Python

• lambdas suck
  
  • statements, not expressions, you are essentially restrictet to a single line
    
  • no local variables in lambdas (although with walrus operation it's a bit better now)
    
  • http://math.andrej.com/2009/04/09/pythons-lambda-is-broken
    
• scoping (i.e. only local and global scopes)
  
• GIL makes the concurrency very annoying at times
  
• no static typing (although it's kinda okay now with annotations and #mypy)
  

related #repl

That’s a part of python’s data model like `len`, `reversed`, `iter`, …:
`datetime` simply defines [`datetime.__format__`](https://docs.python.org/3/library/datetime.html#datetime.date.__format__). The method [is explained here](https://docs.python.org/3/reference/datamodel.html#object.__format__).

https://barahilia.github.io/blog/computers/2017/01/04/to-yield-or-not-to-yield.html

Finally we came to this. To me it is more important that generators make code more elegant. Performance is usually relevant to smaller hot spots.
The idea is simple: in comparison to list interface where computation results are first saved and are returned only at the end, “yielding” could reach faster code and reduce memory usage. As a quick simulation we may use %timeit magic command in IPython with the following two functions:

interpreter – about 10ms
lazy imports
profile slow imports

process remaining arguments

attrs – could be interesting… might solve stupid namedtuple addition etc.
https://github.com/python-attrs/attrs

This gives you the power to use actual classes with actual types in your code instead of confusing tuples or confusingly behaving namedtuples. Which in turn encourages you to write small classes that do one thing well. Never again violate the single responsibility principle just because implementing init et al is a painful drag.

https://twitter.com/isenbaev/status/1108213595920166912

https://stackoverflow.com/questions/27435284/multiprocessing-vs-multithreading-vs-asyncio-in-python-3-4

if io_bound:
    if io_very_slow:
        print("Use Asyncio")
    else:
        print("Use Threads")
else:
    print("Multi Processing")

https://github.com/adamchainz/patchy#patchy

Patch the inner source of python functions at runtime.

https://news.ycombinator.com/item?id=17819138

scrapy shell

with_secrets scrapy shell

from scrapy.http import FormRequest
request=FormRequest(url='http://www.quantified-mind.com/login',formdata={'Email': USER,'Password':PASSWORD,})
fetch(request) – redirects by default

sets response variable

you can do view(response) to view in browser
mind that it might be different from browsing in chrome. because of user agent?

• @contextmanager; prepare; yield exactly once, then shutdown
  
• contextlib.redirect_stdout
  meh. I think it patches sys module? so it redirects sys.stdout calls only; but not all stdout
  

def fname():
import inspect
return inspect.stack()[1][3]

Marshmallow “converts” (deserializes) dicts to SQLAlchemy models or serializes SQLAlchemy models to dicts.
SQLAlchemy is an ORM. It maps database schema (tables) and data to Python objects.
The two packages complement each other. They cannot and do not replace each others’ functionality.
Flask-Marshmallow gives you a convenient interface to Marshmallow under Flask.

https://github.com/karlicoss/dominatepp/blob/master/.travis.yml
hmm, minimalistic travis is not so bad..

https://stackoverflow.com/questions/2295290/what-do-lambda-function-closures-capture

Scoping in Python is dynamic and lexical. A closure will always remember the name and scope of the variable, not the object it's pointing to.

https://docs.python.org/3/library/argparse.html

metavar - A name for the argument in usage messages.

https://python-rapidjson.readthedocs.io/en/latest/benchmarks.html
ok, so loads speedups are not that significant for my purposes

https://stackoverflow.com/questions/12151306/argparse-way-to-include-default-values-in-help/18507871#18507871

Add '%(default)' to the help parameter to control what is displayed.

https://tox.readthedocs.io/en/latest/example/basic.html#integration-with-setup-py-test-command

As the python eco-system rather moves away from using setup.py as a tool entry point it’s maybe best to not go for any setup.py test integration.

