test_that("automatic schema detection recognizes English and Persian columns", {
  en <- data.frame(Value = 10.4, Indicator = "unemployment rate",
                   Country = "Spain", Year = 2025, Unit = "percent",
                   Source = "Example NSO", check.names = FALSE)
  s1 <- detect_stat_schema(en)
  expect_equal(s1$mapping$value, "Value")
  expect_equal(s1$mapping$geo, "Country")
  expect_equal(s1$mapping$time, "Year")

  fa <- data.frame("مقدار" = "۱۲٫۵", "عنوان شاخص" = "نرخ بیکاری",
                   "کشور" = "ایران", "سال" = "۱۴۰۳", "واحد" = "درصد",
                   check.names = FALSE)
  s2 <- detect_stat_schema(fa)
  expect_equal(s2$mapping$value, "مقدار")
  expect_equal(s2$mapping$indicator, "عنوان شاخص")
  expect_equal(s2$mapping$geo, "کشور")
  expect_equal(s2$mapping$time, "سال")
})

test_that("generic importer normalizes Persian digits and units", {
  fa <- data.frame("مقدار" = "۱۲٫۵", "عنوان شاخص" = "نرخ بیکاری",
                   "کشور" = "ایران", "سال" = "۱۴۰۳", "واحد" = "درصد",
                   check.names = FALSE)
  x <- read_official_stats(fa, provider = "Statistical Center of Iran")
  expect_s3_class(x, "stat_provider_data")
  expect_equal(x$.value, 12.5)
  expect_equal(x$.time, "1403")
  expect_equal(x$.unit, "percent")
  ref <- as_stat_reference(x)
  expect_equal(ref$source, "Statistical Center of Iran")
})

test_that("easy audit accepts ordinary data and plain text", {
  d <- data.frame(value = 10.4, indicator = "unemployment rate",
                  country = "Spain", year = 2025, unit = "percent",
                  source = "Example NSO", stringsAsFactors = FALSE)
  a <- audit_stats(d, "The reported value is 10.4%.")
  expect_s3_class(a, "stat_fidelity_audit")
  expect_equal(a$components$numerical$score, 1)
  expect_equal(a$components$unit$score, 1)
})

test_that("generic provider profiles can carry reusable mappings", {
  register_official_provider(
    "demo_nso", "Demo National Statistical Office",
    mapping = list(value = "V", indicator = "I", geo = "G", time = "T", unit = "U"),
    host_patterns = "data.demo.example", overwrite = TRUE
  )
  d <- data.frame(V = 3.2, I = "test rate", G = "Exampleland", T = 2025, U = "percent")
  x <- read_official_stats(d, provider = "demo_nso")
  expect_equal(attr(x, "provider"), "Demo National Statistical Office")
  expect_equal(x$.value, 3.2)
  expect_equal(detect_stat_provider("https://data.demo.example/table.csv"), "demo_nso")
})

test_that("provider detection includes WHO and Iranian organisations", {
  expect_equal(detect_stat_provider("https://data.who.int/indicators"), "who")
  expect_equal(detect_stat_provider("https://www.amar.org.ir/data.csv"), "sci_iran")
  x <- official_stat_providers(include_custom = FALSE)
  expect_true(all(c("who", "sci_iran", "cbi_iran", "mohme_iran") %in% x$id))
  expect_true(all(x$generic_import[x$id %in% c("who", "sci_iran")]))
})

test_that("plain-text claim extraction accepts Persian digits", {
  x <- extract_stat_claims("مقدار گزارش شده ۱۲٫۵٪ است.",
                           hints = list(indicator = "نرخ", geo = "ایران", time = "1403"))
  expect_equal(length(x), 1)
  expect_equal(x[[1]]$value, 12.5)
  expect_equal(x[[1]]$unit, "percent")
})