FeedMe

FeedMe is a deterministic, linear, streaming ingest pipeline with mechanical guarantees around memory, ordering, and failure.

About

FeedMe is a high-performance, streaming data pipeline engine for Rust applications. It provides a linear, deterministic processing model with bounded resource usage, explicit error handling, and comprehensive observability.

Perfect for ETL, log processing, data cleaning, and real-time ingestion pipelines.

Key Guarantees

• Streaming, bounded memory — processes one event at a time; memory usage stays flat
• Deterministic processing — same input + same config → same output
• Structured errors — stage, code, and message for every failure
• Observability — metrics exportable (Prometheus or JSON) without affecting execution
• Extensible — add custom stages via a defined plugin contract

FeedMe Guarantees

FeedMe provides these mechanical guarantees:

• Events are processed strictly in input order
• Memory usage is bounded and input-size independent
• Stages cannot observe shared or mutated state
• Validation failures cannot be silently ignored
• Metrics collection cannot influence execution

Why FeedMe is Not X

FeedMe is intentionally not these things, and that's by design:

Not Distributed (like Vector or Fluent Bit)

FeedMe runs in a single process with no networking or cluster management.

Not Stateful (like traditional ETL tools)

Stages are deterministic: same input + config → same output. No hidden state or concurrency.

Not Async-First (like many modern Rust libraries)

Processing is synchronous by default. Async is an implementation detail for I/O stages.

Not a DSL (like Logstash)

No embedded languages or required config files. Code-first with optional YAML support.

Not a Daemon (like filebeat)

No long-running services or auto-restart. FeedMe is a library you embed in your application.

This focus enables FeedMe's core guarantees while keeping the codebase small and maintainable.

Invariants

FeedMe enforces mechanical behavioral guarantees that are tested via runtime assertions and contract tests. These invariants ensure reliability and prevent regressions.

Installation

Add this to your Cargo.toml:

[dependencies]
feedme = "0.1"
serde_json = "1.0"
regex = "1.0"

Or install via cargo:

cargo add feedme serde_json regex

Quick Start

Here's a 12-line pipeline that:

• ingests logs
• redacts PII
• validates schema
• filters noise
• guarantees determinism
• and fails safely

use feedme::{
    Pipeline, FieldSelect, RequiredFields, StdoutOutput, Deadletter,
    PIIRedaction, Filter, InputSource, Stage
};
use std::path::PathBuf;

fn main() -> anyhow::Result<()> {
    // Create pipeline: select fields → redact PII → require fields → filter → output
    let mut pipeline = Pipeline::new();
    pipeline.add_stage(Box::new(FieldSelect::new(vec![
        "timestamp".into(), "level".into(), "message".into(), "email".into()
    ])));
    pipeline.add_stage(Box::new(PIIRedaction::new(vec!["email".into()])));
    pipeline.add_stage(Box::new(RequiredFields::new(vec!["level".into()])));
    pipeline.add_stage(Box::new(Filter::new(Box::new(|event| {
        event.get("level").and_then(|v| v.as_str()) != Some("debug")
    }))));
    pipeline.add_stage(Box::new(StdoutOutput::new()));

    // Deadletter for errors
    let mut deadletter = Deadletter::new(PathBuf::from("errors.ndjson"));

    // Process input file
    let mut input = InputSource::File(PathBuf::from("input.ndjson"));
    input.process_input(&mut pipeline, &mut Some(&mut deadletter))?;

    // Export final metrics
    println!("Pipeline complete. Metrics:");
    for metric in pipeline.export_json_logs() {
        println!("{}", serde_json::to_string(&metric)?);
    }

    Ok(())
}

Input (input.ndjson):

{"timestamp":"2023-10-01T10:00:00Z","level":"info","message":"User logged in","email":"[email protected]"}
{"level":"debug","message":"Debug info"}
{"message":"Missing level"}

Output (stdout):

{"timestamp":"2023-10-01T10:00:00Z","level":"info","message":"User logged in","email":"[REDACTED]"}

Deadletter (errors.ndjson):

{"error":{"stage":"RequiredFields","code":"MISSING_FIELD","message":"Required field 'level' is missing"},"raw":"{\"message\":\"Missing level\"}"}

Metrics (JSON logs):

{"metric":"events_processed","value":2}
{"metric":"events_dropped","value":1}
{"metric":"errors","value":1}
{"metric":"stage_latencies","stage":"FieldSelect","count":3,"sum":0.05,"min":0.01,"max":0.02}
...

input.process_input(&mut pipeline, &mut deadletter)?;
println!("Example finished — metrics: {:?}", pipeline.export_json_logs());
Ok(())

}

## Determinism Verification

> **Determinism is a core guarantee** — identical runs produce identical outputs.

