How do you achieve this?

[Time-Coder]

That's a wonderful tool!! But that's just a magic for me to understand how you monitor the changable of variable(source code is a little complecated for me).

Do you continuously reading varaible value in a thread or someting else?

[gaogaotiantian]

watchpoints is filled with Python black magic, it was not a very good place to start if you are a beginner. However, there are definitely things that are easy to understand. Starthing with sys.settrace, which allows you to trigger a user-defined callback after executing every line(or bytecode). This function is commonly used for debugging tools and some profiling tools (coveragepy for example). I would assume this solves the first question you would have, which is how I could monitor on values. However, the monitoring itself also has some black magic behind the scene, a simple callback after executing a line is probably good enough within the frame(still, with some black magic), but scope crossing brings other issues.

I agree the source of watchpoints is not the easiest to read, but it's partially due to the nature of the project - it's not an application or a framework, it's a tool to solve mystery :)

[anzhi0708]

Your Bilibili followers need a tutorial episode about the ast module ;-P

[rsemenoff]

I have an app that i converted to flask, and recently added flask-cache to avoid data disappearing between ajax calls, but I dont entirely understand the flask-cache documentation. Pandas is also giving me spurious 'set with copy' error/warning. I also know Pandas made undocumented changes to certain functions I had been using... So I'm dependent on libraries with potentially untrustworthy memory management. On Friday, March 18, 2022, 12:43:40 AM PDT, Time-Coder ***@***.***> wrote: That's a wonderful tool!! But that's just a magic for me to understand how you monitor the changable of variable(source code is a little complecated for me). Do you continuously reading varaible value in a thread or someting else? — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[rsemenoff]

Actually I kind of assumed this is entirely black magic :) I don't know how it works otherwise...but my real problem is my own garbage code that I was trying to avoid refactoring. In the end I fixed my buggy code the old-old-fashioned way. On Friday, March 18, 2022, 11:20:17 AM PDT, gaogaotiantian ***@***.***> wrote: watchpoints is filled with Python black magic, it was not a very good place to start if you are a beginner. However, there are definitely things that are easy to understand. Starthing with sys.settrace, which allows you to trigger a user-defined callback after executing every line(or bytecode). This function is commonly used for debugging tools and some profiling tools (coveragepy for example). I would assume this solves the first question you would have, which is how I could monitor on values. However, the monitoring itself also has some black magic behind the scene, a simple callback after executing a line is probably good enough within the frame(still, with some black magic), but scope crossing brings other issues. I agree the source of watchpoints is not the easiest to read, but it's partially due to the nature of the project - it's not an application or a framework, it's a tool to solve mystery :) — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>