todo

TODO:
* Make a version of the IntervalUnion for when I want intervals to overlap?
    * Or convince myself that I can use sparse segtrees instead? Test it on https://codeforces.com/contest/1061/problem/D
* Permanently


--------------




Make a new function for loading a vector like readvec<ll>(N, A).



I think I can index out of bounds in strings without getting errors :(
    Update: It seems like I can index oob by exactly one.
    Maybe this implementation of std::string has a secret \0 or something?



Rename tsdbg and tsish to repr and to_string in some order?


Quick functions for max/min of a container? Similar to sumv?
    (Does this already exist?)
    (Yeah it definitely does... go remember how?)
    Yeah, it's just *max_element(all(dat)).



Maybe learn about jupyter server?



-----

Try to solve pe502 using generating functions?


--------------


Apparently -u in interact.sh doesn't actually protect against lack of flush in interact.py



terminal.integrated.fontWeightBold ?



pvn1 for V<V<tuple>>


What was that recent problem I talked about with Christine where I suspected FFT but there was actually some DP trick?
Can that method help on Counting Castles?
In general I need some fancy method that helps with a DP where each new entry is some convolution thing on the previous stuff.



For python template setup:

ta = [1, 5, 3]
dbg(f"{ta + ta = }")

Make a snippet that handles this.
//////
Make a new plan where we type an f-string, then remove the letter f, then I handle it nicely within dbg.

https://stackoverflow.com/questions/47339121/how-do-i-convert-a-string-into-an-f-string



ACTUALLY USEFUL:



Maybe I want assert_same(a, b) where it runs dbg() if failing?


macro for declaring N-dimensional vectors!


------

PUZZLE HUNT template:

Jacob suggests: Is it possible to access formatting info through Javascript macros?