English
Share
Sign In
Acceleration of the low-code era
Until now, many of the programming languages we have called 'Code' were ultimately tools for communicating with machines. Think about it. There are foreigners and us. There are ways to communicate freely with them by learning a foreign language, but you can think of it as a faster way for them to learn our language.
In the past, when writing repetitive tasks or scripts, you had to buy a book or take a lecture on a programming language. Now, all you have to do is ask, "I want to do task A. Is there a good way?"
In fact, we have experienced writing the code above. So, how about adding a new feature here?
With the above command we asked for the ability to run it on a regular basis, check the progress, organize it into a csv, etc. Then, we propose a language model as follows.
We will modify the existing code to check the progress and add a function to automatically crawl Naver News every minute and save it to a CSV file. For this we will use Python's time and csv modules.
1.
Use time.sleep(60) to crawl every minute.
2.
We use the csv module to save crawled data to a CSV file.
3.
Prints log messages so you can check progress.
In fact, you can see that a service has been created that crawls and stores news related to ‘artificial intelligence’. So what if an error occurs while making it? Just ask about the error details. If you ask the language model to copy the output value that appears as an error, it will extract an answer that matches the error.
In other words, this not only converts what is said in natural language into code for those who do not know programming, but also provides the following convenient functions for those who write existing code.
Bug fixing: If you find a bug in already written code, you can describe the problem to the language model and ask how to fix it. Models can suggest ways to fix bugs.
Code optimization: Language models can provide ways to improve the performance of your code or make it more efficient. For example, they might suggest a faster algorithm or suggest ways to improve the code structure.
Documentation and commenting: Language models can be used to annotate or document code. You can create comments that explain what each part of your code does.
Recently, Amazon's Q and Meta's Code LLaMA, which are specialized for code writing, have been released, which is expected to accelerate the low-code and no-code market.
ⓒ 2023. Haebom, all rights reserved.
The source is indicated and may be used for commercial purposes with the permission of the copyright holder.