Pout

A collection of handy functions for printing out variables and debugging Python code.

print didn't give enough information while debugging, pprint wasn't much better. I was also getting sick of typing things like: print("var = ", var).

Pout tries to print out variables with their name, and for good measure, it also prints where the pout function was called so you can easily find it and delete it when you're done debugging.

I use pout extensively in basically every python project I work on.

Methods

pout.v(arg1, [arg2, ...]) -- easy way to print variables

example

foo = 1
pout.v(foo)

bar = [1, 2, [3, 4], 5]
pout.v(bar)

should print something like:

foo = 1
(/file.py:line)

bar (4) =
[
        0: 1,
        1: 2,
        2:
                [
                        0: 3,
                        1: 4
                ],
        3: 5
]
(/file.py:line)

You can send as many variables as you want into the call

# pass in as many variables as you want
pout.v(foo, bar, che)

# a multi-line call is also fine
pout.v(
    foo,
    bar
)

pout.h() -- easy way to print "here" in the code

example

pout.h(1)
# do something else
pout.h(2)

# do even more of something else
pout.h()

Should print something like:

here 1 (/file.py:line)

here 2 (/file.py:line)

here N (/file.py:N)

pout.t() -- print a backtrace

Prints a nicely formatted backtrace, by default this should compact system python calls (eg, anything in dist-packages) which makes the backtrace easier for me to follow.

example:

pout.t()

should print something like:

15 - C:\Python27\lib\runpy.py:162
14 - C:\Python27\lib\runpy.py:72
13 - C:\Python27\lib\unittest\__main__.py:12
12 - C:\Python27\lib\unittest\main.py:95
11 - C:\Python27\lib\unittest\main.py:229
10 - C:\Python27\lib\unittest\runner.py:151
09 - C:\Python27\lib\unittest\suite.py:65
08 - C:\Python27\lib\unittest\suite.py:103
07 - C:\Python27\lib\unittest\suite.py:65
06 - C:\Python27\lib\unittest\suite.py:103
05 - C:\Python27\lib\unittest\suite.py:65
04 - C:\Python27\lib\unittest\suite.py:103
03 - C:\Python27\lib\unittest\case.py:376
02 - C:\Python27\lib\unittest\case.py:318
01 - C:\Projects\Pout\_pout\src\test_pout.py:50

        pout.t()

pout.p([title]) -- quick and dirty profiling

example

p("starting profile")
time.sleep(1)
p() # stop the "starting profile" session


# you can go N levels deep
p("one")
p("two")
time.sleep(0.5)
p() # stop profiling of "two"
time.sleep(0.5)
p() # stop profiling of "one"


# you can also use with
with p("benchmarking"):
    time.sleep(0.5)

should print something like:

starting profile - 1008.2 ms
  start: 1368137723.7 (/file/path:n)
  stop: 1368137724.71(/file/path:n)


one > two - 509.2 ms
  start: 1368137722.69 (/file/path:n)
  stop: 1368137723.2(/file/path:n)


one - 1025.9 ms
  start: 1368137722.68 (/file/path:n)
  stop: 1368137723.7(/file/path:n)

pout.x(arg1, [arg2, ...]) -- like pout.v but then will run sys.exit(1)

This just prints out where it was called from, so you can remember where you exited the code while debugging

example:

pout.x()

will print something like this before exiting with an exit code of 1:

exit (/file/path:n)

pout.b([title[, rows[, sep]]]) -- prints lots of lines to break up output

This is is handy if you are printing lots of stuff in a loop and you want to break up the output into sections.

example:

pout.b()
pout.b('this is the title')
pout.b('this is the title 2', 5)
pout.b('this is the title 3', 3, '=')

Would result in output like:

********************************************************************************
(/file/path:n)


****************************** this is the title *******************************
(/file/path:n)


********************************************************************************
********************************************************************************
***************************** this is the title 2 ******************************
********************************************************************************
********************************************************************************
(/file/path:n)


================================================================================
============================= this is the title 3 ==============================
===============================================================================
(/file/path:n)

pout.c(str1, [str2, ...]) -- print info about each char in each str

Kind of like od -c on the command line.

example:

pout.c('this')

will print something like:

Total Characters: 4
t	't'	\u0074	LATIN SMALL LETTER T
h	'h'	\u0068	LATIN SMALL LETTER H
i	'i'	\u0069	LATIN SMALL LETTER I
s	's'	\u0073	LATIN SMALL LETTER S
(/file/path:n)

This could fail if Python isn't compiled with 4 byte unicode support, just something to be aware of, but chances are, if you don't have 4 byte unicode supported Python, you're not doing much with 4 byte unicode.

pout.s(arg1, [arg2, ...]) -- easy way to return pretty versions of variables

Just like pout.v() but will return the value as a string

pout.ss(arg1, [arg2, ...]) -- easy way to return pretty versions of variables without meta information

Just like pout.vv() but will return the value as a string

pout.l([logger_name, [logger_level]]) -- turn logging on just for this context

Turns logging on for the given level (defaults to logging.DEBUG) and prints the logs to stderr. Useful when you just want to check the logs of something without modifying your current logging configuration.

example:

with pout.l():
    logger.debug("This will print to the screen even if logging is off")
logger.debug("this will not print if logging is off")

with pout.l("name"):
    # if "name" logger is used it will print to stderr
# "name" logger goes back to previous configuration

pout.tofile([path])

Routes pout's output to a file (defaults to ./pout.txt)

example:

with pout.tofile():
	# everything in this with block will print to a file in current directory
	pout.b()
	s = "foo"
	pout.v(s)
	
pout.s() # this will print to stderr

Customizing Pout

object magic method

Any class object can define a __pout__ magic method, similar to Python's built in __str__ magic method that can return a customized string of the object if you want to. This method can return anything, it will be run through Pout's internal stringify methods to convert it to a string and print it out.

Console commands

pout json

running a command on the command line that outputs a whole a bunch of json? Pout can help:

$ some-command-that-outputs-json | pout json

pout char

Runs pout.c but on the output from a command line script:

$ echo "some string with chars to analyze" | pout char

Install

Use PIP

pip install pout

Generally, the pypi version and the github version shouldn't be that out of sync, but just in case, you can install from github also:

pip install -U "git+https://github.com/Jaymon/pout#egg=pout"

Make Pout easier to use

When debugging, it's really nice not having to put import pout at the top of every module you want to use it in, so there's a command for that, if you put:

import pout
pout.inject()

Somewhere near the top of your application startup script, then pout will be available in all your files whether you imported it or not, it will be just like str, object, or the rest of python's standard library.

If you don't even want to bother with doing that, then just run:

$ pout inject

from the command line and it will modify your python environment to make pout available as a builtin module, just like the python standard library. This is super handy for development virtual environments.