Let your DuckDB quack SAP!

February 11, 2026 · View on GitHub

Let your DuckDB quack SAP!

Disclaimer: This extension is currently in an experimental state. Feel free to try it out, but be aware that minimal testing and benchmarking were done.

With ERPL, use Enterprise Data in your Data Science and ML pipelines within minutes!

Quick Install

Install with two lines of code in DuckDB (unsigned option must be set):

INSTALL 'erpl' FROM 'http://get.erpl.io';
LOAD 'erpl';

★ What is the ERPL Extension?

The primary objective of this DuckDB extension is to facilitate seamless integration with the SAP data ecosystem. Our approach prioritizes:

Minimal dependencies: Ensuring a lightweight experience.
User-centric design: Intuitive and straightforward usage for DuckDB users.

We focus predominantly on two main use cases:

Data Science & Analytics: Directly accessing data from SAP ERP (via RFC) and SAP BW (via BICS) for interactive analytical tasks.
Data Replication: Efficiently migrating data from SAP ERP or SAP BW into DuckDB for incorporation into DuckDB-centric data workflows.

Please be aware that DataZoo GmbH is the independent developer of this extension and does not hold any affiliation with DuckDB Labs or the DuckDB Foundation. "DuckDB" is a trademark registered by the DuckDB Foundation.

The extension suite currently supports:

Data queries from SAP ERP tables and CDS views (via RFC)
Execution of arbitrary RFC function modules
SAP BW cube and query execution via BICS protocol
BW hierarchy extraction, metadata views, and lineage analysis
Data extraction and replication via SAP ODP (full and delta)
SSH tunneling for secure SAP connections
Virtual catalog mounting of SAP systems via ATTACH

Transparency is our ethos, and in line with this, we are planning a commercial trajectory for the extension, structured as follows:

Community Edition: A gratis version for individual users, both private and commercial, enabling queries from SAP ERP tables.
Commercial License: Designed for businesses developing services or products leveraging our extension. More details can be found in our BSL v1. License.
Enterprise Edition: A premium version offering additional capabilities to query data from SAP Business Warehouse and data replication via SAP ODP.

For inquiries, potential collaborations, or if your curiosity is piqued, connect with us at https://erpl.io.

⚙ Example Usage

A first example demonstrates how to join two SAP tables with an external table. We'll be using the ABAP Flight Reference Scenario, specifically joining the SFLIGHT and SPFLI tables which contain flight and flight schedule details respectively, with an external table WEATHER that holds weather information. We will extract flight information and associated temperatures at departure and arrival cities.

To start with, we assume you have setup DuckDB and already installed the ERPL extension (see below) and have access to an SAP system having the ABAP Flight Reference Scenario data.

First we start DuckDB with the -unsigned - flag set. Then we load the extension and configure the necessary credentials to connect to the SAP system:

LOAD 'erpl';

CREATE PERSISTENT SECRET abap_trial (
    TYPE sap_rfc, 
    ASHOST 'localhost', 
    SYSNR '00', 
    CLIENT '001', 
    USER 'DEVELOPER', 
    PASSWD 'ABAPtr2023#00',
    LANG 'EN'
);

In our case we use the docker based ABAP Platform Trial. The credentials are set by default, details can be found in the documentation of the docker image.

Now we explore the schema of the three tables tables:

SELECT * FROM SAP_DESCRIBE_FIELDS('SFLIGHT');

pos	is_key	field	text	sap_type	length	decimals	check_table	ref_table	ref_field	language
0001	X	MANDT	Client	CLNT	000003	000000	T000			E
0002	X	CARRID	Airline Code	CHAR	000003	000000	SCARR			E
0003	X	CONNID	Flight Connection Number	NUMC	000004	000000	SPFLI			E
0004	X	FLDATE	Flight date	DATS	000008	000000				E
0005		PRICE	Airfare	CURR	000015	000002		SFLIGHT	CURRENCY	E
0006		CURRENCY	Local currency of airline	CUKY	000005	000000	SCURX			E
0007		PLANETYPE	Aircraft Type	CHAR	000010	000000	SAPLANE			E
0008		SEATSMAX	Maximum Capacity in Economy Class	INT4	000010	000000				E
0009		SEATSOCC	Occupied Seats in Economy Class	INT4	000010	000000				E
0010		PAYMENTSUM	Total of current bookings	CURR	000017	000002		SFLIGHT	CURRENCY	E
0011		SEATSMAX_B	Maximum Capacity in Business Class	INT4	000010	000000				E
0012		SEATSOCC_B	Occupied Seats in Business Class	INT4	000010	000000				E
0013		SEATSMAX_F	Maximum Capacity in First Class	INT4	000010	000000				E
0014		SEATSOCC_F	Occupied Seats in First Class	INT4	000010	000000				E

