run a five line shell script like what sea provided. Then run the abbreviated python. One advantage: you can see the filenames. Before you mess with running the python code.

I am assuming you want to complete a long task, so this will get you there the fastest.

I do not see the need for extra quotes or escapes here. By using ./ the file name will not be interpreted as a command option...

It must be in the python part. To illustrate, if I use an equivalent bash script like this:

Code:

for i in {1..5}
do
  curl http://www.google.com > -.txt
  mv ./-.txt $i.txt
done

I get:

Code:

$ ls -- ?.txt
1.txt	2.txt	3.txt	4.txt	5.txt

I tried your suggestion, still I get the same results. The first files is downloaded but not renamed (moved), and then the scripts keeps running giving the message that the files exists. If I keep it running, it will produce the message 3000 times (2999 times) so I have to stop it (^C)

You have suggestions from sea, jim mcnamara, and Scrutinizer. When you said: which suggestion were you talking about? Have you looked at the other suggestions?

What OS are you using?

Please show us the EXACT diagnostic output produced by the failing mv command.