Restructuring the code #5

New Issue

juk0de · 2023-08-17T10:19:37+02:00

juk0de commented

2023-08-17 10:19:37 +02:00

If we'd restructure the code, it would make it easier to change and test it and also improve robustness:

Create modules for different parts (subcommands?) of the code, e. g.:
- tags.py
- message.py
- chat.py
Create classes for the most important components that also implement the required operations:
- Tag / TagLine
  - Tag is a single tag
  - TagLine is a line of tags with a defined prefix
  - Operations: get / add / delete / modify / match tags
- Question / Answer
  - Question is a question string for the AI (with a defined header for the file section)
  - Answer is the response string from the AI (with a defined header for the file section)
  - Operations: extract sourcecode
- Message
  - Message is composed of Question + Answer, optionally a set of Tags and a filename
  - Operations:
    - to_file() : write message to file (i. e. TagLine + Question.Prefix + Question + Answer.Prefix + Answer
    - from_file() : read message parts from file and create a Message instance
    - Others: get / modify the tags, question or answer
- Chat
  - A structured chat history (list of Messages)
  - Operations:
    - Read Messages from DB folder (with filter for file names, tags and sort order)
    - Filter chat history based on message tags
    - Search Questions and Answers for a given string
    - Print chat history (optionally through a pager)
    - Prepare chat history to be used in an AI request
Split the tests into multiple modules
- E. g. one test module per source module

If we'd restructure the code, it would make it easier to change and test it and also improve robustness: * Create modules for different parts (subcommands?) of the code, e. g.: * `tags.py` * `message.py` * `chat.py` * Create classes for the most important components that also implement the required operations: * `Tag` / `TagLine` * `Tag` is a single tag * `TagLine` is a line of tags with a defined prefix * Operations: get / add / delete / modify / match tags * `Question` / `Answer` * `Question` is a question string for the AI (with a defined header for the file section) * `Answer` is the response string from the AI (with a defined header for the file section) * Operations: extract sourcecode * `Message` * `Message` is composed of `Question` + `Answer`, optionally a set of `Tag`s and a filename * Operations: * `to_file()` : write message to file (i. e. `TagLine` + `Question.Prefix` + `Question` + `Answer.Prefix` + `Answer` * `from_file()` : read message parts from file and create a `Message` instance * Others: get / modify the tags, question or answer * `Chat` * A structured chat history (list of `Message`s) * Operations: * Read `Message`s from DB folder (with filter for file names, tags and sort order) * Filter chat history based on message tags * Search `Question`s and `Answer`s for a given string * Print chat history (optionally through a pager) * Prepare chat history to be used in an AI request * Split the tests into multiple modules * E. g. one test module per source module

juk0de self-assigned this 2023-08-17 10:19:46 +02:00

juk0de added reference restructurings

2023-08-17 10:23:32 +02:00

ok referenced a pull request that will close this issue

2023-09-11 13:42:12 +02:00

ChatMasterMind Application Refactor and Enhancement #8

juk0de closed this issue

2023-09-12 07:36:07 +02:00

Sign in to join this conversation.