usage

Open a terminal and start a VIM with servername 'PDB':

$ mvim --servername PDB

Then, from the program to be debugged use it like pdb,

>>> import vimpdb; vimpdb . set_trace()

(the blanks above are needed due to some dumb SVN precommit hook)

When this line is hit, PDB starts and your PDB-VIM should load the source code and highlight the current LOC. Additionally you should get a "PDB" menu, where you can set/toggle breakpoints etc.

how does (should) that work?

vimpdb is a subclass of pdb.Pdb, which hijacks the command loop and listens for UDP packets on port 6666 (blocking). To inform the VIM process what's happening, it uses VIM's remote commands.

On the VIM side there's basically just a VIM script which does the higlight, set up the menu, and defines a function to send UDP messages.

That's all.

bugs

• sometimes the vim script is not sourced correctly
• there are no easy keyboard commands for using the debugger, one needs to use the menu.
• no security
• network settings hard coded