sidrar 0.5.0

• Added an actionable sidrar_limit_error when the SIDRA API rejects a request for exceeding its per-request value limit. The condition remains a sidrar_http_error and records the requested count, limit, minimum number of batches, response body, and request URL (#28).
• Added structured discovery through sidra_catalog(), sidra_metadata(), sidra_periods(), and sidra_locations() using official IBGE JSON endpoints. Identifiers remain character vectors and returned schemas are stable when upstream responses are empty or gain fields.
• Added sidra_query() and sidra_plan() to inspect URLs, explicit dimension cardinalities, and optional limit risk before retrieving values. No service limit is hard-coded.
• Added explicit, sequential batching with sidra_split() and sidra_collect(). Batches must be disjoint, retain their original order, and have identical names and column types before rows are combined.
• Added opt-in collection provenance through sidra_provenance() and opt-in disk caching for normalized catalog and metadata responses. Value responses are never cached by the package.
• Added territorial-view (G) queries, optional extinct units (/u/y), case-insensitive geographic aliases, and supported nNN level codes.
• All HTTP errors now carry status, response body, and URL fields. Transport failures can additionally identify timeout, TLS, DNS, connection, and transient failures while remaining sidrar_http_error conditions.
• Improved local input validation, reset row names after processing API headers, rejected structural URL delimiters and duplicate classifications, and clarified that value_type still applies when api is used.
• Clarified that SIDRA neighborhood identifiers belong to its territorial level and are not census tract identifiers (#8).

sidrar 0.3.0

• Replaced fragile HTML scraping with official JSON endpoints for table descriptors and the aggregate catalog.
• Restored search_sidra() after changes to the SIDRA website. Searches are now case- and accent-insensitive, support multiple non-adjacent terms, and return table codes as names (#24).
• Reworked info_sidra() while preserving its historical list components: table, period, variable, classific_category, and geo.
• Fixed get_sidra(api = ...) for complete official HTTPS URLs and preserved percent-encoded paths. Thanks to @viniciusoike for reporting the problem and proposing a solution in #25 and #26.
• Fixed direct API response parsing so changed or short server responses produce informative package errors instead of failing inside error handling (#16).
• Requests containing /h/n now preserve the first observation instead of treating it as a header.
• Fixed vector geographic filters and the construction of queries containing multiple classifications and partially specified category lists.
• Added value_type to preserve SIDRA special symbols in a character column or alongside the historical numeric value column (#18).
• Added HTTPS status checks, informative API errors, UTF-8 decoding, an identifying user agent, configurable timeouts, and retries for transient failures.
• Added support for intermediary and immediate geographic regions.
• Reduced runtime dependencies to httr, jsonlite, and utils.
• Added unit tests, opt-in live API tests, and continuous integration across current R platforms.

sidrar 0.2.9

• Address SSL error

sidrar 0.2.8

• Minor adjusts to address CRAN warnings

sidrar 0.2.7

• Adjusts in the Vignette to CRAN

sidrar 0.2.6

• Fixed bug in get_sidra (Issue #10)
• Minor change in the Vignette

sidrar 0.2.5

• Fixed bugs in get_sidra (Issue #5, #6)
• “tidyr” and “dplyr” package dependency removed

sidrar 0.2.4

• Fixed bug in search_sidra (Issue #2)
• Better error messages
• Vignette updated

sidrar 0.2.1

• Fixed list-column in resultant data.frame in get_sidra (Issue #1)

sidrar 0.2.0

• New argument in get_sidra (“api”)

sidrar 0.1.1

• Better messages depending of the input arguments

sidrar 0.1.0

• Initial version