use regex::Regex;
use std::sync::LazyLock;
#[derive(Debug, Clone)]
pub struct ClassificationRule {
pub pattern: Regex,
pub n1: String,
pub n2: String,
pub n3: String,
}
pub static CLASSIFICATION_RULES: LazyLock<Vec<ClassificationRule>> = LazyLock::new(|| {
vec![
rule(
r"(?i)juzgado|instrucci[óo]n|tsj|audiencia.*penal",
"PROF",
"PROF-JUD",
"JUD-JUZ",
),
rule(r"(?i)fiscal[íi]a", "PROF", "PROF-FIS", "JUD-FIS"),
rule(
r"(?i)icav|icab|colegio\s+de\s+abogados",
"PROF",
"PROF-COL",
"COL-ABG",
),
rule(r"(?i)procurador", "PROF", "PROF-PROC", "COL-PRO"),
rule(r"(?i)notar[ií]a|notario", "PROF", "PROF-NOT", "COL-NOT"),
rule(
r"(?i)despacho\s+jur[ií]dico|bufete\s+jur[ií]dico|consultor.*legal",
"PROF",
"PROF-CON",
"JUR-CON",
),
rule(
r"(?i)@gva\.es|@xij\.gencat|generalitat",
"INST",
"INST-AUT",
"AUT-VAL",
),
rule(
r"(?i)ayuntamiento|@valencia\.es",
"INST",
"INST-LOC",
"LOC-VAL",
),
rule(
r"(?i)@seg-social\.es|@policia\.es|@mineco\.es|tgss",
"INST",
"INST-EST",
"EST-SS",
),
rule(
r"(?i)ministerio|delegaci[oó]n\s+del\s+gobierno|gobierno\s+de",
"INST",
"INST-EST",
"EST-GOB",
),
rule(
r"(?i)hospital|cl[ií]nica|salud|sanidad|centro\s+de\s+salud",
"SALUD",
"SALUD-SAN",
"SAN-HOSP",
),
rule(
r"(?i)revolut|cofidis|talenom|evobanco",
"FIN",
"FIN-FINTEC",
"FINTEC",
),
rule(
r"(?i)bybit|bit2me|crypto\.com",
"FIN",
"FIN-CRYPTO",
"CRYPTO",
),
rule(
r"(?i)seguros|caser|agencia\.caser",
"FIN",
"FIN-SEG",
"SEGUROS",
),
rule(
r"(?i)axesor|procobro|gescobro",
"FIN",
"FIN-REC",
"RECOBROS",
),
rule(
r"(?i)banco|caixa|bbva|santander|sabadell|bankinter|ing\s|openbank",
"FIN",
"FIN-BAN",
"BANCA",
),
rule(
r"(?i)broker|inversi[oó]n|gestor[ií]a|asesor[ií]a\s+financiera",
"FIN",
"FIN-INV",
"INV",
),
rule(r"(?i)@itti\.es|itti", "FORM", "FORM-FP", "FP"),
rule(r"(?i)unie|universidad", "FORM", "FORM-UNIV", "UNIV"),
rule(
r"(?i)academia|escuela|instituto|colegio\s+\w+|formaci[oó]n",
"FORM",
"FORM-EDU",
"EDU",
),
rule(
r"(?i)gigabyte|sony|mitsubishi|ecodan",
"TEC",
"TEC-HW",
"HW",
),
rule(
r"(?i)cursor|notion|perplexity|clickup|paragraph|neuralnine|bio\.link|mt5|appvillis",
"TEC",
"TEC-SW",
"SW",
),
rule(
r"(?i)orange|digi|one\.com|@orange\.com",
"TEC",
"TEC-COM",
"COM",
),
rule(r"(?i)tidal|spotify|netflix", "TEC", "TEC-ENT", "ENT"),
rule(
r"(?i)hosting|cloud|vps|servidor|dominio",
"TEC",
"TEC-SVC",
"SVC",
),
rule(
r"(?i)hotel|restaurante|bar\s|caf[eé]|hostal",
"HOST",
"HOST-REST",
"REST",
),
rule(
r"(?i)taxi|transporte|log[ií]stica|mensajer[ií]a",
"TRAN",
"TRAN-VEH",
"VEH",
),
rule(
r"(?i)inmobiliaria|constructora|promotora|inmobiliari",
"INMO",
"INMO-INM",
"INM",
),
rule(
r"(?i)limpieza|mantenimiento|seguridad|vigilancia",
"SERV",
"SERV-EXT",
"EXT",
),
rule(
r"(?i)asociaci[oó]n|vecinos|comunidad",
"ASOC",
"ASOC-VEC",
"VEC",
),
]
});
fn rule(pattern: &str, n1: &str, n2: &str, n3: &str) -> ClassificationRule {
ClassificationRule {
pattern: Regex::new(pattern).expect("regex de clasificación inválida"),
n1: n1.to_string(),
n2: n2.to_string(),
n3: n3.to_string(),
}
}