Transcript #375: Pointing at Countries

00:00 >> Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds.

00:05 This is Episode 375, recorded March 19th, 2024.

00:10 I think I got that right. I am Brian Okken.

00:13 >> I'm Michael Kennedy.

00:15 >> Our episode today is sponsored by Scout APM.

00:19 Thank you, Scout. Listen to their section later in the show.

00:22 If you'd like to connect to us, we're on Mastodon and the links are in the show notes,

00:27 but it's @mkennedy, @brianaukin, and @pythonbytes, all on fosstodon.org.

00:33 If you'd like to listen to this show live, we'd love to have you, but it's not necessary.

00:37 You can listen to us later.

00:38 But if you want to listen to us live, go to pythonbytes.fm/live and see all the details and see when the upcoming episode is.

00:47 Michael, what do you got for us?

00:50 >> Well, I would like to start off by telling you about another country,

00:54 maybe one you haven't heard of.

00:55 >> Canada?

00:56 >> You've heard of Canada, you've heard of the US, you've heard of Germany, you've heard of Slovakia,

01:00 you've heard of the Philippines.

01:02 Have you heard of Pi country?

01:04 >> Oh, this should have a song with it.

01:06 In a Pi country.

01:08 >> Does it involve a banjo?

01:12 I don't know what it's called.

01:13 >> I hope not.

01:14 >> I know. I hope not too.

01:16 >> We should have covered this on the 14th.

01:19 >> Pi country, Pi day, all the things.

01:21 So Pi country, it solves something that might not sound like a great big problem

01:26 until you try to deal with this and you're like, are you serious? There are this many things.

01:30 It's similar to dealing with time zones.

01:32 You're like, how complicated can be?

01:34 What are there, 24?

01:35 No. What are their abbreviations?

01:37 What are all the different ways in which they can be abbreviated and so on?

01:40 This is that for country.

01:41 So it's a Python library to access ISO.

01:44 These are standards, country, subdivision, what's a subdivision, language,

01:49 currency, and script definitions, and their translations for all the countries of the world,

01:54 including deleted countries that used to be countries but are not.

01:58 I would have considered those just former countries, but deleted is a very digital way to think of it.

02:03 We just selected that part and we hit command backslash and it really deletes it all the way.

02:11 Anyway, this is really interesting and it's used by 17.9 thousand different projects, which is pretty cool.

02:19 So here's the deal.

02:20 You go in here and somewhere, you can just go into it and you say, pycountry.countries and there's 249 here.

02:27 You get these, not just a list of strings, but you get rich objects back.

02:31 Like the first one is Afghanistan.

02:34 So country, it's a class.

02:36 It has the Alpha 2 code, which is AF, the Alpha 3 code, AFG, and that's the kind of stuff I'm talking about.

02:42 That's like, why are there so many variations?

02:44 You can represent the shorthand version like just two letters, but you can also use three letters or there's numeric versions.

02:51 There's like the full name versus the colloquial name.

02:54 So Afghanistan versus the Islamic Republic of Afghanistan.

02:58 >> Oh, this is cool. I might learn something.

03:00 >> Yeah, exactly. There's a lot more in here.

03:03 It has a dictionary-like access for the variation.

03:07 So you can say, get Alpha 2 equals DE and you'll get good old Deutschland,

03:12 or you could say Alpha 3 equals DEU or the name equals Germany or all these different ways in

03:20 which you can query this or given one, you can obviously just access all these properties.

03:25 Like Brian, you would know that it's the Federal Republic of Germany.

03:29 It has fuzzy searching for country.

03:31 So I can say, give me England.

03:32 It's like, yeah, there's no England.

03:33 What? But there's this thing called United Kingdom, which has its center in England maybe,

03:38 and that's GB and so on.

03:41 So yeah, pretty neat.

03:42 You can say fuzzy search for Cote and you get Cote d'Ivoire.

03:47 Sorry, folks, from there, but it'll even match when it's not exactly the right character.

03:56 Oh, hat, I don't know what that's called.

03:58 Not an umlaut, but a hat-looking thing.

04:00 But you don't have that in the search.

04:02 So this is more, like I said, there's more to this than just a list of countries.

04:06 Then it's historic countries.