(while debugging telegram2org)

https://github.com/PyCQA/pyflakes

Pyflakes makes a simple promise: it will never complain about style, and it will try very, very hard to never emit false positives.
Pyflakes is also faster than Pylint or Pychecker. This is largely because Pyflakes only examines the syntax tree of each file individually. As a consequence, Pyflakes is more limited in the types of things it can check.

https://tox.readthedocs.io/en/latest/example/general.html#interactively-passing-positional-arguments

## Print with @padding
`print("{:>d}{}".fomat(i, a[i]))` prints i aligned to at least two spaces as a decimal.

## Infinite dict
~~~ { .python }
infinite_defaultdict = lambda: defaultdict(infinite_defaultdict)
d = infinite_defaultdict()
d['x']['y']['z'] = 10
~~~

## Web server
to start a webserver to serve files from your current dir: python -m SimpleHTTPServer

@enums

## Enums
from enum import Enum
class Fruit(Enum):
    APPLE = 0
    ORANGE = 1

## Generating @random color
color = "#%06x" % random.randint(0, 0xFFFFFF)


python setup.py install --force

#singleton
@singleton
def singleton(cls):
    instances = {}

    def get_instance():
        if cls not in instances:
            instances[cls] = cls()
    return instances[cls]

return get_instance
TODO tbh, easier with lru_cache(1)?

https://barahilia.github.io/blog/computers/2017/01/04/to-yield-or-not-to-yield.html
hmm. if I decorate function with list constructor, I can get both ease of use and yield in function's body

The best answer in 2020 is to use poetry.
http://poetry.eustace.io

One particularly effective way of handling sub-commands is to combine the use of the add_subparsers() method with calls to set_defaults() so that each subparser knows which Python function it should execute. For example:
This way, you can let parse_args() do the job of calling the appropriate function after argument parsing is complete. Associating functions with actions like this is typically the easiest way to handle the different actions for each of your subparsers. However, if it is necessary to check the name of the subparser that was invoked, the dest keyword argument to the add_subparsers() call will work:

2020 code, no additional packages required, Python 3

https://github.com/pypa/pip/issues/5566#issuecomment-402417467

We disabled the ability to use PEP 508 URL requirements in dependencies, since we don't want that a package installed from PyPI to result in the pip reaching out to an arbitary web URL. Basically, pip install spam should not make pip reach out to anything except PyPI.

https://github.com/pypa/pip/issues/6301

ugh. so I can't have URL dependencies if it's meant to be on pypi???

https://github.com/rstojnic/lazydata

py, more-itertools, zipp, importlib-metadata, pluggy, six, pyparsing, packaging, atomicwrites, wcwidth, attrs, pytest, orger

https://pypi.org/project/backoff/

It is also possible to specify an alternate logger with the logger keyword argument. If a string value is specified the logger will be looked up by name.

We have moved our dependency management from pipenv to poetry and hasn’t been more happier than now working with and publish Python modules.
Now we use pyenv with poetry to run all our projects in Python with venv in project instead of decided by poetry default settings. We changed the settings with:
poetry config settings.virtualenvs.in-project true After this in hgignore or gitignore added .venv folder.
Now everything is the way we wanted, with Pipenv many times faced issues that the project worked with requirements.txt but not in Pipenv. Didn’t have this issue moving to poetry yet.
Also with emacs support and automated deployment tools the predictability of poetry made our team easy to work with and publish Python application modules.

https://github.com/python-poetry/poetry/issues/140

right, poetry can't infer version...

https://stackoverflow.com/questions/37353960/pytest-exits-with-no-error-but-with-collected-0-items

pytest gathers tests according to a naming convention. By default any file that is to contain tests must be named starting with test_ and any function in a file that should be treated as a test must also start with test_.

https://docs.python.org/3/library/traceback.html#tracebackexception-objects

TracebackException Objects

hmm, could use that?

https://stackoverflow.com/questions/19991445/run-an-ols-regression-with-pandas-data-frame

