parser-go

A web server log parser and analysis service written in Go. Parses Combined Log Format access logs, computes traffic summaries, and serves results through an HTTP API with a built-in report viewer.

Install

go install github.com/sagaragas/parser-go/cmd/parsergo@latest

Or build from source:

git clone https://github.com/sagaragas/parser-go.git
cd parser-go
go build -o parsergo ./cmd/parsergo

Usage

Start the service

parsergo serve

Submit a log file

curl -X POST http://127.0.0.1:3120/v1/analyses \
  -F "format=combined" \
  -F "profile=default" \
  -F "file=@/path/to/access.log"

The response includes a job ID and polling URL. Once the job completes, view results at:

• JSON summary: GET /v1/analyses/{id}/summary
• HTML report: http://127.0.0.1:3120/reports/{id}

Docker

docker build -t parsergo .
docker run -p 3120:3120 parsergo

The container defaults to binding 0.0.0.0:3120 so it is reachable from the host via the port mapping. To override, set PARSERGO_ADDR.

Configuration

All configuration is through environment variables:

Variable	Default	Description
PARSERGO_ADDR	127.0.0.1:3120	Bind address
PARSERGO_MAX_UPLOAD_BYTES	10485760 (10 MB)	Maximum upload size
PARSERGO_QUEUE_LIMIT	2	Maximum concurrent analysis jobs
PARSERGO_RETENTION	24h	How long completed jobs are kept

API

Endpoint	Method	Description
/healthz	GET	Liveness check
/readyz	GET	Readiness check (503 during startup)
/v1/analyses	POST	Submit a log file for analysis (202)
/v1/analyses/{id}	GET	Poll job status
/v1/analyses/{id}/summary	GET	JSON analysis results
/v1/analyses/{id}/report	GET	HTML report
/reports	GET	List completed reports
/reports/{id}	GET	View report in browser

Error responses

Status	Meaning
400	Invalid request (bad format, missing fields)
413	Upload too large
415	Unsupported media type
422	Unprocessable input (unsupported format/profile)
429	Queue full, retry after Retry-After seconds
503	Service not ready (during startup)

Project layout

cmd/parsergo/       Service entrypoint
cmd/bench/          Benchmark harness CLI
internal/analysis/  Log parser and aggregation engine
internal/api/       HTTP handlers
internal/job/       Job lifecycle and storage
internal/server/    HTTP server type (minimal)
internal/summary/   Summary types
internal/bench/     Benchmark internals
benchmark/          Scenario definitions and test corpora

Development

go test ./...
go vet ./...

Benchmarks

Go-native benchmarks run on the NASA Kennedy Space Center access logs (1.89M lines, 196 MB). A 10K-line sample is included in the repo:

go test -bench=. -benchmem ./internal/analysis/

On the full NASA dataset (~52 MB/s, ~480K lines/sec). See benchmark/README.md for details.

A cross-language parity harness is also available for comparing against the Python baseline:

BENCH_BASELINE_PYTHON=/path/to/python \
BENCH_LEGACY_REPO=/path/to/web-log-parser \
go run ./cmd/bench run --scenario synthetic-small

License

Apache License 2.0