sfdump 2.8.1

Salesforce data export and archival tool with verification, SQLite database, and Streamlit viewer

Salesforce data export and archival tool. Bulk downloads Attachments, ContentVersions, CSV object data, and invoice PDFs with verification, retry mechanisms, a searchable SQLite database, and a Streamlit web viewer.

Built for organisations whose Salesforce licences may expire – provides confidence that everything has been captured and can be browsed offline.

Features

• Full data export – 44 essential Salesforce objects exported to CSV

• File downloads – Attachments (legacy) and ContentVersions with SHA-256 verification

• Invoice PDFs – bulk download via deployed Apex REST endpoint (sf sins)

• SQLite database – queryable offline database built from CSV exports

• Streamlit viewer – web UI with Document Explorer, record navigator, and search

• Export inventory – single-command completeness check across all categories (sf inventory)

• Idempotent & resumable – all operations support resume; re-run to pick up where you left off

• Verification & retry – SHA-256 integrity checks with automatic retry of failed downloads

Two CLIs

sf	Simple commands for daily use
sfdump	Full CLI with all advanced options

# Simple workflow
sf dump                   # Export everything from Salesforce
sf view                   # Browse exported data in web viewer
sf inventory              # Check export completeness
sf sins                   # Download invoice PDFs
sf status                 # List available exports
sf usage                  # Check API usage / limits

# Advanced
sfdump files              # Export Attachments/ContentVersions
sfdump csv                # Export specific objects to CSV
sfdump verify-files       # Verify file integrity (SHA-256)
sfdump retry-missing      # Retry failed downloads
sfdump inventory -v       # Detailed completeness report
sfdump build-db           # Build SQLite database
sfdump db-viewer          # Launch Streamlit viewer
sfdump docs-index         # Rebuild document indexes

Export Inventory

The inventory command provides a single authoritative answer to “is my export complete?” by inspecting only local files (no API calls):

$ sf inventory

Export Inventory
==================================================
Location:  ./exports/export-2025-03-15

  Category             Status         Expected     Present
  CSV Objects          COMPLETE             44          44
  Attachments          COMPLETE         12,456      12,456
  ContentVersions      COMPLETE            285       1,194
  Invoice PDFs         INCOMPLETE        1,200           0  (1,200 missing)
  Indexes              COMPLETE             11          11
  Database             COMPLETE                  14 tables

  Overall: INCOMPLETE

The inventory checks six categories:

• CSV Objects – all 44 essential Salesforce objects present

• Attachments – legacy file downloads verified against metadata

• ContentVersions – modern file downloads verified against metadata

• Invoice PDFs – FinancialForce/Coda invoice PDFs (c2g__codaInvoice__c)

• Indexes – per-object file indexes and master document index

• Database – SQLite database tables and row counts

Use --json-only for machine-readable output. The manifest is also auto-generated at meta/inventory.json after every sf dump run.

Installation

All Platforms (PyPI)

If you have Python 3.12+ installed:

pip install sfdump

To upgrade:

pip install --upgrade sfdump

Windows (Starting from Nothing)

Option 1: One-Line Install (Recommended)

Open PowerShell (press Win+R, type powershell, press Enter) and paste:

irm https://raw.githubusercontent.com/ksteptoe/sfdump/main/bootstrap.ps1 | iex

This downloads and installs everything automatically.

Option 2: Manual Download

1. Go to https://github.com/ksteptoe/sfdump

2. Click the green “Code” button > “Download ZIP”

3. Extract the ZIP file

4. Double-click install.bat

Detailed Instructions

See INSTALL.md for step-by-step instructions with screenshots and troubleshooting tips.

Requirements

• Windows 10 or 11

• 40 GB free disk space (for Salesforce exports)

• Internet connection

• No administrator rights required

macOS / Linux

git clone https://github.com/ksteptoe/sfdump.git
cd sfdump
pip install -e ".[dev]"

# Or using make
make bootstrap

Quick Start

# First time: configure credentials
sf setup                  # Interactive .env creation
sf test                   # Verify connection

# Export
sf dump                   # Full export (files + CSVs + database)
sf dump --retry           # Export and retry failed downloads

# Browse offline
sf view                   # Launch Streamlit viewer

# Completeness check
sf inventory              # Are we done?

Configuration

Create a .env file in your working directory (or use sf setup):

SF_AUTH_FLOW=client_credentials
SF_CLIENT_ID=<your_consumer_key>
SF_CLIENT_SECRET=<your_consumer_secret>
SF_LOGIN_URL=https://yourcompany.my.salesforce.com

For invoice PDF downloads, a Web Server OAuth flow is also needed:

sfdump login-web          # Opens browser for SSO login

Invoice PDFs

Invoice PDFs are generated on-the-fly by a Visualforce page in Salesforce – they are not stored as files. A deployed Apex REST class (SfdumpInvoicePdf) renders each invoice to PDF server-side.

sfdump login-web          # Authenticate via browser (SSO)
sf sins                   # Download all Complete invoice PDFs
sf sins --force           # Re-download everything

PDFs are saved to {export}/invoices/SIN001234.pdf and indexed for the viewer’s Document Explorer.

Work in Progress

Invoice PDF Pipeline

Status: The Apex REST endpoint is deployed to production and the sf sins command works end-to-end. The remaining gap is that the Web Server OAuth token (sfdump login-web) needs to be refreshed manually. The sf dump orchestrator will attempt invoice PDF downloads automatically if a valid web token exists, but falls back gracefully if not.

What’s deployed in Salesforce:

• Apex class SfdumpInvoicePdf – REST endpoint at /services/apexrest/sfdump/invoice-pdf?id={invoiceId}

• Test class SfdumpInvoicePdfTest – 94% coverage

• These are read-only and harmless; recommend keeping them permanently

Known limitation: The org uses SSO (SAML), so the client_credentials OAuth flow cannot render Visualforce pages. Invoice PDF download requires the Web Server (Authorization Code + PKCE) flow which produces a real user session.

Export Completeness Direction

The sf inventory command is the foundation for a broader completeness guarantee. The direction:

1. Inventory system (done) – offline checks across all six categories, JSON manifest at meta/inventory.json, auto-generated after each export

2. CI integration – use sf inventory --json-only in pipelines to assert export completeness before archival

3. Drift detection – compare inventory manifests over time to detect regressions (e.g. files deleted, database corruption)

4. Archival sign-off – once all categories show COMPLETE, the export can be confidently archived as the authoritative copy of the org’s data

Architecture

Salesforce API --> api.py --> files.py --> CSV + binary exports
                                |
                            verify.py (SHA-256 completeness)
                                |
                            retry.py (failure recovery)
                                |
                          inventory.py (completeness check)
                                |
                           indexing/ (document indexes)
                                |
                     viewer/ + viewer_app/ (SQLite + Streamlit)

Export directory structure:

exports/export-2026-01-26/
    csv/                    # 44 Salesforce objects as CSV
    files/                  # ContentVersion binaries
    files_legacy/           # Attachment binaries
    invoices/               # Invoice PDFs (SIN*.pdf)
    links/                  # Metadata CSVs + file indexes
    meta/
        sfdata.db           # SQLite database
        inventory.json      # Completeness manifest
        master_documents_index.csv

Making Changes & Contributing

This project uses pre-commit, please make sure to install it before making any changes:

pip install pre-commit
cd sfdump
pre-commit install

It is a good idea to update the hooks to the latest version:

pre-commit autoupdate

Don’t forget to tell your contributors to also install and use pre-commit.

Note

This project has been set up using PyScaffold 4.6. For details and usage information on PyScaffold see https://pyscaffold.org/.