I think you can almost do exactly what you thought would be ideal, using the statsmodels package which was one of pandas' optional dependencies before pandas' version 0.20.0 (it was used for a few things in pandas.stats.)

https://github.com/vinta/awesome-python

urwid - A library for creating terminal GUI applications with strong support for widgets, events, rich colors, etc.

import logging
logging.basicConfig()

def bad():
    raise RuntimeError('BAD')

xx = RuntimeError('while running "test"')
ex = None
try:
    bad()
except Exception as e:
    ex = e
    ex.__cause__ = xx
    logging.error('inside!')
    logging.exception(ex)

logging.error('outside! note the empty stacktrace :(')
logging.exception(ex)
logging.error("also __cause__ is ignored even though it's present: '%s'", ex.__cause__)
# I guess it just doesn't use __cause__, only traceback information...

test()

In [20]: ds = "Sep 18 2018 5:48:23 UTC"

In [21]: _TIME_FORMAT = "%b %d %Y %H:%M:%S %Z"

In [22]: datetime.strptime(ds, _TIME_FORMAT)
Out[22]: datetime.datetime(2018, 9, 18, 5, 48, 23)

fucking hell..

https://github.com/encode/httpx

HTTPX is a fully featured HTTP client for Python 3, which provides sync and async APIs, and support for both HTTP/1.1 and HTTP/2.

how to install a dependency as editable

add this to developer guide.. fucking hell

pipenv- like venv - can be used to create virtual envelopes but additionally rolls-in package management and vulnerability checking functionality. Instead of using requirements.txt, pipenv delivers package management via Pipfile. As PyPA endorses pipenv for PACKAGE MANAGEMENT, that would seem to imply pipfile is to supplant requirements.txt.

HOWEVER: pipenv uses virtualenv as its tool for creating virtual envelopes, NOT venv which is endorsed by PyPA as the go-to tool for creating virtual envelopes.

https://pypi.org/project/argparseweb

https://github.com/vinta/awesome-python

Command-line Interface Development
Libraries for building command-line applications.
    Command-line Application Development

https://github.com/GrahamDumpleton/wrapt

It therefore goes way beyond existing mechanisms such as functools.wraps() to ensure that decorators preserve introspectability, signatures, type checking abilities etc. Th

https://github.com/pypa/setuptools_scm/#default-versioning-scheme

def semcal(v):
    tag = v.tag.base_version
    assert len(tag.split('.')) == 2, tag
    v.time
    # tag is
    breakpoint()
    pass


if __name__ == '__main__':
    [pkg] = find_packages('src')
    setup(
        name=pkg,
        # version='0.1.1',
        # use_scm_version=True,
        use_scm_version={
            'version_scheme': semcal,
        },
        setup_requires=['setuptools_scm'],

https://github.com/python-poetry/poetry/issues/222

For now, you can use poetry with tox by using something similar to what I did in pendulum: https://github.com/sdispater/pendulum/blob/master/tox.ini

Although, beware that you may need to change `find_packages()` to `find_namespace_packages()` in your setup.py.
I get burned by that one a lot!

ldisplay_md("hi")
TODO shit. try and work on it?
also provide these as utils..
from IPython.display import display, Latex
display(Markdown("Heat equation:"), heat_eq)
display(Markdown("Boundary conditions:"))
display(Latex(r"\begin{cases} x \\ y \end{cases}"), bc_left, bc_right)
display(Markdown('hi'))
That doesn't look great. `display` is quoting plain

https://github.com/KoffeinFlummi/Chronyk

https://github.com/pypa/sampleproject/blob/master/setup.py

# This field corresponds to the "Description-Content-Type" metadata field:
# https://packaging.python.org/specifications/core-metadata/#description-content-type-optional
long_description_content_type='text/markdown',  # Optional (see note above)

AttributeError: unexpected attribute 'color' to Line, similar attributes are line_color

https://pypi.org/project/pytest-profiling/

Even putting aside it's approach to dependency management, Poetry has a fond place in my heart for making it so much easier to publish to PyPI.
If poetry did nothing but simply the process of publishing and updating python packages, it would still be an amazing tool.

