Exploring endpoints

All discovery functions read a catalog bundled with the package, so they work offline and never touch the network. Use them to find the right endpoint and learn exactly which parameters and fields it supports before you query.

Browse the catalog

catalog() lists every endpoint with its group, title and URL:

import tcepepy as tce

tce.catalog()
#             endpoint     group                           title  url
# 0   ReceitasEstaduais  Receitas  Relação das Receitas Estaduais  ...
# 1  ReceitasMunicipais  Receitas         Receitas Municipais      ...
# ...

Filter by keyword (case-insensitive, matches name or title):

tce.catalog(search="licit")     # procurement (licitação) endpoints
tce.catalog(search="obra")      # public works

Inspect input parameters

params() returns a DataFrame describing the query parameters an endpoint accepts:

tce.params("Contratos")
#               api_name               r_name  required    type        description
# 0       UnidadeGestora      unidade_gestora     False  string    Unidade Gestora
# 1       CodigoEfiscoUG       codigo_efisco_ug    False  string    ...
# ...

• api_name — the exact name expected by the API.
• r_name — the snake_case alias you can pass instead.
• required, type, description — metadata from the catalog.

Either spelling works when querying, so you can copy whichever column you prefer.

Inspect output fields

fields() returns the columns an endpoint produces:

tce.fields("Contratos")
#          name             r_name    type      description
# 0  CodigoContrato  codigo_contrato  string    ...
# ...

The r_name column is the snake_case column name you get back in the result DataFrame (when clean_names=True, the default).

Full endpoint metadata

For programmatic access to everything the catalog knows about an endpoint (raw dict with entrada and saida):

meta = tce.endpoint("Contratos")
meta["grupo"], meta["descricao"]
len(meta["entrada"]), len(meta["saida"])

A discovery helper

Putting it together, a small helper to preview any endpoint:

def preview(name):
    print(tce.catalog(search=name))
    print("\nParameters:\n", tce.params(name))
    print("\nFields:\n", tce.fields(name))

preview("Obras")