04:09 These are the ones. Since I read this, did they change it from historic to deleted?

04:12 I don't know. Historic countries, subdivisions.

04:15 Like what is a subdivision of a country?

04:17 This is like states or providences.

04:20 It's not nearly as strange as it sounds.

04:22 There's scripts also. What is a script?

04:24 It's written language versus spoken language.

04:27 Those might be the same thing, but they might also not be the same, like Latin, for example, right?

04:32 Oh, yeah.

04:33 Currencies are in here.

04:35 Languages are in here.

04:36 Locales are in here.

04:38 So if you do anything with countries, this might be worth looking at.

04:41 That's pretty cool.

04:42 Yeah.

04:43 It's still like hung up on the England isn't a country, but okay.

04:46 According to Pi country, it's not a country.

04:50 Okay.

04:50 And also, so yeah, I'll have to look into this more.

04:54 Is it a subdivision?

04:55 Maybe it's a subdivision.

04:56 And, but also like, where does the word Deutschland fit in for Germany?

05:02 Is that not even listed?

05:04 Yeah. Well, also since it has translations.

05:07 So for example, German, if you ask a German person where they're from, they'll say Deutschland,

05:12 unless they're speaking English to people, right?

05:13 Like they refer to it as Deutschland.

05:15 Yeah.

05:16 But for some reason, we don't agree.

05:17 We don't call it that.

05:18 We call it Germany, right?

05:21 So I don't know.

05:22 It's interesting.

05:23 I suspect if you use the translations maybe, and you said, you know, what is the name,

05:27 the official name of Germany?

05:29 It probably wouldn't say the Federal Republic of Germany.

05:31 We have that translated, right?

05:34 Interesting.

05:35 Yeah.

05:36 Like Switzerland is, Sveits?

05:39 I can't remember exactly, but yeah, it's different.

05:41 It goes by different names in different languages.

05:44 So I think it also has those when it talks about the translations.

05:48 Cool, neat.

05:49 Well, we could point to countries.

05:52 I'm trying to do a transition.

05:54 Or we could point to objects.

05:56 And does Python have pointers?

06:00 You can't really point to a country.

06:01 I don't know.

06:02 Maybe.

06:03 Point to directions.

06:04 Canada's that way.

06:07 Canada's that way?

06:08 No, Canada's this way.

06:10 Oh man.

06:11 But we're probably pointing to different directions.

06:14 Anyway.

06:15 So Ned Batchelder has an article called, does Python have pointers?

06:19 And I'm like, yeah, no, it doesn't.

06:22 But kind of does.

06:24 Anyway, I'm highlighting this because I think it's a really good article

06:28 for somebody that's coming from a language that does have pointers and stuff to Python

06:35 because it's a good entry way into names in Python.

06:38 So the article does Python have pointers, but it's kind of pointing to the idea

06:45 of like talking about the ID function.

06:48 And I know about the ID function really well.

06:52 So if you say ID of an object or a variable name, it tells you like a number for what it points to.

07:01 Like where the, I don't know, it's a representation of the memory address or something.

07:05 That's kind of what a pointer is, right?

07:07 But you can't in C and C++ and Rust and other things, you can dereference it.

07:11 Using that, you can create a variable.

07:14 Using the address, you can create a variable that points to that thing.

07:17 And you cannot do that within Python.

07:19 So it is different, but also kind of everything is a pointer in Python.

07:24 And that's sort of Ned's point is, point, pointer, is that with like just normal object,

07:32 we can say like my var equals 17, and then we can have, if you do my pointer, it doesn't help you any.

07:40 But you can do another variable that points to the same thing and it works.

07:46 It's when, it's just sort of how names work.

07:49 I'm describing this very poorly, but this is a good entry point into, well, how objects work.

07:55 And luckily, Ned also links to a talk he did, names that refer to objects.

08:00 And so he did a talk in 2015, and still the slides are up in the video, but even the slides just going through it,

08:08 it's really good to understand just really how names work.

08:12 We're just, in Python, we just kind of point to things.

08:14 And so I'm gonna link, we're gonna link also to the 2015 article, Python Names and Values.

08:21 And really, this was the trick.

08:24 When I really could grok this thing with Python, then I could understand Python.

08:30 Everything else seems easy after this.

08:32 So good job, Ned.

