Fix generated localisations in non-English languages

[Idhrendur]

Although originating from Vic2 coding, Adjective State look awful outside English, for order (e.g. French should have, e.g., Piémont Français) or missing declension (French Provence Française, German Deutsches Österreich, Spanish Indias Occidentales Españolas). It'd, however, require a gargantuan amount of coding to support:

LANGUAGE	Adjective position	Gender according	Number according	EXAMPLE
English	before	no	no	French Austria
French	after	yes	yes	Autriche Française
German	before	yes	yes	Französisches Österreich
Polish	before	no	no	Franc. Austria
Spanish	after	yes	yes	Austria Francesa
Italian	after	yes	yes	Austria Francese
Swedish	before	no	yes	Franska Österrike
Czech	before	yes	yes	Francouszké Rakousko
Hungarian	before	no	yes	Franciaországi Osztrák
Dutch	before	no	yes	Frans Oostenrijk
Portuguese	after	yes	yes	Áustria Francesa
Russian	before	no	no	Французск. Австрия
Finnish	before	no	no	Ranskan Itävalta

[IohannesIohannium]

Adjective order works correctly. Gender-according is now needed.

For French, Spanish, Italian, Portuguese:

• The converter should check a partial state's category in gender and number (through StateCategories.txt)
• It should then see a yet-to-be-created file called Grammar.txt that will tell it what each category corresponds to grammatically
• Then it would look for the grammatically correct generated owner's adjective
• And it would use that adjective

As an example, the partial state ENG_1979 (Guayane) owned by a democratic France

• ENG_1979 is in category fsac (per StateCategories.txt)
• Category fsac corresponds to _FS (per yet-to-be-created Grammar.txt)
• Owner is FRA, democratic. The adjective to look for starts with FRA_democratic_ADJ_FS
• The localisation is being output in French, so the adjective is FRA_democratic_ADJ_FS_FR (defined internally in the converter)
• The HoI4 state is thus localized as FRA_democratic_ADJ_FS_FR + ` +ENG_1979`
• The "printed" result is therefore Guayane Française

Fallback cases:

• State is in no category: first category of StateCategories.txt is used
• State is in multiple categories: latest category is used
• Category has not been grammatically defined: bare adjective will be used (e.g. FRA_democratic_ADJ)
• Category has been grammatically defined in more ways: ditto

[IohannesIohannium]

Grammar.txt as per now. This must be kept as a configurable, as new state categories can be added arbitrarily.

# Male singular - *Le Danube allemand*
msnc = _MS
msac = _MS
msav = _MS
msnv = _MS

# Male plural - *Les Territoires allemands*
mpac = _MP
mpav = _MP
mpnc = _MP
mpnv = _MP

# Female singular - *L'Autriche allemande*
fsnc = _FS
fsac = _FS
fsav = _FS
fsnv = _FS

# Female plural - *Les Villes allemandes*
fpnc = _FP
fpac = _FP
fpav = _FP
fpnv = _FP