awful programming: feature flags and suffering

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > awful programming: feature flags and suffering

«‹›1115 »

Powerful Two-Hander: Mar 10, 2004; Mods please change my name to "Tooter Skeleton" TIA.

Luigi Thirty posted:

folks, I bring you this thing that appeared on my twitter timeline

visual mumps for the Apple IIgs

https://twitter.com/AndiarSoftware/status/1103283584113262597

huh I thought vaccines protected against mumps yet here we are in yospos talking about mu.... Oh wait no I see now.

# ? Mar 6, 2019 23:07

Adbot: ADBOT LOVES YOU

# ? Apr 24, 2024 18:28

eschaton: Mar 7, 2007; Don't you just hate when you wind up in a store with people who are in a socioeconomic class that is pretty obviously about two levels lower than your own?

Luigi Thirty posted:

folks, I bring you this thing that appeared on my twitter timeline

visual mumps for the Apple IIgs

https://twitter.com/AndiarSoftware/status/1103283584113262597

cursed IDE

pls share your Atari TT stories with us

# ? Mar 7, 2019 04:12

Luigi Thirty: Apr 30, 2006; Emergency confection port.

eschaton posted:

cursed IDE

pls share your Atari TT stories with us

I'm writing a mastodon client in C++ targeting m68k-atari-mint atm

it's only 206kb so far!

# ? Mar 7, 2019 05:49

Luigi Thirty: Apr 30, 2006; Emergency confection port.

Currently: it can make calls to the accounts API, parse the result, and display a couple strings lol

https://twitter.com/LuigiThirty/status/1103539212312293379

# ? Mar 7, 2019 07:51

Corla Plankun: May 8, 2007; improve the lives of everyone

Luigi Thirty posted:

Currently: it can make calls to the accounts API, parse the result, and display a couple strings lol

https://twitter.com/LuigiThirty/status/1103539212312293379

# ? Mar 7, 2019 16:23

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

how do i get at a method/function's arguments in a way where i can fall back from **kwargs to *args? i have this decorator:

code:

def blahblah(key: str):
    def decorator(func):
        @functools.wraps(func)
        def wrapper(*args, **kwargs):
            self = args[0] ## works fine, points to instance of Something
            fileid = kwargs.get('fileid', args[1]) ## IndexError: tuple index out of range
            # here i wanna do some things with key + self + fileid before going to the wrapped method
            return func(*args, **kwargs)
        return wrapper
    return decorator


class Something(object):
    @blahblah('doy')
    def a(fileid, foo, bar):
        pass

    @blahblah('hubba')
    def b(fileid, zoom, pow):
        pass

s = Something()
s.a(fileid='blah', foo=1, bar=2)
s.b(fileid='blah', zoom=3, pow=4)
s.a('blah', 5, 6)
s.b('blah', 7, 8)

i want the a/b methods to be callable with both fileid as a named or an unnamed argument, but args[1] above fails even though it should only be called if 'fileid' not in kwargs...?

Carthag Tuek fucked around with this message at 13:59 on Mar 8, 2019

# ? Mar 8, 2019 13:55

Volte: Oct 4, 2004; woosh woosh

args[1] has to be evaluated in order to even call kwargs.get with it as an argument - it's not short circuited.

code:

fileid = kwargs.get('fileid')
if fileid is None:
  fileid = args[1]

# ? Mar 8, 2019 14:05

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

ahh thanks, that makes sense

# ? Mar 8, 2019 14:18

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

it works! wrote up a little helper for it :cool:

code:

def get_arg(args, kwargs, name, index):
    val = kwargs.get(name, None)
    if val is None:
        val = args[index]
    return val

fileid = get_arg(args, kwargs, 'fileid', 1)

# ? Mar 8, 2019 16:29

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

this is probably more robust re possible "None" arguments

code:

def get_arg(args, kwargs, name, index):
    if name in kwargs:
        return kwargs[name]
    else:
        return args[index]

# ? Mar 8, 2019 16:43

FlapYoJacks: Feb 12, 2009

Krankenstyle posted:

this is probably more robust re possible "None" arguments
code:
def get_arg(args, kwargs, name, index):
    if name in kwargs:
        return kwargs[name]
    else:
        return args[index]

You don't need that else though?

Edit:

code:

def get_arg(args, kwargs, name, index):
    if name in kwargs:
        return kwargs[name]
    return args[index]

Works just fine. :colbert:

# ? Mar 8, 2019 16:49

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

right

also i considered going

code:

return kwargs[name] if name in kwargs else args[index]

but im not a huge fan of pythons dumbass lack of ternary operator, i mean this is just ugly

Carthag Tuek fucked around with this message at 16:59 on Mar 8, 2019

# ? Mar 8, 2019 16:57

kalel: Jun 19, 2012

ratbert90 posted:

You don't need that else though?

