A data live-capture tool for Twitter.
Uses the Twitter API to live capture tweets that match certain keywords. It dumps these tweets into an SQLite database, and optionally prints them to STDOUT.
tweetcap.pl [--config path] --db path --keywords keywords [--monitor] [--no-rt]
--config
Use this to specify a different authentication file.
If not specified, tweetcap will use 'auth.conf' in the current directory.
Authentication files should return a Perl hash containing OAuth/Twitter API keys and tokens. You need to get these yourself.
The file should look like this:
return
{
token => "<FILL THIS IN>",
token_secret => "<FILL THIS IN>",
consumer_key => "<FILL THIS IN>",
consumer_secret => "<FILL THIS IN>"
};
--db
Specifies the SQLite database. Currently mandatory. Set this to something in /tmp if you really don't need it.
If the database file doesn't exist, it will be created and the table initialised.
If the database does exist, all Tweets will be appended to the end of the table.
--keywords
A list of keywords to filter.
Can be one word, or a list of terms. Must be in quotes.
--monitor
Print Tweets to STDOUT in realtime.
--no-rt
Omit quote retweets (i.e. any tweet that begins with 'RT .)
Where do I even start.
- The main problem is with outputting wide characters in Tweets, inasmuch as it doesn't, at least not well.