SELECT * FROM SAP_DESCRIBE_FIELDS('SPFLI');

pos	is_key	field	text	sap_type	length	decimals	check_table	ref_table	ref_field	language
0001	X	MANDT	Client	CLNT	000003	000000	T000			E
0002	X	CARRID	Airline Code	CHAR	000003	000000	SCARR			E
0003	X	CONNID	Flight Connection Number	NUMC	000004	000000				E
0004		COUNTRYFR	Country Key	CHAR	000003	000000	SGEOCITY			E
0005		CITYFROM	Departure city	CHAR	000020	000000	SGEOCITY			E
0006		AIRPFROM	Departure airport	CHAR	000003	000000	SAIRPORT			E
0007		COUNTRYTO	Country Key	CHAR	000003	000000	SGEOCITY			E
0008		CITYTO	Arrival city	CHAR	000020	000000	SGEOCITY			E
0009		AIRPTO	Destination airport	CHAR	000003	000000	SAIRPORT			E
0010		FLTIME	Flight time	INT4	000010	000000				E
0011		DEPTIME	Departure time	TIMS	000006	000000				E
0012		ARRTIME	Arrival time	TIMS	000006	000000				E
0013		DISTANCE	Distance	QUAN	000009	000004		SPFLI	DISTID	E
0014		DISTID	Mass unit of distance (kms, miles)	UNIT	000003	000000				E
0015		FLTYPE	Flight type	CHAR	000001	000000				E
0016		PERIOD	Arrival n day(s) later	INT1	000003	000000				E

DESCRIBE SELECT * FROM 'WEATHER.csv';

column_name	column_type	null
FLDATE	DATE	YES
COUNTRY	VARCHAR	YES
CITY	VARCHAR	YES
TEMPERATURE	DOUBLE	YES
CONDITION	VARCHAR	YES

The actual example joins this three tables:

SELECT 
  f.CARRID,
  f.CONNID,
  f.FLDATE,
  s.CITYFROM as CITY_FROM,
  ROUND(w_from.TEMPERATURE, 1) as TEMP_FROM,
  s.CITYTO as CITY_TO,
  ROUND(w_to.TEMPERATURE, 1) as TEMP_TO,
  FROM sap_read_table('SFLIGHT') AS f
  JOIN sap_read_table('SPFLI') AS s 
      ON (f.MANDT = s.MANDT AND f.CARRID = s.CARRID AND f.CONNID = s.CONNID)
  JOIN "WEATHER.csv" AS w_from
      ON (f.FLDATE = w_from.FLDATE AND s.COUNTRYFR = w_from.COUNTRY AND s.CITYFROM = w_from.CITY)
  JOIN "WEATHER.csv" AS w_to
      ON (f.FLDATE = w_to.FLDATE AND s.COUNTRYTO = w_to.COUNTRY AND s.CITYTO = w_to.CITY)
  ORDER BY 1, 2, 3
  LIMIT 25

This SQL query performs the following operations:

Retrieves flight details from SFLIGHT using ERPL's sap_read_table, aliasing it as f.
Again using ERPL's sap_read_table we join SPFLI (aliased as s) on MANDT, CARRID, and CONNID to get the flight's city of origin and destination.
Incorporates two instances of an external weather data CSV file, w_from and w_to, matching on flight date and respective cities' country and name for departure and arrival.
Rounds the temperature data to one decimal place for readability.
Orders the results by CARRID, CONNID, and FLDATE.
Limits the output to the first 25 rows for a concise view.

