1:- module(eorscope_employer_costs, 2 [ countries/1, % -Countries 3 country/2, % +Iso, -Country 4 regions/1, % -Regions 5 countries_in_region/2, % +Region, -Countries 6 contributions/2, % +Iso, -Lines 7 assumptions/2 % +Iso, -Assumptions 8 ]). 9:- use_module(library(apply)). 10:- use_module(library(lists)). 11:- use_module(library(pairs)). 12:- use_module(library(error)). 13:- use_module(library(readutil)).

EOR Scope dataset of statutory employer costs (76 countries)

Read-only lookup of the EOR Scope open dataset (CC BY 4.0), embedded at compile time from the CSV files in the pack's data/ directory. No dependencies, no network calls.

The library returns the figures as exported and computes nothing. Employer cost totals are the ones published in the dataset; the library does not recompute them. For a cost at another salary, use the calculator at https://eorscope.com/eor-cost-calculator/.

Records are dicts:

country{iso, name, region, currency, example_salary_usd, example_role, example_place, basic_share_of_gross, employer_cost_pct, employer_cost_pct_printed, total_declared_floor, total_declared_ceiling, employer_cost_usd_month, total_excludes, fx_local_per_usd, fx_date, last_reviewed}
contribution{iso, country, currency, line_type, line_id, line_name, rate, base, base_factor, base_floor_annual_local, base_cap_annual_local, flat_annual_local, min_gross_annual_local, max_gross_annual_local, extra_kind, extra_value, secondary_source, thirteenth_month, total_declared_floor, total_declared_ceiling, in_total, applies, notes, source_name, source_url, source_checked_at}
assumption{iso, id, label, value, notes}

Text is a string, integers and floats are numbers, booleans are the atoms true and false, dates are date(Y,M,D) and an empty optional value is the atom null.

A country with total_declared_floor true publishes an "at least" figure, not a total. Under total_declared_ceiling the modelled contribution base is the highest the law allows, and lines with in_total false come on top. A contribution's rate applies to its base times base_factor (1 unless the law builds the base from more than the salary, as Mexico's integrated base does), before base_floor_annual_local and base_cap_annual_local.

Sources and method: https://eorscope.com/methodology/. The figures are a cost comparison, not legal or tax advice.

license: - Code MIT; embedded data (data/) CC BY 4.0, see data/LICENSE.

86countries_in_region(Region, Countries) :- 87 must_be(text, Region), 88 string_lower(Region, Key), 89 findall(C, ( summary_row(_, C), 90 get_dict(region, C, R), 91 string_lower(R, Key) 92 ), Countries).

108assumptions(Iso, Assumptions) :- 109 iso_key(Iso, Key), 110 findall(A, assumption_row(Key, A), Assumptions). 111 112iso_key(Iso, Key) :- 113 must_be(text, Iso), 114 string_upper(Iso, Key). 115 116 117 /******************************* 118 * PARSING * 119 *******************************/

