Metadata-Version: 2.4
Name: trickkiste
Version: 0.3.5
Summary: Random useful stuff
Project-URL: Repository, https://github.com/frans-fuerst/trickkiste
Author-email: Frans Fürst <frans.fuerst+om-office@protonmail.com>
Requires-Python: <4,>=3.10.4
Requires-Dist: python-dateutil>=2.9.0.post0
Description-Content-Type: text/markdown

# trickkiste

`trickkiste` is a collection of convenience functionality I use quite often that
are too small to give them their own project on the one side but prone to
changes due to changes to the Python ecosystem to just copy&paste them and leave
them as they are.

So they are here now, enjoy.

Beware: most of this is not part of any standard library for a reason. That
reason might be there are better built in solutions which are harder (but more
flexible) to use or I just didn't find them back then. In most cases you're
advised to build them your own (to avoid dependencies and bugs).


## Content

Here is an overview of what's in here, see the docstings for a better
description (some even have a doctest!):

### misc

* `throw(BaseException)`: functional `raise` for closed expressions
* `md5from(Path)`: convenient wrapper returning an MD5 file checksum
* `cwd(Path)`: path changing context manager
* `process_output(cmd)`: short form for 'run a command and return it's output as one string'
* `asyncify`: decorator turning a sync method or function into an async one
* `async_chain`: like `chain` but async
* `async_filter`: like `filter` but async

* `date_str(timestamp, [datefmt])`: convenience wrapper for `strftime`
* `def date_from(timestamp: float | str)`: : convenience wrapper for `strptime`
* `dur_str(seconds, [fixed])`: short human readable representation of a duration
* `age_str(now, age, [fixed])`: like `dur_str` but with an age context
* `parse_age(string: str)`: inverse of `age_str`

* `smart_split(string: str, delimiter: str = ",")`: like `split` but parenthesis aware!
* `multi_replace(string: str, *substitutions: tuple[str, str])`: short form for chainging `replace`
* `compact_dict(mapping, [maxlen], [delim])`: create a short one-string map
* `def split_params(string: str)`:  inverse of `compact_dict`

### logging

* `apply_common_logging_cli_args(ArgumentParser)`: enrich a given parser to recognize default stuff like `--verbose`, etc.
* `setup_logging`: sets up a `rich` enhanced logger with useful stuff like call stack and posix thread id
* `logging` filters: internally used by `setup_logging` but can be re-used of course
    - `markup_escape_filter`: escapes (accidental) markup elements in the message
    - `thread_id_filter`: adds the (posix) thread id
    - `callstack_filter`: adds a short call stack
    - `logger_name_filter`: adds the logger name (`rich` edition)
    - `logger_funcname_filter`: adds the function name (`rich` edition)
    - `logger_levelname_filter`: adds the level name (`rich` edition)

### Textual stuff / TuiBaseApp

* `TuiBaseApp`: like `textual.App` but with convenience stuff mostly regarding logging
* `RichLogHandler`: a logging handler for a `textual` widget
* `LockingRichLog`: locks your log view scroll position until you scroll to the bottom
* `HeatBar`: a colorful heat bar widget for things like CPU monitoring

### std_suppress

`std_suppress` (with it's own shell entry point `suppress`) can be prefixed
to a command and acts like adding `>/dev/null 2>&1` but will print the output
if the command needs 'longer than expected' something bad happens (i.e.
nonzero exit).
Use it in scripts where you don't need the output to spam your log but you still
want to know if something unusual happend.


## License

See [License.md](License.md).


## Contribution

### Initialize

```bash
git clone https://github.com/frans-fuerst/trickkiste --recurse-submodules
cd trickkiste
uv run pre-commit install
```

### Manually run checks and fixes

Depending on your workflow / IDE you _might_ want to stay in charge over what's
being executed in order to check and fix code validity. For a number of
scenarios there is a bunch of scripts located in the `dev` submodule which just
wrap manual calls to what would be done by `pre-commit`. Some actually run
`pre-commit` for consistency, some don't in order to stay flexible regarding
command line arguments, e.g. for unit-testing using `pytest`.

*run all checks which would be executed on commit, but on unstaged stuff, too*
```bash
dev/check-validity
```

*do this in a loop (convenience wrapper around `inotifywait`)*
```bash
dev/watchandcheck
```

*run all checks separately*

```bash
dev/check-python-doctests
dev/check-python-formatting
dev/check-python-linting
dev/check-python-systemtests
dev/check-python-typing
dev/check-python-unittests
dev/check-shellscripts
dev/check-yaml-linting
```

*apply fixes for automatically fixable stuff*

```bash
dev/fix-python-formatting
dev/fix-python-linting
```

### Version bump and publishing

```bash
uv version --bump <patch|minor|major>
uv lock --upgrade
uv build

# do manual tests here

git commit -m"Bump version [..]"
git push
uv publish --token <TOKEN>
```