vetusware.sh

#!/bin/bash

# This is just a temporary script to parse a range of pages from
# vetusware.com

set -eo pipefail

clear

out_f="${HOME}/vetusware_links-${RANDOM}.txt"
out_f_sort="${HOME}/vetusware_links-sorted-${RANDOM}.txt"

touch "$out_f" "$out_f_sort"

declare -A regex

regex[type]='<p class=\"item_type\"><a href=\".*\">.*<\/a>.*<a href=\".*\">.*<\/a><\/p>.*<p class=\"item_lang\">.*<\/p>'
regex[desc]='^.*<p class=\"item_description\">(.*)<\/p>.*<p><a href=\"(https:\/\/vetusware.com\/download\/.*)\">download<\/a><\/p>.*$'
regex[link]='^.*<p><a href=\"(https:\/\/vetusware.com\/download\/.*)\">download<\/a><\/p>.*$'
regex[name]='^.*<h3>.*<strong>(.*)</strong>.*</h3>.*$'

n_line=0

for n in {1..30}; do
    tmp_f="/dev/shm/vetusware_p${n}-${RANDOM}.txt"

    curl -s -o "$tmp_f" "https://vetusware.com/category/OS/?cat=1&page=${n}"

    while read line; do
        (( n_line = n_line + 1 ))

        if [[ $line =~ ${regex[type]} ]]; then
            line_name=$(( n_line + 1 ))
            line_desc=$(( n_line + 2 ))

            sed -Ei "${line_name}s|${regex[name]}|name: \1|" "$tmp_f"

            line_desc_tmp=$(sed -n "${line_desc} p" "$tmp_f")

            if [[ $line_desc_tmp =~ ${regex[desc]} ]]; then
                sed -Ei "${line_desc}s|${regex[desc]}|desc: \1 link: \2|" "$tmp_f"
            else
                sed -Ei "${line_desc}s|${regex[link]}|link: \1|" "$tmp_f"
            fi

            sed -n "${line_name},${line_desc} p" "$tmp_f" >> "$out_f"
        fi
    done <"$tmp_f"

    rm "$tmp_f"
    n_line=0
done

regex[name]='^name: '
regex[desc]='^desc: '
n_line=0

while read line; do
    (( n_line = n_line + 1 ))

    if [[ $line =~ ${regex[name]} ]]; then
        line_desc=$(( n_line +1 ))

        sed -n "${n_line},${line_desc} p" "$out_f" | grep -iqF -e source -e dos -e 'os/2' -e windows

        if [[ $? -eq 0 ]]; then
            sed -n "${n_line},${line_desc} p" "$out_f" >> "$out_f_sort"
            echo >> "$out_f_sort"
        fi
    fi
done <"$out_f"

rm "$out_f"