100 Days of Python, Projects 32-40 #100DaysofCode

Things are ramping up a bit again here.  Now we get all sorts of fancy API calls for real time data.  This will definitely be useful for later personal projects, along with the email notifications.  I will probably be running dozens of scripts on a Pi or my web server by this time next year.  This is the sort of stuff I really dig, keeping notice of what’s going on with things.

As usual, my Github Repository for all this code is here.

Day 32 – Sendmail (Birthday and Quote Email-er)

So, I am sure on some level, I was aware that “Python can probably send an email”, but I’ve never really looked into it much.  But actually making it happen is a lot cooler, and it’s simple enough that frankly, I will probably build a class that I can stick in other projects, later.

Also fun, I already have a paid web host that can run cron jobs, or even just a bunch of Raspberry Pis, so anything I made on this next section of projects, I can easily automate to actually use right off.

The bigger thing I actually learned how to work in this chapter, is .gitignore.  Which was not part of the lesson.  I set up a fresh email account for sending mail with these scripts and it’s using easily revoked App Passwords, but since I am posting my work to Github, there is a risk of accidentally posting a file with actual authentication information to the public.  Fortunately, my experience in PHP kicked in immediately with how to handle log in data properly.  For starters, it all goes in it’s own separate file and imported.  This makes it easy to edit later if needed.  It also means, I can tell Git to ignore these files when posting, using the .gitignore file.  

Because I am keeping all of the projects (so far) in one repository, it did become tedious to .gitignore every copy of auth.py, I found that adding “**/auth.py”, which is a bit of bash shell script syntax, is enough to omit all iterations of auth.py, automatically, across all sub-directories.

So my local copy runs fine, but the online copies are all missing the auth.py files.  Easy.  And if I fuck it up for some reason well, you can’t log into the gmail account with the app token, so the account is safe.  It’s used only for this purpose, so nothing of value is on the email of that account, maybe someone could send emails as that account, but chances are I would realize I screwed up and immediately revoke and reset the App Password.

Day 33 – ISS Notifier

It’s too bad the International Space Station (ISS) is getting de-orbited soon, this nice little tool will, unfortunately, be useless soon.  Anyway, this lesson was more work with API calls.  The app pulls the current position of the ISS, in latitude and longitude, then compares it to the lat/lon entered for the app (the user’s position).  It also makes a second API call to check when Sunrise and Sunset are locally, to find out if it’s currently night time out, since you can’t see anything during the day.

If it’s night time, and it’s within 5 degrees Lat/Lon, of the user, it sends an email to notify that the ISS is potentially visible overhead.

I actually took this one and put it on my web server running every ten minutes, so we’ll see if I ever get any emails from it.

We also made a silly little Kanye Quote app, that pulls and populates Kanye West quotes to a quote box with a little cartoon Kanye.

Day 34 – Quiz Game Revisited

I called it, for the most part.  Back on Day 17, we made a simple text Quiz Game.  And the way it was put together, I was pretty sure it would come back to get a GUI in a future lesson, and here it is.  Additionally, it got a nice API call to the OpenTrivia API to get fresh questions every run.  

We also configured the Tkinter interface as a class, instead of as part of the main code, which was different, but not super tricky.  

I want to go back and add a title screen with options for topics, number of questions, and difficulty later.  I did make a few simple modifications.  I didn’t like the Red/Green for Incorrect/Correct, so mine just says “Correct” or “Incorrect”.  I also added a final splash screen of the final score.  Additionally, I use a try/except in the data gathering code, and left the original data in the file.  If the API fails to load, it will load the original tech questions I built for it.

Day 35 Rain Notifier

So, the main purpose for this app was to use apps which require authentication.  Also, to send a text message instead of an email, using Twilio.

Getting the weather wasn’t too tricky, except that the suggested API seemed really invested in getting a credit card from me.  I’m not adverse to this, and it looks like I would get 1000 free calls per month, way more than I would need at about 33 per day, but I opted to just use an alternative API.  The alternative I believe will only work for a bit before the “free trial” runs out, so this won’t be as useful in the long term without revision.  

The point is more the practice anyway.  There seemed to be a LOT of potential responses that included rain, so I made sure the code would loop through and check for any of them.  I didn’t use them, but I also picked out the responses that related to Snowy weather as well.  

I also used an alternative method for the sending a text.  Somewhat tangentially related, a few months ago, I played, most of this little Programming Practice game called “TwilioQuest”, which is put out by Twilio.  I skipped the API portion of that entirely.  It seems that because I created my account while using a VPN, it became instantly locked.  I decided it wasn’t worth the hassle of unlocking it at the time, and it’s come up again for the exercises in this class.  

