Dataiku Gantt Chart Plugin

Technical Specification

Plugin ID: gantt-chart Current Version: 0.1.2 Last Updated: 2025-12-20

This is a living document describing the current state of the plugin. For version history, see CHANGELOG.md. For release-specific notes, see plan/releases/.

1. Overview

1.1 Purpose

A native Gantt chart visualization for Dataiku DSS that appears alongside built-in chart types, enabling project managers and analysts to visualize task timelines directly from any dataset.

1.2 Capabilities

Native Integration: Appears in Charts tab → "Other" section
Offline Operation: Zero external network dependencies (works in air-gapped environments)
Full-Featured: Task dependencies, progress tracking, color coding, 7 view modes
Production Quality: Handles edge cases gracefully with clear error feedback

1.3 Scope Boundaries

Included:

Read-only visualization of task timelines
Dependency arrows, progress bars, category colors
Interactive view mode switching and navigation

Not Included:

Write-back to dataset (read-only visualization)
Real-time collaboration features
Export to MS Project or other formats
Custom task creation UI

1.4 Performance

Renders 1,000 tasks in under 3 seconds
Default maxTasks limit prevents browser overload
Works identically in online and air-gapped instances

2. User Stories

ID	As a...	I want to...	So that...	Status
US1	Data Analyst	Select task/start/end columns from any dataset	I can quickly visualize project timelines	Done
US2	Project Manager	See task dependencies as connecting arrows	I understand the critical path	Done
US3	Team Lead	Color tasks by assignee or category	I can identify workload distribution	Done
US4	Executive	Switch between day/week/month views	I can see both details and big picture	Done
US5	User in secure environment	Use the chart without internet	My air-gapped instance works fully	Done

3. Architecture

3.1 Plugin Type

Webapp-based Custom Chart — Uses the chart field in webapp.json to register in the Charts tab rather than the Webapps menu.

3.2 Component Diagram

┌─────────────────────────────────────────────────────────────────┐
│                        Dataiku DSS                               │
│  ┌───────────────────────────────────────────────────────────┐  │
│  │                    Charts Tab                              │  │
│  │  ┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────────────────┐  │  │
│  │  │ Bar     │ │ Line    │ │ Scatter │ │ Other           │  │  │
│  │  └─────────┘ └─────────┘ └─────────┘ │ ┌─────────────┐ │  │  │
│  │                                       │ │ Gantt Chart │◄┼──┼──┤ Plugin
│  │                                       │ └─────────────┘ │  │  │
│  │                                       └─────────────────┘  │  │
│  └───────────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────┘

┌─────────────────────────────────────────────────────────────────┐
│                    Plugin Structure                              │
│                                                                  │
│  ┌──────────────┐    ┌──────────────┐    ┌──────────────────┐  │
│  │   webapp.json │    │  backend.py  │    │     app.js       │  │
│  │              │    │              │    │                  │  │
│  │  • chart     │    │  • /get-tasks│    │  • Fetch data    │  │
│  │    config    │───▶│  • /get-config│───▶│  • Init Gantt    │  │
│  │  • leftBar   │    │              │    │  • Handle events │  │
│  │    Params    │    │              │    │                  │  │
│  └──────────────┘    └──────┬───────┘    └────────┬─────────┘  │
│                             │                      │            │
│                             ▼                      ▼            │
│  ┌──────────────┐    ┌─────────────┐    ┌──────────────────┐   │
│  │  python-lib  │    │  Dataiku    │    │  Frappe Gantt    │   │
│  │  ganttchart/ │◄───│  Dataset    │    │  (bundled)       │   │
│  └──────────────┘    └─────────────┘    └──────────────────┘   │
└─────────────────────────────────────────────────────────────────┘

3.3 Business Logic Modules

Module	Purpose
`date_parser.py`	Parse dates from multiple formats to ISO
`color_mapper.py`	Map categorical values to CSS color classes
`dependency_validator.py`	Detect and break circular dependencies (DFS)
`task_transformer.py`	Orchestrate DataFrame → Frappe Gantt task transformation

4. Data Contracts

4.1 Task Object Schema

The backend transforms dataset rows into this JSON structure:

interface Task {
  id: string;           // Required: Unique identifier
  name: string;         // Required: Display label (or ID if name not configured)
  start: string;        // Required: ISO date "YYYY-MM-DD"
  end: string;          // Required: ISO date "YYYY-MM-DD"
  progress?: number;    // Optional: 0-100
  dependencies?: string;// Optional: Comma-separated task IDs
  custom_class?: string;// Optional: CSS class for styling
}

4.2 API Endpoints

GET /get-tasks

Returns task data transformed from the dataset.

Query Parameters:

config (JSON string): Webapp configuration
filters (JSON string): Dataiku filters from UI

Response (Success):

{
  "tasks": [Task, ...],
  "metadata": {
    "totalRows": 1500,
    "displayedRows": 1000,
    "skippedRows": 500,
    "skipReasons": {
      "invalid_dates": 400,
      "start_after_end": 100
    },
    "warnings": ["Circular dependency: A→B→A broken"]
  },
  "colorMapping": {
    "Dev": "bar-blue",
    "QA": "bar-green"
  }
}

Response (Error):

{
  "error": {
    "code": "COLUMN_NOT_FOUND",
    "message": "Column 'start_date' not found in dataset",
    "details": {"column": "start_date"}
  }
}

GET /get-config

Returns Frappe Gantt configuration derived from webapp config.