08:34 - Yeah, Ned always has good writing.

08:36 I'm subscribed to his RSS feed.

08:38 I think it's interesting to consider whether Python, you know, the article, does Python have pointers?

08:43 Like, boy, howdy, does it?

08:44 (both laughing)

08:45 - Yeah, yeah.

08:46 - Like, it has more pointers than C++, which is insane.

08:50 It doesn't have a wider variety of ways to like dereference them, you know, as stars,

08:55 like casting a void star, star to something else.

08:57 But everything, literally everything in Python is a pointer.

09:00 Even numbers are pointers.

09:02 Whereas in C++, numbers usually are value types, right?

09:06 And you can have stuff that's on the stack that's not a pointer, just has the value.

09:09 In Python, there is no way to have just a value.

09:12 Every, everything is a pointer, often pointers to pointer.

09:15 You know, you're traversing something to something, right?

09:17 It's like, I'm going to find the dictionary of the class, and then I'm following that to where it points to

09:23 to get to the value.

09:24 So it's worth spending some time on, even though you don't ever see a star

09:30 or an ampersand in the context of like juggling pointers in Python.

09:34 - Yeah, and like vectors or lists in Python are not, they're not the start of a chunk of memory

09:42 that represents all of your items.

09:43 That's not what's going on.

09:45 It's something completely different.

09:47 Forget that.

09:48 So, yeah.

09:49 - Indeed, indeed.

09:51 Very good, very good.

09:52 - Well, changing gears a little bit, we'd like to thank Scout APM for sponsoring this episode.

09:58 - Yes, we would.

10:00 Let me tell you real quick about Scout APM.

10:04 They're big supporters of Python Bytes, so we appreciate that very much.

10:07 So if you are tired of spending hours trying to find the root cause of issues

10:12 impacting your performance, then you owe it to yourself to check out Scout APM.

10:16 They're a leading Python application performance monitoring tool, APM, that helps you identify and solve performance abnormalities

10:24 faster and easier.

10:26 Scout APM ties bottlenecks such as memory leaks, slow database queries, background jobs,

10:31 and the dreaded N+1 queries that you can end up if you do lazy loading in your ORM,

10:36 and then you say, "Oh no, why is it so slow?

10:39 "Why are you doing 200 database queries "for what should be one?" So you can find out things like that.

10:43 And it links it back directly to source code, so you can spend less time in the debugger

10:47 and healing logs and just finding the problems and moving on.

10:50 And you'll love it because it's built for developers by developers.

10:53 It makes it easy to get set up.

10:55 Seriously, you can do it in less than four minutes, so that's awesome.

10:58 And the best part is the pricing is straightforward.

11:02 You only pay for the data that you use with no hidden overage fees or per seat pricing.

11:07 And I just learned this, Brian, they also have, they provide the pro version

11:12 for free to all open source projects.

11:14 So if you're an open source maintainer and you want to have Scout APM for that project,

11:19 just shoot them a message or something on their pricing page about that.

11:22 So you can start your free trial and get instant insights today.

11:26 Visit pythonbytes.fm/scout.

11:28 The link is in your podcast player show notes as well.

11:31 And please use that link.

11:32 Don't just search for them because otherwise they don't think you came from us.

11:36 And then they'd stop supporting the show.

11:37 So please use our link pythonbytes.fm/scout.

11:40 Check them out.

11:41 It really supports the show.

11:43 - Yes, it does.

11:44 - Thank you.

11:45 It does, it sure does.

11:45 Now, Brian, let's talk about ingestion, data ingestion.

11:50 - Okay.

11:51 So here's that open source CLI tool.

11:54 It's not exactly Python focused, but certainly useful for Python people.

11:59 Ingestor, and it's straight out of the web 2.0 days 'cause it's dropping some of the vowels.

12:05 It says copy data between any source and any destination thinking database type things.

12:12 So what you do is you just say ingestor, ingest.

12:15 Source database connection is whatever.

12:18 Source table is whatever.

12:20 And then the destination is where it goes.

12:23 And all of a sudden now you have that.

12:24 So do you want your Postgres data to appear over in BigQuery?

12:29 Do you want your MongoDB collection to show up in Postgres?

12:33 One CLI command done.