Fortunately, I can’t just send text messages by email.  Specifically, sending to (PhoneNumber)@txt.att.net will send an email to that number, on AT&T.  I am pretty sure this works on other carriers, but the to domain is of course going to be different.  The easiest way to find this is to send a “text” to an email address, since the phone number will be the “From” address.

In the end, I don’t really WANT texts anyway, so If I use this, I’ll just set it to send me an email instead anyway.  In fact, going forward, for these projects, I plan to just omit the Twilio/Text part and substitute in a regular email.  I just, prefer those. I even broke out my emailing code into it’s own class, which has been extremely useful for follow up projects since I can just drop that code into a new project, import it, then make a “Mailer” which I can call with the text of the message to send emails.

Day 36 – Stock Trading News Alert App

Just a more complex version of the “APIs with Authentication” as well as this lesson’s first “Do it all on your own” project with bare bones instructions.  It also involves two APIs, one for the Stock Prices and one for News Stories.

The end project, if a tracked stock price changes more than 5% up or down from yesterday and the day before yesterday, look up the top news stories about that stock, and email them to the user (originally text).  The overall project wasn’t that complicated, though I did make mine a bit more complicated than it needed to be.  Specifically, for the Stock Data, I created a function that accounts for the weekends when calculating yesterday and the day before yesterday.  It wasn’t hard, but, I guess the API data essentially already does this if you just read in sequence instead of by date.  

Oh well.  Both methods work fine and I used the dates in the News Story algorithm.

i also converted the whole system to work in OOP, so I could create several “Stock” classes for several companies, which were cataloged in an array and it would cycle through each one, maybe on a Cron Job and oh shit never-mind, the free version of the stock API is extremely limited on calls per minute.

Also oh well.  I don’t really follow stocks that closely anyway.

Day 37 – Habit Tracking App

I went ahead and put this one down on my “List of programs to improve later”, specifically, add a GUI, maybe the ability to manage multiple habits.  It’s a neat idea and concept and I like the API backend it used.  The real “hold up” is that I kind of already use a pretty robust Habit Tracking App that also does moods, called Daylio. I’ve been using it for…  it says 1740 day is my streak, which is almost 5 years.  Note to self, make a Daylio Blog Post in about a few months.

This is straight habit tracking/making though, without the mood.  So it still may be useful for tracking quantities over just straight “ifs” like Daylio is tracking.

Day 38 – Fitness Tracking with Google Sheets

Similar to last lesson, I already have apps that do this, though I am pretty sure most of the point here is to learn how to use the Google Sheets API.  Which actually will be useful for me, because I already have a few older, jankey, programs I have written that use Google Sheets as a back end, but make POST requests through Google Forms.  Connecting directly would be way way cleaner.

The more interesting part was working with an API that does natural language processing for work outs.  I kept trying to feed it weird things and it did a good job of translating it into the sheet dump.  Like “I climbed Mount Everest for 50 minutes” was found as “Rock Climbing”.  It also filled in for “I ran a marathon”.  Which was interesting.  It seems to have some basic idea of lengths to use.

Day 39 and 40 – Flight Deal Finder Capstone Project

I skipped the Day 39/40 Project.  Unfortunately there aren’t even any videos to watch over, but there is code to look over, so I did that. I literally, never, every fly.  And I don’t see myself every really having any use for this.  My wife isn’t going to fly, for a variety of reasons.  The only way I might fly, is if my job required it, but that’s going to be covered by my job, through their own travel management website.

Also the API wanted my payment information.  I’m not super keen on that one, especially because it seemed to want to connect to a bank account and not just some throw away CC number or an easily revoked Paypal connection.  

I updated the missing IATA Codes in the Google Sheet using Sheetly, and it’s PUT request API, which was the main “actually new” bit here.  I didn’t have the API for getting real IATA codes (see above) so I just took the first three letters for each city, upper cased them, and had the API put them into the sheet.

Also, Sheety has a limit on it’s free tier of 200 calls per month, and I’ve probably already used about half of them just testing the IATA Codes and the previous day’s code. Looking through the comments, it seems like everyone kind of disliked this project.

And this feels like a good breaking point for this section’s post.  The “Intermediate+” section isn’t done, but the next few lessons switch to a general over view of the web and HTML, before getting into other APIs.  I expect these next few lessons to go by quickly as well.  I am already well versed in HTML/PHP/CSS/etc, and there don’t really seem to be projects associated with the lessons.  I do plan to watch through them (mostly) to see if there is anything I can glean from them.

