NEWS

GetFREData 1.2.0

Populated cd_cvm across all tables returned by get_fre_data2() (mapped from cia_aberta via document ID with CNPJ fallback), ensuring companies_cvm_codes filters all tables in the output.

Added companies_cvm_codes argument to get_fre_data2(), allowing users to filter results by one or more CVM company codes.
Cleaned table names in the output list by stripping the redundant fre_cia_aberta_ prefix (e.g. capital_social instead of cia_aberta_capital_social).
Added verbose logical parameter to get_fre_data2() to allow controlling or suppressing informative console output.
Added parsed RDS caching in cache_folder/parsed-rds/ so subsequent calls for previously parsed years load instantly from cached RDS files without re-unzipping or re-parsing CSVs.
Added comprehensive input validation to get_fre_data2(): type and range checks for first_year/last_year, verbose, a warning when requesting future years, and type checking for companies_cvm_codes.
Added _pkgdown.yml for website configuration and package-level documentation (?GetFREData).

Fixed a crash in the download helper where file.size() was called on a non-existent file after a failed download attempt.
Download failures now raise a clear error (cli::cli_abort()) instead of silently returning FALSE, which previously led to confusing downstream errors.
Suppressed readr's CSV reading progress and speed messages (progress = FALSE, suppressMessages()) during dataset parsing.

Removed the accidentally exported my_fix_cols() function from the public API. This was an internal helper and should not have been exported.

Expanded unit test coverage with offline tests for CSV parsing helpers (fix_single_name(), read_single_csv(), fix_df()), input validation, and deprecation stubs.
Updated GitHub Actions workflow to @v4 (actions/checkout@v4).
Resolved R CMD check NOTE regarding undefined global variable .data in filter_by_company(), reaching a clean Status: OK (0 ERRORS, 0 WARNINGS, 0 NOTES).

Removed ~1,100 lines of dead code from the deprecated get_fre_data() XML pipeline: fcts_xml.R, fct_read_fre_zip_file.R, fct_get_fre_links.R, and fct_utils.R. The get_fre_data() function remains as a minimal deprecation stub pointing users to get_fre_data2().
Dropped XML and magrittr from package dependencies.
Migrated all remaining %>% (magrittr pipe) usage to the native R pipe (|>).

Added clean_fre_cache() utility to allow users to clear down local raw ZIP and parsed RDS files.
Suppressed expected coercion warnings during CSV file parsing and column format conversions in get_fre_data2().
Updated all user documentation, vignettes, and roxygen example references to use the active get_fre_data2() function instead of deprecated get_fre_data().

added new function get_fre_data2(), which downloads data from CVM website (and not b3)

fixed issue with directory of ftp files with FRE links. Now, the folder is dest_file <- file.path(cache_folder, paste0('ftp_zip_raw-', Sys.Date()), indexed by day of query. This should facilitate updating the data.