12:35 - Oh, that's pretty cool.

12:36 - It's pretty cool, right?

12:37 - Yeah.

12:38 - So it's a command line application that allows ingesting or copying data from any source into any database.

12:45 I mean, this guy probably should be a star on any.

12:50 How about many, many data sources?

12:52 - Okay, yeah.

12:53 - Yeah, so if we jump over here to their docs where like their homepage is kind of just

12:59 to catch your interest.

13:01 So you can copy, I'll find the sources and tell you about them in a second.

13:03 But it says there's three ways to copy stuff over.

13:07 You can do an append, which is kind of cool.

13:09 Like if you've got multiple data sources that are localized saving of data

13:14 and you want to put them all into one place.

13:17 Like if you had a bunch of SQLite files from different locations, say you've got a bunch of,

13:23 I don't know, a bunch of IoT things, or you got a bunch of servers and they each keep some data in a SQLite file

13:30 and you want to put that into one big database, right?

13:33 You could do append.

13:34 Or you could say merge.

13:35 I'm guessing if the primary key is the same, it just skips it, maybe it doesn't update to that thing

13:40 if the fields are different, I don't know.

13:42 Or you can just wipe it free and insert.

13:45 So like a clean replacement, delete and insert.

13:49 - Yeah. - Yeah, pretty cool.

13:50 And then somewhere down here, supported sources.

13:53 Here's where I was joking about the any.

13:54 Supports Postgres and it has, some of them are bidirectional source or destination.

13:59 Some are just source, some are just destination.

14:01 Postgres, BigQuery, Snowflake, Redshift, Databricks, DiffDB, Microsoft SQL Server,

14:05 and CSV files are all bidirectional.

14:08 CSV files are bidirectional, okay?

14:10 But Mongo, Oracle, SQLite, and MySQL are only sources.

14:15 So I don't know why MongoDB is not a destination 'cause it's incredibly easy to just have like a flat file,

14:21 a flat table, no like nested stuff.

14:24 I know if you're copying from a relational database, but yeah, whatever.

14:27 Still pretty cool, I think, to be able to use this.

14:30 So if you got just data you wanna keep in sync or migrate over, you're thinking of writing Python code

14:34 to do it, you might not have to write any code at all.

14:36 - Well, that's pretty cool.

14:37 I don't know what Snowflake is.

14:39 It's a great name.

14:39 - Yeah, we're getting out of that season right now.

14:41 It's starting to be warm and summer's coming, but it'll be back in the fall, like maybe December.

14:46 No.

14:47 - That's funny, cool.

14:48 Well, yeah, very interesting.

14:50 Speaking of command lines, I ran across David Lord's article, "Make your terminal nice."

14:56 And this is just sort of a little shout out because so apparently David is using,

15:02 he's using Phish, which I haven't heard about for a while.

15:06 So a nicer prompt with Starship and a nicer shell with Phish.

15:10 And this isn't really a detailed article, but it's just sort of like things can be nicer.

15:16 And I was reminded about this because I think a long time ago we were talking with,

15:21 anyway, somebody else about Starship.

15:25 And I was like, I should try Starship.

15:28 And I tried it and for some reason it didn't work for me.

15:31 But I'm like, this, I think it's time to start again.

15:33 And so I looked into both of these.

15:36 I tried Phish and I do really appreciate that it says finally a command line shell for the 90s,

15:42 which is awesome and funny.

15:45 And it's also in Comic Sans, which I ironically appreciate.

15:50 But the scripting language is a little different and I've got a lot of customizations for Bash

15:56 and Z Shell is very compatible with Bash.

15:59 So I usually use either Bash or Z Shell.

16:02 And I think I'm gonna stick with that.

16:04 But Starship, Starship's amazing.

16:07 And so I switched to Starship and the, when you go to Starship, there's an installation guide

16:13 and really it was so easy.

16:15 It was, there's a macOS line.

16:18 You just curl install it or you can homebrew.

16:22 Think I did the homebrew thing.

16:23 Anyway, but then how do you use it?

16:26 There's a usage of just, so I was on Z Shell.

16:30 So you just had to add one line to your Z Shell RC, but there's instructions on how to do it within Bash

16:37 or Phish or PowerShell or, oh cool, it works with PowerShell, that's pretty neat.

