SPipe EngineeringandClaude Opus 5.5
e3f386addf
- 데모 v2 이후 개발 PC의 변경을 옮김 (이 저장소의 이름 변경·README 정리는 유지) - 서버 기본 DEM, DEM 등고선, 속성 테이블 범위 선택, WNTR 교차 검증 시험 포함 - 새 라이브러리: ezdxf(Python), maplibre-contour(npm) - 변경 내역: docs/SPipe_패치내역_도면출력_2026-11-06.txt Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Ubuntu deployment
The repository is private. Authenticate on the server using gh auth login
(GitHub.com, HTTPS, web browser), then run:
gh repo clone tpstn23-afk/SPipe
cd SPipe
COMPOSE_PARALLEL_LIMIT=1 docker compose build
docker compose up -d
curl --fail http://127.0.0.1/api/health
Docker Engine and Compose must already be installed. Check free disk space
before building; the target server originally had about 4 GB free.
Set SPIPE_PORT=8080 when port 80 is already occupied.
The API builds the official EPANET v2.3.5 Linux library and runs a Net1 analysis while building. Source: https://github.com/OpenWaterAnalytics/EPANET/releases/tag/v2.3.5 Only the web/API dependencies are installed; blocksys is a separate CLI. The web and API share one origin through nginx. Only the web port is published. The configuration serves HTTP; domain/TLS setup is a separate server step. Map service allowed domains may need to include the deployed address.
Validation: local Windows regression suite; Docker build and live server checks still required (Docker is unavailable on the development PC).