100 Days of Python, Projects 23-31 #100DaysofCode

In case anyone is keeping track (which they are not), you might notice that one, I’ve not been posting these posts in “real time”.  I’ve been writing them later, after the fact.  I kind of hope to change that.  Also, I started on September 12th, and my last post was on September 28th, which is 16 days, though the last post went through Day 22.  I’m actually moving faster than “100 Days” on my “100 Days of Code”.

This is intentional, I want to do Advent of Code again this year, starting December 1st. So I want to hopefully finish this up before December, so I am not overlapping these two daily level coding projects.  I just, don’t have the time to do both.  Based on my calculations, I need to get 18 projects ahead to end on November 30th.

Anyway, this round I want to wrap up the “Intermediate” level projects.  This one may be a bit boring, just because it is a lot more “Basic Data Analysis” projects than “Cool Retro Game remakes”.

Day 23 – Turtle Crossing or Crossey Turtle

I was a little disappointed because the description sounded like this was going to be a “remake” of Frogger, only with Turtles.  Instead it’s a remake-ish, of Crossey Road.  Which is slightly less exciting.  I feel like I could probably make my own Frogger, and I will probably put that on a ToDo list somewhere to forget about because all my ToDo lists are 1000 items long….

Anyway, the job is to get the turtle across by avoiding the cars, which randomly spawn on one side of the screen.  Each time the turtle makes it, the cars start moving slightly faster.  It was an interesting project to develop, and like Snake, I cleaned it up a bit from the instructor’s code by aligning the cars and turtle to lines.  The main this this does it helps to remove the ugly overlap with a pure random spawn.  I considered adding some code to stop them from spawning on top of each other in the same line but decided I didn’t care enough to bother.  

The trickiest part of this was adjusting the delays and spawn timer so things felt right.  With a spawn timer too low, then cars were just pouring out.  With it too high then the road is just wide open.  Another issue I had briefly, when the cars would speed up, only the currently on screen cars would speed up, everything new would still run slow.

In general I don’t really care for this game, primarily I think because the game actually gets easier the faster the cars get because the gaps become huge and easy to pass through.  It could be adjusted a bit probably to spawn more cars at higher speeds which may fix this.

Day 24 – Improved Snake Game and Mail Merge

Today’s lesson had to do with file I/O.  I’m actually already pretty familiar with this, a lot of the scripts I have written in my free time take a file in, manipulate the data in some way, and spit a file out.  

Step one was to modify the Snake Game to have a persistent high score written to a file.  Which was neat.

The second was a simple Mail Merge exercise.  Take in a letter, take in a list of names, output a series of letters with the names inserted.

Day 25 – US State naming Game

The exercise today was an introduction to Pandas, which is a tool that makes working with large data sets much easier.  In the first round of practice, we pull data from a large csv of Squirrel Data and count Squirrel Colors from here: https://data.cityofnewyork.us/Environment/2018-Central-Park-Squirrel-Census-Squirrel-Data/vfnx-vebw which was silly and fun.

The second was a little learning game where the player names states of the United States.  Each correct guess shows on the map.  When you type “Exit” you get a file back of which states were missed.  One of the interesting things here was adding a background image to the Turtle Screen.  I also tried to build it in a way that it would be easy to say, swap out the background image and data set for European Countries, or anything similar.

On a total side note, I got all 50 states on my first try.  I am pretty sure I know all the capitols still as well.

Day 26 – List and Dictionary Comprehension and NATO Alphabet Translator

Probably one of the quickest programming days but longer lessons page with several simple exercises.  It also felt like one of the more “necesary in the long run” lessons.  Basically, the whole lesson was about better ways to iterate through data.  Single line codes like “list = [item for item in other_list]”.

The end project was very short, pull in a CSV of NATO Phonetics using Pandas, then one line iterate it into a dictionary.  Ask for a user input, loop through the user input to make a list of words, output the list.

Day 27 – Miles to Kilometers Conversion Project and TKinter

So on the surface, a Miles to Kilometers converter isn’t that exciting.  In fact, it’s a straight forward multiplication/division that even the most beginner level coder could whip up.  The point of this lesson was more of an introduction to TKinter.  

The end result is a little window based converter, that in my case, goes both ways, so you can do Miles to KM or KM to Miles, which is probably boring but wasn’t part of the requirement for the project.

Day 28 – Pomodoro Timer

The object of this lesson was to build a Pomodoro Method Timer.  I don’t super follow “productivity” methods but I guess the Pomodoro Method is to use a timer, work for a period, take a short break, repeat this pattern a few times, then take a longer break.  I don’t know if these times are set in stone, but for the project we used 25 minutes of work, 5 minutes of break, repeated 3 times, and then 20 minutes of long break.

The actual lesson here is how to create a TKinter window that is “constantly running”.  While able to take some input, and manipulating various bits of the UI.  It counts down a timer, it adds check marks every round, it had a start and reset button.  

