This patch updates the notification thread to send all packets
through the notification plugins. The plugins themselves need to
do smart filter to not reply to every packet. This allows for
more interesting plugins.
Also fixed an issue with the messages tab in the admin ui, not
showing all of the recieved packets. The messages tab now also
sees all the packets that aprsd recieves.
This patch adds a new optional feature called Watch list.
Aprsd will filter IN all aprs packets from a list of callsigns.
APRSD will keep track of the last time a callsign has been seen.
When the configured timeout value has been reached, the next time
a callsign is seen, APRSD will send the next packet from that callsign
through the new notification plugins list.
The new BaseNotifyPlugin is the default core APRSD notify based plugin.
When it gets a packet it will construct a reply message to be sent
to the configured alert callsign to alert them that the seen callsign
is now on the APRS network.
This basically acts as a notification that your watched callsign list is
available on APRS.
The new configuration options:
aprsd:
watch_list:
# The callsign to send a message to once a watch list callsign
# is now seen on APRS-IS
alert_callsign: NOCALL
# The time in seconds to wait for notification.
# The default is 12 hours.
alert_time_seconds: 43200
# The list of callsigns to watch for
callsigns:
- WB4BOR
- KFART
# Enable/disable this feature
enabled: false
# The list of notify based plugins to load for
# processing a new seen packet from a callsign.
enabled_plugins:
- aprsd.plugins.notify.BaseNotifyPlugin
This patch also adds a new section in the Admin UI for showing the
watch list and the age of the last seen packet for each callsing since
APRSD startup.
This patch adds the dumping out of a flattened config to the log
at startup. This is helpful for seeing what aprsd server is actually
using for config entries at startup and since it's in the log, you can
reference it.
This patch adds usage of update_checker to check to make sure the
version of APRSD being launched is the latest version. Also added a
call to upate_checker as part of the KeepAlive thread. It will
call update_check every hour. If there is no aprsd connectivitity,
the update check will silently fail.
This patch reworks the stats object dict and includes more data.
Also includes aprsis last update timestamp (from last recieved message).
This is used to help determine if the aprsis server connection is still
alive and well.
This patch adds the aprsd-lnav.json formatting file.
This is useful when you want to tail the logfile with the lnav
log tailing app.
http://lnav.org/
To install the aprsd-lnav.json formatter
1) install lnav
2) lnav -i aprsd-lnav.json
3) lnav -C -- just to test it out
The next time you launch aprsd do it with this
aprsd server --loglevel DEBUG | lnav
This patch also updates the logging output from the flask
web service to 1) disable flask web url logging and 2)
use the same output format as the rest of the app.
This patch moves the default log format string and date format string
to the config file, so users can format the logs as they see fit.
The default log format also includes the file and line number that
posted the log entry.
The new entries in the config are here:
aprsd:
logformat: "String here"
dateformat: "string here"
This patch fixes the CTRL-C signal_handler.
This patch also adds the new Messages WEB UI page
as well as the save url, which are both behind an
http basic auth.
The flask web service now has users in the config file
aprsd:
web:
users:
admin: <password>
This patch adds the stats object to collect statistics of
the running server. This also optionally adds the ability
to run a flask web service on a port to use as a keepalive
healthcheck.
This patch reorganizes the config file layout and options
to make more logical sense as well as make it more readable.
This breaks backwards compatibility.
This patch adds the openweathermap weather plugin.
Also adds a new config option to set the overall
units setting from imperial (default) to metric.
to change it add the following to the ~/.config/aprsd/aprsd.yaml
...
aprsd:
units: metric
This patch adds 2 new time plugins to allow admins to use their
opencagedata APIkey or openweathermap API key to fetch the timezone
from the lat/lon GPS coordinates for the callsign requesting the time.
This will enable fetching the time local to the ham radio's last beacon,
and not time local to the aprsd server instance running. If the
location is not found, then the timezone will default to UTC.
The 2 new plugins are
- aprsd.plugins.time.TimeOpenCageDataPlugin
Fetches timezone from lat/lon using the opencagedata api that can be
found here: https://opencagedata.com/dashboard#api-keys
This requires a new ~/.config/aprsd/aprsd.yml entry to specify the
api key.
opencagedata:
apiKey: <the api key hash here>
- aprsd.plugins.time.TimeOWMPlugin
Fetches the timezone from lat/lon using the openweathermap api
that can be found here: https://home.openweathermap.org/api_keys
This requires a new ~/.config/aprsd/aprsd.yml entry to specify the
api key.
openweathermap:
apiKey: <the api key hash here>
This patch adds a new add_config_comments() function in utils.py
that allows you to insert a comment string in a raw_yaml string
that's already been created from the yaml.dump() call.
