wget link from random line in file

Question

I'm trying to download a website that uses semi-predictable urls, meaning the url will always end with a random five character alphanumeric string. I created a file with crunch with the random string by using the command:

crunch 5 5 abcdefghijklmnopqrstuvwxyz123456789 > possible_links

Then I created a bash file to call the lines and wget the links:

#!/bin/bash
FILE=possible_links
while read line; do
        wget -q --wait=20 www.ghostbin.com/paste/${line}
done < $FILE

but obviously it is going to go to aaaaa, then aaaab, then aaaac, aaaad, etc etc, is there a way to make it go to random lines?

`but obviously it is going to go to aaaaa` with out `done<"$FILE"` huh? — sjsam, May 29 '16 at 04:09
@sjsam yea i did a copy paste and didnt grab the < $FILE it doesnt have quotes though — Matt, May 29 '16 at 04:12
http://stackoverflow.com/questions/448005/whats-an-easy-way-to-read-random-line-from-a-file-in-unix-command-line may have some useful thoughts. — larsks, May 29 '16 at 04:16
@Matt : My solution is specific to this problem, the link \@larsks pointed out give you many general solutions. Do have a look :-) — sjsam, May 29 '16 at 04:36
@sjsam i did, i was mid modifying the script when you posted your answer :) — Matt, May 29 '16 at 04:44

sjsam · Accepted Answer · 2016-05-29T04:50:36.413

Use mktemp --dry-run option :

#!/bin/bash
while true # or specify a count using something like while [ $count -le 20 ]
do
rand_str="$(mktemp --dry-run XXXXX)" # 5 Xs for five random characters
  wget -q --wait=20 www.ghostbin.com/paste/${rand_str}
# if you use count increment count ie do '((count++))' else you get infinite loop
done

General solution (for n random characters)

str=$(printf "%-10s" "X") # here n=10
while condition
do
rand_str=$(mktemp --dry-run ${str// /X}) 
.
.

wget link from random line in file

1 Answers1