Response:

{
  "view_mode": "Week",
  "view_mode_select": true,
  "bar_height": 30,
  "bar_corner_radius": 3,
  "column_width": 45,
  "padding": 18,
  "readonly": true,
  "popup_on": "click",
  "today_button": true,
  "scroll_to": "today",
  "holidays": {"var(--g-weekend-highlight-color)": "weekend"}
}

4.3 Error Codes

Code	Meaning	User Action
`DATASET_NOT_SPECIFIED`	No dataset selected	Select a dataset
`DATASET_NOT_FOUND`	Dataset doesn't exist or no access	Check permissions
`EMPTY_DATASET`	Dataset is empty or all rows filtered	Add data or adjust filters
`COLUMN_NOT_FOUND`	Configured column missing	Re-select columns
`NO_VALID_TASKS`	All rows had invalid data	Verify data quality
`INVALID_CONFIGURATION`	Invalid config parameters	Check parameter values
`INTERNAL_ERROR`	Unexpected exception	Check logs, report bug

5. Configuration Parameters

5.1 Data Column Parameters (7)

Parameter	Type	Mandatory	Purpose
`idColumn`	DATASET_COLUMN	Yes	Unique task identifier
`nameColumn`	DATASET_COLUMN	No	Task display name (falls back to ID)
`startColumn`	DATASET_COLUMN	Yes	Task start date
`endColumn`	DATASET_COLUMN	Yes	Task end date
`progressColumn`	DATASET_COLUMN	No	Completion percentage (0-100)
`dependenciesColumn`	DATASET_COLUMN	No	Comma-separated predecessor IDs
`colorColumn`	DATASET_COLUMN	No	Categorical column for color coding

5.2 View Settings (3)

Parameter	Type	Default	Options
`viewMode`	SELECT	Week	Hour, Quarter Day, Half Day, Day, Week, Month, Year
`viewModeSelect`	BOOLEAN	true	Show dropdown to change view
`scrollTo`	SELECT	today	today, start, end

5.3 Appearance (4)

Parameter	Type	Default	Range
`barHeight`	INT	30	15-60
`barCornerRadius`	INT	3	0-15
`columnWidth`	INT	45	20-100
`padding`	INT	18	5-40

5.4 Behavior (4)

Parameter	Type	Default	Purpose
`readonly`	BOOLEAN	true	Disable all editing
`popupOn`	SELECT	click	click, hover
`todayButton`	BOOLEAN	true	Show "Today" navigation
`highlightWeekends`	BOOLEAN	true	Visual weekend distinction

5.5 Performance (1)

Parameter	Type	Default	Purpose
`maxTasks`	INT	1000	Limit rows for performance (0 = unlimited)

Total: 19 parameters (currently all in left sidebar)

6. Edge Cases & Error Handling

6.1 Data Quality Issues

Scenario	Behavior
Empty dataset	Error: "Dataset is empty or all rows filtered out"
No valid date rows	Error with count of skipped rows and reasons
Start > End date	Skip row, include in skipReasons
Null/empty task name	Use Task ID as fallback, or generate "Task {idx}"
Null task ID	Generate ID: "task_{row_idx}"
Duplicate task IDs	Append suffix: "id_1", "id_2"
Invalid dependency reference	Remove invalid reference, keep valid ones
Circular dependencies	Detect via DFS, break cycle, log warning
Progress outside [0,100]	Clamp to range (negative→0, >100→100)
Invalid progress value	Omit progress field for that task

6.2 Date Parsing

Supported input formats (auto-detected):

ISO 8601: 2024-01-15, 2024-01-15T00:00:00
Pandas Timestamp objects
Python datetime objects
Unix timestamps (seconds, within 1970-2100 range)
Various string formats via pd.to_datetime

Output: Always YYYY-MM-DD

7. File Structure

gantt-chart/
├── plugin.json                    # Plugin metadata (v0.0.2)
├── CHANGELOG.md                   # Version history
├── plugin-spec.md                 # This document
├── README.md                      # User documentation
│
├── python-lib/ganttchart/         # Business logic (DSS-independent)
│   ├── __init__.py
│   ├── date_parser.py             # Date parsing strategies
│   ├── color_mapper.py            # Category → color mapping
│   ├── dependency_validator.py    # Cycle detection (DFS)
│   └── task_transformer.py        # Main transformation orchestrator
│
├── webapps/gantt-chart/
│   ├── webapp.json                # Chart configuration (19 params)
│   ├── backend.py                 # Flask endpoints
│   └── body.html                  # HTML container
│
├── resource/
│   ├── frappe-gantt.umd.js        # Bundled library (~47KB)
│   ├── frappe-gantt.css           # Bundled styles (~6KB)
│   ├── frappe-gantt.es.js         # ES module version
│   ├── license.txt                # MIT license (Frappe Gantt)
│   └── webapp/
│       ├── app.js                 # Frontend logic
│       ├── style.css              # Component styles + colors
│       └── dku-helpers.js         # Backend communication helper
│
├── tests/python/unit/             # 77 unit tests
│   ├── conftest.py                # Test fixtures
│   ├── test_date_parser.py        # 28 tests
│   ├── test_color_mapper.py       # 18 tests
│   ├── test_dependency_validator.py # 16 tests
│   └── test_task_transformer.py   # 15 tests
│
├── code-env/python/spec/
│   └── requirements.txt           # pandas, numpy
│
└── plan/                          # Development documentation
    ├── releases/                  # Release notes per version
    └── *.md                       # Phase guides

