File format

[IohannesIohannium]

How the dictionary file should look like (example with table):

English	en_article	en_class	en_adj	French	fr_article	fr_class	fr_adj	German	de_article	de_class	de_adj	Spanish	es_article	es_class	es_adj	Russian	ru_class	ru_adj	Polish	pl_class	pl_adj	pl_adj_a	Korean	Chinese	Czech	cz_class	cz_adj	cz_adj_a	Italian	it_article	it_class	it_adj	Hungarian	hu_class	Dutch	nl_article	nl_class	nl_adj	Portuguese	pt_article	pt_class	pt_adj
England		en_s	English	Angleterre	L'	fr_fsav	anglaise	England		de_n	englisch	Inglaterra	La	es_fsa	inglesa	Англия	ru_f	английск	Anglia	pl_fs	angielsk	angielsc	잉글랜드	英国	Anglie	cz_fs	anglick	angličt	Inghilterra	L'	it_fsav	inglese	Anglia	hu_s	Engeland		nl_n	Engels	Inglaterra		pt_fsn	inglesa
United States	The	en_pl	American	États Unis	Les	fr_mpav	américaine	Vereinigte Staaten	Die	de_pl	amerikanisch	Estados Unidos	Los	es_mpa	americana	Соединённые Шта́ты	ru_pl	американск	Stany Zjednoczone	pl_mp	amerykańsk	amerykańsc	미국	美国	Spojené Státy	cz_mp	americk	američt	Stati Uniti	Gli	it_mpac	americana	Egyesült Államok	hu_p	Verenigde Staten	Die	nl_pl	Amerikaans	Estados Unidos	Os	pt_mpa	americana
Chicago		en_s	Chicagoan	Chicago		fr_fsn	chicagoise	Chicago		de_n	chicagoerisch	Chicago		es_fsn	chicagüense	Чикаго	ru_m	чикагск	Chicago	pl_ns	chicagowsk	chicagowsc	시카고	芝加哥	Chicago	cz_ns	chicagsk	chicagšk	Chicago		it_fsn	cicaghese	Chicago	hu_s	Chicago		nl_n	Chicagoer	Chicago		pt_fsn	chicaguense

Explanation:

• (Language name): Localized name in said language
• _article: The article used with the country name, space included ("The ", not "The")
• _class: Grammar category. Starts with the language two-letter code (e.g. en) and may have gender (m, f, n), number (s, p/pl), use of article (a, n), starting with vowel or consonant (v, c)
• _adj: The adjective this country uses. Notes:
  • French, Spanish, Italian, Portuguese: feminine singular form used
  • Russian, German, Dutch: form devoid of ending used
  • Polish, Czech: form devoid of ending used, although having a separate _adj_a field for nom. pl. m. consonant change

[IohannesIohannium]

Outputs for CK3

• ⟨title⟩: ⟨Language⟩
• ⟨title⟩_adj: ⟨lang⟩_adj
• ⟨title⟩_article: ⟨lang⟩_article (for languages where this exists)

Outputs for EU4

• ⟨tag⟩: ⟨lang⟩_article⟨Language⟩
• ⟨tag⟩_ADJ: ⟨lang⟩_adj
• As a country flag in file history, output all ⟨lang⟩_classes

Outputs for Vic2

• ⟨tag⟩: ⟨lang⟩_article⟨Language⟩
• ⟨tag⟩_ADJ: ⟨lang⟩_adj

Outputs for HoI4

• ⟨tag⟩_⟨ideology⟩: ⟨Language⟩
• ⟨tag⟩_⟨ideology⟩_DEF: ⟨lang⟩_article⟨Language⟩
• ⟨tag⟩_⟨ideology⟩_ADJ: ⟨lang⟩_adj
• As a country flag in file history, output all ⟨lang⟩_classes
• For Polish localization only, ⟨tag⟩_⟨ideology⟩_ADJ_PL: ⟨lang⟩_adj_a, along with a (yet to be defined) special way of triggering the custom loc

[IohannesIohannium]

Special things:

• hu_adj is set to be ⟨Hungarian⟩i, in all cases
• ko_adj is set to be ⟨Korean⟩ in all cases, zh_adj to be ⟨Chinese⟩

How the dictionary file should work:

• Converter looks up English localization for a polity (tag, title, etc.)
• IF the game is EU4 or Vic2, it'll try to match it to a composed entry of en_article English (e.g. The United States)
• ELSE it'll try to match it to an entry in English (e.g. United States)
• THEN, it'll consider the polity to match that row and appropriately output localisation (even if it overrides any other input)
  This last thing means that if a game localises "United States" in French as "EE.UU.", the converter will disregard that and output "États Unis" anyway.

Matching is case-insensitive, diacritic-sensitive.

Output will be restricted to the languages a game actually has (we'll skip Vic2 Finnish column because Finnish is impossible to handle) and will not overwrite vanilla localisations (if a game already has RUS_ADJ as localized, the converter will NOT output RUS_ADJ)

[IohannesIohannium]

Grammar category will be used to define:

• Adjective agreement (La France est grande / Les États Unis sont grands)
• Verb agreement (La France est grande / Les États Unis sont grands)
• Output of articles in games where the name does not naturally have them in the polity's bare localization (La France est grande / Les États Unis sont grands)

Every other grammar shenanigan such as adjective declension (français, française, ...), polity name's declension (Франция, Франции, ...) is to be done separately through the use of respective adj_rules.txt and decl_rules.txt files

[IohannesIohannium]

As an update, most columns that have articles separate have been removed, as articles can be inferred from categories anyway.
Czech, Hungarian and Dutch have also been dropped.

[IohannesIohannium]

DROPPED: now resorting to simply giving the converter grammatical categories for each string (e.g. telling the converter that the French string Géorgie is in category fsac, that the Spanish string Puerto Rico is in category msn, etc.)