Add support for campaign status

[kubaau]

Adds support for Roman-style (video) campaigns and the world campaign (video).

Fixed world campaign scripts to unlock the same continents as it was in S2.

Related to issues #1639 and #1177

[Spikeone]

Either I'm blind or the implementation is somewhere else? What is rttr:SetCampaignChapterCompleted good for?

[kubaau]

Either I'm blind or the implementation is somewhere else? What is rttr:SetCampaignChapterCompleted good for?

It doesn't make a difference in Roman-style campaigns, but it marks a continent as conquered in the world campaign. It could also be used for other campaign types as well if we wanted to show what was completed, especially if someone designs a campaign whose progression is not linear.

Edit: For roman-style campaigns it shows the "You have successfully completed chapter X." screen after ending the game.

[Flamefire]

Awesome, the last bit for campaigns! Review inline.

General hint: Try to make best use of the type system. E.g. "unsigned char" is/should be a (text-)character (although only since C++11 introduced uint8_t) And using this narrow width number-type as parameters likely introduces overhead (registers are wider so often need to be masked) so it is even faster to just use unsigned there.
Similar for string which should be text. It can be stored in the config as a string for brevity but converted during load/save to a more appropriate type, especially when (scoped) enums can be used.

In general I like the changes, thanks! :)

[Flamefire]

So the default is to enable the first 2 chapters? Why that? Wouldn't (only) the first one be a more sensible default?
I'm also a bit confused as you seem to be using indices and 1-based values for chapters

Maybe stick to indices and use an optional instead of zero

[Hirotaro]

In the original S2, I also had first 2 missions unlocked at start. Maybe because 1st one is actually a tutorial without AI opponent?

Just rechecked with GOG S2 Gold Edition.

[kubaau]

Rework in progress - converting to draft, will notify when ready for review. Resolving some conversations since they will not be relevant.

[jarvis-jhk]

Thanks for this PR — we merged it into our fork (jarvis-jhk/s25coop, a coop-campaign fork) and found a few things while testing it; feel free to take any of it: jarvis-jhk@07bf3205a

• The world mission scripts use 1-based chapter ids while chaptersEnabled / isChapterPlayable use map indices, so winning Europe marks Africa as conquered. Converted the scripts to 0-based.
• luaFolder = "<RTTR_RTTR>/CAMPAIGNS/WORLD" breaks on case-sensitive file systems (the folder is campaigns/world); reverted to "".
• A CONFIG.INI without a [campaigns] section (every existing install) failed Load() and was replaced by the defaults (also the settings.size() < SECTION_NAMES.size() check). Made the section optional.
• toSaveString resizes with '\0', so untouched chapters between two known ones are written as NUL bytes. Now padded with '0'.
• IsWinnerHuman indexes players by bestTeam & (1 << i) (Flamefire's comment); it now checks the winner actually announced (team or single player).
• The victory screen shows the 0-based index ("chapter 0"); the pending state is reset in Game::Start, and settings are saved in ShowMenu when progress changed.

Tests we added: settings roundtrip/old config/save format, a check that every shipped mission script completes its own chapter index, and CheckObjective winner cases in testLua.