A daemon for monitoring filesystem events with inotify on Linux and run tasks like filesystem operations (copy, move or delete), a shell command or custom async python methods.
It is possible to schedule tasks with a delay, which can then be canceled again in case a canceling event occurs. A useful example for this is to run tasks only if a file has not changed within a certain amount of time.
pyinotifyd offers great flexibility through its dev-op configuration approach, which enables you to do almost anything you want.
# install pyinotifyd with pip
pip install pyinotifyd
# install service files and config
pyinotifyd --install
# uninstall service files and unmodified config
pyinotifyd --uninstallThe following init systems are supported.
# start the daemon at boot time
systemctl enable pyinotifyd.service
# start the daemon immediately
systemctl start pyinotifyd.service# start the daemon at boot time
rc-update add pyinotifyd default
# start the daemon immediately
rc-service pyinotifyd start The config file /etc/pyinotifyd/config.py is written in python syntax. pyinotifyd reads and executes its content, that means you can write your custom async python methods directly into the config file.
The basic idea is to instantiate one or multiple schedulers and map specific inotify events to schedulers with the help of event maps. Then, watch the given paths for events and run tasks as defined in the event maps.
pyinotifyd has different schedulers to schedule tasks with an optional delay. The advantages of using a scheduler are consistent logging and the possibility to cancel delayed tasks. Furthermore, schedulers have the ability to differentiate between files and directories.
Schedule a custom python method job with an optional delay in seconds. Skip scheduling of tasks for files and/or directories according to files and dirs arguments. If there already is a scheduled task, re-schedule it with delay. Use logname in log messages. All additional modules, functions and variables that are defined in the config file and are needed within the job, need to be passed as dictionary to the TaskManager through global_vars. If you want to limit the scheduler to run only one job at a time, set singlejob to True.
All arguments except for job are optional.
# Please note that pyinotifyd uses pythons asyncio for asynchronous task execution.
# Do not run anything inside the custom python method that blocks the daemon.
#
# Bad: time.sleep(10)
# Good: await asyncio.sleep(10)
import asyncio
import logging
async def custom_job(event, task_id):
await asyncio.sleep(10)
logging.info(f"{task_id}: execute example task: {event}")
task_sched = TaskScheduler(
job=custom_job,
files=True,
dirs=False,
delay=0,
logname="sched",
global_vars=globals(),
singlejob=False)Schedule a shell command cmd. Replace {maskname}, {pathname} and {src_pathname} in cmd with the actual values of occuring events. This scheduler is based on TaskScheduler and has the same optional arguments.
# Please note that **{src_pathname}** is only present for IN_MOVED_TO events and only
# in the case where the IN_MOVED_FROM events are watched too.
# If it is not present, the command line argument will be an empty string.
shell_sched = ShellScheduler(
cmd="/usr/local/bin/task.sh {maskname} {pathname} {src_pathname}")Move, copy or delete files and/or directories following the list of rules, the first matching rule is executed.
This scheduler is based on TaskScheduler and has the same optional arguments.
A rule holds an action (move, copy or delete) and a regular expression src_re. The action will be executed if src_re matches the path of an event. In case where action is copy or move, use dst_re as destination path. Subgroups and/or named-subgroups may be used in src_re and dst_re.
Automatically create possibly missing sub-directories if auto_create is set to True. Set the mode and ownership of moved or copied files/directories and newly created sub-directories to filemode and dirmode. Override destination files if override is set to True.
If action is delete, delete non-empty directories if rec is set to True.
move_rule = FileManagerRule(
action="move",
src_re="^/src_path/(?P<path>.*).to_move$",
dst_re="/dst_path/\g<path>",
auto_create=False,
rec=False,
filemode=None,
dirmode=None,
user=None,
group=None,
override=False)
delete_rule = FileManagerRule(
action="delete",
src_re="^/src_path/(?P<path>.*).to_delete$",
rec=False)
file_sched = FileManagerScheduler(
rules=[move_rule, delete_rule])Map specific events to one or multiple schedulers. Ignore the event if the scheduler is set to None. Use Cancel to cancel a scheduled task within a scheduler.
This is an example which schedules tasks for newly created files if they are not modified, moved or deleted within the delay time of the scheduler.
