orcaflex-model-sanitization

name	orcaflex-model-sanitization
version	1.0.0
category	engineering
description	Sanitize OrcaFlex models by stripping client-identifiable references, converting binary .dat to YAML .yml, and organizing into the reference model library.
type	reference
invocation	/orcaflex-model-sanitization
applies-to	["claude","codex","gemini"]
auto_generated	true
source	incoming/digitalmodel/orcaflex-model-sanitization
promoted	"2026-02-18T00:00:00.000Z"
capabilities	[]
requires	[]
see_also	[]
tags	[]

OrcaFlex Model Sanitization Skill

Description

Sanitize OrcaFlex models from client projects by stripping identifiable references (project names, vessel names, operator codes, user/machine metadata), converting binary .dat to YAML .yml, deduplicating, and organizing into the reference model library.

When to Use

• Importing OrcaFlex models from external/client sources
• Converting .dat (binary) to .yml (YAML) format
• Stripping client-identifiable references before committing
• Organizing models into the docs/modules/orcaflex/ library structure
• Deduplicating model collections (hash-based)

Sanitization Pipeline

Overview

Source .dat/.yml → Dedup → Convert → Sanitize Text → Strip Metadata → Organize → Legal Scan → Extract Spec

Step 1: Inventory & Dedup

• Walk source directory for .dat and .yml files
• Compute SHA-256 hash for each file
• Skip files with identical content (keep first occurrence)
• Exclude: oversized files (>50MB), non-model files (CAD, reports, scripts)

Step 2: Format Conversion (.dat → .yml)

import OrcFxAPI
model = OrcFxAPI.Model(str(dat_path))
model.SaveData(str(output_yml_path))

Gotchas:

• SaveData() embeds User:, Machine:, File: metadata in YAML header
• SaveData() exports ALL properties including dormant ones
• Binary .dat files may reference external DLLs or data files
• Large models (>50MB) may timeout — skip and log

Step 3: Text-Based Sanitization

Apply regex replacements using a mapping table:

SANITIZATION_MAP = {
    "ClientProject": "generic_project_a",
    "VesselName": "installation_vessel_01",
    # ... sorted by key length descending to avoid partial matches
}

# Sort keys longest-first
sorted_keys = (SANITIZATION_MAP.keys(), key=, reverse=)

text = Path(yml_file).read_text(encoding=, errors=)
 key  sorted_keys:
    replacement = SANITIZATION_MAP[key]
    text = text.replace(key, replacement)