PURVIEW CLI v1.3.8 - Microsoft Purview Automation & Data Governance

LATEST UPDATE v1.3.8 (November 3, 2025):

🔗 Advanced Lineage Features & Column-Level Mapping

New Lineage Capabilities:

[NEW] Column-level lineage with multi-target support (1 source → N targets)

[NEW] Direct lineage (UI-style) - No visible Process entity

[NEW] Dual-mode CSV import - Automatic detection of Process vs Direct lineage

[NEW] Column mapping in direct relationships - Granular data flow tracking

[NEW] Enhanced error handling with SSL retry strategies

New CLI Commands:
pvw lineage create-column   # Column-level lineage (Process-based)
pvw lineage create-direct   # Direct lineage (UI-style, no Process)
pvw lineage list-column     # List column lineages
pvw lineage delete-column   # Delete column lineage
What's New:

✅ Column mapping visible in Purview UI

✅ Compatible with manual UI lineage creation

✅ Type validation (prevent invalid lineage)

✅ Batch CSV import with 5 sample files

✅ Comprehensive documentation & examples

Full Release Notes v1.3.8 | v1.3.3 Release Notes | Migration Guide

What is PVW CLI?

PVW CLI v1.3.8 is a modern, full-featured command-line interface and Python library for Microsoft Purview. It enables automation and management of all major Purview APIs with 86% Unified Catalog API coverage (45 of 52 operations).

Key Capabilities

Unified Catalog (UC) Management - 86% Complete

Complete governance domains, glossary terms, data products, OKRs, CDEs
Relationships API - Link data products/CDEs/terms to entities and columns
Query APIs - Advanced OData filtering with multi-criteria search
Policy Management - Complete CRUD for governance and RBAC policies
Custom Metadata & Attributes - Extensible business metadata and attributes

Data Operations

Entity management (create, update, bulk, import/export)
Lineage operations with interactive creation and CSV import
Advanced search and discovery with fixed suggest/autocomplete
Business metadata with proper scope configuration

Automation & Scripting

Bulk Operations - Import/export from CSV/JSON with dry-run support
Scriptable Output - Multiple formats (table, json, jsonc) for PowerShell/bash
80+ usage examples and 15+ comprehensive guides
PowerShell integration with ConvertFrom-Json support

Legacy API Support

Collection and account management
Data product management (legacy compatibility)
Classification, label, and status management

The CLI is designed for data engineers, stewards, architects, and platform teams to automate, scale, and enhance their Microsoft Purview experience.

NEW: MCP Server for AI Assistants

[NEW] Model Context Protocol (MCP) server enables LLM-powered data governance workflows!

Natural language interface to Purview catalog
20+ tools for AI assistants (Claude, Cline, etc.)
Automate complex multi-step operations
See mcp/README.md for setup instructions

What's New in Recent Releases

v1.3.8 (November 3, 2025) - Advanced Lineage Features

Column-Level Lineage & Direct Relationships:

Column-level lineage with multi-target support (1→N)
Direct lineage creation (UI-style, no visible Process)
Dual-mode CSV import with automatic type detection
Column mapping in direct relationships
Enhanced error handling with SSL retry strategies

New Commands:

pvw lineage create-column   # Column lineage (Process-based)
pvw lineage create-direct   # Direct lineage (UI-style)
pvw lineage list-column     # List column lineages
pvw lineage delete-column   # Delete lineage

CSV Import Examples:

# Direct lineage with column mapping
source_entity_guid,target_entity_guid,relationship_type,column_mapping
guid1,guid2,direct_lineage_dataset_dataset,"[{""Source"":""ID"",""Sink"":""ID""}]"

Full v1.3.8 Release Notes

v1.2.5 (October 30, 2025) - 86% UC API Coverage

Version 1.2.5 achieves 86% coverage of the Microsoft Purview Unified Catalog API with 35 new operations:

Resource Type	Coverage	Operations	Status
Business Domains	100%	5/5	✅ Complete
Data Products	90%	9/10	⚠️ 1 missing (Facets)
Glossary Terms	73%	8/11	⚠️ 3 missing
Objectives & Key Results	92%	11/12	⚠️ 1 missing
Critical Data Elements	90%	9/10	⚠️ 1 missing
Policies	100%	5/5	✅ Complete
Relationships	100%	6/6	✅ Complete
Query	100%	4/4	✅ Complete
Custom Metadata	100%	5/5	✅ Complete
Custom Attributes	100%	5/5	✅ Complete
TOTAL	86%	45/52	🎯 A- Grade

🚀 New APIs Implemented

Relationships API (6 operations)

# Link data product to entity
pvw uc dataproduct link-entity --id <dp-id> --entity-id <guid>

# Link CDE to column
pvw uc cde link-entity --id <cde-id> --entity-id <guid> --column-qualified-name "..."

Query APIs (4 operations)

# Advanced OData filtering
pvw uc term query --domain-ids "finance" --status Approved --top 50

# Multi-criteria search with pagination
pvw uc dataproduct query --keywords "customer,revenue" --skip 10 --top 25

Policy Management (5 operations)

# Complete policy CRUD
pvw uc policy list
pvw uc policy create --payload-file policy.json
pvw uc policy update --id <policy-id> --payload-file updated.json

Custom Metadata (5 operations)

# Business metadata via Atlas API
pvw uc custom-metadata import --file metadata.csv
pvw uc custom-metadata add --guid <entity-guid> --name "BusinessConcept"

Custom Attributes (5 operations)

# Extensible attribute definitions
pvw uc custom-attribute create --name "Department" --type String
pvw uc custom-attribute list

🔧 Major Fixes & Improvements

Lineage Management Overhaul - Complete rewrite with interactive PowerShell script, real entity support, and proper Process entities
Search API Fixed - Resolved HTTP 400 errors in suggest and autocomplete endpoints
Business Metadata Scope - Fixed Business Concept attributes on Glossary Terms with proper applicableEntityTypes
Architecture Refactoring - Unified endpoints dictionary, zero hardcoded URLs, complete consistency

📚 Documentation (3,500+ lines)

15+ new guides including relationships, query APIs, lineage creation, business metadata
80+ usage examples across all new features
Complete API coverage gap analysis
Roadmap to 100% with implementation plans

View Full Release Notes

Getting Started

Follow this short flow to get PVW CLI installed and running quickly.