# ? Mar 8, 2019 17:00

FlapYoJacks: Feb 12, 2009

Krankenstyle posted:

right

also i considered going
code:
return kwargs[name] if name in kwargs else args[index]
but im not a huge fan of pythons dumbass lack of ternary operator, i mean this is just ugly

* Rushes in with heavy breathing * HAVE YOU SEEN PYTHONS LIST COMPREHENSION?!?!?!

SciFiDownBeat posted:

It's unnecessary and stupid. :colbert:

# ? Mar 8, 2019 17:00

kalel: Jun 19, 2012

semicolons are unnecessary in javascript but I still use them like a good boy scout :angel:

# ? Mar 8, 2019 17:05

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

lol at all the code golf-y alternatives here:
https://en.wikipedia.org/wiki/%3F:#Python

# ? Mar 8, 2019 17:08

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

ratbert90 posted:

* Rushes in with heavy breathing * HAVE YOU SEEN PYTHONS LIST COMPREHENSION?!?!?!

one of them is actually a tuple with an index (because true/false evaluate to 1/0): (args[index], kwargs[name])[name in kwargs] :barf:

# ? Mar 8, 2019 17:10

cinci zoo sniper: Mar 15, 2013

Krankenstyle posted:

this is probably more robust re possible "None" arguments
code:
def get_arg(args, kwargs, name, index):
    if name in kwargs:
        return kwargs[name]
    else:
        return args[index]

what you want is

code:

def get_arg(args, kwargs, name, index):
    if kwargs:
        return kwargs.get(name) if kwargs.get(name) is not None else args[index]
    return args[index]

"if name in kwargs" will throw attributerror if kwargs is none

cinci zoo sniper fucked around with this message at 17:13 on Mar 8, 2019

# ? Mar 8, 2019 17:11

cinci zoo sniper: Mar 15, 2013

3 notes:

this is going to get ugly if it is permissible for a kwarg value to be None
if you have imbalanced input data (e.g. majority of kwargs is like this or like that) then its better to rewrite this via try/except
if you can guaranteed that all kwargs values will be truthy then you can reduce "is not None" to basic truthiness check

# ? Mar 8, 2019 17:15

Symbolic Butt: Mar 22, 2009; (_!_); Buglord

SciFiDownBeat posted:

semicolons are unnecessary in javascript but I still use them like a good boy scout

iirc there are some really weird gotchas if you don't use semicolons so it's not a bad habit really

# ? Mar 8, 2019 17:22

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

cinci zoo sniper posted:

what you want is
code:
def get_arg(args, kwargs, name, index):
    if kwargs:
        return kwargs.get(name) if kwargs.get(name) is not None else args[index]
    return args[index]
"if name in kwargs" will throw attributerror if kwargs is none

wouldnt this work just as well then?

code:

def get_arg(args, kwargs, name, index):
    if kwargs and name in kwargs:
        return kwargs[name]
    return args[index]

# ? Mar 8, 2019 17:22

Aramoro: Jun 1, 2012

Symbolic Butt posted:

iirc there are some really weird gotchas

fixed for all of JavaScript.

# ? Mar 8, 2019 17:25

cinci zoo sniper: Mar 15, 2013

Krankenstyle posted:

wouldnt this work just as well then?

code:

def get_arg(args, kwargs, name, index):
    if kwargs and name in kwargs:
        return kwargs[name]
    return args[index]

yea, this would. {'foo': None} would get you a None, if that matters. it's what i originally wrote replying to you, but then i brain sharted somewhere, this was a hellweek at $job and i was completely fried by like tuesday afternoon

# ? Mar 8, 2019 17:42

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

ya I'm fine with a None showing up if that was indeed what th method was called with

# ? Mar 8, 2019 17:48

cinci zoo sniper: Mar 15, 2013

Krankenstyle posted:

ya I'm fine with a None showing up if that was indeed what th method was called with

you're all set then, kwargs conditional should not have any meaningful failure modes remaining. args[index] can go sideways - i'd at least do some index check in a "better safe than sorry" fashion

# ? Mar 8, 2019 17:54

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

good call tho I doubt it'll happen. better safe as you say

# ? Mar 8, 2019 17:57

cinci zoo sniper: Mar 15, 2013

Krankenstyle posted:

good call tho I doubt it'll happen. better safe as you say

if you doubt it will happen, wrap what you have into a try/except IndexError block.

# ? Mar 8, 2019 18:01

cinci zoo sniper: Mar 15, 2013

as everyone who has followed my database posts itt might imagine, im really into "defensive coding" by now :v:

# ? Mar 8, 2019 18:02

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

cinci zoo sniper posted:

if you doubt it will happen, wrap what you have into a try/except IndexError block.

on second thought, i want my app to blow up if the argument is missing cause theres no reasonable way to go forward without a fileid anyway

# ? Mar 8, 2019 18:15

pokeyman: Nov 26, 2006; That elephant ate my entire platoon.

