Puneet Tripathi

My name is Puneet, and I have 7 years of experience. I have worked in multiple companies, both service-based and product-based companies. I have experience with Python over the past 7 years, and I've worked with several Python frameworks, including Django and Flask, as well as creating REST APIs. I've also worked with AWS services. If I'm talking about AWS services, I have experience with different platforms and services, such as EC2 instances, S3 buckets, AWS API gateways, and Lambda functions. In fact, I've worked on a particular application that was entirely dependent on Lambda functions and API gateways. So, apart from that, if I'm discussing handling the production system, I have approximately 5 to 6 years of experience handling bugs and issues in production, deploying code in the testing system, and then moving to production demos. I've also worked with databases such as MySQL and PostgreSQL. Currently, I'm working with PostgreSQL. I have experience with GitHub, and that's about it.

K. So there are some Python libraries for finances, like NumPy and Pandas. Mostly, they are used for finance. We need to write our own packages to utilize because many cases, the Pandas library for decimal precision is limited. And, but Pandas are a bit larger in size. So while deploying, we are having issues. So, but they are very fast. Like Mumbai and Bandra, both are very fast. And, if we are talking about sci-fi, there are other Python libraries also there. But if we are talking about these two are the particular libraries that we are using.

Yeah. AC, stands for consistency, isolation, and durability. So when we're talking about basic properties, the atomicity since it's the process of one transaction. So that transaction used to be atomic in the case. Either it would happen, or it would not happen. So if we're talking about that, consistency seems like any changes in the database should be persistent and consistent. We're doing the change in any of the transactions that suppose some other person cannot change until and unless that transaction is completed. Because if someone would do that, the changes I'm writing would be inconsistent. And someone would read that before writing, then there would be inconsistency. And isolation seems like the changes should be isolated from each other. And durability says it should be durable. The changes we're doing should be durable in the database.

Yeah, so if the question is about integrating that, the Python and PostgreSQL or any other SQL. So, Probably the best way to do that is any ORM. Like, even though if we are not using the XANGO ORM, because if you are using the XANGO, by default, it provides the XANGO ORM, objects relational mappers. So, apart from that, like, This SQL Alchemy is the best tool that we can use to integrate that.

Okay, this is the connection object. So, this is the connection class. A single pattern should be used here because if you will not use that, it would create multiple connection pools. Even though some of the connections are being unused, it would create another connection rule. So, to avoid that, we need to use this in the singleton pattern. And we need to make one class here and utilize the current object. If the current object is already there, the object has already been created, then we need to return that current object. If it is not, then only it would create the object. So we need to create a function that would create a new connection. Because the constructor class is used to create a new connection each and every time.

Actually, I belong from the Python background, so probably I'm not a good person to answer this question. I have learned some of the JavaScript questions and about JavaScript, but this question belongs to managing server-side logic on the mode test. So I think I'm not the right person to answer that.

Yeah. And exception class, we need to log that. That's an option, and we need to raise the exception with raise e. So that the exception is raised. Apart from that, the fetch URL and there should be multiple conditions like which kind of that. Okay. So already fetched data. We need to check whether it is successful or failed. Based on their code, I guess. So that will be the better approach here. We need to get the response code, the HTTP code, and then respond back.

Yeah. It's the recursion function. And this function needs to calculate the Fibonacci series. This function needs to calculate that for a given number, it will give the Fibonacci number. 3, for example. But it would give only one number, so I need to enter into the loop.

There are a couple of Python frameworks. If I'm talking with a lightweight framework, I would prefer Python Flask. Apart from that, there are the Python Zango frameworks also there. And if I'm creating APIs, Fast API is already for async services. It's already for async services. So, if we are talking about the high responsiveness of the web application, it would depend upon the frameworks and where we are hosting that. That would give the high responsiveness web application where we are deploying that. And if we need to give the high responsiveness, we need to deploy the same application with high availability, different zones, like AWS and all. So that depends upon the configuration. Most probably, the Python Zango framework is mostly widely used because it's all packaged already in there. We can use Flask also because it's lightweight. If the framework has advantages and disadvantages, they're written in both one. So, probably it depends upon the application. If there are smaller applications, we need to use the Flask one. And if we are very complex and there are very interdependability in that application, we need to use the Zango or the first one.

Yeah. So if we use any kind of transaction, by default, PostgreSQL and MySQL and other databases used to manage consistency. And when we're talking about data consistency, we need to manage it from our end. So suppose, even if it's time-series data, we need to get the pipeline and make sure that no data is missed when inserting it. So, to suppose one of the data source systems failed, we need to make a retry or hit that source from the particular timestamp, and store those values. Then again, we can store those values. We cannot lose that data. Then we can hit the APIs again and get that data again.