parse_purchases.sh

#!/bin/bash

# This script retrieves ratings and rating counts from a user's Itch.io library.

i=1
rm -v json/my-purchases*.json
while true; do
echo "Reading purchases page $i."

# Use your web browser's dev tools to copy this link as curl (POSIX) address:
# https://itch.io/my-purchases?page=1&format=json
# Then insert the headers in the following command:
curl -s 'https://itch.io/my-purchases?page='$i'&format=json' ***headers go here*** >json/my-purchases$i.json

sleep 1
if grep '"num_items":0' json/my-purchases$i.json; then
rm -v json/my-purchases$i.json
break
fi
((i++))
done

# Notes
# - Using find because it can sort file names numerically. So file1, file2, file10 instead of file1, file10, file2.
# - The awk command prints unique lines only. So like sort -u, but without sorting.
cat $(find json -name "*.json") | tr -d '\\' | grep -o 'https://[^.]*\.itch\.io/[^/]*' | awk '!x[$0]++' > "url_list.txt"

rm -v parse_purchases.txt

for url in $(cat "url_list.txt")
do

pathname=$(echo "$url" | cut -d / -f 4)

mkdir -p "html"
filename="html/$pathname.html"

if test ! -e "$filename"; then
echo "$pathname"
wget -O "$filename" -q "$url"
sleep 1
fi

ratingCount=$(grep -oP 'ratingCount":\d*' "$filename" | cut -d : -f 2)
if test -z "$ratingCount"
then ratingCount="0"
fi

ratingValue=$(grep -oP 'ratingValue":"[^"]*' "$filename" | cut -d '"' -f 3)
if test -z "$ratingValue"
then ratingValue="0.0"
fi

echo -e "$ratingValue\t$ratingCount\t$pathname\t$url" >>parse_purchases.txt

done