Install (from PyPI):

pip install pvw-cli

For the bleeding edge or development:

pip install git+https://github.com/Keayoub/Purview_cli.git
# or for editable development
git clone https://github.com/Keayoub/Purview_cli.git
cd Purview_cli
pip install -r requirements.txt
pip install -e .

Set required environment variables (examples for cmd, PowerShell, and pwsh)

Windows cmd (example):

set PURVIEW_ACCOUNT_NAME=your-purview-account
set PURVIEW_ACCOUNT_ID=your-purview-account-id-guid
set PURVIEW_RESOURCE_GROUP=your-resource-group-name
set AZURE_REGION=  # optional

PowerShell (Windows PowerShell):

$env:PURVIEW_ACCOUNT_NAME = "your-purview-account"
$env:PURVIEW_ACCOUNT_ID = "your-purview-account-id-guid"
$env:PURVIEW_RESOURCE_GROUP = "your-resource-group-name"
$env:AZURE_REGION = ""  # optional

pwsh (PowerShell Core - cross-platform, recommended):

$env:PURVIEW_ACCOUNT_NAME = 'your-purview-account'
$env:PURVIEW_ACCOUNT_ID = 'your-purview-account-id-guid'
$env:PURVIEW_RESOURCE_GROUP = 'your-resource-group-name'
$env:AZURE_REGION = ''  # optional

Authenticate

Run az login (recommended), or
Provide Service Principal credentials via environment variables.

Try a few commands:

# List governance domains
pvw uc domain list

# Search
pvw search query --keywords="customer" --limit=5

# Get help
pvw --help
pvw uc --help

For more advanced usage, see the documentation in doc/ or the project docs: https://pvw-cli.readthedocs.io/

Quick Start Examples

v1.3.8 - Column-Level Lineage

# Create column-level lineage (Process-based)
pvw lineage create-column \
  --process-name "ETL_Sales_Transform" \
  --source-table-guid "9ebbd583-4987-4d1b-b4f5-d8f6f6f60000" \
  --target-table-guids "c88126ba-5fb5-4d33-bbe2-5ff6f6f60000" \
  --column-mapping "ProductID:ProductID,Name:Name"

# Create direct lineage (UI-style, no visible Process)
pvw lineage create-direct \
  --source-guid "9ebbd583-4987-4d1b-b4f5-d8f6f6f60000" \
  --target-guid "c88126ba-5fb5-4d33-bbe2-5ff6f6f60000" \
  --column-mapping "ProductID:ProductID,Name:Name,Amount:TotalAmount"

# Import lineage from CSV (automatic type detection)
pvw lineage import samples/csv/lineage_with_columns.csv

# List column lineages
pvw lineage list-column --format table

# Delete column lineage
pvw lineage delete-column --process-guid <guid> --force

v1.2.5 - Relationships API

# Link data product to SQL table
pvw uc dataproduct link-entity \
  --id "dp-sales-2024" \
  --entity-id "4fae348b-e960-42f7-834c-38f6f6f60000" \
  --type-name "azure_sql_table"

# Link CDE to specific column
pvw uc cde link-entity \
  --id "cde-customer-email" \
  --entity-id "ea3412c3-7387-4bc1-9923-11f6f6f60000" \
  --column-qualified-name "mssql://server/db/schema/table#EmailAddress"

# List all linked entities
pvw uc dataproduct list-entities --id "dp-sales-2024"

v1.2.5 - Query APIs

# Query terms by domain and status
pvw uc term query --domain-ids "finance,sales" --status Approved --top 50

# Query data products with keywords
pvw uc dataproduct query --keywords "customer,revenue" --skip 0 --top 25

# Query CDEs by domain with pagination
pvw uc cde query --domain-ids "compliance" --orderby "name" --top 100

v1.2.5 - Policy Management

# List all policies
pvw uc policy list

# Create new policy
pvw uc policy create --payload-file policy-rbac.json

# Update existing policy
pvw uc policy update --id "policy-001" --payload-file updated.json

v1.2.5 - Custom Metadata

# Import business metadata from CSV
pvw uc custom-metadata import --file business_concept.csv

# Add metadata to entity
pvw uc custom-metadata add \
  --guid "4fae348b-e960-42f7-834c-38f6f6f60000" \
  --name "BusinessConcept" \
  --attributes '{"Department":"Sales"}'

# Create custom attribute
pvw uc custom-attribute create --name "Department" --type String

Overview

PVW CLI v1.3.8 is a modern command-line interface and Python library for Microsoft Purview, enabling:

MCP Server - Natural language interface for AI assistants (Claude, Cline)
Advanced data catalog search and discovery
Bulk import/export of entities, glossary terms, and lineage
Real-time monitoring and analytics
Automated governance and compliance
Extensible plugin system

Installation

You can install PVW CLI in two ways:

From PyPI (recommended for most users):
```
pip install pvw-cli
```
Directly from the GitHub repository (for latest/dev version):
```
pip install git+https://github.com/Keayoub/Purview_cli.git
```

Or for development (editable install):

git clone https://github.com/Keayoub/Purview_cli.git
cd Purview_cli
pip install -r requirements.txt
pip install -e .

