Release history.

Added

• add version synchronization and build scripts
• web-ui: implement subsystem selection in download steps and enhance navigation
• enhance SIM filename parsing and update FTP date range checks
• enhance settings and directory management
• new React+FastAPI web UI, fix double-nested datasus_db storage path
• Update storage paths to use 'datasus_db' for Parquet files and enhance compatibility with legacy structures
• Add SIM descriptive mappings and enhance VIEW creation for categorical fields
• Add new fields to SIM schema and update CID array transformation for maternal cause
• Implement custom BOOLEAN mappings and enhance CID array transformation for SIM subsystem
• Enhance SIM data processing with CID array transformation and subsystem-specific mappings
• Enhance SIM data processing with IDADE field transformation
• Add upload command for MotherDuck integration
• Implement TODO.md improvements (tasks 1-3)
• Update TODO with new features and enhancements
• Implement Unicode support and enhance logging in CLI and Web Interface
• Integrate progress callback with Web Interface
• Implement TODO improvements (CID cleanup, logs, cancellation, progress)
• I8: Implement memory-aware processing for large datasets
• I7: Improve Query page data dictionary with types and null stats
• I6: Improve Status page with tables instead of charts
• I4: Add download-estimate command to preview file sizes
• I3: Improve DBC conversion with CSV output and single file support
• I5: Rename CLI commands run→pipeline, download→download-only
• I2: Configure Streamlit to skip email prompt on first run
• I1: Rename project from pydatasus to datasus-etl
• Refactor project structure and enhance Web Interface functionality
• Update dataframe display settings to use 'stretch' width for improved layout
• Add CID validation transform for SQL data processing
• Enhance SQL transformation and validation, update web date inputs
• web: Enhance SQL templates and editor functionality in web UI
• web: Improve web UI for health researchers
• Enhance Bash permissions with additional commands for modular transformations
• Use original DBC filename for exported files (Melhoria 04)
• Add --output-format option for CSV export (Melhoria 03)
• Add --raw option for export without type conversions (Melhoria 05)
• Add modular SQL transform system (Melhoria 07)
• Add TODO.md for pipeline performance improvements and enhancements
• Update permissions and ignore virtual environment files
• Add SIM pipeline and base architecture (Melhoria 04)
• Add pre-download report with confirmation (Melhoria 03)
• Improve CLI UX with required params and examples (Melhoria 02)
• Add automatic cleanup of temporary DBC/DBF files (Melhoria 01)
• Add git reset command to permissions in settings.local.json
• Add IBGE municipality enrichment (Melhoria 02)
• Add automatic cleanup of temporary DBC/DBF files (Melhoria 01)
• Update .gitignore to include additional file types and add new IBGE report
• Add Streamlit web interface command to CLI
• Add Streamlit web interface (Melhoria 07)
• Add unit and integration tests (Melhoria 06)
• Add incremental update support (Melhoria 05)
• Modularize dataset configs with base class (Melhoria 04)
• Create datasus CLI with typer (Melhoria 03)
• Add source_file column to track origin DBC file (Melhoria 02)
• Add subsystem folder structure (Melhoria 01)
• Add TODO.md to .gitignore to prevent tracking of TODO file
• Add Hive-partitioned Parquet export with canonical schema
• Enhance configuration and processing pipeline with new constants and improved SQL transformations
• Add SQL auto-generation helpers to SIHSUS schema
• Add SIHSUS Parquet schema definition with DuckDB types
• Replace TABWIN with datasus-dbc for cross-platform support
• Add Phase 2 optimizations and comprehensive examples
• Optimize pipeline with DuckDB streaming and SQL transformations

Fixed

• nuitka: compile package dir (not __main__.py); Windows 4-part version
• web-ui: drop postcss.config.js from tsconfig.node includes, add bun.lock
• Correct MotherDuck upload to use direct connection
• Use subprocess + log file for Web Interface terminal output
• Use module-level dict for thread-safe progress updates
• Use st.session_state + st.rerun() for Web Interface progress
• Use threading for real-time progress updates in Web Interface
• Update PyPI metadata with expanded keywords and classifiers (Melhoria 06)
• Update repository URL in README for cloning instructions
• Delete empty dbc/ and dbf/ directories after cleanup
• Remove typer[all] extra to fix pip installation warning
• Use rglob for recursive cleanup of DBC/DBF files
• Correct IBGE code extraction and remove duplicate columns
• Use rglob for recursive cleanup of DBC/DBF files

Performance

• Optimize data transfer using Parquet instead of row-by-row inserts
• Add adaptive DBF insertion strategy for optimal performance

Changed

• Replace progress bar with terminal output in Web Interface
• Remove redundant output of generated files in basic usage example
• Remove PHASE3_SUMMARY.md as it is no longer needed
• Remove check_transform.py script as it is no longer needed
• Clean up export logging and remove unused date extraction - Remove file size control at 512MB per file (unsupported in Duckdb)

Documentation

• I8: Add RAM analysis report for pipeline memory optimization
• Add research documents for improvements 01, 02, 09, 10
• Add example scripts for new features (Melhoria 08)
• Update README.md with complete usage examples (Melhoria 03)
• Add schema usage examples
• Remove all CSV references and legacy migration guide
• Add Phase 3 summary and update Phase 2 references
• Update README for optimized architecture
• Add comprehensive Phase 2 summary
• Add deprecation warnings to legacy modules
• Add usage example and optimization summary

Other

• update version to 0.1.2
• (feat): Try the first cloud build
• Remove outdated documentation and improve user experience with new tutorial and web UI enhancements
• fixÇ commit web-ui lib modules
• Better UI
• Fix datasus CLI
• Switch storage from duckdb to parquet
• Refactor DataSUS ETL: Replace Parquet with DuckDB integration
• Update README.md to correct descriptions, enhance CLI usage examples, and add memory-aware processing details
• Rename project from "pydatasus" to "datasus-etl" and update related documentation; add DuckDB CLI integration in the command line interface.
• Fix in Paulo's house
• Enhance ParquetQueryEngine and SQLTransformer for improved schema handling and error logging
• Update examples to remove CSV references and reset version to 0.1.0
• Remove deprecated code and unused dependencies (v2.0)
• Init repo
• Initial commit