README

Many organizations run large complex software packages and flexibility is not necessarily their strong point. Simple customer specific adjustments or process changes are very time-consuming and expensive. Low-code development platforms or bots promise to bring back flexibility, but can themselves be a closed ecosystem. Datapool is a lightweight open source web application that gives control back to the user or smaller organizational units within a company. Datapool can carry out periodic data crunching with processes defined at team or department level. Datapool can also be used to bridge temporary gaps, for testing processes as well as for production.

Datapool was originally developed to process invoices (pdf documents) within a patent department, in which all invoice data is extracted, analyzed and compared with the patent files. The invoice data is then processed in conjunction with UNYCOM and SAP. In this production environment, Datapool has been used to process approx. 1000 invoices per month. The data is compared with approx. 200,000 cost data records and 100,000 patent files. Processing takes place 1-2 times per month.

Sample application

Moving data between the packages is even more challenging. This framework aims to fill the gap between the big software packages such as SAP and e.g. UNYCOM in the setup of an IPR or patent department. The software is designed to relieve people from mindless repetitive tasks, allowing them to focus on the valuable tasks.

The figure shows a typical application example in a company software setup including SAP and UNYCOM. UNYCOM is used by patent departments of larger enterprises. UNYCOM manages patent files including cost records. There can be a substantial amount of incoming invoices. The payment is usually dealt with by SAP but the invoice data (content) as well as the documentation of the payment made through SAP needs to end up in the correct UNYCOM patent case.

This example requires the following steps:

Parsing: content extraction from the invoice. SAP relevant data as well as patent case specific data.
Matching an SAP accounting record with the patent case.
Mapping: adjusting data formats and types to create a UNYCOM compatible dataset.

The Datapool framework can achieve exactly this in a very transparent manner!

Features

Very lightweight web-application with data-processing, calendar, user account and rights management, a multimedia app and forum
Data sources can be media-files, pdf-documents, spreadsheet-files either uploaded manually or downloaded from an email inbox
External devices can provide data or files through a client interface
The result of the processing can be spreadsheet-files, zip-files, emails or SMS-messages
Data processing can be controlled manually or by trigger derived from values or calendar events
Processes can be easily designed and adopted via a graphical user interface
Processes can easily be exported or imported to other systems running Datapool

Hosting the web-application

Requirements

This software is designed to run on a server, i.e. the user interface is the web browser. It requires PHP 8+ and a database. Depending on the application requirements access to an email account might be required.

Installation

For the installation and creation of the first user account please refer to the video below.

Choose your target directory on your web server or your computer and run composer composer create-project sourcepot/datapool {add your target directory here}. This will create, among other things, the ../www/-subdirectory, which is the www-root and should be accessible via the network, i.e. from a client web browser.
Create a database and a corresponding database user. Set the database collation to utf8mb4_unicode_ci.
Call the webpage through a web browser. This will create an error message since the database access needs to be set up. (Check the error log which can be found in the ../debugging/-subdirectory. Each error generates a JSON-file containing the error details.)
Calling the webpage creates the file ../setup/Database/connect.json which contains the database access credentials. Use a text editor to update or match the credentials with the database user credentials.
Refresh the webpage. This will create an initial admin user account. The user credentials of this account can be found in ../setup/User/initAdminAccount.json. Register your own account and use the initial admin user account to change your own account to admin access level. Delete the initial admin user account after you have set up your own admin account.
Make sure only the ../www/-subdirectory is visible to the public.

Example installation using Composer on a notebook computer running MS Windows, XAMPP server and MariaDB:

2023-10-29_Add_web_page_to_www-Directory_and_create_admin_account_720p.mp4

Initial adjustments

After you have set up your admin account you should login and update the webmaster email address Admin → Admin → Page settings → EmailWebmaster. Allways use the ✓ button the save changes.

Architecture

Datapool is based on an object collection or oc, i.e. a collection objects instantiated from PHP-classes in the ../php/ folder. The object collection is created by the constructor of class ../php/Root.php each time the web-application is called. ../php/Root.php provides the collection to all instantiated classes which implement the method init(array $oc). Typically the classes have a private property oc which is set/updated by the init method of the class.

The configuration file ../setup/objectList.csv determines the order of creation of objects. With the private property registerVendorClasses of class ../php/Root.php vendor classes can be added to the object collection. Otherwise, an instance of a vendor class can be created within the source code when required.

Web page creation

The following flowchart shows the sequence of object instantiations, method calls and content creation.

Any class which implements the App interface must provide a run method. The run method defines the app specific menu item, the app visibility and the method adds the app specific web page content. The following figure shows the run method of the calendar app SourcePot\Datapool\GenericApps\Calendar->run().

Data category Apps, e.g. Invoices

Data apps use the DataExplorer class SourcePot\Datapool\Foundation\DataExplorer. The data explorer provides a blank canvas to create data crunching processes graphically. This is done by adding canvas elements and by configuring their properties. A canvas element is a view of a database table. The database table view applies a selector Content → Selector (see the figure below). Features can be added to the canvas element such as File upload (e.g. for invoice documents, email etc.), pdf-parser and/or a processor. There is a set of basic processors to e.g. match, map or forward entries. There are also a basic processors to create pdf-documents, to send emails or SMS.

The DataExplorer has two modes: view and edit The figure below shows how to togle between view and edit mode. In edit mode each canvas element can be dragged, selected or deleted. To change canvas element properties the canvas element needs to be selected by clicking on the diamond shaped red button of the respective canvas element.

sourcepot / datapool

Maintainers

Details