Requirements

Python 3.8+
Azure CLI (az login) or Service Principal credentials
Microsoft Purview account

Getting Started

Install
```
pip install pvw-cli
```

Set Required Environment Variables

# Required for Purview API access
set PURVIEW_ACCOUNT_NAME=your-purview-account
set PURVIEW_ACCOUNT_ID=your-purview-account-id-guid
set PURVIEW_RESOURCE_GROUP=your-resource-group-name

# Optional
set AZURE_REGION=  # (optional, e.g. 'china', 'usgov')

Authenticate
- Azure CLI: az login
- Or set Service Principal credentials as environment variables

Run a Command

pvw search query --keywords="customer" --limit=5

See All Commands
```
pvw --help
```

Authentication

PVW CLI supports multiple authentication methods for connecting to Microsoft Purview, powered by Azure Identity's DefaultAzureCredential. This allows you to use the CLI securely in local development, CI/CD, and production environments.

1. Azure CLI Authentication (Recommended for Interactive Use)

Run az login to authenticate interactively with your Azure account.
The CLI will automatically use your Azure CLI credentials.

2. Service Principal Authentication (Recommended for Automation/CI/CD)

Set the following environment variables before running any PVW CLI command:

AZURE_CLIENT_ID (your Azure AD app registration/client ID)
AZURE_TENANT_ID (your Azure AD tenant ID)
AZURE_CLIENT_SECRET (your client secret)

Example (Windows):

set AZURE_CLIENT_ID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
set AZURE_TENANT_ID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
set AZURE_CLIENT_SECRET=your-client-secret

Example (Linux/macOS):

export AZURE_CLIENT_ID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
export AZURE_TENANT_ID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
export AZURE_CLIENT_SECRET=your-client-secret

3. Managed Identity (for Azure VMs, App Services, etc.)

If running in Azure with a managed identity, no extra configuration is needed. The CLI will use the managed identity automatically.

4. Visual Studio/VS Code Authentication

If you are signed in to Azure in Visual Studio or VS Code, DefaultAzureCredential can use those credentials as a fallback.

Note:

The CLI will try all supported authentication methods in order. The first one that works will be used.
For most automation and CI/CD scenarios, service principal authentication is recommended.
For local development, Azure CLI authentication is easiest.

For more details, see the Azure Identity documentation.

Output Formats & Scripting Integration

PVW CLI supports multiple output formats to fit different use cases - from human-readable tables to machine-parseable JSON.

Output Format Options

All list commands now support the --output parameter with three formats:

table (default) - Rich formatted table with colors for human viewing
json - Plain JSON for scripting with PowerShell, bash, jq, etc.
jsonc - Colored JSON with syntax highlighting for viewing

PowerShell Integration

The --output json format produces plain JSON that works perfectly with PowerShell's ConvertFrom-Json:

# Get all terms as PowerShell objects
$domainId = "59ae27b5-40bc-4c90-abfe-fe1a0638fe3a"
$terms = py -m purviewcli uc term list --domain-id $domainId --output json | ConvertFrom-Json

# Access properties
Write-Host "Found $($terms.Count) terms"
foreach ($term in $terms) {
    Write-Host "  • $($term.name) - $($term.status)"
}

# Filter and export
$draftTerms = $terms | Where-Object { $_.status -eq "Draft" }
$draftTerms | Export-Csv -Path "draft_terms.csv" -NoTypeInformation

# Group by status
$terms | Group-Object status | Format-Table Count, Name

Bash/Linux Integration

Use jq for JSON processing in bash:

# Get domain ID
DOMAIN_ID="59ae27b5-40bc-4c90-abfe-fe1a0638fe3a"

# Get term names only
pvw uc term list --domain-id $DOMAIN_ID --output json | jq -r '.[] | .name'

# Count terms
pvw uc term list --domain-id $DOMAIN_ID --output json | jq 'length'

# Filter by status
pvw uc term list --domain-id $DOMAIN_ID --output json | jq '.[] | select(.status == "Draft")'

# Group by status
pvw uc term list --domain-id $DOMAIN_ID --output json | jq 'group_by(.status) | map({status: .[0].status, count: length})'

# Save to file
pvw uc term list --domain-id $DOMAIN_ID --output json > terms.json

Examples by Command

# Domains
pvw uc domain list --output json | jq '.[] | .name'

# Terms  
pvw uc term list --domain-id "abc-123" --output json
pvw uc term list --domain-id "abc-123" --output table   # Default
pvw uc term list --domain-id "abc-123" --output jsonc   # Colored for viewing

# Data Products
pvw uc dataproduct list --domain-id "abc-123" --output json

Migration from Old --json Flag

Old (deprecated):

pvw uc term list --domain-id "abc-123" --json

New (recommended):

pvw uc term list --domain-id "abc-123" --output json    # Plain JSON for scripting
pvw uc term list --domain-id "abc-123" --output jsonc   # Colored JSON (old behavior)

Required Purview Configuration

Before using PVW CLI, you need to set three essential environment variables. Here's how to find them:

🔍 How to Find Your Purview Values

1. PURVIEW_ACCOUNT_NAME

This is your Purview account name as it appears in Azure Portal
Example: kaydemopurview

2. PURVIEW_ACCOUNT_ID

This is the GUID that identifies your Purview account for Unified Catalog APIs
Important: For most Purview deployments, this is your Azure Tenant ID

Method 1 - Get your Tenant ID (recommended):

Bash/Command Prompt:

az account show --query tenantId -o tsv

PowerShell:

az account show --query tenantId -o tsv
# Or store directly in environment variable:
$env:PURVIEW_ACCOUNT_ID = az account show --query tenantId -o tsv

Method 2 - Azure CLI (extract from Atlas endpoint):
```
az purview account show --name YOUR_ACCOUNT_NAME --resource-group YOUR_RG --query endpoints.catalog -o tsv
```
Extract the GUID from the URL (before -api.purview-service.microsoft.com)
Method 3 - Azure Portal:
1. Go to your Purview account in Azure Portal
2. Navigate to Properties → Atlas endpoint URL
3. Extract GUID from: https://GUID-api.purview-service.microsoft.com/catalog

3. PURVIEW_RESOURCE_GROUP

The Azure resource group containing your Purview account
Example: fabric-artifacts

