factorylint 0.2.0

A CLI tool to lint Azure Data Factory resources for naming convention compliance

🏭 FactoryLint

FactoryLint is a Python CLI tool for linting Azure Data Factory (ADF) resources to ensure they follow consistent, enforceable naming conventions.

It validates pipelines, datasets, linked services, triggers, parameters, and variables using a fully configurable rules file, making it ideal for CI/CD pipelines (Azure DevOps, GitHub Actions) and team-wide governance.

---

✨ Features

• ✅ Lint ADF resources:
  • Pipelines
  • Datasets
  • Linked Services
  • Triggers
  • Pipeline & Dataset Parameters
  • Pipeline Variables
• ⚙️ Fully configurable rules via YAML or JSON
• 🧠 Automatic ADF resource type detection
• 📊 Clear, colorized terminal output
• 💾 Machine-readable JSON report output
• 🚀 Designed for CI/CD usage
• 🛠 Simple, predictable CLI interface

---

📦 Installation

Install from PyPI

pip install factorylint

Local development install

git clone https://github.com/DimaFrank/FactoryLint.git
cd FactoryLint
pip install -e .

🚀 Usage

Initialize project (optional)

Creates the .adf-linter directory used for repo

factorylint init

Lint ADF resources

Run linting against a directory containing ADF resources.

factorylint lint --config ./factorylint.yml --resources .

Option	Description
--config	Path to rules configuration file (YAML or JSON)
--resources	Root directory containing ADF resources
--fail-fast	Stop on first error

🗂 Expected Folder Structure

FactoryLint automatically scans these subfolders under --resources:

pipeline/
dataset/
linkedService/
trigger/

Each folder may contain nested subdirectories.

📝 Configuration

The configuration file defines naming and validation rules for each ADF resource type.

Supported formats: YAML (.yml, .yaml) or JSON

The config is validated before linting starts

Invalid configs fail the run immediately (CI-safe)

The default config file name is factorylint.yml.

All rules live under a top-level resources: key. Parameters and variables are configured at the top level.

Naming rules reference

Resource	Structure	Example
Pipeline	PL_<PROJECT>_<ENTITY>_<ACTION>	PL_MES_ORDERS_INGEST
Dataset	DS_<SOURCE>_<DETAIL...>_<FORMAT>	DS_ADLS_CUSTOMERS_PARQUET
Linked Service	LS_<ABBR>_<DESCRIPTION...>	LS_ADLS_BRONZE
Trigger	TR_<TYPE>_<FREQUENCY>_<DESCRIPTION...>	TR_SCHEDULE_DAILY_MES_IOT
Parameter	p_<name> (lowercase)	p_start_date
Variable	v_<name> (lowercase)	v_row_count

Example factorylint.yml

resources:

  pipelines:
    enabled: true
    naming:
      prefix: "PL_"
      case: "upper"
      separator: "_"
      pattern: "^PL_[A-Z0-9_]+$"
      min_separated_parts: 4
      max_separated_parts: 6
      allowed_actions:
        - INGEST
        - TRANSFORM
        - ORCHESTRATE
        - LOAD
        - COPY
        - EXTRACT
        - PROCESS
      description_required: true
      ignore_folder: "_Config"

  datasets:
    enabled: true
    naming:
      prefix: "DS"
      separator: "_"
      case: "upper"
      pattern: "^[A-Z0-9_]+$"
      required_source_position: 2
      allowed_source_abbreviations:
        "Azure Data Lake Storage Gen2": ADLS
        "Azure SQL Database": ASQL
        "SAP Table": SAPT
        "File System": FILE
      allowed_formats:
        - CSV
        - PARQUET
        - TABLE
        - JSON
        - EXCEL
        - AVRO
        - BIN
        - XML
        - ORC
      min_separated_parts: 4
      max_separated_parts: 8

  linked_services:
    enabled: true
    naming:
      prefix: "LS_"
      case: "upper"
      pattern: "^[A-Z0-9_]+$"
      separator: "_"
      min_separated_parts: 2
      max_separated_parts: 5
      allowed_abbreviations:
        - ADLS
        - ASQL
        - KV
        - ADBR
        - FILE

  triggers:
    enabled: true
    naming:
      prefix: "TR_"
      separator: "_"
      case: "upper"
      pattern: "^[A-Z0-9_]+$"
      min_separated_parts: 4
      max_separated_parts: 7
      allowed_types:
        - SCHEDULE
        - EVENT
      allowed_frequencies:
        - HOURLY
        - DAILY
        - NIGHTLY
        - WEEKLY
        - MONTHLY

parameters:
  enabled: true
  naming:
    prefix: "p_"
    case: "lower"
    pattern: "^[a-z_][a-z0-9_]*$"

variables:
  enabled: true
  naming:
    prefix: "v_"
    case: "lower"
    pattern: "^[a-z_][a-z0-9_]*$"

📊 Output

Terminal output

FactoryLint provides clear, colorized feedback:

❌ pipeline/PL_MES_ORDERS_RUN.json
   - Pipeline 'PL_MES_ORDERS_RUN' has invalid action 'RUN'. Allowed: ['INGEST', 'TRANSFORM', ...]
   - Parameter 'StartDate' must start with prefix 'p_'
✅ pipeline/PL_MES_ORDERS_INGEST.json
✅ dataset/DS_ADLS_CUSTOMERS_PARQUET.json

JSON report

All linting errors are saved to:

.adf-linter/linter_results.json

Example:

{
  "pipeline/PL_MES_ORDERS_RUN.json": [
    "Pipeline 'PL_MES_ORDERS_RUN' has invalid action 'RUN'. Allowed: ['INGEST', 'TRANSFORM', ...]"
  ]
}

🔁 CI/CD Usage (Azure DevOps example)

- task: UsePythonVersion@0
  inputs:
    versionSpec: '3.x'

- script: |
    pip install factorylint
    factorylint lint --config ./factorylint.yml --resources .
  displayName: 'Run FactoryLint'

• Exit code 1 if errors are found

• Perfect for gating PRs and enforcing standards

🧠 Design Principles

• ❌ No hardcoded paths

• ❌ No assumptions about project layout outside --resources

• ✅ Fully installable CLI

• ✅ Deterministic behavior in CI

• ✅ Clear separation of CLI and core logic

📝 License

This project is licensed under the MIT License. See the LICENSE file for details.