hpr_transcripts/hpr1824.txt

Episode: 1824
Title: HPR1824: I'm Learning Some Python
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr1824/hpr1824.mp3
Transcribed: 2025-10-18 09:45:29

---

This is HPR episode 1,824 entitled, I'm learning some Python, it is hosted by John Culp and
is about 32 minutes long, the summary is, I discuss how I use Python and some of the
cool modules and libraries that I've found.
This episode of HPR is brought to you by an honesthost.com, get 15% discount on all shared
hosting with the offer code HPR15, that's HPR15, better web hosting that's honest and
fair at an honesthost.com.
Hey everybody, this is John Culp and Lafayette Louisiana and today I'm going to talk about
something that's been occupying quite a lot of my time lately and that is Python.
I'm learning some Python and I'm not really learning it in any systematic kind of way,
that's kind of not really how I do things but what I'm doing is learning how to use it
in a way that directly benefits me every day and if you've listened to my previous episodes
you'll know that I use voice commands quite a lot and so one of the things I'm doing
is converting a whole lot of the voice command actions that I used to do in Bash over to
Python if it's appropriate.
I have actually started an online course in Python but I find that I will watch a lesson
and then do the exercises and then what I'll do is think of a way that it applies more
specifically to me and so I'll go work on my own project and this is how I've always learned
stuff like this. When I was learning Bash it didn't really benefit me very much to learn
abstract concepts. It was much more helpful if I had an immediate need and I could learn
how to do something that would address that need directly rather than learning abstract
concepts.
The biggest project that I've done in Python so far was to convert the build scripts for
the School of Music websites from Bash over to Python. There was an episode I did maybe
a year ago or two years ago I don't remember really when it was but I talked about overhauling
the School of Music website and part of that process was not only cleaning up all the
content and everything but also coming up other way to have I essentially built my own
content management system using Bash where the main content of every page on the site and again
I was faced with a situation where I did not have a content management system all I had was
FTP access to a static web server so I could just dump files there and so everything on the website
is done in static HTML and I came up with a system whereby I could keep the main contents or the
contents of the main part of each page as a separate file and then build each page around it
using elements of a template so there would be a top navigation bar the main header with the
University logo a left navigation bar a footer and a tagline all of these things would be kept
separate from the content of the page and so I would if I ever had to make any edits to a page
I would edit the contents file and then build the page around it and this worked very well
in Bash but what I discovered was that if I wanted anyone else to do any kind of work on the website
then my life was difficult because other people did not have the same computer setup that I have
or the same skills so what I wanted was a way for me to set up a build environment for particularly
one colleague who I trust has decent computer skills and it was not scared of going to a command line
and running a command to build a web page I wanted to be able to do this in a cross-platform way
and so I converted my build script over to Python very happy to say that I've tested the build script
on my Linux machine on my little Mac over there on my desk and also on my son's Windows 8 laptop
and it works great on all three so I now have some I don't know prevention for the
hopefully unlikely event that I'm hit by a bus and there's nobody left who can
build the school music website and maintain it so anyway that was actually a fairly major
project let me see if I can find it here so I did a couple of
saying where is it I don't hang on I don't have the right script what I did was I have a separate
Python file where all of the the real business of the build is done and then
I'm sorry to where the heck is it make it means okay is that okay yeah here so I build a like a
function I suppose that will take care of all the sanity checks and assemble the template around
the contents page and that's all as a function in one file and then I call that from another file
that just has a couple of lines and it that says to go use the function in here so I essentially
built my own library I guess I says import this file and then run that command on it
and I do such things as what do I do some of the path manipulations this is one of the hardest
things for me to get my head around was how to deal with paths in Python you have to do it I mean
you can do it in a similar way that you did in bash but if you do it that way then you're you're
crippling the whole point of the of using a language like Python which is to make a cross platform
compatible and so instead of putting you know slash home slash user slash documents or whatever
to get to a file you have to do it in such a way that whether on windows linux or mac it will
understand what you want and do things the right way and so I learned how to do the os dot path dot
join method to join various things together I learned how to expand the user the the current
logged in user and stuff like that so anyway I sorted out how to deal with paths that was very
confusing but I'm starting to get the hang of it now and I'm getting more comfortable and I see
the logic behind it so I'm glad that I took the time to learn how to do that the right way the
build script really would not have worked on windows I don't think if I had not done it this way so
it was good that I was forced to do that I also have let's see one of the difficulties with the
system that I set up was that each page on the website needs to have a different title in the
you know the title is the part that appears in the the top bar of your browser the the page title
it also needed sometimes to have different left navigation images and you know you don't want
every page on the site to look exactly the same I wanted some flexibility and so the way I did
that in bash was to have at the top of every contents file a series of comments that held the
values that I wanted so if there's a left navigation image a page title a left nav image caption if