📋 Setting the Variables

Windows Command Prompt:

set PURVIEW_ACCOUNT_NAME=your-purview-account
set PURVIEW_ACCOUNT_ID=your-purview-account-id
set PURVIEW_RESOURCE_GROUP=your-resource-group

Windows PowerShell:

$env:PURVIEW_ACCOUNT_NAME="your-purview-account"
$env:PURVIEW_ACCOUNT_ID="your-purview-account-id" 
$env:PURVIEW_RESOURCE_GROUP="your-resource-group"

Linux/macOS:

export PURVIEW_ACCOUNT_NAME=your-purview-account
export PURVIEW_ACCOUNT_ID=your-purview-account-id
export PURVIEW_RESOURCE_GROUP=your-resource-group

Permanent (Windows Command Prompt):

setx PURVIEW_ACCOUNT_NAME "your-purview-account"
setx PURVIEW_ACCOUNT_ID "your-purview-account-id"
setx PURVIEW_RESOURCE_GROUP "your-resource-group"

Permanent (Windows PowerShell):

[Environment]::SetEnvironmentVariable("PURVIEW_ACCOUNT_NAME", "your-purview-account", "User")
[Environment]::SetEnvironmentVariable("PURVIEW_ACCOUNT_ID", "your-purview-account-id", "User")
[Environment]::SetEnvironmentVariable("PURVIEW_RESOURCE_GROUP", "your-resource-group", "User")

Debug Environment Issues

If you experience issues with environment variables between different terminals, use these debug commands:

Command Prompt/Bash:

# Run this to check your current environment
python -c "
import os
print('PURVIEW_ACCOUNT_NAME:', os.getenv('PURVIEW_ACCOUNT_NAME'))
print('PURVIEW_ACCOUNT_ID:', os.getenv('PURVIEW_ACCOUNT_ID'))
print('PURVIEW_RESOURCE_GROUP:', os.getenv('PURVIEW_RESOURCE_GROUP'))
"

PowerShell:

# Check environment variables in PowerShell
python -c "
import os
print('PURVIEW_ACCOUNT_NAME:', os.getenv('PURVIEW_ACCOUNT_NAME'))
print('PURVIEW_ACCOUNT_ID:', os.getenv('PURVIEW_ACCOUNT_ID'))
print('PURVIEW_RESOURCE_GROUP:', os.getenv('PURVIEW_RESOURCE_GROUP'))
"

# Or use PowerShell native commands
Write-Host "PURVIEW_ACCOUNT_NAME: $env:PURVIEW_ACCOUNT_NAME"
Write-Host "PURVIEW_ACCOUNT_ID: $env:PURVIEW_ACCOUNT_ID" 
Write-Host "PURVIEW_RESOURCE_GROUP: $env:PURVIEW_RESOURCE_GROUP"

Search Command (Discovery Query API)

The PVW CLI provides advanced search using the latest Microsoft Purview Discovery Query API:

Search for assets, tables, files, and more with flexible filters
Use autocomplete and suggestion endpoints
Perform faceted, time-based, and entity-type-specific queries

v1.3.8 Improvements:

Fixed suggest and autocomplete API payload format (removed empty filter causing HTTP 400 errors)
Enhanced collection display with robust type checking and fallback logic
All search commands validated and working correctly (query, browse, suggest, find-table)

CLI Usage Examples

Multiple Output Formats

# 1. Table Format (Default) - Quick overview
pvw search query --keywords="customer" --limit=5
# → Clean table with Name, Type, Collection, Classifications, Qualified Name

# 2. Detailed Format - Human-readable with all metadata  
pvw search query --keywords="customer" --limit=5 --detailed
# → Rich panels showing full details, timestamps, search scores

# 3. JSON Format - Complete technical details with syntax highlighting (WELL-FORMATTED)
pvw search query --keywords="customer" --limit=5 --json
# → Full JSON response with indentation, line numbers and color coding

# 4. Table with IDs - For entity operations
pvw search query --keywords="customer" --limit=5 --show-ids
# → Table format + entity GUIDs for copy/paste into update commands

Search Operations

# Basic search for assets with keyword 'customer'
pvw search query --keywords="customer" --limit=5

# Advanced search with classification filter
pvw search query --keywords="sales" --classification="PII" --objectType="Tables" --limit=10