16:43 For all you Windows people out there.

16:45 But so what you get, what I really like is this notion of the directory is right there listed above your command.

16:55 And then listing what I really like is what my get branch is my directory, which version of Python I'm using.

17:05 The example here shows a rust, but it shows which version of Python you're using.

17:10 And then also what virtual environment you're in.

17:12 So having, and I didn't configure that at all.

17:15 That's just the default.

17:16 Then I'm like, that's good.

17:17 I'm good with that.

17:19 So I haven't configured it at all.

17:20 Just the default right out of the box.

17:22 It works really good.

17:23 So I'm pretty happy so far.

17:25 Anyway. - Nice.

17:26 Yeah, I think there's a lot of options here, right?

17:28 There's on my Z Shell, there's Starship.

17:31 There's all the different shells.

17:33 These days I'm using warp, warp.dev, which is pretty excellent.

17:36 I talked about that on the show before.

17:38 - I kind of forgot about warp.

17:39 We should try it.

17:39 - Yeah, warp is cool.

17:41 I think the big thing, the big message people should take away from this is,

17:45 don't just take what your operating system gives you.

17:48 It's really quite bad.

17:50 It's not just not great, it's bad.

17:53 And a lot of people who are like, I don't really like, I'm not really a terminal person,

17:58 or I'm not a CLI person.

18:00 Like that's because they didn't spend 30 minutes once getting into it.

18:05 Like, oh, actually, oh, that's nice.

18:06 And like you pointed out, like, oh, well, is there a virtual environment active?

18:09 What is its version?

18:11 And then all these other things, are there Git changes?

18:13 Am I in a Git repository at all?

18:14 Are there changes, et cetera?

18:16 Like all those little, every one of those is just, oh, that's nice too.

18:19 And it really adds up.

18:21 - Well, and I've always been one to, I mean, for my career, usually juggling multiple projects,

18:28 multiple branches on projects and everything.

18:32 And I've kind of been okay knowing what I'm doing.

18:35 However, recently I just have leveled up that of using way more branches, way more projects,

18:43 and just having that information for me right there to make sure that I know what version of Python I'm using,

18:49 what branch I'm on.

18:50 And having that just ever present is like, oh my God, I don't have to think about that anymore.

18:55 Awesome, it's just right there.

18:56 So I just hit that pain point where having that information is helping me.

19:01 And so I appreciate it.

19:03 So yeah. - Very nice.

19:04 - Cool. - Very nice.

19:06 - How about any extras?

19:07 Do you have any extras for us?

19:08 - I'm feeling extra today.

19:09 I got a few extras for folks.

19:11 First of all, I spoke about UV, or no, Brian, you spoke about UV and I just jumped on.

19:15 We both wanted to talk about it, but you had it first, I believe.

19:18 - Yeah.

19:19 - So it was officially your topic on the show.

19:20 However, and you had Charlie Marsh on Testing Code.

19:25 And so I just published an episode of Talk Python, having a conversation with Charlie about UV

19:32 and a little bit of rough and just life in general.

19:34 But I have a big backlog of shows over there that I'm working to release

19:39 'cause I've recorded faster than weekly release cadence.

19:43 But this one seemed very timely.

19:45 So I put it to the front of the queue so people can check that out.

19:48 - I can't wait to listen to.

19:50 - Yeah, it was fun.

19:51 Also follow up here is there's this really interesting YouTube channel.

19:56 I don't know if it's in general interesting, just the one I watched is called Dark Matter.

20:01 And they have like highly produced video versions of what would be a podcast,

20:08 which is kind of interesting.

20:10 So they had the one with David DHH talking about the Leaving the Cloud, Cloud Fusion.

20:15 But just like, if you look at the video side of it, it's like, could be a documentary sort of thing.

20:20 So anyway, that's interesting.

20:21 We're checking out.

20:22 You wanna follow up on all my Leaving the Cloud talk I did.

20:25 For those who like to live out near the front of the leading edge, Python 3.13 Alpha 5 is available

20:34 and people can go check that out.

20:36 So it's not quite beta.

20:38 And once it hits beta, there's no more changes, but right now there's still potentially breaking changes.

20:43 Things could be added, things could be removed, so on.

