{"id":28348065,"url":"https://github.com/macieklesiczka/azof","last_synced_at":"2026-03-02T00:34:23.587Z","repository":{"id":277982064,"uuid":"934095175","full_name":"MaciekLesiczka/azof","owner":"MaciekLesiczka","description":"Lakehouse with time travel","archived":false,"fork":false,"pushed_at":"2026-02-11T14:45:15.000Z","size":99168,"stargazers_count":4,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2026-02-11T22:40:12.964Z","etag":null,"topics":["datafusion","datalake","lakehouse","parquet","rust-lang"],"latest_commit_sha":null,"homepage":"","language":"Rust","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/MaciekLesiczka.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-02-17T09:11:30.000Z","updated_at":"2026-02-11T14:45:21.000Z","dependencies_parsed_at":"2025-02-17T11:27:55.258Z","dependency_job_id":"cb6bd046-ab0a-44ec-9acd-0d8d903ed509","html_url":"https://github.com/MaciekLesiczka/azof","commit_stats":null,"previous_names":["macieklesiczka/bazof","macieklesiczka/azof"],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/MaciekLesiczka/azof","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MaciekLesiczka%2Fazof","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MaciekLesiczka%2Fazof/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MaciekLesiczka%2Fazof/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MaciekLesiczka%2Fazof/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/MaciekLesiczka","download_url":"https://codeload.github.com/MaciekLesiczka/azof/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MaciekLesiczka%2Fazof/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29988054,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-03-01T22:42:38.399Z","status":"ssl_error","status_checked_at":"2026-03-01T22:41:51.863Z","response_time":124,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["datafusion","datalake","lakehouse","parquet","rust-lang"],"created_at":"2025-05-27T18:09:56.799Z","updated_at":"2026-03-02T00:34:23.229Z","avatar_url":"https://github.com/MaciekLesiczka.png","language":"Rust","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Azof\n\n[![Rust CI](https://github.com/MaciekLesiczka/azof/actions/workflows/rust.yml/badge.svg)](https://github.com/MaciekLesiczka/azof/actions/workflows/rust.yml)\n[![Crates.io](https://img.shields.io/crates/v/azof.svg)](https://crates.io/crates/azof)\n[![Documentation](https://docs.rs/azof/badge.svg)](https://docs.rs/azof)\n[![License](https://img.shields.io/crates/l/azof.svg)](LICENSE)\n\nQuery tables in object storage as of event time.\n\nAzof is a lakehouse format with time-travel capabilities that allows you to query data as it existed at any point in time, based on when events actually occurred rather than when they were recorded.\n\n\n# Practical Example\n\nThe `test-data/financials` table contains historical financial information about US public companies, including revenue, income, and share counts. The data is organized by fiscal quarter end dates - when financial results became official - rather than when the data was recorded in the system.\n\nWith Azof, you can query this data as it existed at any specific point in time:\n\n```sql\nSELECT key as symbol, revenue, net_income\n  FROM financials\n    AT ('2019-01-17T00:00:00.000Z') -- point-in-time snapshot of available financial data\n WHERE industry IN ('Software')\n ORDER BY revenue DESC\n LIMIT 5\n```\n\nThis query returns the top 5 software companies by revenue, using only financial data that was officially reported as of January 17, 2019.\n\n\n## What Problem Does Azof Solve?\n\nLakehouse formats typicaly allow time travel based on when data was written (processing time). Azof instead focuses on **event time** - the time when events actually occurred in the real world. This distinction is crucial for:\n\n- **Late-arriving data**: Process data that arrives out of order without rewriting history\n- **Consistent historical views**: Get consistent snapshots of data as it existed at specific points in time\n- **High cardinality datasets with frequent updates**: Efficiently handle use cases involving business processes (sales, support, project management, financial data) or slowly changing dimensions\n- **Point-in-time analysis**: Analyze the state of your data exactly as it was at any moment\n\n## Key Features\n\n- **Event time-based time travel**: Query data based on when events occurred, not when they were recorded\n- **Efficient storage of updates**: Preserves compacted snapshots of state to minimize storage and query overhead\n- **Hierarchical organization**: Uses segments and delta files to efficiently organize temporal data\n- **Tunable compaction policy**: Adjust based on your data distribution patterns\n- **SQL integration**: Query using DataFusion with familiar SQL syntax\n- **Integration with object storage**: Works with any object store (local, S3, etc.)\n\n\n## Project Structure\n\nThe Azof project is organized as a Rust workspace with multiple crates:\n\n- **azof**: The core library providing the lakehouse format functionality\n- **azof-cli**: A CLI utility demonstrating how to use the library\n- **azof-datafusion**: DataFusion integration for SQL queries\n\n## Getting Started\n\nTo build all projects in the workspace:\n\n```bash\ncargo build --workspace\n```\n\n## Using the CLI\n\nThe azof-cli provides a command-line interface for interacting with azof:\n\n```bash\n# Scan a table (current version)\ncargo run -p azof-cli -- scan --path ./test-data --table table0\n\n# Scan a table as of a specific event time\ncargo run -p azof-cli -- scan --path ./test-data --table table0 --as-of \"2024-03-15T14:30:00\"\n\n# Generate test parquet file from CSV\ncargo run -p azof-cli -- gen --path ./test-data --table table2 --file base\n```\n\n## DataFusion Integration\n\nThe azof-datafusion crate provides integration with Apache DataFusion, allowing you to:\n\n1. Register Azof tables in a DataFusion context\n2. Run SQL queries against Azof tables\n3. Perform time-travel queries using the AsOf functionality\n\n### Example\n\n```rust\nuse azof_datafusion::context::ExecutionContext;\n\nasync fn query_azof() -\u003e Result\u003c(), Box\u003cdyn std::error::Error\u003e\u003e {\n    let ctx = ExecutionContext::new(\"/path/to/azof\");\n\n    let df = ctx\n        .sql(\n            \"\n    SELECT key as symbol, revenue, net_income\n      FROM financials\n        AT ('2019-01-17T00:00:00.000Z')\n     WHERE industry IN ('Software')\n     ORDER BY revenue DESC\n     LIMIT 5;\n     \",\n        )\n        .await?;\n\n    df.show().await?;\n\n    Ok(())\n}\n```\n\nRun the example:\n\n```bash\ncargo run --example query_example -p azof-datafusion\n```\n\nIf you install the CLI with `cargo install --path crates/azof-cli`, you can run it directly with:\n\n```bash\nazof-cli scan --path ./test-data --table table0\n```\n\n## Project Roadmap\n\nAzof is under development. The goal is to implement a data lakehouse with the following capabilities:\n\n* Atomic, non-concurrent writes (single writer)\n* Consistent reads\n* Schema evolution\n* Event time travel queries \n* Handling late-arriving data\n* Integration with an execution engine\n\n### Milestone 0\n\n- [x] Script/tool for generating sample kv data set\n- [x] Key-value reader\n- [x] DataFusion table provider\n\n### Milestone 1\n\n - [x] Multiple columns support\n - [x] Data Types columns support\n - [x] Projection pushdown\n - [x] Projection pushdown in DataFusion table provider\n - [x] DataFusion table provider with AS OF operator\n - [ ] Single row, key-value writer\n - [ ] Document spec\n - [ ] Delta -\u003e snapshot compaction\n - [ ] Metadata validity checks\n\n### Milestone 2\n- [ ] Streaming in scan\n- [ ] Schema definition and evolution\n- [ ] Late-arriving data support\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmacieklesiczka%2Fazof","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmacieklesiczka%2Fazof","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmacieklesiczka%2Fazof/lists"}