# Pagination through large result sets
pvw search query --keywords="SQL" --offset=10 --limit=5

# Autocomplete suggestions for partial keyword
pvw search autocomplete --keywords="ord" --limit=3

# Get search suggestions (fuzzy matching)
pvw search suggest --keywords="prod" --limit=2

**IMPORTANT - Command Line Quoting:**
```cmd
# [OK] CORRECT - Use quotes around keywords
pvw search query --keywords="customer" --limit=5

# [OK] CORRECT - For wildcard searches, use quotes
pvw search query --keywords="*" --limit=5

# ❌ WRONG - Don't use unquoted * (shell expands to file names)
pvw search query --keywords=* --limit=5
# This causes: "Error: Got unexpected extra arguments (dist doc ...)"

# Faceted search with aggregation
pvw search query --keywords="finance" --facetFields="objectType,classification" --limit=5

# Browse entities by type and path
pvw search browse --entityType="Tables" --path="/root/finance" --limit=2

# Time-based search for assets created after a date
pvw search query --keywords="audit" --createdAfter="2024-01-01" --limit=1

# Entity type specific search
pvw search query --keywords="finance" --entityTypes="Files,Tables" --limit=2

Usage Scenarios

Daily browsing: Use default table format for quick scans
Understanding assets: Use --detailed for rich information panels
Technical work: Use --json for complete API data access
Entity operations: Use --show-ids to get GUIDs for updates

Python Usage Example

from purviewcli.client._search import Search

search = Search()
args = {"--keywords": "customer", "--limit": 5}
search.searchQuery(args)
print(search.payload)  # Shows the constructed search payload

Test Examples

See tests/test_search_examples.py for ready-to-run pytest examples covering all search scenarios:

Basic query
Advanced filter
Autocomplete
Suggest
Faceted search
Browse
Time-based search
Entity type search

Unified Catalog Management (NEW)

PVW CLI now includes comprehensive Microsoft Purview Unified Catalog (UC) support with the new uc command group. This provides complete management of modern data governance features including governance domains, glossary terms, data products, objectives (OKRs), and critical data elements.

🎯 Feature Parity: Full compatibility with UnifiedCatalogPy functionality.

See doc/commands/unified-catalog.md for complete documentation and examples.

Quick UC Examples

Governance Domains Management

# List all governance domains
pvw uc domain list

# Create a new governance domain
pvw uc domain create --name "Finance" --description "Financial data governance domain"

# Get domain details
pvw uc domain get --domain-id "abc-123-def-456"

# Update domain information
pvw uc domain update --domain-id "abc-123" --description "Updated financial governance"

Glossary Terms in UC

# List all terms in a domain
pvw uc term list --domain-id "abc-123"
pvw uc term list --domain-id "abc-123" --output json    # Plain JSON for scripting
pvw uc term list --domain-id "abc-123" --output jsonc   # Colored JSON for viewing

# Create a single glossary term
pvw uc term create --name "Customer" --domain-id "abc-123" --description "A person or entity that purchases products"

# Get term details
pvw uc term show --term-id "term-456"

# Update term
pvw uc term update --term-id "term-456" --description "Updated description"

# Delete term
pvw uc term delete --term-id "term-456" --confirm

📦 Bulk Import (NEW)

Import multiple terms from CSV or JSON files with validation and progress tracking:

# CSV Import - Preview with dry-run
pvw uc term import-csv --csv-file "samples/csv/uc_terms_bulk_example.csv" --domain-id "abc-123" --dry-run

# CSV Import - Actual import
pvw uc term import-csv --csv-file "samples/csv/uc_terms_bulk_example.csv" --domain-id "abc-123"

# JSON Import - Preview with dry-run
pvw uc term import-json --json-file "samples/json/term/uc_terms_bulk_example.json" --dry-run

# JSON Import - Actual import (domain_id from JSON or override with flag)
pvw uc term import-json --json-file "samples/json/term/uc_terms_bulk_example.json"
pvw uc term import-json --json-file "samples/json/term/uc_terms_bulk_example.json" --domain-id "abc-123"

Bulk Import Features:

[OK] Import from CSV or JSON files
[OK] Dry-run mode to preview before importing
[OK] Support for multiple owners (Entra ID Object IDs), acronyms, and resources
[OK] Progress tracking with Rich console output
[OK] Detailed error messages and summary reports
[OK] Sequential POST requests (no native bulk endpoint available)

CSV Format Example:

name,description,status,acronym,owner_id,resource_name,resource_url
Customer Acquisition Cost,Cost to acquire new customer,Draft,CAC,<guid>,Metrics Guide,https://docs.example.com
Monthly Recurring Revenue,Predictable monthly revenue,Draft,MRR,<guid>,Finance Dashboard,https://finance.example.com

JSON Format Example:

{
  "terms": [
    {
      "name": "Data Lake",
      "description": "Centralized repository for structured/unstructured data",
      "domain_id": "your-domain-id-here",
      "status": "Draft",
      "acronyms": ["DL"],
      "owner_ids": ["<entra-id-object-id-guid>"],
      "resources": [{"name": "Architecture Guide", "url": "https://example.com"}]
    }
  ]
}

Important Notes:

⚠️ Owner IDs must be Entra ID Object IDs (GUIDs), not email addresses
⚠️ Terms cannot be "Published" in unpublished domains - use "Draft" status
[OK] Sample files available: samples/csv/uc_terms_bulk_example.csv, samples/json/term/uc_terms_bulk_example.json
📖 Complete documentation: doc/commands/unified-catalog/term-bulk-import.md

🗑️ Bulk Delete (NEW)

Delete all terms in a domain using PowerShell or Python scripts:

# PowerShell - Delete all terms with confirmation
.\scripts\delete-all-uc-terms.ps1 -DomainId "abc-123"

# PowerShell - Delete without confirmation
.\scripts\delete-all-uc-terms.ps1 -DomainId "abc-123" -Force

# Python - Delete all terms with confirmation
python scripts/delete_all_uc_terms_v2.py --domain-id "abc-123"

# Python - Delete without confirmation
python scripts/delete_all_uc_terms_v2.py --domain-id "abc-123" --force

Bulk Delete Features:

[OK] Interactive confirmation prompts (type "DELETE" to confirm)
[OK] Beautiful progress display with colors
[OK] Success/failure tracking per term
[OK] Detailed summary reports
[OK] Rate limiting (200ms delay between deletes)
[OK] Graceful error handling and Ctrl+C support

Data Products Management

# List all data products in a domain
pvw uc dataproduct list --domain-id "abc-123"

# Create a comprehensive data product
pvw uc dataproduct create \