The output of this query will provide a comprehensive view of the flights, including their departure and arrival cities, and the corresponding temperatures, thus offering valuable insights for flight operations analysis.

| CARRID   |   CONNID | FLDATE              | CITY_FROM   |   TEMP_FROM | CITY_TO       |   TEMP_TO |
|:---------|---------:|:--------------------|:------------|------------:|:--------------|----------:|
| AA       |     0017 | 2016-11-15 00:00:00 | NEW YORK    |        28.3 | SAN FRANCISCO |      19.8 |
| AA       |     0017 | 2017-02-03 00:00:00 | NEW YORK    |        18.9 | SAN FRANCISCO |      17.2 |
| AA       |     0017 | 2017-04-24 00:00:00 | NEW YORK    |        14.7 | SAN FRANCISCO |      16.2 |
| AA       |     0017 | 2017-07-13 00:00:00 | NEW YORK    |        16.8 | SAN FRANCISCO |      22.5 |
| AA       |     0017 | 2017-10-01 00:00:00 | NEW YORK    |        14.6 | SAN FRANCISCO |      28.3 |
| AA       |     0017 | 2017-12-20 00:00:00 | NEW YORK    |        13   | SAN FRANCISCO |      21.4 |
| AZ       |     0555 | 2016-11-15 00:00:00 | ROME        |        20.6 | FRANKFURT     |      24.2 |
| AZ       |     0555 | 2017-02-03 00:00:00 | ROME        |        13.1 | FRANKFURT     |      24.1 |
| AZ       |     0555 | 2017-04-24 00:00:00 | ROME        |        20.7 | FRANKFURT     |      24.6 |
| AZ       |     0555 | 2017-07-13 00:00:00 | ROME        |        14   | FRANKFURT     |      16.3 |
| AZ       |     0555 | 2017-10-01 00:00:00 | ROME        |        20.9 | FRANKFURT     |      15.3 |
| AZ       |     0555 | 2017-12-20 00:00:00 | ROME        |        23.8 | FRANKFURT     |      24.7 |
| AZ       |     0789 | 2016-11-15 00:00:00 | TOKYO       |        29.7 | ROME          |      20.6 |
| AZ       |     0789 | 2017-02-03 00:00:00 | TOKYO       |        23.4 | ROME          |      13.1 |
| AZ       |     0789 | 2017-04-24 00:00:00 | TOKYO       |        28.5 | ROME          |      20.7 |
| AZ       |     0789 | 2017-07-13 00:00:00 | TOKYO       |        19.4 | ROME          |      14   |
| AZ       |     0789 | 2017-10-01 00:00:00 | TOKYO       |        19.8 | ROME          |      20.9 |
| AZ       |     0789 | 2017-12-20 00:00:00 | TOKYO       |        15.6 | ROME          |      23.8 |
| DL       |     0106 | 2016-11-13 00:00:00 | NEW YORK    |         9.1 | FRANKFURT     |      13.8 |
| DL       |     0106 | 2017-02-01 00:00:00 | NEW YORK    |        24.3 | FRANKFURT     |      23.9 |
| DL       |     0106 | 2017-04-22 00:00:00 | NEW YORK    |        15.8 | FRANKFURT     |      23.5 |
| DL       |     0106 | 2017-07-11 00:00:00 | NEW YORK    |        20.7 | FRANKFURT     |      30.6 |
| DL       |     0106 | 2017-09-29 00:00:00 | NEW YORK    |        16.3 | FRANKFURT     |      19.1 |
| DL       |     0106 | 2017-12-18 00:00:00 | NEW YORK    |        14.4 | FRANKFURT     |      21.8 |
| JL       |     0407 | 2016-11-17 00:00:00 | TOKYO       |        20.3 | FRANKFURT     |      20.4 |

This example can also be found in its Python, R, or NODEJS version in the examples folder.

Using ATTACH for Catalog-Style Access

Instead of calling sap_read_table() directly, you can attach an SAP system as a DuckDB catalog. This lets you query SAP tables with standard SQL syntax:

ATTACH '' AS sap (TYPE sap_rfc, SECRET 'abap_trial');

-- Query tables directly via catalog
SELECT * FROM sap."/DMO/FLIGHT";

-- Join across SAP tables
SELECT f.CARRID, f.CONNID, f.FLDATE
FROM sap."/DMO/FLIGHT" AS f
JOIN sap."/DMO/CARRIER" AS c ON f.CARRID = c.CARRIER_ID;

You can optionally restrict which tables are accessible using the TABLES option:

ATTACH '' AS sap (TYPE sap_rfc, SECRET 'abap_trial', TABLES '/DMO/FLIGHT,/DMO/CARRIER');

➜ Obtaining the ERPL Extension

Introduction

Building extensions for DuckDB can be challenging due to the varying C++ compiler and library ecosystem. This variability often leads to incompatibilities between locally built extensions and the centrally distributed DuckDB binary, primarily due to differences in the Application Binary Interface (ABI).

