Finalized dblp and tested the entire pipeline.

# Convert XML to bracket notation.

./dblp_to_bracket.py

# Remove 'www' entries.

awk '!/{www{key{homepages/' dblp.bracket > dblp_no_www.bracket

# Sort the dataset.

./../utilities/sort_dataset.sh dblp.bracket

./../utilities/sort_dataset.sh dblp_no_www.bracket > dblp_no_www_sorted.bracket

# Tidy up.

# rm *xml*

# rm *.dtd

rm *xml*

rm *.dtd

rm dblp.bracket

rm dblp_no_www.bracket

#!/bin/bash

# Delete all downloaded files.

rm *xml*

rm *.dtd

# NOTE: We substract the escaped brackets because they're part of node labels.

# 

cat $input | awk '{print gsub("{","{")-gsub("\\\\{","\\{"), $0}' | \

sort -n --buffer-size=4G | cut -d' ' -f2-

sort -n --buffer-size=10G | cut -d' ' -f2-