  --name "Customer Analytics Dashboard" \
  --domain-id "abc-123" \
  --description "360-degree customer analytics with behavioral insights" \
  --type Analytical \
  --status Draft

# Get detailed data product information
pvw uc dataproduct show --product-id "prod-789"

# Update data product (partial updates supported - only specify fields to change)
pvw uc dataproduct update \
  --product-id "prod-789" \
  --status Published \
  --description "Updated comprehensive customer analytics" \
  --endorsed

# Update multiple fields at once
pvw uc dataproduct update \
  --product-id "prod-789" \
  --status Published \
  --update-frequency Monthly \
  --endorsed

# Delete a data product (with confirmation)
pvw uc dataproduct delete --product-id "prod-789"

# Delete without confirmation prompt
pvw uc dataproduct delete --product-id "prod-789" --yes

Objectives & Key Results (OKRs)

# List objectives for a domain
pvw uc objective list --domain-id "abc-123"

# Create measurable objectives
pvw uc objective create \
  --definition "Improve data quality score by 25% within Q4" \
  --domain-id "abc-123" \
  --target-value "95" \
  --measurement-unit "percentage"

# Track objective progress
pvw uc objective update \
  --objective-id "obj-456" \
  --domain-id "abc-123" \
  --current-value "87" \
  --status "in-progress"

Critical Data Elements (CDEs)

# List critical data elements
pvw uc cde list --domain-id "abc-123"

# Define critical data elements with governance rules
pvw uc cde create \
  --name "Social Security Number" \
  --data-type "String" \
  --domain-id "abc-123" \
  --classification "PII" \
  --retention-period "7-years"

# Associate CDEs with data assets
pvw uc cde link \
  --cde-id "cde-789" \
  --domain-id "abc-123" \
  --asset-id "ea3412c3-7387-4bc1-9923-11f6f6f60000"

Health Monitoring (NEW)

Monitor governance health and get automated recommendations to improve your data governance posture.

# List all health findings and recommendations
pvw uc health query

# Filter by severity
pvw uc health query --severity High
pvw uc health query --severity Medium

# Filter by status
pvw uc health query --status NotStarted
pvw uc health query --status InProgress

# Get detailed information about a specific health action
pvw uc health show --action-id "5ea3fc78-6a77-4098-8779-ed81de6f87c9"

# Update health action status
pvw uc health update \
  --action-id "5ea3fc78-6a77-4098-8779-ed81de6f87c9" \
  --status InProgress \
  --reason "Working on assigning glossary terms to data products"

# Get health summary statistics
pvw uc health summary

# Output health findings in JSON format
pvw uc health query --json

Health Finding Types:

Missing glossary terms on data products (High)
Data products without OKRs (Medium)
Missing data quality scores (Medium)
Classification gaps on data assets (Medium)
Description quality issues (Medium)
Business domains without critical data entities (Medium)

Workflow Management (NEW)

Manage approval workflows and business process automation in Purview.

# List all workflows
pvw workflow list

# Get workflow details
pvw workflow get --workflow-id "workflow-123"

# Create a new workflow (requires JSON definition)
pvw workflow create --workflow-id "approval-flow-1" --payload-file workflow-definition.json

# Execute a workflow
pvw workflow execute --workflow-id "workflow-123"

# List workflow executions
pvw workflow executions --workflow-id "workflow-123"

# View specific execution details
pvw workflow execution-details --workflow-id "workflow-123" --execution-id "exec-456"

# Update workflow configuration
pvw workflow update --workflow-id "workflow-123" --payload-file updated-workflow.json

# Delete a workflow
pvw workflow delete --workflow-id "workflow-123"

# Output workflows in JSON format
pvw workflow list --json

Workflow Use Cases:

Data access request approvals
Glossary term certification workflows
Data product publishing approvals
Classification review processes

Integrated Workflow Example

# 1. Discover assets to govern
pvw search query --keywords="customer" --detailed

# 2. Create governance domain for discovered assets
pvw uc domain create --name "Customer Data" --description "Customer information governance"

# 3. Define governance terms
pvw uc term create --name "Customer PII" --domain-id "new-domain-id" --definition "Personal customer information"

# 4. Create data product from discovered assets
pvw uc dataproduct create --name "Customer Master Data" --domain-id "new-domain-id"

# 5. Set governance objectives
pvw uc objective create --definition "Ensure 100% PII classification compliance" --domain-id "new-domain-id"

Entity Management & Updates

PVW CLI provides comprehensive entity management capabilities for updating Purview assets like descriptions, classifications, and custom attributes.

Entity Update Examples

Update Asset Descriptions

# Update table description using GUID
pvw entity update-attribute \
  --guid "ece43ce5-ac45-4e50-a4d0-365a64299efc" \
  --attribute "description" \
  --value "Updated customer data warehouse table with enhanced analytics"

# Update dataset description using qualified name
pvw entity update-attribute \
  --qualifiedName "https://app.powerbi.com/groups/abc-123/datasets/def-456" \
  --attribute "description" \
  --value "Power BI dataset for customer analytics dashboard"

Bulk Entity Operations

# Read entity details before updating
pvw entity read-by-attribute \
  --guid "ea3412c3-7387-4bc1-9923-11f6f6f60000" \
  --attribute "description,classifications,customAttributes"

# Update multiple attributes at once
pvw entity update-bulk \
  --input-file entities_to_update.json \
  --output-file update_results.json

Column-Level Updates

# Update specific column descriptions in a table
pvw entity update-attribute \
  --guid "column-guid-123" \
  --attribute "description" \
  --value "Customer unique identifier - Primary Key"

# Add classifications to sensitive columns
pvw entity add-classification \
  --guid "column-guid-456" \
  --classification "MICROSOFT.PERSONAL.EMAIL"

Discovery to Update Workflow

# 1. Find assets that need updates
pvw search query --keywords="customer table" --show-ids --limit=10

# 2. Get detailed information about a specific asset
pvw entity read-by-attribute --guid "FOUND_GUID" --attribute "description,classifications"

# 3. Update the asset description
pvw entity update-attribute \
  --guid "FOUND_GUID" \
  --attribute "description" \
  --value "Updated description based on business requirements"