FeedMe guarantees deterministic output for identical inputs. Verify this with:

```bash
cargo run --example 09_complex_pipeline > run1.out
cargo run --example 09_complex_pipeline > run2.out
# On Unix: sha256sum run1.out run2.out
# On Windows: certutil -hashfile run1.out SHA256 && certutil -hashfile run2.out SHA256

The hashes should match, proving deterministic behavior.

Examples

Messy Input → Clean Output

Given messy.ndjson:

{"timestamp":"2023-10-01T10:00:00Z","level":"info","message":"User logged in","email":"[email protected]"}
{"level":"info","message":"Missing timestamp"}
{invalid json}

Run:

cargo run --example 01_redact_validate_deadletter

Processed output (samples/processed.ndjson):

{"timestamp":"2023-10-01T10:00:00Z","level":"info","message":"User logged in","email":"[REDACTED]"}

Deadletter (errors logged with context):

{"error":{"stage":"Input_File","code":"PARSE_ERROR","message":"expected value at line 1 column 1"},"raw":"{invalid json}"}

Metrics:

{"metric":"events_processed","value":1}
{"metric":"events_dropped","value":1}
{"metric":"errors","value":1}
{"metric":"stage_latencies","stage":"PIIRedaction","count":1,"sum":0.1,"min":0.1,"max":0.1}

Available Examples

FeedMe includes 12 comprehensive examples covering:

• Data Cleaning: PII redaction, field validation
• Compliance: GDPR/CCPA handling
• Monitoring: Bounded metrics, observability
• ETL/Ingestion: Multi-stage pipelines
• Error Resilience: Deadletter queues
• Extensibility: Custom stages and plugins

Run any example:

cargo run --example <number>_<description>

See examples/ for the full list.

🏗️ API Reference

• Pipeline - Core processing pipeline
• Stage - Processing stage trait
• InputSource - Data input sources
• Event - Data event structure
• Metrics - Observability metrics

Full documentation: docs.rs/feedme

Performance

FeedMe is designed for high-throughput, low-latency data processing:

• Bounded memory usage — no unbounded buffering regardless of input size
• Efficient streaming — minimal allocations with ownership transfer
• Zero-overhead observability — metrics collection doesn't affect execution
• Horizontal scalability — linear processing model scales across cores/processes

🛡️ Invariants

FeedMe enforces mechanical behavioral guarantees that are tested via runtime assertions and contract tests. These invariants ensure reliability and prevent regressions. Key guarantees include:

• Metrics purity: Exporting metrics doesn't affect pipeline state
• Drop counting rules: Events are only counted as dropped under specific conditions
• Latency recording: Successful stage execution always records timing
• Directory determinism: Same directory input produces identical output

See src/ppt_invariant_contracts.rs for the complete contract test suite.

🚫 Non-Goals

• Distributed processing
• Network I/O (except stubbed HTTP_Post)
• Persistent storage
• Query languages
• Compression or encryption

📋 API Overview

Stage Contract

pub trait Stage {
    fn execute(&mut self, event: Event) -> Result<Option<Event>, PipelineError>;
    fn name(&self) -> &str;
    fn is_output(&self) -> bool { false }  // true if consumes event
}

Execution Semantics

• Some(event): Pass to next stage
• None: Drop (filtered); if is_output(), consumed
• Err: Stop pipeline, error with attribution

Core Types

• Pipeline: Add stages, process events, export metrics
• Event: JSON data + optional metadata
• InputSource: Stream from stdin/file/directory
• PipelineError: Categorized errors (Parse/Transform/Validation/Output/System)

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Development Setup

git clone https://github.com/Michael-A-Kuykendall/feedme.git
cd feedme
cargo build
cargo test

Determinism Verification

FeedMe guarantees deterministic output for identical inputs. Verify this with:

cargo run --example 09_complex_pipeline > run1.out
cargo run --example 09_complex_pipeline > run2.out
# On Unix: sha256sum run1.out run2.out
# On Windows: certutil -hashfile run1.out SHA256 && certutil -hashfile run2.out SHA256

The hashes should match, proving deterministic behavior.

Code of Conduct

This project follows a code of conduct to ensure a welcoming environment for all contributors. See CODE_OF_CONDUCT.md for details.

Sponsors

FeedMe is supported by our amazing sponsors. See SPONSORS.md for details.

License

Licensed under the MIT License. See LICENSE for details.

Made with ❤️ by Michael Kuykendall