20:45 But in this thing I'm linking to, there's a nice list of what's some of the improvements.

20:49 Like for example, how exciting a preliminary experimental JIT was added,

20:53 by the way, just like throwing that out there as if it's nothing.

20:56 Your Python will be compiled.

20:59 And then finally, this one comes to us from Pat Decker, Target Tech.

21:04 So the tech blog at Target.

21:06 This is really cool.

21:07 Said, "Target posted something on their blog," which I didn't know they had a tech blog,

21:12 so that's I guess news, but cool for them.

21:15 "Announcing Target's open source fund." - Oh, cool. - Yeah.

21:18 And it says, "We created Target's open source fund "as a commitment to support open source projects we use.

21:22 "In short, Target now directly donates to projects "through an internal process that prioritizes

21:28 "representation of Target's technology team." Perfect, this is perfect.

21:32 Like it says, look, we're basing this on some of the prior work that's out there, and in fact, the success of Sentry

21:40 and Microsoft's open source funds have inspired them to go further.

21:45 And then it lists projects like, you just gave a shout out to Homebrew, open API generator, SDK man,

21:50 MockK for Kotlin and HTTP 4K.

21:55 So it looks like they do a lot of stuff with Kotlin.

21:58 So they're supporting that, but whatever, right?

22:00 You support what you use.

22:01 I think that's cool.

22:02 - Yeah, I think, yeah, every company.

22:04 I mean, every big company, even companies you don't even think use open source

22:08 use open source, so.

22:10 - Yes, exactly.

22:11 Yeah, there was a quote they gave in here that said, "The Linux Foundation estimates that 90% of the code

22:16 "in production is of open source origin." Probably using libraries based on open source,

22:22 I would imagine.

22:23 The fact you use Flask would make your stuff of open source origin, I would imagine.

22:27 - Yeah, and I know that there are some wrinkles in how the money often gets distributed,

22:33 but some money getting distributed is better than none.

22:36 So that's a good thing.

22:37 - I'll say one more thing just to encourage other companies, people within other companies to do this.

22:42 We recognize that it's our responsibility to engage across a broad spectrum of activities.

22:47 This includes contributing our own software to open source, joining communities and patching software used by others,

22:52 plus funding with money the others they mentioned.

22:55 That's pretty cool.

22:56 - Yeah, cool.

22:58 Well, the only extra I've got is that I have been working on a side project that I'm starting to ramp up a little bit,

23:05 and I've decided to have a start in a SaaS project.

23:10 And I've started, thought I would get a little headstart by using Corey Zhu's SaaS Pegasus.

23:18 So I talked with him just the other day.

23:21 So his interview will be on Python test.

23:25 But as I go along, I'll be reporting how it's going.

23:30 And yeah, I'm just excited to get started.

23:33 - Nice, you're gonna live blog it, blog the progress.

23:36 - Yeah, it's just gonna be slow going 'cause there's a lot more going on in my life,

23:42 but I wanna get it off the ground.

23:44 - Indeed.

23:45 I know the feeling.

23:47 - How about something funny?

23:49 - Yes, let's talk about something funny.

23:50 So this comes to us from DevHumor, directly submitted by admin.

23:54 It's got like a kind of a cybersecurity thing here, Brian.

23:58 So like social engineering, like, hey, I'm from IT, you're gonna get a pop-up on your authenticator

24:04 that says, we're just checking some systems.

24:07 So if you don't mind just approving that.

24:10 I know it's your personal email, but we're just working on it for you.

24:13 So anyway, here's the joke.

24:15 If a person who tries to trick others into giving them personal or sensitive information

24:20 is a social engineer, does that mean that the person who tries to prevent that is an anti-social engineer?

24:26 - Well, I mean, I've met a lot of cybersecurity people and I, maybe. (laughs)

24:33 I think it might be an overgeneralization, but yeah.

24:37 - Yeah, it's funny.

24:38 Anyway, are you an anti-social engineer?

24:40 This time it's a good thing.

24:42 - I know just who I need to send this to.

24:45 This is awesome.

24:46 - Perfect.

24:47 - Awesome, thanks.

24:49 Well, that was a fun episode and thank you everybody for listening and watching

24:53 and everything and thank you, Michael.

24:55 - Yep, you bet.