8. Testing

8.1 Unit Test Coverage

Module	Tests	Status
date_parser	28	Passing
color_mapper	18	Passing
dependency_validator	16	Passing
task_transformer	15	Passing
Total	77	100%

8.2 Run Tests

cd /opt/dataiku/dss_design/plugins/dev/gantt-chart
PYTHONPATH=python-lib:$PYTHONPATH python3 -m pytest tests/python/unit/ -v

9. Implementation Status

9.1 Completed Features

9.2 Not Implemented

9.3 Known Issues

Priority	Issue	Status
High	OOM risk with large datasets (full DataFrame load)	Open
Medium	In-memory filtering (should push to data engine)	Open
Low	pandas `infer_datetime_format` deprecation warning	Open

Appendix A: Color Palette

.bar-blue    { fill: #3498db; }  /* Primary */
.bar-green   { fill: #2ecc71; }  /* Success */
.bar-orange  { fill: #e67e22; }  /* Warning */
.bar-purple  { fill: #9b59b6; }  /* Info */
.bar-red     { fill: #e74c3c; }  /* Danger */
.bar-teal    { fill: #1abc9c; }
.bar-pink    { fill: #e91e63; }
.bar-indigo  { fill: #3f51b5; }
.bar-cyan    { fill: #00bcd4; }
.bar-amber   { fill: #ffc107; }
.bar-lime    { fill: #cddc39; }
.bar-gray    { fill: #607d8b; }  /* Default for null/NaN */

Cycles for >12 categories.

Appendix B: Frappe Gantt Options Mapping

Frappe Option	Plugin Parameter	Implemented
`view_mode`	`viewMode`	Yes
`view_mode_select`	`viewModeSelect`	Yes
`bar_height`	`barHeight`	Yes
`bar_corner_radius`	`barCornerRadius`	Yes
`column_width`	`columnWidth`	Yes
`padding`	`padding`	Yes
`readonly`	`readonly`	Yes
`popup_on`	`popupOn`	Yes
`today_button`	`todayButton`	Yes
`scroll_to`	`scrollTo`	Yes
`holidays`	Derived from `highlightWeekends`	Yes
`language`	—	No
`lines`	—	No
`date_format`	—	No
`show_expected_progress`	—	No

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Dataiku Gantt Chart Plugin

Technical Specification

1. Overview

1.1 Purpose

1.2 Capabilities

1.3 Scope Boundaries

1.4 Performance

2. User Stories

3. Architecture

3.1 Plugin Type

3.2 Component Diagram

3.3 Business Logic Modules

4. Data Contracts

4.1 Task Object Schema

4.2 API Endpoints

GET /get-tasks

GET /get-config

4.3 Error Codes

5. Configuration Parameters

5.1 Data Column Parameters (7)

5.2 View Settings (3)

5.3 Appearance (4)

5.4 Behavior (4)

5.5 Performance (1)

6. Edge Cases & Error Handling

6.1 Data Quality Issues

6.2 Date Parsing

7. File Structure

8. Testing

8.1 Unit Test Coverage

8.2 Run Tests

9. Implementation Status

9.1 Completed Features

9.2 Not Implemented

9.3 Known Issues

Appendix A: Color Palette

Appendix B: Frappe Gantt Options Mapping

FilesExpand file tree

plugin-spec.md

Latest commit

History

plugin-spec.md

File metadata and controls

Dataiku Gantt Chart Plugin

Technical Specification

1. Overview

1.1 Purpose

1.2 Capabilities

1.3 Scope Boundaries

1.4 Performance

2. User Stories

3. Architecture

3.1 Plugin Type

3.2 Component Diagram

3.3 Business Logic Modules

4. Data Contracts

4.1 Task Object Schema

4.2 API Endpoints

GET /get-tasks

GET /get-config

4.3 Error Codes

5. Configuration Parameters

5.1 Data Column Parameters (7)

5.2 View Settings (3)

5.3 Appearance (4)

5.4 Behavior (4)

5.5 Performance (1)

6. Edge Cases & Error Handling

6.1 Data Quality Issues

6.2 Date Parsing

7. File Structure

8. Testing

8.1 Unit Test Coverage

8.2 Run Tests

9. Implementation Status

9.1 Completed Features

9.2 Not Implemented

9.3 Known Issues

Appendix A: Color Palette

Appendix B: Frappe Gantt Options Mapping