128parse_csv(Text, Name, Records) :- 129 string_codes(Text, Cs0), 130 ( Cs0 = [0xFEFF|Cs] -> true ; Cs = Cs0 ), 131 csv_records(Cs, Name, Records), 132 ( Records = [Head|Rows] 133 -> length(Head, N), 134 forall(member(R, Rows), 135 ( length(R, M), 136 M == N 137 -> true 138 ; length(R, M), 139 format(atom(Msg), 'expected ~d fields, got ~d', [N, M]), 140 csv_fail(Name, Msg) 141 )) 142 ; csv_fail(Name, empty) 143 ). 144 145csv_records([], _, []) :- !. 146csv_records(Cs, Name, Records) :- 147 ( eol(Cs, Rest) % blank line 148 -> csv_records(Rest, Name, Records) 149 ; csv_record(Cs, Name, Fields, Rest), 150 Records = [Fields|More], 151 csv_records(Rest, Name, More) 152 ). 153 154eol([0'\r,0'\n|R], R) :- !. 155eol([0'\n|R], R) :- !. 156eol([0'\r|R], R). 157 158csv_record(Cs, Name, [F|Fs], Rest) :- 159 csv_field(Cs, Name, F, Cs1), 160 ( Cs1 = [0',|Cs2] 161 -> csv_record(Cs2, Name, Fs, Rest) 162 ; Cs1 == [] 163 -> Fs = [], Rest = [] 164 ; eol(Cs1, Rest) 165 -> Fs = [] 166 ). 167 168csv_field([0'"|Cs], Name, Field, Rest) :- 169 !, 170 quoted(Cs, Name, Codes, Rest), 171 ( ( Rest == [] ; Rest = [C|_], memberchk(C, `,\r\n`) ) 172 -> string_codes(Field, Codes) 173 ; csv_fail(Name, 'extraneous character after a closing quote') 174 ). 175csv_field(Cs, Name, Field, Rest) :- 176 unquoted(Cs, Name, Codes, Rest), 177 string_codes(Field, Codes). 178 179quoted([], Name, _, _) :- 180 csv_fail(Name, 'unterminated quoted field'). 181quoted([0'",0'"|Cs], Name, [0'"|T], Rest) :- !, 182 quoted(Cs, Name, T, Rest). 183quoted([0'"|Cs], _, [], Cs) :- !. 184quoted([C|Cs], Name, [C|T], Rest) :- 185 quoted(Cs, Name, T, Rest). 186 187unquoted([], _, [], []). 188unquoted([C|Cs], Name, Codes, Rest) :- 189 ( memberchk(C, `,\r\n`) 190 -> Codes = [], Rest = [C|Cs] 191 ; C == 0'" 192 -> csv_fail(Name, 'bare quote in an unquoted field') 193 ; Codes = [C|T], 194 unquoted(Cs, Name, T, Rest) 195 ). 196 197csv_fail(Name, Msg) :- 198 throw(error(syntax_error(csv(Name, Msg)), _)). 199 200% Field conversion. Column types follow the dataset's datapackage.json. 201 202field(Row, Col, Value) :- 203 ( memberchk(Col-V0, Row) 204 -> Value = V0 205 ; throw(error(existence_error(csv_column, Col), _)) 206 ). 207 208str(Row, Col, S) :- field(Row, Col, S). 209 210int(Row, Col, I) :- 211 field(Row, Col, S), 212 ( number_string(I, S), integer(I) -> true ; bad(Col, integer, S) ). 213 214num(Row, Col, F) :- 215 opt(Row, Col, F), 216 ( F == null -> bad(Col, number, "") ; true ). 217 218opt(Row, Col, F) :- 219 field(Row, Col, S), 220 ( S == "" 221 -> F = null 222 ; catch(number_string(N, S), _, fail) 223 -> F is float(N) 224 ; bad(Col, number, S) 225 ). 226 227bool(Row, Col, B) :- 228 field(Row, Col, S), 229 ( memberchk(S, ["true", "TRUE", "True", "1"]) -> B = true 230 ; memberchk(S, ["false", "FALSE", "False", "0"]) -> B = false 231 ; bad(Col, boolean, S) 232 ). 233 234day(Row, Col, D) :- 235 field(Row, Col, S), 236 ( S == "" 237 -> D = null 238 ; split_string(S, "-", "", [Ys, Ms, Ds]), 239 string_length(Ys, 4), string_length(Ms, 2), string_length(Ds, 2), 240 number_string(Y, Ys), number_string(M, Ms), number_string(Day, Ds), 241 integer(Y), integer(M), integer(Day), 242 catch(date_time_stamp(date(Y, M, Day, 0, 0, 0, 0, -, -), Stamp), _, fail), 243 stamp_date_time(Stamp, date(Y, M, Day, _, _, _, _, _, _), 0) 244 -> D = date(Y, M, Day) 245 ; bad(Col, date, S) 246 ). 247 248bad(Col, Type, S) :- 249 throw(error(type_error(Type, Col=S), _)). 250 251read_rows(File, Rows) :- 252 prolog_load_context(directory, Dir), 253 atomic_list_concat([Dir, '/../data/', File], Path), 254 read_file_to_string(Path, Text, [encoding(utf8)]), 255 parse_csv(Text, File, [Head|Recs]), 256 findall(Row, ( member(Rec, Recs), pairs_keys_values(Row, Head, Rec) ), Rows). 