Rip threads from a Tapatalk forum into a Sqlite3 database
Tapatalk recently* took over Yuku, which, while it had it's problems, still looked like a proper PHPBB host. Then came Tapatalk, with its promises of being mobile friendly, which of course means "modern UX with MOAR WHITESPACE and gray-on-gray", hiding tons of details from users, 33 external stylesheets, 18 external JS files, and 5 tons of flax.
If you feel like moving on from such amazing software, you are (not) alone. This is a way to extract the HTML content of each forum post into a Sqlite database. Each post's content, sig, and metadata are stored separately, on the assumption that you're going to do something to move it into another forum. It might be on different software, but at least you have a DB to keep it organized, right?
Or are you just put off by the constant stream of 503s and 99 luftbaloon error messages? Do you want a backup of your forum in case it hits the fan yet again? This software also for you.
* for certain values of "recent". The appear to be eating other forum software companies because it's a low margin industry.
cpanm
installed, just do: cpanm Mojo::UserAgent Carp DBI Date::Manip
. Seriously Mojolicious' Mojo::DOM is great, I learned a lot from this project.cpanm
installed, you can cd
into the install directory and run:
cpanm --installdeps .cpan
interactive shell and
install Mojo::UserAgent
install Carp
install DBD::SQLite
install Date::ManipRead the first part of the file taparip.pl
. You'll need to specify
the URL of your site, where you want to save your database, what thread
ids you want to download, as a list or range. There's some more stuff,
but it's documented in the file.
If you're having trouble generating the schema file, you may have luck doing: sqlite3 --init schema.sql But this should happen automatically in most versions.
perl taparip.pl
Will download and rip from the forum based on the configuration you provided above, and fill the data into the database. Many painless, much simple wow.
perl taparip.pl thread1.html thread2.html ...
Instead of downloading, this will just extract the data from files you've already downloaded. If someone has already ripped the forum with other software, this is probably the way to go.
This only grabs textual content of threads, along with its sigs. It's in HTML format, not PHPBB format. Anything that's metadata about the posts or threads gets saved -- post time, number of times edited, author, thread, forum, etc.
Images and avatars are not downloaded. While each thread gets a forum ID, it does not save the names of forums. Nor does it gather thread-level information like last edited/posted, but hey, you can calculate that yourself from the post data. That's why I gave you the data in an RDBMS.
Well... OK, the Tapatalk ToS say that you are not allowed to harvest or scrape data from their website for any purpose not expressly allowed by them. They then state that you are solely responsible for backing up data, and WE ARE ALLCAPS TOTES INDEMNIFIED IF WE LOSE ANYTHING BECAUSE IT'S ALL UP TO YOU. Anyway IANAL so this is not legal advice but it sounds to me like they are expressly allowing it for backup purposes. Also it's not like they have a robots.txt that matches what their ToS says. Also also did you agree to these terms when you signed up for Yuku? Probably not. (Terms of Service: They make me violate them, no matter who they are.)
Brent Laabs bslaabs@gmail.com
MIT License
By using taparip, hereafter "the software", you agree to eat one (1) cookie.