Rank the users you follow on Twitter by their Followcost (Bash)

Babylonian-neckbeard.txt

# This _will_ break if you follow over 500 accounts since it only does single requests.
# Also, I know nothing about parsing JSON; consider yourself warned.
# PS: If you follow over 500 accounts you are insane.
# Cross-platform: Linux! BSD! OSX! Solaris! Anything but Windows PowerShell!
# Requires: pcregrep (you'll need to install 'pcre' if you're on OSX)

## Part 1: Brain Pain & JSON

# Fetches user info, dumps it in following-json.txt if you want an aneurysm.
# Prints who you follow in following.txt then puts line-separated Followcost reports in userFC-json.txt
================================================================
curl -G 'https://api.twitter.com/1/friends/ids.json' -d 'screen_name=HEY_BABE_PUT_YOUR_USERNAME_HERE' | pcregrep -o "(?<=\[).+?(?=\])" | curl -G 'https://api.twitter.com/1/users/lookup.json' --data-urlencode user_id@- | tee following-json.txt | pcregrep -o --buffer-size 1m "(?<=\"screen_name\"\:\").+?(?=\")" | tee following.txt | sed -r 's/(.+)/url=http\:\/\/followcost.com\/\1.json/g' | curl -K- | sed -r 's/\}\{/\}\n\{/g' > userFC-json.txt

# Same thing, but you only get the Followcost JSON output.
================================================================
curl -G 'https://api.twitter.com/1/friends/ids.json' -d 'screen_name=HEY_BABE_PUT_YOUR_USERNAME_HERE' | pcregrep -o "(?<=\[).+?(?=\])" | curl -G 'https://api.twitter.com/1/users/lookup.json' --data-urlencode user_id@- | pcregrep -o --buffer-size 1m "(?<=\"screen_name\"\:\").+?(?=\")" | sed -r 's/(.+)/url=http\:\/\/followcost.com\/\1.json/g' | curl -K- | sed -r 's/\}\{/\}\n\{/g' > userFC-json.txt


## Part 2: Sorting!

# Sort by recent FollowCost
sort -grt: +8 -9 userFC-json.txt

# Sort by all-time FollowCost
sort -grt: +1 -2 userFC-json.txt

# Sort by @reply index
sort -grt: +7 -8 userFC-json.txt

# Print only usernames from the line-separated JSON - 'suitable' for filtering sort output.
# Eventually, I'll write an awful kludge to cut the crap of 'fat-free' JSON & show just names and one number.
SortedOutput | pcregrep -o "(?<=\"username\"\:\").+?(?=\")" 

Component snippets.txt

## Component Snippets

# Fetches who you follow, then resolves their UIDs, yielding a huge wad of JSON.
curl --get 'https://api.twitter.com/1/friends/ids.json' --data 'screen_name=HEY_BABE_PUT_YOUR_USERNAME_HERE' | pcregrep -o "(?<=\[).+?(?=\])" | curl --get 'https://api.twitter.com/1/users/lookup.json' --data-urlencode user_id@- > brainpain.txt

# Parse usernames out of the JSON blob. 1 per line.
pcregrep -o --buffer-size 1m "(?<=\"screen_name\"\:\").+?(?=\")" following-json.txt > following.txt

# Generate FC API URLs - the 'url=' part is needed because curl is a frustrating utility. (wget > curl)
sed -r 's/(.+)/url=http\:\/\/followcost.com\/\1.json/g' following.txt | curl -K- > userFC-json.txt

# Split FC JSON blob 1 user per line
sed -ri 's/\}\{/\}\n\{/g' userFC-json.txt