event_map = {
"IN_ACCESS": None,
"IN_ATTRIB": None,
"IN_CLOSE_NOWRITE": None,
"IN_CLOSE_WRITE": task_sched,
"IN_CREATE": task_sched,
"IN_DELETE": Cancel(task_sched),
"IN_DELETE_SELF": Cancel(task_sched),
"IN_IGNORED": None,
"IN_MODIFY": Cancel(task_sched),
"IN_MOVE_SELF": None,
"IN_MOVED_FROM": Cancel(task_sched),
"IN_MOVED_TO": task_sched,
"IN_OPEN": None,
"IN_Q_OVERFLOW": None,
"IN_UNMOUNT": Cancel(task_sched)}
# It is possible to instantiate an event map with a default scheduler set for every event
event_map = EventMap(default_sched=task_sched)The following events are available:
- IN_ACCESS: a file was accessed
- IN_ATTRIB: a metadata changed
- IN_CLOSE_NOWRITE: an unwritable file was closed
- IN_CLOSE_WRITE: a writable file was closed
- IN_CREATE: a file/directory was created
- IN_DELETE: a file/directory was deleted
- IN_DELETE_SELF: a watched item itself was deleted
- IN_IGNORED: raised when a watch is removed, probably useless for you
- IN_MODIFY: a file was modified
- IN_MOVE_SELF: a watched item was moved, currently its full pathname destination can only be known if its source and destination directories were both watched. Otherwise, the file is still being watched but you cannot rely anymore on the given path attribute event.path
- IN_MOVED_FROM: a file/directory in a watched directory was moved from another specified watched directory. Can trace the full move of an item when IN_MOVED_TO is available too, in this case if the moved item is itself watched, its path will be updated (see IN_MOVE_SELF)
- IN_MOVED_TO: a file/directory was moved to another specified watched directory (see IN_MOVE_FROM)
- IN_OPEN: a file was opened
- IN_Q_OVERFLOW: the event queue overflown. This event is not associated with any watch descriptor
- IN_UNMOUNT: when backing filesystem was unmounted. Notified to each watch of this filesystem
pyinotifyd requires you to define a variable called pyinotifyd within the config file, which contains an instance of the Pyinotifyd class. Set the optional list of watches and the shutdown_timeout. Pyinotifyd will wait shutdown_timeout seconds for pending tasks to complete before shutdown. Use logname in log messages.
pyinotifyd = Pyinotifyd(
watches=[],
shutdown_timeout=30,
logname="daemon")A watch connects the path to an event_map. Automatically add a watch on each sub-directories in path if rec is set to True. If auto_add is True, a watch will be added automatically on newly created sub-directories in path. All events for paths matching one of the regular expressions in exclude_filter are ignored. If the value of exclude_filter is a string, it is assumed to be a path to a file from which the list of regular expressions will be read.
# Add a watch directly to Pyinotifyd.
pyinotifyd.add_watch(
path="/src_path",
event_map=event_map,
rec=False,
auto_add=False,
exclude_filter=["^/src_path/subpath$"])
# Or instantiate and add it
w = Watch(
path="/src_path",
event_map=event_map,
rec=False,
auto_add=False,
exclude_filter=["^/src_path/subpath$"])
pyinotifyd.add_watch(watch=w)Pythons logging framework is used to log messages (see https://docs.python.org/3/howto/logging.html).
The following loglevels are available:
- DEBUG
- INFO
- WARNING
- ERROR
- CRITICAL
# Configure global loglevel
setLoglevel(INFO)
# Configure loglevel per logname.
setLoglevel(INFO, logname="daemon")Send log messages to the local syslog server.
# Enable logging to local syslog server (/dev/log).
# Use *address* to specify a different target.
enableSyslog(loglevel=INFO, address="/dev/log")
# Enable syslog per logname
enableSyslog(lglevel=INFO, name="daemon")import logging
async def custom_job(event, task_id):
logging.info(f"{task_id}: execute example task: {event}")
task_sched = TaskScheduler(
job=custom_job,
files=True,
dirs=True)
event_map = EventMap(
default_sched=task_sched)
pyinotifyd = Pyinotifyd()
pyinotifyd.add_watch(
path="/src_path",
event_map=event_map,
rec=True,
auto_add=True)shell_sched = ShellScheduler(
cmd="/usr/local/sbin/task.sh {pathname}",
files=True,
dirs=False)
event_map = {
"IN_WRITE_CLOSE": shell_sched}
pyinotifyd = Pyinotifyd()
pyinotifyd.add_watch(
path="/src_path",
event_map=event_map,
rec=True,
auto_add=True)move_rule = FileManagerRule(
action="move",
src_re="^/src_path/(?P<path>.*)\.to_move$",
dst_re="/dst_path/\g<path>",
auto_create=True,
filemode=0o644,
dirmode=0o755)
copy_rule = FileManagerRule(
action="copy",
src_re="^/src_path/(?P<path>.*)\.to_copy$",
dst_re="/dst_path/\g<path>",
auto_create=True,
filemode=0o644,
dirmode=0o755)
delete_rule = FileManagerRule(
action="delete",
src_re="^/src_path/(?P<path>.*)\.to_delete$",
rec=False)
file_sched = FileManagerScheduler(
rules=[move_rule, copy_rule, delete_rule],
delay=60,
files=True,
dirs=False)
event_map = {
"IN_CLOSE_WRITE": file_sched,
"IN_CREATE": file_sched,
"IN_DELETE": Cancel(file_sched),
"IN_DELETE_SELF": Cancel(file_sched),
"IN_MODIFY": Cancel(file_sched),
"IN_MOVED_FROM": Cancel(file_sched),
"IN_MOVED_TO": file_sched,
"IN_UNMOUNT": Cancel(file_sched)}
# Please note that the shutdown timeout should be greater than the greatest scheduler delay,
# otherwise pending tasks may get cancelled during shutdown.
pyinotifyd = Pyinotifyd(shutdown_timeout=35)
pyinotifyd.add_watch(
path="/src_path",
event_map=event_map,
rec=True,
auto_add=True)