Speed of execution is similar to CPython for most operations.

https://logzero.readthedocs.io/en/latest/

logzero.setup_logger(name=None, logfile

https://github.com/marshmallow-code/marshmallow
zoopla/zoopla/schemas.py

PandaPy has the speed of NumPy and the usability of Pandas - https://news.ycombinator.com/item?id=22142869

https://github.com/pytest-dev/pytest-xdist/issues/385

Isn't this enough?
py.test -m sequential
py.test -m "non sequential" -n8

https://github.com/PyCQA/pylint/issues/2388

Thanks for the tip. I actually had that setting before and still keep it on. --extension-pkg-whitelist=numpy
However, after upgrading to later versions of python and pylint and astroid I no longer have the issue even with the numpy whitelist setting.
Python 3.7.1
pylint 2.3.1
astroid 2.2.5

https://github.com/PyCQA/pylint/issues/617#event-2701192212

Search pylint configuration in setup.cfg and pyproject.toml

https://stackoverflow.com/questions/5071601/how-do-i-use-regex-in-a-sqlite-query

With python, assuming con is the connection to SQLite, you can define the required UDF by writing:
con.create_function('regexp', 2, lambda x, y: 1 if re.search(x,y) else 0)

Pydantic checks that each key I expected exists and is of the correct type. So now all the assumptions I’ve made about the shape of the data is a runtime contract, expressed in native python. If in production it turns out that description is, in fact, optional I’ll get a very descriptive error and the code will fail early.

d for d in sorted(pkg_resources.working_set, key=lambda p: p.project_name)]

Exhaustiveness Checking

use else clause with NoReturn, then mypy assists in refining

wow! * (same plot) vs + (different plots) is pretty clever!
also I like how multiplication doesn't commute so the 'first' plot wins the layout parameters

https://github.com/naiquevin/pipdeptree

Docum

https://github.com/kirankoduru/scrapy-programmatically

tests/integration_test.py::test_example_config
  /Users/runner/runners/2.169.1/work/promnesia/promnesia/.tox/py3/lib/python3.7/site-packages/hug/use.py:185: DeprecationWarning: invalid escape sequence \d
    status_code = int("".join(re.findall("\d+", response.status)))

https://news.ycombinator.com/item?id=22787332

mjw1007 28 minutes ago [-]
One variant of this is to use Python syntax but parse it with the `ast` module rather than evaluating it.
This avoids the security problems and the risk that someone will start to write programs inside the configuration file, but gives you the reasonably nice and powerful syntax.

ython Fire is a library for automatically generating command line interfaces (CLIs) from absolutely any Python object.

why is it not including files at all??

import sys
# make sure it doesn't need pytest in prod
if 'pytest' in sys.modules:
    import pytest # type: ignore
else:
    from unittest.mock import Mock
    pytest = Mock()

pkgutil.extend_path(path, name)

    Extend the search path for the modules which comprise a package. Intended use is to place the following code in a package’s __init__.py:

    from pkgutil import extend_path
    __path__ = extend_path(__path__, __name__)

    This will add to the package’s __path__ all subdirectories of directories on sys.path named after the package. This is useful if one wants to distribute different parts of a single logical package as multiple directories.

https://www.python.org/dev/peps/pep-0593/

https://dev.to/wemake-services/simple-dependent-types-in-python-4e14

Now, we are going to combine our new knowledge about Literal and @overload together to solve our problem with open. At last!

As one point of comparison, SymPy is comically slow compared to Sage. This is mostly because SymPy is purely Python; Sage on the other hand uses a derivative of GiNaC [1] for its internal symbolic expression representation, and then multiple external libraries for non-trivial operations. Symbolic transformations are mostly Maxima [3]. Sage literally converts expressions to strings, pipes them through a Maxima process, and then parses the result back. This is still much faster than the pure Python SymPy.
There is an effort to speed up SymPy core, SymEngine [3], but it's been in development for years now, and still isn't integrated into SymPy. Not sure why.
Case in point: 'expand("(2+3x+4x*y)^60")' takes 5 seconds with SymPy; Sage does the same in 0.02 seconds.

