This a little brother database module and supporting tools for using the addresses stored in mutt's header cache as an address book.
The motivation for doing this is that to push emails through lbdb-fetchaddr you need to pull the whole email from the server. This is fine if you are using offlineimap or a local Maildir, but if you are using imap normally, it will mean pulling a lot of data. Since mutt already pulls this data to build the index view, and caches it, you can extract the information from there.
- python2.7
- pytc, Python bindings for tokyo cabinet
To install:
$ sudo python setup.py install
To use, first you need to build the database.
$ lbdb-hcache-extract-addresses
As long as the mutt header cache is in the default place, this will be all you need. If you have to specify another location, use the --hcache option.
To use the lbdb, you need to tell it where it can find the lbdb-hcache module and tell it to use the lbdb-hcache method. The path to the lbdb-hcache should be
{sys.prefix}/share/lbdb-hcache/lbdb_modules.
Most likely this will be
/usr/local/share/lbdb-hcache/lbdb_modules.
Add this to $MODULES_PATH in ~/.lbdbrc and tell it to use the lbdb-hcache method.
MODULES_PATH="$MODULES_PATH /usr/local/share/lbdb-hcache/lbdb_modules"
METHODS="m_hcache"
SORT_OUTPUT=no
It is advisable to disable sorting on the output, so that the most commonly seen addresses are shown at the top.
If the database is in a location different to the default, this can be specified with HCACHE_ADDRESS_DB in ~/.lbdbrc.
This package was developed against the debian mutt package. As such, the offsets in the tokyocabinet entries are those used by this package. In particular, the size of the HEADER structure depends on whether USE_POP, USE_IMAP and MIXMASTER. In debian they all are.
Additionally, the number of fields to read for each address changes whether EXACT_ADDRESS is defined. In debian it is not.