有些急性子

有些急性子

有些急性子
jike
HomeArchives

What is programming thinking - taking a small project as an example

7
AI Translation
This post is translated from Chinese into English through AI.View Original
AI-generated summary
The document discusses several key characteristics of programming thinking, including: 1. **Problem Decomposition**: Breaking down large problems into smaller, manageable parts. 2. **MVP (Minimum Viable Product)**: Creating a basic version of a solution to validate direction and build confidence. 3. **Step-by-Step Testing**: Testing each part of the process to avoid having to start over if issues arise. 4. **Automation**: Utilizing programs for repetitive tasks instead of relying on human effort. 5. **Logging**: Keeping detailed logs to track changes and facilitate problem-solving. An example is provided using the "Didi" ride-hailing service, where the author faced challenges in copying trip data from a PDF to an expense report system. By analyzing the requirements, they created a program using AI tools to automate the extraction of trip details into a structured Excel format. The process involved building an MVP, conducting step-by-step tests, and making adjustments based on output requirements. Ultimately, the automation significantly reduced the complexity of the task. The author emphasizes that programming thinking is a valuable approach in various work contexts, aiding in problem-solving and efficiency. The document concludes with a personal note about the importance of documenting thoughts and experiences.

1. Characteristics of Programming Thinking#

  • Problem Decomposition

    • Break down large problems into smaller ones that can be solved step by step.

  • Build MVP

    • "Get it running" first. A preliminary viable solution can validate our direction and also increase confidence.

  • Step-by-Step Testing

    • Test each step correspondingly, rather than waiting until all work is done; otherwise, discovering issues may require starting over.

  • Automation

    • Repetitive and bulk work is not a strong suit for humans; it's more suitable to be handled by programs.

  • Logging

    • Memory is unreliable. Only logs can faithfully record all information, making it easier to locate problems.

2. Example of Summarizing Trip Data with "Didi Taxi Trip Receipt"#

2.1 Problem Decomposition#

Problem: When filling out the reimbursement form, I found that I needed to copy trip and amount information from the trip receipt to the BOS system one by one, which was too cumbersome, and copying content from PDF easily led to errors.

Upon reflection, shouldn't this be the kind of work that programs excel at? Why should I make it difficult for myself?

Just do it:

Analyze the requirements: Extract corresponding time, location, amount, and other information from the trip receipt PDF, and then output it to a fixed template Excel sheet.

All PDF formats are fixed, and the Excel format to be exported is also fixed. Therefore, once the program is written, it can adapt to all Didi trip receipts.

2.2 Build MVP#

I used the AI programming tool Trea to build the code more quickly. This is the first AI IDE produced by ByteDance, which deeply understands Chinese development scenarios.

First, create a folder to store the code files and place a Didi trip receipt PDF file in the folder. Then open this folder in Trea.

Next, switch the dialog box on the right side of Trea to Builder mode and enter the following command.

Extract the corresponding start time, start location, end location, and amount information from the trip receipt PDF, and then output the corresponding content to the terminal.

First, build an MVP that can retrieve the corresponding data. Clearly, this small requirement is a minor case for AI. In less than a minute, it wrote the preliminary code and produced the desired output. This made it much easier for me to copy the required data from the terminal. Of course, this is definitely not the final result I want.

2.3 Step-by-Step Testing#

Next, I input another command to the AI:

I want to save the output content as an Excel file.

Quickly, the AI completed its task and realized my vision.

However, this is not the end; I have specific requirements for the format of the Excel file. Next, I directly placed the standard Excel sample file I needed in the folder and told the AI:

Please adjust the format of the Excel file you output based on the Excel file I provided.

Soon, the desired Excel format was achieved.

However, there were still some aspects of the output Excel that did not meet my needs; I wanted the entries in the output Excel to be sorted in chronological order by start date and time. I made another request to the AI:

Sort the entries in the output Excel in chronological order by start date and time.

Thus, through step-by-step testing and adjustments, I finally achieved the "mini-program" I needed.

2.4 Automation#

This program serves as an automated way to free me from the tedious task of filling out trip receipts. Although it took me several minutes of continuous adjustments, consuming time for reimbursement filling, it will certainly save me a lot of trouble in future tasks.

2.5 Logging#

In programming, the git tool serves as a good logging mechanism, with corresponding log records for each code update.

This article is also a log of my attempt this time.

3. Conclusion#

Programming thinking should be an important mindset we need to possess in our work. It is not only used for writing code; the ability to decompose problems allows us to tackle "big problems"; MVP enables us to build preliminary models for solving issues; logging facilitates our review of the steps taken to solve problems and helps us summarize experiences better...

4. Aside#

I haven't written anything for a long time; I still need to exercise my brain and record something.

Loading...
Ownership of this post data is guaranteed by blockchain and smart contracts to the creator alone.