Recommended Approach

To ensure compatibility and ease of use, we follow a build process similar to that of the DuckDB team. Our advice is to start with the pre-compiled binaries available in our GitHub releases. For those interested in building the extension themselves, our development instructions provide detailed guidance.

Binary Selection

The assets in each release follow this naming convention:

erpl-${DUCKDB_VERSION}-extension-{OS}-{ARCH}.tar.gz

Choose the binary that matches your usage scenario. The table below summarizes the available binaries for various platforms and use cases:

Usage	Operating System	Architecture
DuckDB CLI	Linux	amd64_gcc4
Python	Linux	amd64
R	Linux	amd64
Julia	Linux	amd64
Node.js	Linux	amd64
DuckDB CLI	Windows	amd64
Python	Windows	amd64
R	Windows	amd64
Julia	Windows	amd64
Node.js	Windows	amd64
DuckDB CLI	OSX	amd64
Python	OSX	amd64
R	OSX	amd64
Julia	OSX	amd64
Node.js	OSX	amd64
DuckDB CLI	OSX	arm64
Python	OSX	arm64
R	OSX	arm64
Julia	OSX	arm64
Node.js	OSX	arm64

💻 Installing the ERPL Binaries

Introduction

Installation of the ERPL extension is straightforward. Please note that this extension is independent of the DuckDB Foundation and DuckDB Labs, meaning the binaries are unsigned. Consequently, DuckDB must be initiated with the -unsigned flag. Detailed instructions on this process can be found in the DuckDB documentation.

Installation Steps

Enable Unsigned Extensions in DuckDB: Set the -unsigned flag as described in the DuckDB documentation.

Install and Load the ERPL Extension:

INSTALL 'path/to/erpl.duckdb_extension';
LOAD 'erpl';

Confirmation of Successful Installation

Upon successful installation and loading, the extension will output messages similar to:

-- Loading ERPL Trampoline Extension. --
(The purpose of the extension is to extract dependencies and load the ERPL implementation)
Saving ERPL SAP dependencies to '/home/user/.duckdb/extensions/v1.2.0/linux_amd64' and loading them ... done
ERPL RFC extension extracted and saved to /home/user/.duckdb/extensions/v1.2.0/linux_amd64.
ERPL Tunnel extension extracted and saved to /home/user/.duckdb/extensions/v1.2.0/linux_amd64.
ERPL BICS extension extracted and saved to /home/user/.duckdb/extensions/v1.2.0/linux_amd64.
ERPL ODP extension extracted and saved to /home/user/.duckdb/extensions/v1.2.0/linux_amd64.
ERPL RFC extension installed and loaded.
ERPL TUNNEL extension installed and loaded.
ERPL BICS extension installed and loaded.
ERPL ODP extension installed and loaded.
ERPL extensions loaded. For instructions on how to use them, visit https://erpl.io

Understanding the Extension Loading Process

The ERPL extension is composed of a trampoline and multiple sub-extensions:

Trampoline Extension (erpl): Extracts SAP NetWeaver RFC SDK libraries and the sub-extension binaries from its embedded payload.
Sub-extensions: The actual functional parts — erpl_rfc (RFC connectivity), erpl_tunnel (SSH tunneling), and optionally erpl_bics (SAP BW) and erpl_odp (data replication).

The trampoline extracts dependencies into the DuckDB extension folder, then installs and loads each sub-extension. Post-installation, the directory ~/.duckdb/extensions/<version>/<platform> will contain:

erpl.duckdb_extension         # Trampoline
erpl_rfc.duckdb_extension     # RFC connectivity
erpl_tunnel.duckdb_extension  # SSH tunneling
erpl_bics.duckdb_extension    # SAP BW (Enterprise Edition)
erpl_odp.duckdb_extension     # ODP replication (Enterprise Edition)
libicudata.so.50              # SAP SDK dependencies
libicui18n.so.50
libicuuc.so.50
libsapnwrfc.so
libsapucum.so

Function Reference

The complete API reference is in API_REFERENCE.md. Below is a summary.