necessary an image title you know various things that would change from one page to another I
stored them in comments in a sort of more or less strictly formatted way at the top so that when
used the bash script it would use like I think it would just grip for the for the string that I
wanted and then it would strip out the other stuff and return only the value that I needed in Python
I kept those same comments at the top of the page only I would search the page for those and then
turn the resulting matches into lists and then just grab the elements from the list that would work
in there and so it's actually a little bit more elegant than the way it was done in bash and I think
probably more repeatable and predictable anyway it works and so that that was also something that
took a little bit of time to sort out I also had to learn how to make Python read from existing text
files and do various things to them that would include things like the the navigation bar on the
left side the header material and anywhere where there was something that might change from one
page to another like the page title and whatnot I put in what do I call them variables I guess
where in the the text file that I'm reading from there would be a curly brace zero closed curly
brace kind of thing and then when I open the file I tell it to format those curly braces using
the following variables and it so it plugs in the right values wherever there is one of those
little curly brace with an index number and it let me see one of the really cool libraries
that I found while I was doing this was one to smarten the punctuation I ran into a problem when
I first was running the build script on windows in that there was a certain unicode character that
it choked on every time made the build fail only certain pages had the character it was left I
think it was the the smart quotation like the unicode left double quote and the unicode right double
choked on those every time on windows and so I needed a different way to deal with the smartening
of punctuation so that you'd have the little curly quotes on either side of words
and I found I'm starting to get the hang of this whole
libraries and modules way of looking at things I'm starting anytime I need to learn how to do
something the first thing I do is check to whether see whether someone else has already done this
and has a module that I can download and that will do it magically and not surprisingly when I
checked for smartening punctuation I found that there is a library in python that will do this and
it's called smarty pants and one of the reasons I really like this module is because of the name it's
great to be able to have the word smarty pants in your code and not only is it funny it actually does
something really really cool and so now in all of the source files I don't put any kind of smart
quotes I don't do either the html entities where you do like the ampersand LDQUO semi-colon which
would do the left double quote nor do I put a unicode double quote I just put double straight
quotes and then when I the last step in the build of each page is to smarten the punctuation using
the smarty pants library and so all of that straight quotes everything is turned into smart quotations
automatically as part of the build process and this works wonderfully really really like the smarty
pants library okay and so the other thing that I had to figure out how to do was to automate the
building of the entire website from that and for that I had to learn how to iterate through
sub-directories and files within those sub-directories and that that took some thinking and some
working out but I got it in the end so anyway that that was my biggest project in python by far is
the creation of the build scripts in python for the school of music website and once I had those
set up I did the same thing for my personal website as well so I'm very very pleased with that
really stoked to be able to let my colleague do his own build script because he had been working on
some of the pages but in order to make it so that he could kind of preview the pages that he was
working on he would change URLs and stuff inside and so that when he sent them back to me I had to go
through and fix back all of the things that he changed and that was not good I did not like having
to do that so now he can work on the same files that I do and build them and preview them in the same
way and then just send me the modified contents files which I will then build and push over to the
web server all right so that was the biggest thing that I learned let's see what's next on my list
here so in conjunction with bladder one of the things I do the most is manipulation of text I will
select some text speak a command and it does various things to it whether it be putting HTML tags
around it capitalizing it correcting certain frequently misspelled words that come up when I
do my little dictation box kinds of things or whatever and so a lot of these things I've been
switching over to Python and I'm finding that Python handles them very well critical to this was
the discovery of the piper clip library piper clip is a cross-platform clipboard module for Python
and so I've found that when I select some text there are a couple of steps here the the basic workflow
is selecting some text speaking a command and the bladder command will have two parts to it
actually three parts the first part is running a virtual keystroke to do control plus C
which will copy the the selected text into the clipboard then it runs my Python script which will
take whatever is in the clipboard and do something to it either put something on either side of it
like a HTML tags or mark down formatting or something like that or it will run some library like
the title case library to capitalize all of the words that need capitalizing in a smart way for
titles and then after it's done that it sends it back to the clipboard and then the last part of
the bladder command is that it does a virtual control v which pace it back into my document
this all worked fine under bash but I'm finding that it works better in Python
if only because it's it's more cross-platform compatible and this is not only between Linux and
Windows or Mac for example but also just within different graphical libraries in my own desktop
here I use I'm running Debian with open box but one of the apps that I use the most is caliber
when I'm editing ebooks and I found that caliber does not play well with the tool that I had been
using to type out the results of all of these transformations the xvkbd package for whatever reason
