Repeating zsh brace expansion values in zsh to download multiple files using wget2

Create a function which accepts <brace-expansion-value> as an argument. Then your function can re-use that argument as many times as you like:

get_one() { # $1 is a number 1..6000 wget2 -q https://example.com/"$1"x"$1" } 

Now call that function once for each integer you want to iterate over.

Another alternative would be to avoid using brace expansion and construct your wget directly, using jot:

jot -w 'wget2 -q https://example.com/%dx%1$d' 6000 | ( head -5; echo ...; tail -5 ) wget2 -q https://example.com/1x1 wget2 -q https://example.com/2x2 wget2 -q https://example.com/3x3 wget2 -q https://example.com/4x4 wget2 -q https://example.com/5x5 ... wget2 -q https://example.com/5996x5996 wget2 -q https://example.com/5997x5997 wget2 -q https://example.com/5998x5998 wget2 -q https://example.com/5999x5999 wget2 -q https://example.com/6000x6000 

Then pipe that to a shell.

If multiple calls to wget2 isn't your bag, then use jot solely for creating the URLs and pipe the URLs to xargs(1):

jot -w 'https://example.com/%dx%1$d' 6000 | xargs wget2 -q 

And indeed, as @muru points out, you could reduce that to:

jot -w 'https://example.com/%dx%1$d' 6000 | wget2 -qi- 

You can't use just brace expansion for that.

One approach would be to do the equivalent of perl's:

exec "wget2", map {"https://example.com/${_}x$_"} 1..6000 

In zsh, to do that mapping, there are a few options

set -o extendedglob (){wget ${@/(#m)*/https:\/\/example.com\/${MATCH}x$MATCH}} {1..6000} 

Or the equivalent with the csh-style :s modifier which avoids having to escape the / (but is currently quite buggy when quoting is involved):

set -o extendedglob -o histsubstpattern (){wget $@:s@(#m)*@https://example.com/${MATCH}x$MATCH} {1..6000} 

Or:

urls=() printf -v urls 'https://example.com/%1$sx%1$s' {1..6000} wget2 $urls 

Though building that same list with a loop is about as short:

urls=() for ((i=1;i<=6000;i++)) urls+=https://example.com/${i}x${i} wget2 $urls 

(even shorter with for i ({1..6000}) though that's less flexible and means building the whole list of numbers in memory and then iterate on it).

On most systems, those 6000 URLs are likely to make for too long a command line for the execve() system call though, so you'd need to split it, either with xargs or zsh's zargs, though here, combined with printf, xargs would be more handy:

printf 'https://example.com/%1$sx%1$s\0' {1..6000} | xargs -0 wget2 

(essentially the same as @JimL's except it uses the builtin printf instead of BSD jot and by using NUL as delimiter we allow the arguments to contain anything).

Or here using wget2's -i which expects a newline-delimited list (OK for a list of URLs which can't contain newlines):

printf 'https://example.com/%1$sx%1$s\n' {1..6000} | wget2 -i-