🏅 Top traders

50 points out of 50

Found the bug, made the last commit.

Resolves to 10-12

11h

7/8 passed. 1h till deadline.

In the last case i have a small bug, where \n in the message stream "overflows" into the next set of data. And there is likely a mistake in the test too, because the test expects a message of length 6 characters, which contains only one symbol.

10h

4/8 tests passed. Out of those 3 did not pass, because program does not manage to read the socket again. One fails because times out.

So even though it is 4 failed tests, it is actually 2 mistakes.

9h

2 out of 8 tests passed.

5 hours till deadline

8h

None of the tests executes. I had carriaer return error, I had symbolic links not being able to read error.

Will probabbly delete every test and git clone them from zero and start setting from the beginning.

6h till deadline.

7:40

The code of the tests prepared by the teacher outputs for me a lot of errors out of the scope of my python program. But it worked for other students.

That means I am executing the tests incorrectly.

6h total.

I finished everything with python code, now need to search for mistakes in the architecture and make it run. Started loading tests into wsl, but it takes long: installing many things and learning new stuff.

5h total.

Half of a method is left before testing, but need to clean the code, some snippets are too similar and should be made into functions.

4h total.

I have 2 methods left. Then insert about 12 exceptions associated with those methods. Then load everything into Linux and learn how to do automated tests.

And then a huge portion of fixing everything.

3h total.

Done the core loop of the child process. Now doing the branching part: different methods depending on different requests.

After that I will fill the code with try-except to raise all the exceptions i am supposed to do.

No idea whether I am supposed to prevent several processes trying to open the same file. It will be solved by operation system, one of them will succeed and the other will sent an error to the respective user, but it is not explicitly stated whether this is the needed behaviour.

I think it's fine. Let it be

40m+80m=120m

Doing really slowly and calmly, trying to understand every bit. Everything feels managable. I could have done the same progress in 20 minutes by copying lines from documentations, but not understanding then.

Main process ready. That is only because child processes will be doing all the talking to the client and file system.

40m in. Reading task, setting up environment. Does not feel too difficult, I am 90% sure that after I implement the general tcp listening writing the responses themselves will be easy.