This patch adds the new required aprs.fi api key. This key is used
by 2 of the core plugins, locationPlugin and weatherPlugin.
You must set the apiKey in the config, or aprsd won't start.
This commit adds the new send-message command for sending messages.
This also redoes the logging of sent/rx'd packets to a single method
which is syncrhonized, so we don't get intermixed log messages for
packets.
Also adds email address validation during startup, and
optionally disables the validation via a command line switch. without
email validation for production running aprsd, emails sent can turn up
garbage and cause issues when those emails are received by aprsd
message processing as invalid content.
This branch refactors the majority of main.py out into individual
modules to compartmentalize the code. Migrated all email related
features unti email.py, sending of messages into messaging.py
Also refactored all of the socket code to use aprslib for all APRS-IS
communication as well as message/packet processing.
Moved the email command into it's own Plugin.
This patch adds the new APRSD Command Plugin architecture.
All Comand plugins must implement the same object API, which includes
plugin object is subclass of APRSDPluginBase
version attribute
command_regex attribute
command method
When an APRS command is detected, then the regex is run against
the command. If the command_regex matches, then the plugin's
command() method will be called. If the command() method returns
a string, then that string is sent as a reply to the APRS caller.
A new aprs.yml config section is added to support selecting
which plugins to enable.
If you want all plugins enabled, then omit "enabled_plugins" entirely
from the aprs section of the config.
To load custom plugins:
1) create a directory with an __init__.py file
2) Add a plugin.py file that contains your plugin
Look at the exmaples directory for an example plugin.
This patch includes lots of changes to tox environment for
automatically detecting pep8 failures, which can cause python2 vs
python3 failures after install.
The following tox commands have been added
tox -efmt-check - This checks the python syntax and formatting
tox -efmt - Automatically fixes python syntax formatting that
fmt-check complains about.
tox -etype-check - check on types
tox -elint - flake8 run
This patch also changes where the default config file is located.
The new location is ~/.config/aprsd/aprsd.yml
You can now also specify a custom config file on the command line
with the -c or --config option as well.
This patch adds support for tox. Tox is used to run various
python compliance tests. This enables pep8 tests, as well as python2
and python3 compatibility as well as coverage and documentation
building.
This patch reads the SMTP settings from the config.yml now.
Also added a logfile entry to the aprs: section of the config.yml
so the logfile can be placed anywhere.
This patch adds the port to the aprs: section of the config.yml
as well as fixes a possible issue with the user telnet auth command
where the user is a string and the port is an int. python can't
concatonate a string with an int.
This patch completes the migration to using a config.yml file.
~/.aprsd/config.yml is now required and all options for callsign,
imap, aprs user, passwords are in the config. If there is no existing
~/.aprsd/config.yml file, then the app will output a sample config
and exit.
This patch also adds a global logging facility that allows logging all
commands to aprsd.log as well as stdout. You can disable logging to
stdout by adding --quiet on the command line. You can specify the log
level with --loglevel INFO. By default the log level is DEBUG.
This patch also updates some formatting issues and small refactoring
to ensure that the logging facility and config is read prior to starting
any network connections and/or services.
This patch adds support to read a ~/.aprsd/config.yml file.
If one doesn't exist, it puts out an example yaml string to stdout
that can be copied into a file and edited.
Since this patch adds a new external requirement (pyyaml) you need
to re-install the app for dev with
pip install -e .
This patch does some refactoring of the code and the directory
structure to conform to the needs of a pypi project.
The python code now lives in the aprsd directory so it acts like a real
python package that can be installed/included/used.
The aprsd.py is now aprds/main.py
This patch also adds support for using pbr, which enables a consistent
bin install that you can then call as 'aprsd' from the command line.
To use this as a developer you should create a virtualenv
virtualenv .venv
source .venv/bin/activate
pip install -e .
now you can edit the aprds/main.py and then test it by immediately
running aprsd from the command line.
The -e option for pip allows you to install the package as an editable
package in the .venv, so you can hack on it and not need to re-install
every time you make a change.