# 4. Verify the update
pvw search query --keywords="FOUND_GUID" --detailed

Lineage CSV Import & Management

PVW CLI provides powerful lineage management capabilities including CSV-based bulk import for automating data lineage creation.

Lineage CSV Import

Import lineage relationships from CSV files to automate the creation of data flow documentation in Microsoft Purview.

CSV Format

The CSV file must contain the following columns:

Required columns:

source_entity_guid - GUID of the source entity
target_entity_guid - GUID of the target entity

Optional columns:

relationship_type - Type of relationship (default: "Process")
process_name - Name of the transformation process
description - Description of the transformation
confidence_score - Confidence score (0-1)
owner - Process owner
metadata - Additional JSON metadata

Example CSV:

source_entity_guid,target_entity_guid,relationship_type,process_name,description,confidence_score,owner,metadata
dcfc99ed-c74d-49aa-bd0b-72f6f6f60000,1db9c650-acfb-4914-8bc5-1cf6f6f60000,Process,Transform_Product_Data,Transform product data for analytics,0.95,data-engineering,"{""tool"": ""Azure Data Factory""}"

Lineage Commands

# Validate CSV format before import (no API calls)
pvw lineage validate lineage_data.csv

# Import lineage relationships from CSV
pvw lineage import lineage_data.csv

# Generate sample CSV file with examples
pvw lineage sample output.csv --num-samples 10 --template detailed

# View available CSV templates
pvw lineage templates

Available Templates

basic - Minimal columns (source, target, process name)
detailed - All columns including metadata and confidence scores
qualified_names - Use qualified names instead of GUIDs

Workflow Example

# 1. Find entity GUIDs using search
pvw search find-table --name "Product" --schema "dbo" --id-only

# 2. Create CSV file with lineage relationships
# (use the GUIDs from step 1)

# 3. Validate CSV format
pvw lineage validate my_lineage.csv
# Output: SUCCESS: Lineage validation passed (5 rows, 8 columns)

# 4. Import to Purview
pvw lineage import my_lineage.csv
# Output: SUCCESS: Lineage import completed successfully

Advanced Features

GUID Validation: Automatic validation of GUID format with helpful error messages
Process Entity Creation: Creates intermediate "Process" entities to link source→target relationships
Metadata Support: Add custom JSON metadata to each lineage relationship
Dry-Run Validation: Validate CSV format locally before making API calls

For detailed documentation, see: doc/guides/lineage-csv-import.md

Data Product Management (Legacy)

PVW CLI also includes the original data-product command group for backward compatibility with traditional data product lifecycle management.

See doc/commands/data-product.md for full documentation and examples.

Example Commands

# Create a data product
pvw data-product create --qualified-name="product.test.1" --name="Test Product" --description="A test data product"

# Add classification and label
pvw data-product add-classification --qualified-name="product.test.1" --classification="PII"
pvw data-product add-label --qualified-name="product.test.1" --label="gold"

# Link glossary term
pvw data-product link-glossary --qualified-name="product.test.1" --term="Customer"

# Set status and show lineage
pvw data-product set-status --qualified-name="product.test.1" --status="active"
pvw data-product show-lineage --qualified-name="product.test.1"

Core Features

Unified Catalog (UC): Complete modern data governance (NEW)

# Manage governance domains, terms, data products, OKRs, CDEs
pvw uc domain list
pvw uc term create --name "Customer" --domain-id "abc-123"
pvw uc objective create --definition "Improve quality" --domain-id "abc-123"

Discovery Query/Search: Flexible, advanced search for all catalog assets
Entity Management: Bulk import/export, update, and validation

Glossary Management: Import/export terms, assign terms in bulk

# List all terms in a glossary
pvw glossary list-terms --glossary-guid "your-glossary-guid"

# Create and manage glossary terms
pvw glossary create-term --payload-file term.json

Lineage Operations: Lineage discovery, CSV-based bulk lineage import/export

# Import lineage relationships from CSV
pvw lineage import lineage_data.csv

# Validate CSV format before import
pvw lineage validate lineage_data.csv

# Generate sample CSV file
pvw lineage sample output.csv --num-samples 10

Monitoring & Analytics: Real-time dashboards, metrics, and reporting
Plugin System: Extensible with custom plugins

API Coverage and Support

PVW CLI provides comprehensive automation for all major Microsoft Purview APIs, including the new Unified Catalog APIs for modern data governance.

Supported API Groups

Unified Catalog: Complete governance domains, glossary terms, data products, OKRs, CDEs management [OK]
- Health Monitoring: Automated governance health checks and recommendations [OK] NEW
- Workflows: Approval workflows and business process automation [OK] NEW
Data Map: Full entity and lineage management [OK]
Discovery: Advanced search, browse, and query capabilities [OK]
Collections: Collection and account management [OK]
Management: Administrative operations [OK]
Scan: Data source scanning and configuration [OK]

API Version Support

Unified Catalog: Latest UC API endpoints (September 2025)
Data Map: 2024-03-01-preview (default) or 2023-09-01 (stable)
Collections: 2019-11-01-preview
Account: 2019-11-01-preview
Management: 2021-07-01
Scan: 2018-12-01-preview

For the latest API documentation and updates, see:

If you need a feature that is not yet implemented, please open an issue or check for updates in future releases.

Sample Files & Scripts

PVW CLI includes comprehensive sample files and scripts for bulk operations:

Bulk Import Samples

CSV Samples: samples/csv/uc_terms_bulk_example.csv (8 sample terms)
JSON Samples:
- samples/json/term/uc_terms_bulk_example.json (8 data management terms)
- samples/json/term/uc_terms_sample.json (8 business terms)
Lineage CSV Samples: samples/csv/lineage_example.csv - Multiple lineage relationships with metadata

Lineage Documentation

Comprehensive Guide: doc/guides/lineage-csv-import.md - Complete lineage CSV import documentation
- CSV format specification with required/optional columns
- Command examples for validate, import, sample, templates
- Workflow recommendations and troubleshooting
- Advanced scenarios with metadata and multiple transformations

Bulk Delete Scripts