257 258summary(Row, C) :- 259 str(Row, "iso", Iso), 260 str(Row, "country", Name), 261 str(Row, "region", Region), 262 str(Row, "currency", Currency), 263 int(Row, "example_salary_usd", Salary), 264 str(Row, "example_role", Role), 265 str(Row, "example_place", Place), 266 opt(Row, "basic_share_of_gross", Basic), 267 num(Row, "employer_cost_pct", Pct), 268 num(Row, "employer_cost_pct_printed", Printed), 269 bool(Row, "total_declared_floor", Floor), 270 bool(Row, "total_declared_ceiling", Ceiling), 271 num(Row, "employer_cost_usd_month", USDMonth), 272 str(Row, "total_excludes", Excludes), 273 num(Row, "fx_local_per_usd", FX), 274 day(Row, "fx_date", FXDate), 275 day(Row, "last_reviewed", Reviewed), 276 C = country{iso:Iso, name:Name, region:Region, currency:Currency, 277 example_salary_usd:Salary, example_role:Role, 278 example_place:Place, basic_share_of_gross:Basic, 279 employer_cost_pct:Pct, employer_cost_pct_printed:Printed, 280 total_declared_floor:Floor, total_declared_ceiling:Ceiling, 281 employer_cost_usd_month:USDMonth, total_excludes:Excludes, 282 fx_local_per_usd:FX, fx_date:FXDate, last_reviewed:Reviewed}. 283 284contribution(Row, L) :- 285 str(Row, "iso", Iso), 286 str(Row, "country", Country), 287 str(Row, "currency", Currency), 288 str(Row, "line_type", Type), 289 str(Row, "line_id", ID), 290 str(Row, "line_name", Name), 291 opt(Row, "rate", Rate), 292 str(Row, "base", Base), 293 opt(Row, "base_factor", Factor), 294 opt(Row, "base_floor_annual_local", Floor0), 295 opt(Row, "base_cap_annual_local", Cap), 296 opt(Row, "flat_annual_local", Flat), 297 opt(Row, "min_gross_annual_local", MinGross), 298 opt(Row, "max_gross_annual_local", MaxGross), 299 str(Row, "extra_kind", Kind), 300 opt(Row, "extra_value", Extra), 301 bool(Row, "secondary_source", Secondary), 302 bool(Row, "thirteenth_month", Thirteenth), 303 bool(Row, "total_declared_floor", Floor), 304 bool(Row, "total_declared_ceiling", Ceiling), 305 bool(Row, "in_total", InTotal), 306 str(Row, "applies", Applies), 307 str(Row, "notes", Notes), 308 str(Row, "source_name", SourceName), 309 str(Row, "source_url", SourceURL), 310 day(Row, "source_checked_at", Checked), 311 L = contribution{iso:Iso, country:Country, currency:Currency, 312 line_type:Type, line_id:ID, line_name:Name, rate:Rate, 313 base:Base, base_factor:Factor, 314 base_floor_annual_local:Floor0, 315 base_cap_annual_local:Cap, flat_annual_local:Flat, 316 min_gross_annual_local:MinGross, 317 max_gross_annual_local:MaxGross, extra_kind:Kind, 318 extra_value:Extra, secondary_source:Secondary, 319 thirteenth_month:Thirteenth, total_declared_floor:Floor, 320 total_declared_ceiling:Ceiling, in_total:InTotal, 321 applies:Applies, notes:Notes, source_name:SourceName, 322 source_url:SourceURL, source_checked_at:Checked}. 323 324assumption(Row, A) :- 325 str(Row, "iso", Iso), 326 str(Row, "id", ID), 327 str(Row, "label", Label), 328 num(Row, "value", Value), 329 str(Row, "notes", Notes), 330 A = assumption{iso:Iso, id:ID, label:Label, value:Value, notes:Notes}. 331 332% The data are read once, when this file is compiled, and become facts. 333 334data_clauses(Clauses) :- 335 read_rows('country_summary.csv', SRows), 336 maplist(summary, SRows, Cs), 337 map_list_to_pairs([C, K]>>get_dict(iso, C, K), Cs, Pairs0), 338 keysort(Pairs0, Pairs), 339 findall(summary_row(K, C), member(K-C, Pairs), S), 340 read_rows('country_contributions.csv', CRows), 341 findall(contribution_row(K, L), 342 ( member(R, CRows), contribution(R, L), get_dict(iso, L, K) ), Ls), 343 read_rows('country_assumptions.csv', ARows), 344 findall(assumption_row(K, A), 345 ( member(R, ARows), assumption(R, A), get_dict(iso, A, K) ), As), 346 append([S, Ls, As], Clauses). 347 348term_expansion(eorscope_data, Clauses) :- 349 data_clauses(Clauses). 350 351eorscope_data