The Qt console is a very lightweight application that largely feels like a terminal, but provides a number of enhancements only possible in a GUI, such as inline figures, proper multi-line editing with syntax highlighting, graphical calltips, and much more. The Qt console can use any Jupyter kernel.

cell-var-from-loop -- can detect bad lambda capturing
 pylint --disable='C,fixme,'

But for big projects with lots of subcommands, written in Python, using argparse may be a better idea, especially if you need to do advanced stuff for your parsing, like configuring common options in one function, allow plug-ins to change the command line API, and so on …

https://stackoverflow.com/questions/27907373/dynamically-change-slots-in-python-3

Normally, an object's attributes are stored in a dict, which requires a fair bit of memory itself. If you are creating millions of objects then the space required by these dicts becomes prohibitive. __slots__ informs the python machinery that makes the class object that there will only be so many attributes refered to by instances of this class and what the names of the attributes will be. Therefore, the class can make an optimisation by storing the attributes directly on the instance rather than in a dict. It places the memory for the (pointers to the) attributes directly on the object, rather than creating a new dict for the object.

In the spirit of HN contrarian comments, I'd say the closest thing spiritually to a modern-day HyperCard is PySimpleGUI: https://github.com/PySimpleGUI

https://github.com/pschanely/CrossHair

https://github.com/lihaoyi/macropy

https://logbook.readthedocs.io/en/stable/features.html#advantages-over-logging

https://patsy.readthedocs.io/en/latest/formulas.html

The formula language is actually fairly simple once you get the hang of it, but if you’re ever in doubt as to what some construction means, you can always ask Patsy how it expands.

https://chriswarrick.com/blog/2018/07/17/pipenv-promises-a-lot-delivers-very-little/

Requirements Files described most simply, are just a list of pip install arguments placed into a file.
Whereas install_requires defines the dependencies for a single project, Requirements Files are often used to define the requirements for a complete Python environment.
Whereas install_requires requirements are minimal, requirements files often contain an exhaustive listing of pinned versions for the purpose of achieving repeatable installations of a complete environment.
Whereas install_requires requirements are “Abstract”, i.e. not associated with any particular index, requirements files often contain pip options like --index-url or --find-links to make requirements “Concrete”, i.e. associated with a particular index or directory of packages. 1
Whereas install_requires metadata is automatically analyzed by pip during an install, requirements files are not, and only are used when a user specifically installs them using pip install -r.

https://github.com/kynan/nbstripout

pretty good comparizon

def building_for_pypi() -> bool:
    return sys.argv[1] == 'sdist'
   for_pypi = building_for_pypi()
           'my': [] if for_pypi else ['HPI @ git+https://github.com/karlicoss/my.git'],
       },

https://twitter.com/karlicoss/status/1238791294547353601

@dataclass
(init=False)
class FancyPath:
    path: Path
    mtime: float
    def __init__(self, path: Path):
        self.path = path
        self.mtime = path.stat().st_mtime

right, I suppose it's type to use dataclasses

http://python-notes.curiousefficiency.org/en/latest/python_concepts/import_traps.html#the-init-py-trap

Use parens (...) for fluent APIs

You can toggle the scroll window in the main menu of the notebook
Cell -> Current Outputs -> Toggle Scrolling

https://news.ycombinator.com/item?id=21902827

Not really. Enable something like MyPy or PyType on a big enough codebase with zero explicit annotation and it'll already find plenty of bugs and unhandled cases from the inferred types alone. Some of these are stuff a strong IDE may catch too (using the wrong function name, passing wrong number of args, etc), but some other ones are actually deeper in the code.