Overview by Extension

Extension	Functions	Description
erpl_rfc	`sap_read_table`, `sap_rfc_invoke`, `sap_show_tables`, `sap_describe_fields`, `sap_rfc_describe_function`, `ATTACH ... TYPE sap_rfc`, and more	SAP RFC connectivity, table reads, function calls, metadata, ATTACH catalog
erpl_bics	`sap_bics_show`, `sap_bics_begin/rows/columns/filter/result`, `sap_bics_hierarchy`, `sap_bics_meta_`, `sap_bics_lineage_*`	SAP BW queries, hierarchies, metadata views, lineage analysis
erpl_odp	`sap_odp_show*`, `sap_odp_describe`, `sap_odp_read_full`, `sap_odp_preview`	SAP ODP data extraction and replication
erpl_tunnel	`tunnel_create`, `tunnel_close`, `tunnel_close_all`, `tunnels()`	SSH tunneling for secure SAP connections

Common Functions Quick Reference

Function	Example	Description
`sap_read_table`	`SELECT * FROM sap_read_table('SFLIGHT')`	Read SAP table data with filter/column pushdown
`sap_rfc_invoke`	`SELECT * FROM sap_rfc_invoke('STFC_CONNECTION', {'REQUTEXT': 'Hi'})`	Call any RFC function module
`sap_show_tables`	`SELECT * FROM sap_show_tables(TABLENAME='FLIGHT')`	Search SAP tables by name/description
`sap_describe_fields`	`SELECT * FROM sap_describe_fields('SFLIGHT')`	Get field metadata for an SAP table
`sap_bics_show_cubes`	`SELECT * FROM sap_bics_show_cubes()`	List available SAP BW cubes
`sap_odp_read_full`	`SELECT * FROM sap_odp_read_full('BW', 'MY_ODP')`	Full extraction from ODP source
`ATTACH`	`ATTACH '' AS sap (TYPE sap_rfc)`	Mount SAP system as virtual DuckDB database

For full parameter details, return types, and advanced usage see API_REFERENCE.md.

Tracking

Overview

Our extension automatically collects basic usage data to enhance its performance and gain insights into user engagement. We employ Posthog for data analysis, transmitting information securely to the European Posthog instance at https://eu.posthog.com via HTTPS.

Data Collected

Each transmitted request includes the following information:

Extension Version
DuckDB Version
Operating System
System Architecture
MAC Address of the Primary Network Interface

Event Tracking

Data is transmitted under these circumstances:

Extension Load: No extra data is sent beyond the initial usage information.
Function Invocation: The name of the invoked function is sent. Note: Function inputs/outputs are not transmitted.
Error Occurrence: The error message is transmitted.

User Configuration Options

Users can control tracking through these settings:

Enable/Disable Tracking:

SET erpl_telemetry_enabled = TRUE; -- Enabled by default; set to FALSE to disable tracking

Posthog API Key Configuration (usually unchanged):

SET erpl_telemetry_key = 'phc_XXX'; -- Pre-set to our key; customizable to your own key

This approach ensures transparency about data collection while offering users control over their privacy settings.

License

The ERPL extension is licensed under the Business Source License (BSL) Version 1.1. The BSL is a source-available license that gives you the following permissions:

Allowed:

Copy, Modify, and Create Derivative Works: You have the right to copy the software, modify it, and create derivative works based on it.
Redistribute and Non-Production Use: Redistribution and non-production use of the software is permitted.
Limited Production Use: You can make production use of the software, but with limitations. Specifically, the software cannot be offered to third parties on a hosted or embedded basis.
Change License Rights: After the Change Date (five years from the first publication of the Licensed Work), you gain rights under the terms of the Change License (MPL 2.0). This implies broader permissions after this date.

Not Allowed:

Offering to Third Parties on Hosted/Embedded Basis: The Additional Use Grant restricts using the software in a manner that it is offered to third parties on a hosted or embedded basis.
Violation of Current License Requirements: If your use does not comply with the requirements of the BSL, you must either purchase a commercial license or refrain from using the software.
Trademark Usage: You don't have rights to any trademark or logo of Licensor or its affiliates, except as expressly required by the License.

Additional Points:

Separate Application for Each Version: The license applies individually to each version of the Licensed Work. The Change Date may vary for each version.
Display of License: You must conspicuously display this License on each original or modified copy of the Licensed Work.
Third-Party Receipt: If you receive the Licensed Work from a third party, the terms and conditions of this License still apply.
Automatic Termination on Violation: Any use of the Licensed Work in violation of this License automatically terminates your rights under this License for all versions of the Licensed Work.
Disclaimer of Warranties: The Licensed Work is provided "as is" without any warranties, including but not limited to the warranties of merchantability, fitness for a particular purpose, non-infringement, and title.

This summary is based on the provided license text and should be used as a guideline. For legal advice or clarification on specific points, consulting a legal professional is recommended, especially for commercial or complex use cases.