SciFiDownBeat posted:

yes, but when you don�t have a compiler telling you �hey dipshit you haven�t returned anything from all code paths� then I like to code in a way that I�m less likely to forget to return something

SciFiDownBeat posted:

semicolons are unnecessary in javascript but I still use them like a good boy scout

this is way more confusing than the other thing and is not a good comparison

# ? Mar 8, 2019 19:04

fritz: Jul 26, 2003

ratbert90 posted:

It's unnecessary and stupid.

so are my posts but that's enver stopped me

# ? Mar 8, 2019 20:14

kalel: Jun 19, 2012

pokeyman posted:

yes, but when you don’t have a compiler telling you “hey dipshit you haven’t returned anything from all code paths” then I like to code in a way that I’m less likely to forget to return something

I agree in practice but it's a two line utility function for pete's sake

# ? Mar 8, 2019 20:23

Captain Foo: May 11, 2004; we vibin'
we slidin'
we breathin'
we dyin'

SciFiDownBeat posted:

I agree in practice but it's a two line utility function for pete's sake

whatever you're coding is now doomed to be mission critical line of business code

# ? Mar 8, 2019 22:02

Sapozhnik: Jan 2, 2005; Nap Ghost

Captain Foo posted:

whatever you're coding is now doomed to be mission critical line of business code

College stoner me: "Like, the universe isn't straight-up malevolent, it's just totally indifferent to us man"
Working programmer me: "I dunno, maybe there's a case to be made that it actually is malevolent"

# ? Mar 9, 2019 00:20

hackbunny: Jul 22, 2007; I haven't been on SA for years but the person who gave me my previous av as a joke felt guilty for doing so and decided to get me a non-shitty av

ctps: writing unit tests for the first time in my life and I'm exhausted. this is taking me an ungodly amount of time, but I'm anxious about hitting all the corner cases. anyone has examples handy of "good" unit tests for some simple real world library? I suspect I'm maybe overdoing it a little

# ? Mar 9, 2019 01:00

Luigi Thirty: Apr 30, 2006; Emergency confection port.

me, 3 hours ago: this C GEM library sucks, I want object oriented controls

me, 5 minutes ago: I think I�ve accidentally invented WinApi common controls but for the Atari ST

# ? Mar 9, 2019 01:05

4lokos basilisk: Jul 17, 2008

hackbunny posted:

ctps: writing unit tests for the first time in my life and I'm exhausted. this is taking me an ungodly amount of time, but I'm anxious about hitting all the corner cases. anyone has examples handy of "good" unit tests for some simple real world library? I suspect I'm maybe overdoing it a little

not sure if it is possible to hit all corner cases or even most of them before you need unit tests on the unit tests

seems to be a thing where practice will help you figure out the balance

personally I try hitting only the corner cases that I think might be occurring actually or the cases that could be difficult to detect and debug later on

# ? Mar 9, 2019 01:36

gonadic io: Feb 16, 2011; >>=

hackbunny posted:

ctps: writing unit tests for the first time in my life and I'm exhausted. this is taking me an ungodly amount of time, but I'm anxious about hitting all the corner cases. anyone has examples handy of "good" unit tests for some simple real world library? I suspect I'm maybe overdoing it a little

check ur coverage

(no really, it points out when you miss branches)

# ? Mar 9, 2019 01:41

4lokos basilisk: Jul 17, 2008

also I think unit tests are meant to force you write the software in easily testable layers so you can replace the layers later on if you need to without messing with tests

ie like when you replace your thing which handles db interaction then you can focus on only that and know that whatever you do in the code as long as tests succeed you should be good

# ? Mar 9, 2019 01:44

Adbot: ADBOT LOVES YOU

# ? Apr 24, 2024 18:28

Private Speech: Mar 30, 2011; I HAVE EVEN MORE WORTHLESS BEANIE BABIES IN MY COLLECTION THAN I HAVE WORTHLESS POSTS IN THE BEANIE BABY THREAD YET I STILL HAVE THE TEMERITY TO CRITICIZE OTHERS' COLLECTIONS

IF YOU SEE ME TALKING ABOUT BEANIE BABIES, PLEASE TELL ME TO

EAT. SHIT.

i've been wondering if unit testing in embedded/kernel/IoT space can be useful, there's an itch to try but when you don't even compile in all of the standard library into the firmware to save space it's fairly hard to design tests which would actually be meaningful and still run on the device

and most of the bugs come from a complex interplay of PCBs/firmware/components/third-party software anyway

obviously hardware testing/black box testing is important, but that's another kettle of fish entirely

e: there's the design verification route of running in a simulator maybe, but I don't know if that wouldn't bring in even more issues than it would solve

Private Speech fucked around with this message at 02:19 on Mar 9, 2019

# ? Mar 9, 2019 02:06

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > awful programming: feature flags and suffering

«‹›1115 »