So already, with zero annotations, you already get value out, let alone once you type a few tricky variables that are harder for Python to track.

tests in a separate dir vs tests in a subpackage

https://plot.ly/python/splom/

https://github.com/box/flaky

https://github.com/lambdalisue/jupyter-vim-binding

Yes. I've written a few systems using SQLAlchemy that see 50-100k writes  and 200k or so reads per second, and as long as you're using the Core everything works fine. Also allows you to use the ORM for ease-of-development on bits that don't need performance.

SQLAlchemy is probably the one library in which I've never reached the "wait, I need to do something and can't" point. Every single weird edge case I've had to handle has been supported either natively or by being able to access the raw interfaces directly in a clean way. Fantastic library, can't recommend it more highly.

Also, the lead dev is extremely responsive on the google group and is more than happy to answer questions.

https://python-packaging.readthedocs.io/en/latest/command-line-scripts.html#the-console-scripts-entry-point

https://en.wikipedia.org/wiki/Hy

https://lisp-journey.gitlab.io/pythonvslisp
don't feel like it's very objective comparison.
lots of things are picked etc
although I'm not very positive towards lisp either

https://www.reddit.com/r/Python/comments/69wym8/when_do_you_us_beautiful_soup_or_scrapy/

why would you never use BS
The selectors system lack in flexibility, it is slow and its HTML parser is not really good for the modern world of opening a page in a browser and looking at what's there to know what to do.

https://pandas.pydata.org/pandas-docs/stable/reference/frame.html

https://pandas.pydata.org/pandas-docs/stable/getting_started/comparison/comparison_with_sql.html#compare-with-sql-join

https://www.reddit.com/r/Python/comments/akgctg/theres_now_a_dead_simple_way_to_slap_a_web/

def wrapped(path: Path):
    try:
        it = ex(path)
        # ugh. keeping yeild in the try section is not ideal, bt
        yield from it
    except Exception as e:
        breakpoint()

I guess it's somewhat useful to have multiple stages

raw providers emit dataframe without any index in case of reporting dateless errors (if that makes sense)
then consumer can filter out errors and present them and then choose their own index and aggregateion

https://www.quora.com/Which-one-is-better-Bokeh-or-Plotly

Compare Bokeh with Plotly
1. Easy to install. The process is very similar to Plotly.
2. Bokeh plot is not as interactive as Plotly.
3. Limitation on drawing string value on plot. It looks like only the bar chart can take the string values. Date time string has to convert to DateTime type first in order to be on non-bar plots such as scatter, line, etc.
4. It is not easy to overlay different type of plots on one figure.
5. Less examples online.
6. Less integrated with Panda than Plotly.

https://www.reddit.com/r/Python/comments/5nwk9r/plotly_vs_bokeh_vs/

The last advantage I found that plotly has over bokeh is that almost everything in plotly is a dict. This is really useful because I've found certain plotly functions are really slow for some reason or can't be called when you use the multiprocessing module because of pickling issues. Not a problem, in the end they generate dicts under the hood, you can do the same yourself and still pass those dicts to plotly. Also calling help() on plotly objects tells you what keys and values you need to add to your dict.

e.g. unpicling requires exactly same modules etc? might be hard if you want to unpickle elsewhere

https://stackoverflow.com/a/25244948/706389

https://pbpython.com/python-vis-flowchart.html

https://bokeh.pydata.org/en/latest/docs/user_guide/quickstart.html#userguide-quickstart

https://stackoverflow.com/questions/43727278/how-to-set-readable-xticks-in-seaborns-facetgrid

The seaborn.pointplot is not the right tool for this plot.

plotting a lot of data via seaborn

https://python-graph-gallery.com/45-control-color-of-each-marker-seaborn/

scatter_kws={'facecolors':df['color']}

https://stackoverflow.com/questions/44496383/format-x-axis-on-chart-created-with-pandas-plot-method

ax = df.plot(kind='bar')
x_labels = df.index.strftime('%b')
ax.set_xticklabels(x_labels)