xvkbd and q5 I think these are the problems they they do not work well together and so it's
impossible using xvkbd to execute any kind of uppercase or shifted characters and so I would speak a
command like I don't know ht header 2 is one of my commands that would put ht or they would
put htmail second-level header tags around the selected text and what I would come up with instead
of the less than h2 greater than thing would be comma h2 period because it could not execute the
shift and this is kind of a problem but what I found is that by using the Piper Clip library I can
actually have it go into the clipboard and paste it for whatever reason I was able to use under
bash I could use x clip to get text out of the clipboard and then process it but when I tried to
use x clip to put it back into the clipboard it wouldn't go like it wouldn't go in such a way
where I could do a virtual control v to make it paste back into the document so I was left with the
option of having it virtually type out the results and so that that didn't work very well but
using Piper Clip it seems to handle this much better it will take what's in the clipboard do
whatever you want to it and then send it back to the clipboard in such a way that the control v
keystroke sticks it back in your document so I do this for lots and lots of things
at first I was having a separate python script for every for every kind of operation that I wanted
so I would have one for putting paragraph tags one for putting bold tags for header two tags for
header three tags so I had like a dozen different python scripts that would just put each one
do a different kind of tag and then it occurred to me that I could streamline this whole thing by
having my having one python script that would accept a command line argument and that argument
would be whatever tag that I was going to put and so when I run the blather command instead of doing
p tags.py for paragraph tags it would just do I have a separate command now called html tags
which takes one argument and for paragraphs I would do html tags.py space p and that p
is then grabbed by the script and used as the element that's going to be put inside of the html tags
likewise I could put span like if I wanted to put span tags I would just run html.py
space span and it would use that whatever word or character I put after that whether it's h2
or h3 block quote whatever it is it will just stick that inside html tags and then put that around
the text that I have selected so that works really really well just putting tags around stuff
um I have another one for for wrapping a little bit more complicated things around it I call it
the wrap.py and what I do is I wrap something around the selected text and this one takes two
command line arguments the first one is whatever you want to appear before the selected text and the
second argument is what you want to appear after the select a text normally these have to go
inside single quotes so that you are sure that it's all treated as a single argument and that one
works very well too I do that when if I want not only to have for example put not only paragraph tags
around something but to also include something like a class designation like p and then class
equals no indent or something of that sort that has that can't really be handled as well
by the html tag script okay so that is how I deal with putting tags in there now I also
when I'm editing ebooks I have to do a lot of stripping out of superfluous tags and I found an
excellent Python library for this which is an extremely powerful one and one that I've only just
begun to to understand how to use but it's called beautiful soup this is beautiful soup for
specifically are also called BS4 and beautiful soup is a library that allows you like um
I know surgical like access to the html structure and so whereas I used to
do things like grip and said substitutions to try to substitute out one tag for another or to
strip out tags with beautiful soup there's a way simply to tell it what kind of tag you're looking for
and what to do with it and it just does it and it doesn't have to do any of the um hidden
mist kind of searches that I used to do so I really really like beautiful soup and I'm hopeful that
I can learn how to use it more right now I'll do things like strip out span tags or strip out
direct formatting that's done and where like inside a paragraph tag you'll see style equals and
then a whole string of style elements and these are the kinds of things that are done or that end
up there when someone takes a word process or document and it makes an ebook out of it when they
don't really understand how to use a word processor in the first place and so they do all kinds of
direct formatting they clearly have not listened to a hookah's series on how to use a word processor
but you can strip out a lot of that kind of stuff using beautiful soup and the unwrap method
let's see so that's from my clean html script and I have various ones to strip out particular html
tags and again like I run a no tag script that takes one command line argument which will be whatever
tag that I want to strip out of the selected text and it will do the unwrap method on that and
get rid of that tag let's see so another way that I use this is to insert snippets of text
by using voice commands and these snippets could be anything from very very short things like
I don't know CSS rules or things like that to entire templates for lily pond files or html pages
or things like that and so I have a template.py script where I have a whole list of my frequently
used templates right now there are 14 of them and each I have as a comment after each one an index
number so that I can easily see which index number I use and when I run the template.py script
I use one command line argument that tells which item from the list to use so for a
a single staff in lily pond I'll use template.py space one and that argument number one is used
as the index to grab the right file from the list of templates and it will copy that again this
is something that is copied into the clipboard by the piper clip library and then it's pasted into
my text editor using a virtual control v. Also I made a script that helps me a lot in my editing
of the counterpoint book that I've been working on which is I've been going through and identifying
all of the prelude's fugues inventions and whatever by JS Bach that the author uses as musical
examples because all he does through the whole book is just as Bach he says this example is by Bach
but he doesn't tell you what it is or where it comes from or anything and so part of my job as an