This one was actually pretty tricky, and I kept ending up with essentially several layered running timers because I was using the window.after function improperly at first.  Otherwise it’s just a loop that runs through some conditionals to see where it’s at in the loop.  I did manage to do like 90% before watching the videos, but it took me an extra session of code work to puzzle it out.  The one bit I wasn’t sure on was how to make the Reset button actually top the timer loop, which wasn’t complex, but I did have to get that one from the class.

Day 29 – Password Manager Part 1

So, I really liked this project, and I need to find out how to make exe files out of these pieces of code.  I actually would definitely use this project regularly, especially after adding some additional features, which may in fact come in tomorrow’s part 2.  This was also another project that I essentially completely built before actually watching the lessons.  Which I suppose means the previous lessons on how to use TKinter worked well.  I will say, the most annoying part of TKInter is just how finickey the positioning can get,  The instructor likes using Grid() to place but personally, I think I like using Place(), but the problem is, I could get caught up tweaking Place() for hours.

This project so far has combined several previous projects, from the Password Generator, to using TKinter, to writing output files.  I suspect tomorrow will add in even more with Reading Files, and working with “organized data” in the process.  I could also see using Ciphers to actually encrypt the output data file to something that’s not readable without a master password.

My favorite part was revisiting the “Password Generator” code created back on Day 5.  In addition aot simplifying it with some list comprehension, I also converted it to be an Object Oriented Class, that could be imported and called. Not super difficult but I still felt pretty proud of that one. It doesn’t really add a lot of benefit, but it was more just an exercise for my own ability.  I also made it easy to change the default email address by putting it in a separate config style file that I may work with more later.

Day 30 – Password Manager Part 2

Most of this day’s lesson was using error handling, specifically, “try:”, “except:”, “else:” and “finally:”.  I kind of figured there was a better way of handling “random errors” than painstakingly considering every option, but I’ve never really had a need to look into it too super deeply.  Usually for error handling, I’d just stick an “if” into a loop, and if the “if” fails, it just keeps looping until the input is good.  

I did this a lot early on, on my own accord, for many of the text inputs.  I would make an array of “valid entries”, often something like:

valid = [“yes”, “y”, “no”, “n”]

while answer not in valid:

    answer = inout(“Yes or no?: ).lower()

Which honestly is probably still plenty valid for small choice selections.

I was a bit disappointed that the end result didn’t even make a token attempt to encode the output data, storing passwords in plain text is a really really really really really really bad idea.  Oh well, future project.  

All in all, this is definitely the most complex project so far, and it’s possible the most complex single project I’ve done ever code wise.  I’ve done some elaborate PHP/HTML/CSS web stuff, but those were always more, ongoing projects with sub-projects tacked on to them.

Day 31 – Flashcard App

Another one that I definitely will use once I make it work in a stand alone fashion in the future.  This app was quite a bit simpler than the Password Manager but I’ve found that so far, at the end of each section, there has been a complex capstone project, then a few simpler ones.  The bigger purpose of this project was learning to manipulate JSON and CSV data better, in addition to making a simple GUI.  

This app reads some data in, in this case, language words, but it could be any set of data.  Then it shows the foreign word, for three seconds, before showing the English word.  If you guessed it correctly, you hit the green box, it not, hit the X.  It’s kind of an honor system thing, but if you are trying to learn a language, why would you cheat yourself like that.

It also removes “learned words” from the pool of possible cards, permanently.  Well, or until you reset it by deleting the save files.

I’ve written a few times about my goal to learn some languages besides English by 2030.  I’ve gotten alright at Spanish, and I’ve been working on Norwegian.  The class provided a word bank for French but also explained how to make your own word bank, which I did.  Making a new word bank was not required, but I know absolutely zero French, and it was hard to tell if the cards were accurate.  So I built a Norwegian Word Bank, and will definitely build a Spanish one.  The only problem is, I went with the “Top 10,000 words”, and you only really need the “Top 2-3000 words” to start grasping the language.  So my custom word bank helped, but not as much as it could have, because I started getting some really lesser used words.  It’s easy enough to trim it off though, I can just open the data file and delete the bottom 9,000 lines.

And so this wraps up the “Intermediate” lessons.  The next session is “Intermediate+”.  Based on the daily topic headers, it looks like this next section is going to delve way deeper into using APIs to gather and manipulate web services, which should be a lot of fun.  I’m definitely learning more and I plan to revisit some of my own personal projects to make them much better once I’ve finished with the course.  A lot of my old data manipulation involved lists and a lot of if/else statements inside for loops.  Which is messy, and probably slow.  A lot of the new tools I’ve learned will really help.