Skip to content

Commit

Permalink
pequena correcao na restricao de certas abreviacoes no final da string
Browse files Browse the repository at this point in the history
  • Loading branch information
dhersz committed Apr 5, 2024
1 parent 2a3aae0 commit 2e316cd
Show file tree
Hide file tree
Showing 2 changed files with 11 additions and 10 deletions.
17 changes: 9 additions & 8 deletions R/padronizar_bairros.R
Original file line number Diff line number Diff line change
Expand Up @@ -76,7 +76,7 @@ padronizar_bairros <- function(bairros) {
"\\bD\\.? INDUSTRIAL\\b" = "DISTRITO INDUSTRIAL",
"\\bS\\.? INDUSTRIAL\\b" = "SETOR INDUSTRIAL",
"\\b(P\\.? INDUSTRIAL|PARQUE IN)\\b\\.?" = "PARQUE INDUSTRIAL",
"\\bLOT(EAME?)?\\b\\.?[^$]" = "LOTEAMENTO",
"\\bLOT(EAME?)?\\b\\.?(?!$)" = "LOTEAMENTO",
"^LT\\b\\.?" = "LOTEAMENTO",
"\\bZN\\b\\.?" = "ZONA",
"^Z\\b\\.?" = "ZONA",
Expand All @@ -102,8 +102,8 @@ padronizar_bairros <- function(bairros) {
"^SIT\\b\\.?" = "SITIO",
"^CAM\\b\\.?" = "CAMINHO",
"\\bCERQ\\b\\.?" = "CERQUEIRA",
"\\bCONS\\b\\.?[^$]" = "CONSELHEIRO",
"\\bPROL\\b\\.?[^$]" = "PROLONGAMENTO",
"\\bCONS\\b\\.?(?!$)" = "CONSELHEIRO", # CONS COMUN => CONSELHO COMUNITARIO, provavelmente
"\\bPROL\\b\\.?(?!$)" = "PROLONGAMENTO",

# titulos
"\\bSTO\\b\\.?" = "SANTO",
Expand All @@ -128,23 +128,24 @@ padronizar_bairros <- function(bairros) {
"\\bS\\.? (PAULO|VICENTE|FRANCISCO|DOMINGOS?|CRISTOVAO)\\b" = "SAO \\1",

"\\bALMTE\\b\\.?" = "ALMIRANTE",
"\\bMAL\\b\\.?[^$]" = "MARECHAL",
"\\bMAL\\b\\.?(?!$)" = "MARECHAL",
"\\bSGTO?\\b\\.?" = "SARGENTO",
"\\bCEL\\b\\.?" = "CORONEL",
"\\bBRIG\\b\\.?" = "BRIGADEIRO",
"\\bTEN\\b\\.?" = "TENENTE",
"\\bBRIGADEIRO (F\\.?|FARIA) (L|LIMA)\\b\\.?" = "BRIGADEIRO FARIA LIMA",

"\\bPRES(ID)?\\b\\.?[^$]" = "PRESIDENTE",
# consertar esse presidente
"\\bPRES(ID)?\\b\\.?(?!$)" = "PRESIDENTE",
"\\bGOV\\b\\.?" = "GOVERNADOR", # pode acabar com GOV. - e.g. ilha do gov.
"\\bPREF\\b\\.?[^$]" = "PREFEITO",
"\\bDEP\\b\\.?[^$]" = "DEPUTADO",
"\\bPREF\\b\\.?(?!$)" = "PREFEITO",
"\\bDEP\\b\\.?(?!$)" = "DEPUTADO",

"\\bDR\\b\\.?" = "DOUTOR",
"\\bDRA\\b\\.?" = "DOUTORA",
"\\bPROF\\b\\.?" = "PROFESSOR",
"\\bPROFA\\b\\.?" = "PROFESSORA",
"\\bPE\\b\\.[^$]" = "PADRE",
"\\bPE\\b\\.(?!$)" = "PADRE",

"\\bD\\b\\.? (PEDRO|JOAO|HENRIQUE)" = "DOM \\1",
"\\bI(NF)?\\.? DOM\\b" = "INFANTE DOM",
Expand Down
4 changes: 2 additions & 2 deletions R/padronizar_logradouros.R
Original file line number Diff line number Diff line change
Expand Up @@ -193,9 +193,9 @@ padronizar_logradouros <- function(logradouros) {
"\\bSEN\\b\\.?" = "SENADOR",
"\\bPREF\\b\\.?" = "PREFEITO",
"\\bDEP\\b\\.?" = "DEPUTADO",
"\\bVER\\b\\.?[^$ ]" = "VEREADOR",
"\\bVER\\b\\.?(?!$)" = "VEREADOR",
"\\bESPL?\\.? (DOS )?MIN(IST(ERIOS?)?)?\\b\\.?" = "ESPLANADA DOS MINISTERIOS",
"\\bMIN\\b\\.?[^$ ]" = "MINISTRO",
"\\bMIN\\b\\.?(?!$)" = "MINISTRO",

# abreviacoes
"\\bUNID\\b\\.?" = "UNIDADE",
Expand Down

0 comments on commit 2e316cd

Please sign in to comment.