PowerShell: scripts/delete-all-uc-terms.ps1 - Full-featured with confirmation prompts
Python: scripts/delete_all_uc_terms_v2.py - Rich progress bars and error handling

Test Scripts

PowerShell: scripts/test-json-output.ps1 - Validates JSON output parsing

Jupyter Notebooks

samples/notebooks (plus)/unified_catalog_terms_examples.ipynb - Complete examples including:
- Examples 10-16: Bulk import demonstrations
- Code generation for CSV/JSON files
- Dry-run and actual import examples
- Term verification workflows

Documentation

Core Documentation

Main Documentation: doc/README.md
Unified Catalog: doc/commands/unified-catalog.md
Bulk Import Guide: doc/commands/unified-catalog/term-bulk-import.md
Data Products: doc/commands/data-product.md

Quick Reference

API Coverage: All major Purview APIs including Unified Catalog, Data Map, Discovery, Collections
Authentication: Azure CLI, Service Principal, Managed Identity support
Output Formats: Table (default), JSON (plain), JSONC (colored)
Bulk Operations: Import/export terms from CSV/JSON, bulk delete scripts

Recent Updates (October 2025)

Bulk Term Import/Export

Import multiple terms from CSV or JSON files
Dry-run mode for validation before import
Support for owners (Entra ID GUIDs), acronyms, resources
Progress tracking and detailed error reporting
100% success rate in testing (8/8 terms)

PowerShell & Scripting Integration

New --output parameter with table/json/jsonc formats
Plain JSON works with PowerShell's ConvertFrom-Json
Compatible with jq, Python json module, and other tools
Migration from deprecated --json flag

Bulk Delete Scripts

PowerShell script with interactive confirmation ("DELETE" to confirm)
Python script with Rich progress bars
Beautiful UI with colored output
Success/failure tracking per term
Rate limiting (200ms delay)

Critical Fixes (v1.3.8)

Search API Suggest/Autocomplete: Fixed HTTP 400 errors by removing empty filter objects from payload
Collection Display: Enhanced collection name detection with proper fallback logic (isinstance checks)
Owner ID Format: Must use Entra ID Object IDs (GUIDs), not email addresses
Domain Status: Terms cannot be "Published" in unpublished domains - use "Draft"
Error Validation: Enhanced error handling shows actual API responses
Windows Console Compatibility: All emoji removed for CP-1252 encoding support

Key Features Summary

Unified Catalog (UC) - Complete Management

Governance domains, glossary terms, data products
Objectives & Key Results (OKRs), Critical Data Elements (CDEs)
Health monitoring and workflow automation
Full CRUD operations with smart partial updates

Bulk Operations

CSV/JSON import with dry-run validation
PowerShell and Python bulk delete scripts
Progress tracking and error handling
Sample files and templates included

Multiple Output Formats

Table format for human viewing (default)
Plain JSON for PowerShell/bash scripting
Colored JSON for visual inspection

Automation & Integration

Azure CLI, Service Principal, Managed Identity auth
Works in local development, CI/CD, and production
Compatible with PowerShell, bash, Python, jq

Comprehensive Documentation

Complete API coverage documentation
Jupyter notebook examples
Troubleshooting guides
Sample files and templates

Contributing & Support

Documentation: Full Documentation
Issue Tracker: GitHub Issues
Email Support: [email protected]
Repository: GitHub - Keayoub/Purview_cli

License

See LICENSE file for details.

PVW CLI v1.3.8 empowers data engineers, stewards, and architects to automate, scale, and enhance their Microsoft Purview experience with powerful command-line and programmatic capabilities.

Latest in v1.3.8:

Fixed Search API suggest/autocomplete (HTTP 400 errors resolved)
Enhanced collection display with robust fallback logic
Comprehensive search command validation
Bulk term import/export with dry-run support
PowerShell integration with plain JSON output
Multiple output formats and beautiful progress tracking

Name		Name	Last commit message	Last commit date
Latest commit History 221 Commits
.github		.github
doc		doc
purviewcli		purviewcli
releases		releases
samples		samples
scripts		scripts
tests		tests
tools		tools
.gitignore		.gitignore
README.md		README.md
pyproject.toml		pyproject.toml
quick_start_guide.bat		quick_start_guide.bat
requirements.in		requirements.in
requirements.txt		requirements.txt

Keayoub/pvw-cli

Folders and files

Latest commit

History

Repository files navigation

PURVIEW CLI v1.3.8 - Microsoft Purview Automation & Data Governance

What is PVW CLI?

Key Capabilities

NEW: MCP Server for AI Assistants

What's New in Recent Releases

v1.3.8 (November 3, 2025) - Advanced Lineage Features

v1.2.5 (October 30, 2025) - 86% UC API Coverage

🚀 New APIs Implemented

🔧 Major Fixes & Improvements

📚 Documentation (3,500+ lines)

Getting Started

Quick Start Examples

v1.3.8 - Column-Level Lineage

v1.2.5 - Relationships API

v1.2.5 - Query APIs

v1.2.5 - Policy Management

v1.2.5 - Custom Metadata

Overview

Installation

Requirements

Getting Started

Authentication

1. Azure CLI Authentication (Recommended for Interactive Use)

2. Service Principal Authentication (Recommended for Automation/CI/CD)

3. Managed Identity (for Azure VMs, App Services, etc.)

4. Visual Studio/VS Code Authentication

Output Formats & Scripting Integration

Output Format Options

PowerShell Integration

Bash/Linux Integration

Examples by Command

Migration from Old --json Flag

Required Purview Configuration

🔍 How to Find Your Purview Values

1. PURVIEW_ACCOUNT_NAME

2. PURVIEW_ACCOUNT_ID

3. PURVIEW_RESOURCE_GROUP

📋 Setting the Variables

Debug Environment Issues

Search Command (Discovery Query API)

CLI Usage Examples

Multiple Output Formats

Search Operations

Usage Scenarios

Python Usage Example

Test Examples

Unified Catalog Management (NEW)

Quick UC Examples

Governance Domains Management

Glossary Terms in UC

Data Products Management

Objectives & Key Results (OKRs)

Critical Data Elements (CDEs)

Health Monitoring (NEW)

Workflow Management (NEW)

Integrated Workflow Example

Entity Management & Updates

Entity Update Examples

Update Asset Descriptions

Bulk Entity Operations

Column-Level Updates

Discovery to Update Workflow

Lineage CSV Import & Management

Lineage CSV Import

CSV Format

Lineage Commands

Available Templates

Workflow Example

Advanced Features

Data Product Management (Legacy)

Example Commands

Core Features

API Coverage and Support

Supported API Groups

API Version Support

Packages