Comment:

pandas date ticks

i.e. default got username set to "", which gets in the way of using token

https://stackoverflow.com/questions/12151306/argparse-way-to-include-default-values-in-help

parser = argparse.ArgumentParser(
    # ... other options ...
    formatter_class=argparse.ArgumentDefaultsHelpFormatter)

• [2019-07-24] I know all of that but maybe good for newbies
  

https://github.com/google/pytype

https://news.ycombinator.com/item?id=10185290

I had a simpler case that I couldn't figure out using docopt. I was trying to add in tls flags into docker-compose to match docker. I could not get a combination working where you can just specify --tls to mean tls is true while still accepting --tls=true|false|0|1.

Why not Arrow?

Arrow is the most popular datetime library for Python right now, however its behavior and API can be erratic and unpredictable.

Another way is ipython %edit magic command and $EDITOR set to vim and using %autoreload

ast.literal_eval

https://buttondown.email/hillelwayne/archive/why-python-is-my-favorite-language/

What I need is an “X for Y”, but better. Something like To Ruby from Python or Go for Python Programmers only cover the core. They don’t cover the peripheral libraries or the lagangue ecosystem. Hyperpolyglot is better, but doesn’t talk much about the ecosystem. It also doesn’t compare across language families, which could be very helpful here.

Module for decorators, wrappers and monkey patching.

To power its capabilities, Typer internally stands on top of Click, which is very well-known and battle-tested. This means that it can leverage all its benefits, community and plugins, while starting simple with less boilerplate code and growing more complex as you need.

Features of Nbdev
    - Automatically generate docs from Jupyter notebooks. These docs are searchable and automatically hyperlinked to appropriate documentation pages by introspecting keywords you surround in backticks.
    - Utilities to automate the publishing of pypi and conda packages including version number management.
    - A robust, two-way sync between notebooks and source code, which allow you to use your IDE for code navigation or quick edits if desired.
    - Fine-grained control on hiding/showing cells: you can choose to hide entire cells, just the output, or just the input. Furthermore, you can embed cells in collapsible elements that are open or closed by default.
    - Ability to write tests directly in notebooks without having to learn special APIs. These tests get executed in parallel with a single CLI command. You can even define certain groups of tests such that you don't have to always run long-running tests.
    - Tools for merge/conflict resolution with notebooks in a human readable format.
    - Continuous integration (CI) comes setup for you with GitHub Actions

Running Python code in a subprocess with a time limit

Typer, build great CLIs. Easy to code. Based on Python type hints.

My code lives in a read-only partition mounted into a Docker container and therefore the build fails:

There's a very good reason the traceback is not stored in the exception; because the traceback holds references to its stack's locals, this would result in a circular reference and (temporary) memory leak until the circular GC kicks in. (This is why you should never store the traceback in a local variable.)

there is github topics, github description, readme, pypi readem, readthedocs readme, classifiers etc

Agree. I now have a sudden urge to write a guide how to start from very simple (i.e. system interpreter, no packaging, minimal verification) and how to progress further on. And how to decide what tools do you even need to consider and at which stage.

In hindsight as a somewhat experienced Python programmer I can see how each of these tools can be helpful – but if I was a beginner I would be completely overwhelmed.

but I have no idea how to communicate it.
some problems are just obscure to write a proper blog post or something
self-answering stackoverflow questions??

https://stackoverflow.com/questions/3472430/how-can-i-make-setuptools-install-a-package-thats-not-on-pypi
why python sucks: dependencies

- name: Publish a Python distribution to PyPI
  uses: pypa/gh-action-pypi-publish@master
  with:
    user: __token__
    password: ${{ secrets.pypi_password }}

so much indirection!!

The autoimp module imports all available Python modules automatically at the Python interactive prompt, similarly to Matlab:

pyenv install 3.8.0
pyenv shell 3.8.0
https://noamelf.com/posts/my-python-setup-for-2020/

PyPy 2.6 introduced vmprof, a very-low-overhead statistical profiler. The standard, non-statistical cProfile is also supported, and can be enabled without turning off the JIT. We do recommend vmprof anyway because turning on cProfile can distort the result (sometimes massively, though hopefully this should not be too common).

