Twitter markov thing for OlegDB.

Generating Corpus

I scraped all of the feeds I wanted by hand with chrome and just saved them in a file. Then:

./prepare_html.sh
./prepare_step2.py && wc -l prepared3.txt && cat prepared3.txt | sort | uniq > prepared4.txt
cp prepared4.txt redone_final.txt

hashtags.txt was generated with the following:

grep -o '#\w\+' prepared2.html | sort | uniq -c | sort -g -r > hashtags.txt

Running

You'll need to put your sick twitter creds in local_settings.py. Then you can just run main.py and it'll tweet for you. If you just want to see some example output, set debug=True in main.py and it'll print what it wants to tweet.

What is it?

This is a twitter bot designed to market OlegDB in the best way we know how: By generating markov text from the tweets of other database projects! You can see the corpus in redone_final.txt.