editor I feel is to identify all of these specifically each one of those pieces will have a bunch of
information that is pertinent about it like the prelude number two in whatever key BWV number
this and it was hard to keep track of all of these things because every piece has a different BWV
listing number and so what I did was made a Python script that has all of the information available
as a list and then to insert let's say I realized that this is invention number eight from the
two-part inventions what I do is in my text editor all I do is type the number eight and then speak
the command invention title and it knows to select that number eight and use that as the command
line argument for finding the right string of text from my list and then it pace it into my
document for me it's really really magical because I'd never make mistakes anymore about BWV numbers
or keys or anything of that sort absolutely wonderful so this Python is beautifully suited for
this using lists and the piper clip thing to stick in this clipboard and it's starting to run
long here so I don't want to go too much further with this but finally I use Python for fun stuff
too a lot of the things that I used to do with you know interacting with my laptop using bladder
like I would ask it what time it is or what day it is or what should I have for dinner or what for
breakfast or if I tell it thank you then it used to choose from a list of commands it like it would
use the shuff command and shuffle all of the possible thank you responses that I had told it
and then choose the top one and then run it through eSpeak well in Python I found that there is
a library called pi ttsx which allows you to get at the system text-to-speech engine whether it
be on windows linux or mac and you can tell it how fast to speak and all you have to do you know
you tell it you set up a couple of things like you get it started with the init method and then
you tell it the rate that you want it to speak you know this is optional it'll it'll speak at the
default rate if you don't tell it but then you just tell it what you want it to say and it will say
it and this works on mac windows linux and so in that way it's nice to have it across platform so
what I do is I I made up various Python scripts that will use the random library and choose a random
integer and use that to figure out which item from a long list of things to bring back and speak to
me so it's fun you know I like to use Python and the pi ttsx things to have my computer talk to
I wrote this one called coffee dot pi I'm really proud of this one and maybe I'll put a link to the
YouTube video where I demonstrate this but I ask my computer whether it has had too much coffee
and first it does a coin flip using the random dot rand int method it chooses a random integer
between zero and one for a coin flip and if it turns up zero that means it's already had its coffee
and I set the speech rate or no if it comes up zero that means it has not had coffee yet
and so it's really sluggish and slow so I set the speech rate at only 90 and it it talks very slowly
like this and I have a whole series of things that it might say and then if it turns up one on
the coin flip then that means it has had its coffee so the rate is set really fast and it speaks
that the response is super duper fast so I get to use a conditional statement there and I get to
use lists and yeah it's fun I've also used the beautiful soup thing to create a major league
baseball xm radio schedule for me I find I used to I have a way to do this with bash and I actually
have my raspberry pie run it every single day and it emails me a list of all the games that are
going to be on the radio and what channels they're on but it's it's a very much hidden mis proposition
with making sure that I get the right schedule for today I have to search for the word today and
then search for the word tomorrow and then leave out everything else and then do a little bit of
formatting what I found with beautiful soup is that I can use the url git
I don't remember exactly what it is but I I grab the content of the page and then I use beautiful
soup to extract only the one table I want and it goes right there with laser like precision
grabs that one table for today's games and then I can do some formatting I actually have it build
a little HTML page and then put it in place on my server so that I can go to this one major league
baseball page on my server and see today's games and it runs every hour so that by the end of the
day it only shows the one or two games that are still left to happen so I really like that also
got a weather script that uses the pi w api pi weather api library and it gets weather from weather
and so you have to set you have to tell it the code for your area minus us la 0261 for
Lafayette Louisiana and when you grab the current conditions from there it brings back a dictionary
and some lists in there with all kinds of weather information and I've set up in my script just
the things I want like I wanted to tell me currently it's either cloudy or sunny or whatever
tells me how many degrees it is the humidity and tells me the perspective high for today and the
chance of rain and then it gives me a tomorrow's high and tomorrow's chance of rain as well
and then it actually speaks that to me using the system text-to-speech engine so I use the
pi ttsx library for this script as well anyway I see that I've hit the 30 minute mark so I should
probably stop now but anyway I hope you guys have enjoyed hear me talk about how I use Python
again I'm not really a programmer and so I don't write programs so much as I do just little
things that help me get by and help me have fun so yep that's it I will talk to you guys some other
time maybe with more about Python if I actually learn anything else thanks bye
you've been listening to hecka public radio at hecka public radio dot org we are a community
podcast network that releases shows every weekday Monday through Friday today's show like all our
shows was contributed by an hbr listener like yourself if you ever thought of recording a podcast
and click on our contributing to find out how easy it really is hecka public radio was found
by the digital dog pound and the infonomican computer club and it's part of the binary revolution
at binrev.com if you have comments on today's show please email the host directly leave a comment
on the website or record a follow-up episode yourself unless otherwise stated today's show is
released on the creative comments attribution share a live 3.0 license