Stat based reloader.
The reloading is triggered based on the difference of the modified status times of the watched files. For this, we need to know how many files we need to track as well as how often we should track them for changes. Once we have this information we can then calculate the modified time and later trigger the reloading.
reloader = StatReloader(interval=5.0)
reloader.run()
# Create a reloader context instance
reloader = StatReloader()
with reloader:
reloader.run()
Note
The current implementation is based on the stat based changes
detected by the stat(). This can be further
improved using the watchdog.observer. But it seems that it
increases the overhead. Whereas this implementation is minimal
and roughly works the same.
Changed in version 1.0.0: In previous development version “track”, “ignore_patterns” and
“ignore_dirs” were used as means of named arguments for
controlling the behavior of the StatReloader. Now
added support for kwargs instead.
Restart the execution in a new Python interpreter with same arguments.
Run the function in an independent Python interpreter.
Live debugging decorator.
Decorate function that needs to be reloaded on change. This function
internally calls run_with_reloader().
func – Function to run with live reloading.
track – Iterable of absolute paths of both python & non-python files to track. Defaults to None.
ignore_patterns – Iterable of path patterns to ignore. The patterns can be regular expressions or just names of the file(s) and directories to be ignored. Defaults to None.
ignore_dirs – Directories to skip while watching. This does not expect the full absolute path of the directories, only the directory names will do. Defaults to None.
interval – Seconds to sleep between reloading. The less the duration, more aggressively it will track for changes. Defaults to 1.0 sec.
@debug
def func(*args, **kwargs):
...
@debug(track=["/home/.bashrc", "/home/.bash_profile"])
def func(*args, **kwargs):
...
Note
Well, you can add non-python files to track but doing so might increase the CPU utilization. It is more advisable to track only python files e.g. py, pyc or pyw.