Counting default vs custom Gravatars

check-gravatar.sh

#!/bin/bash

set -euo pipefail

mkdir -p out

while read -r hash; do
    if [ ! -f "out/$hash" ]; then
        curl -I -s -o "out/$hash" "http://www.gravatar.com/avatar/$hash"
        echo "Fetched: $hash"
    else
        echo "Skipped: $hash"
    fi
done <email_hash_list.txt

count-gravatar.sh

#!/bin/bash

set -euo pipefail

TOTAL_COUNT=$(find out | wc -l | xargs)
echo "Total Gravatar files:"
echo "    $TOTAL_COUNT"

# Two known different default avatars with lengths of 2669 and 2637 bytes.
# Can check with:   ack -h "Content-Length" out/* | sort | uniq -c
DEFAULT_COUNT=$(ack -l "Content-Length: (2669|2637)" out/ | wc -l | xargs)
echo "Default Gravatars found:"
echo "    $DEFAULT_COUNT"

echo "Percentage custom Gravatar set vs default Gravatar:"
echo "    $(echo "scale=3; (1 - $DEFAULT_COUNT / $TOTAL_COUNT) * 100" | bc -l) (custom)"
echo "    $(echo "scale=3; $DEFAULT_COUNT / $TOTAL_COUNT * 100" | bc -l) (default)"

sample output over 10k real user emails

$ ./check-gravatar.sh
...
$ ./count-gravatar.sh
Total Gravatar files:
    10565
Default Gravatars found:
    8351
Percentage custom Gravatar set vs default Gravatar:
    21.000 (custom)
    79.000 (default)