Tables Editor — Visual SQL Pipeline Builder

No-Code Database Queries

We built a visual SQL query editor that transforms complex database operations into an intuitive node graph. Instead of writing SQL by hand, operators and configurators connect colored nodes — and the system generates optimized queries automatically.

The Tables Editor showing a complete data pipeline: blue PostgreSQL Query nodes (data sources), orange Filter and Time Bucket nodes (transformations), purple Group By and Derived Table nodes (aggregation), green Sort node (output ordering), and red WinCC OA Tags node (live data integration). The left palette shows all available node types; the bottom panel previews the generated SQL in real time.

How It Works

Node Types

Category	Nodes	Purpose
Data Sources	PostgreSQL Query, Named Query, WinCC OA Tags, CSV Import, Static Data	Bring data into the pipeline
Transforms	Filter Rows, Select Columns, Rename Columns, Sort, Limit/Offset, Expression	Reshape and filter
Aggregation	Group By, Time Bucket	Aggregate with SUM, AVG, COUNT, MIN, MAX
Joins	All SQL join types	Combine multiple data sources
Output	Sort, Derived Table	Final result presentation

Column Propagation

This is one of the smartest features we implemented: automatic schema inference through the pipeline. When you connect a PostgreSQL Query node to a Filter node, the Filter automatically knows which columns are available. Add a Group By after that, and it knows the result schema. This means:

• No manual column mapping
• Errors detected at design time, not runtime
• Downstream nodes always show valid options

SQL-to-Nodes Parser

We also built the reverse direction — paste an existing SQL query and the system decomposes it into a visual node graph. This enables importing legacy queries into the visual editor for modification and maintenance.

Live Preview & Generated SQL

At the bottom of the editor, the system shows the generated SQL query in real time. Click "Run" and results display immediately. This gives full transparency — you can see exactly what the visual pipeline produces and verify it.

Pipeline Storage & Workflow Orchestration

Pipelines can be saved, loaded, and composed into multi-step workflows with dependency resolution:

• Topological sort (Kahn's algorithm) for execution ordering
• Circular dependency detection with clear error messages
• Parameter passing between steps — results from step N feed into step N+1
• Transactional execution — each step runs in its own transaction

From Editor to Live Dashboard — No Code

The real power of the Tables Editor is what comes after designing the pipeline. Every saved pipeline becomes available as a data source for the live application — without writing a single line of code.

Report Node Integration

In the Menu Editor, a Report node can reference any pipeline by name. The Edit Report dialog shows all available pipelines — both simple queries and multi-step workflows (prefixed with [WF]):

The Edit Report dialog in the Menu Editor: selecting a Workflow Pipeline from a dropdown of 14 available pipelines — including "Demo - Industrial Orders", "Building Energy Monthly Analytics", "OEE Weekly Trend", "Water Compliance Summary", and multi-step workflows like "[WF] Production OEE → Line → Shift Detail". The $DP variable inheritance is visible at the top ($DP_Secondo = Valvola).

The Result: Production-Ready Tables

Once a pipeline is connected via a Report node, the TablePostgre widget renders the data automatically. Here are examples of tables created entirely through the visual pipeline — zero SQL written by hand, zero backend code:

A "City Overview → Plant Details" master-detail dashboard: the left table shows 21 Italian cities with population, total plants, and water flow metrics — all pulled from a visual pipeline. Clicking a row loads the plants for that city in the right panel. No code was written to create this view.

The same pipeline with all columns visible: CITY_ID, City, Region, Population, Total Plants, Water Flow, Energy KWh, AVG Efficiency, Total Alarms — and the detail panel showing Plant ID, Plant Name, and Plant Type. All data, sorting, and aggregation configured visually in the Tables Editor.

Complete No-Code Path

The workflow from design to dashboard is entirely visual:

1. Build the query visually in the Tables Editor (drag nodes, connect wires)
2. Save it as a named pipeline
3. Reference the pipeline from a Report node in the Menu Editor
4. The TablePostgre widget executes it and displays the results

No SQL written. No backend code. Just visual connections.