[TF2] Fix Sappers becoming indestructible under certain circumstances

[bakugo]

Fixes ValveSoftware/Source-1-Games#6689

The Spy's Sapper has a 0.1s build time between being placed on a building and becoming active. During this time, Engineers can still upgrade the building. Whenever a building's model changes (which happens during upgrades depending on the building type), build points are reset, and an attempt is made to re-attach any previously attached Sappers to the new build points, but for some reason, the code responsible will refuse to re-attach a Sapper that is still in the process of being built.

As a result, the build point that the Sapper was previously attached to will be empty, at least until the building swaps models again, at which point it will be re-attached correctly since the 0.1s build time has passed. However, while the build point is empty, the Spy can actually place a second Sapper, and if he does, there will be 2 Sappers attached to the same building. Once the next model change happens (when the upgrade finishes, in the case of the Sentry), the first Sapper will be reattached, but since there's only one build point, the second Sapper will be destroyed, setting the sapped building's m_bHasSapper to false, which re-enables the building and prevents the Engineer from damaging the first Sapper (it can be removed by picking up the building).

Fixed by removing the condition preventing re-attachment of Sappers that are not yet fully built. There doesn't really seem to be any reason for this condition to exist.

[henke37]

I take it that rewriting the m_bHasSapper variable to a function that directly does the check is out of scope and/or undesirable?

[bakugo]

I take it that rewriting the m_bHasSapper variable to a function that directly does the check is out of scope and/or undesirable?

Not undesirable, but probably out of scope, yeah.

The code behind buildings and building attachments is a mess of legacy cruft that I'm pretty sure was inherited from TF2: Invasion, which had a much bigger focus on buildings, and could benefit greatly from being at least partially refactored (and it's definitely not the only system in the game that this is true for), but I assume that simpler targeted fixes like this are more likely to be merged due to being much easier to review and less likely to cause regressions.