If your Python decorator unintentionally changes the signatures of my callables or doesn’t work with class methods, it’s broken and should be fixed. Sadly most decorators are broken because the web is full of bad advice.

There’s also ipdb but importing and calling it directly only works in the terminal, not in notebooks (see this issue).

In other words, poetry uses pyproject.toml to replace setup.py, requirements.txt, setup.cfg, MANIFEST.in and the newly added Pipfile.

import pdb; pdb.pm() starts a post-mortem debugger. When you get an error in an interactive session (such as a Jupyter notebook), this will jump to the scene of the error, let you inspect local variables, and let you walk up and down the stack.

egg-info in local directory makes pip think package is installed

I hate easy-install.pth

shit. I guess this explains why it was finding mypkgs…

acropy is excellent when one wants to use syntactic macros. One problem I had was that, for my scripts, I cannot directly use macropy and other syntax modifiers since they rely on the import hooks, which means that you can not use them directly on the script file. You have to import them separately, which makes their use cumbersome.

This was my project to work around this issue. It works, but is very hacky, and requires adding a new codec to the Python installation.

HoloViews is an open-source Python library designed to make data analysis and visualization seamless and simple. With HoloViews, you can usually express what you want to do in very few lines of code, letting you focus on what you are trying to explore and convey, not on the process of plotting.

e.g. when I was writing pytest tests/core/test_pandas.py vs pytest tests/core/pandas.py

Python 3.9 2020 UPDATE
You can just use them together (taken from the 3.9 docs):
class G:
    @classmethod
    @property
    def __doc__(cls):
        return f'A doc for {cls.__name__!r}'

import pandas as pd
dfs = pd.read_html(url)

wow. can scrape off table elements?

could be like re.Pattern, or with automatic cache?

reasonable list (aka why python logging sucks by default)

Itamar Turner-Trauring 2, in the comment section on dev.to gave an interesting example:

handling accidental variable capture in closures

focus on performance?

find ./my | entr flake8 --ignore=E402,E501,E741,W503,E266,E302,E305,E203,E261,E252,E251,E221,W291,E225,E303,E702,E202,F841,E731,E306,E127 E722,E231 my | grep -v __NOT_HPI_MODULE__

good tutorial

Seems like a more sensible alternative to black

You're looking for postmortem mode:

try:
    code_that_may_raise_exception()
except Exception:
    import pdb; pdb.post_mortem()

This will break on exception and drop you into a debugger right where it is being raised from

nice! a bit magical, but works exactly as described

• no overhead about creating new files
  
• no need to worry about pythonpaths etc
  

how to:

• pytest imports in the function itself
  
• defensive decorator?
  
• fixtures still work! like tmp_path
  

TODO mention doctests? pros/cons. how to set up common env for doctests?

>> unaware.replace(tzinfo=ZoneInfo('localtime'))
datetime.datetime(2020, 10, 31, 12, 0, tzinfo=zoneinfo.ZoneInfo(key='localtime'))
>>> str(_)
'2020-10-31 12:00:00+01:00'
pip install backports.zoneinfo

Although strptime can parse the 'Z' character to UTC, fromisoformat is faster by ~ x40 (see also: A faster strptime):

You can use the py_modules argument instead of the packages argument to list single file modules.
See https://docs.python.org/3/distutils/setupscript.html#listing-individual-modules

progress bars, nice

PythonQL is implemented as a special encoding in a normal python script. When this encoding is specified, the pythonql preprocessor is run, which converts the pythonql syntax to pure python.
So you should have a line in the beginning of your script:
#coding: pythonql
result = [ select y for x in [1,2,3] let y = x**2 ]

whoa

*.egg-link files that contain the name of a built or development egg, to support symbolic linking on platforms that do not have native symbolic links.