xOBSE 22.10beta causes console spam for Lovers with PK script xx0090ea

[Hotdog3223]

This is a script that just does a simple check to see if OBSE is actually running or not.

Deleting the script itself in xEdit stops the spam. Older versions of xOBSE did not have this spam problem.

[llde]

Revert. The beta has some issues.

[calebtt]

What is it spamming?

[Hotdog3223]

It just keeps spamming the script id & null, it doesn't actually have any impact on the mod's functionality but it spams the console with that over and over endlessly

[llde]

Yeah I think I mistook an hook for another in 5c31c42

BTW @calebtt Seems that there is an issue for 0d71ba4 that can cause corruption and exploding cosaves.

[llde]

@calebtt Seems that using std::array<char,dim>
instead of a direct array cause this corruption. Do you have any idea why?

EDIT: IT's actually a bit different:
SetElementString(arrayID, newKey, std::string(buffer.cbegin(), buffer.cend()));
if I create the std::string using an iterator ()either const cbegin/cend or normal begin/end) the resulting cosave is partially corrupted and explode in size.
If I create the string using the std::Array data() inner pointer then it's correct and with the right size (940KB (ok) vs 14 MB (wrong))
Do you have an idea on what's is going wrong? @calebtt
I tried to print the string content but they are equals

[llde]

Actually in this case it seems the script have an issue (maybe it wasn't recompiled properly?) as even in MessageLogger with 22.9 I have
2023/09/15 10:12:42 [005170B1] [WARNING] Unable to find function definition for command 0 in script 'xLoversCheckObse2QuestScript'.

Now the last beta have actually an issue where the message buffer wasn't saved properly or was corrupted, causing the message to be strange

[calebtt]

I'll try to look into that

[calebtt]

@calebtt Seems that using std::array<char,dim> instead of a direct array cause this corruption. Do you have any idea why?

EDIT: IT's actually a bit different: SetElementString(arrayID, newKey, std::string(buffer.cbegin(), buffer.cend())); if I create the std::string using an iterator ()either const cbegin/cend or normal begin/end) the resulting cosave is partially corrupted and explode in size. If I create the string using the std::Array data() inner pointer then it's correct and with the right size (940KB (ok) vs 14 MB (wrong)) Do you have an idea on what's is going wrong? @calebtt I tried to print the string content but they are equals

iterators would iterate over the entire array, I think using .data works because you have it null padded after the string's contents. As long as you initialize the array to zero.

Oh I see we're actually setting a 0 value at the string's length.

[calebtt]

Instead of the iterators I think it should be .data()