Frappe Report Expert

This skill provides comprehensive guidance for working with Frappe reports, their structure, creation workflow, and best practices.

Overview

Frappe provides a powerful reporting framework with multiple report types for different use cases:

• •Report Builder: Visual report builder without code (uses DocType fields)
• •Query Report: SQL-based reports with direct database queries
• •Script Report: Python-based reports with full programmatic control (most flexible)
• •Custom Report: Customized version of an existing report

This skill focuses primarily on Script Reports as they are the most commonly created programmatically and offer the most flexibility.

Quick Reference

Report Types Comparison

Standard Script Report Structure

Every Script Report consists of three files:

{module}/report/{report_name}/
├── __init__.py (empty)
├── {report_name}.json (metadata)
├── {report_name}.py (execute function)
└── {report_name}.js (filters and client-side logic)

Core Concepts

The Execute Function

The Python file must contain an execute(filters=None) function:

def execute(filters=None):
    columns, data = [], []
    # Your logic here
    return columns, data

Return value:

• •Returns a tuple: (columns, data) or extended (columns, data, message, chart, report_summary, skip_total_row)
• •columns: List of column definitions
• •data: List of rows (each row is a list or dict)

Column Format

Columns can be defined as strings or dictionaries:

String format:

columns = [
    "Name:Data:150",
    "Amount:Currency:120",
    "Customer:Link/Customer:200"
]

Dictionary format:

columns = [
    {
        "label": "Name",
        "fieldname": "name",
        "fieldtype": "Data",
        "width": 150
    }
]

Common fieldtypes:

• •Text: Data, Small Text, Text, Long Text, Text Editor, HTML Editor, Markdown Editor, Code
• •Numeric: Int, Long Int, Float, Currency, Percent
• •Date/Time: Date, Datetime, Time, Duration
• •Relationships: Link (requires options), Dynamic Link
• •Boolean: Check
• •Special: Attach, Attach Image, Signature, Color, Barcode, Rating, Icon, Geolocation, Phone, Autocomplete, JSON, Password, Read Only

See references/column-fieldtypes.md for complete field type reference.

Data Format

Data rows can be lists or dictionaries:

List format (matches column order):

data = [
    ["ID-001", 1000, "Customer A"],
    ["ID-002", 2000, "Customer B"],
]

Dictionary format (uses fieldnames):

data = [
    {"name": "ID-001", "amount": 1000, "customer": "Customer A"},
    {"name": "ID-002", "amount": 2000, "customer": "Customer B"},
]

Filters

Filters are defined in the JavaScript file:

frappe.query_reports["Report Name"] = {
    filters: [
        {
            fieldname: "company",
            label: __("Company"),
            fieldtype: "Link",
            options: "Company",
            reqd: 1
        },
        {
            fieldname: "from_date",
            label: __("From Date"),
            fieldtype: "Date",
            default: frappe.datetime.add_months(frappe.datetime.get_today(), -1)
        }
    ]
};

See references/filter-fieldtypes.md for complete filter reference.

Creating a Script Report

Quick Start

1. •Create Report via Desk: Navigate to Report DocType, create new with type "Script Report"
2. •Implement Execute Function: Edit the generated .py file with your logic
3. •Define Filters: Edit the generated .js file with filter definitions
4. •Test: Run bench migrate and navigate to /app/query-report/Your Report

See references/report-creation-workflow.md for detailed step-by-step guide.

Common Use Cases

Simple List Report

def execute(filters=None):
    return get_columns(), frappe.get_list(
        "DocType",
        fields=["name", "status", "amount"],
        filters=filters
    )

def get_columns():
    return [
        "Name:Link/DocType:150",
        "Status:Data:100",
        "Amount:Currency:120"
    ]

Report with Calculations

Add calculated fields in Python before returning data.

Grouped/Hierarchical Reports

Use indent field and return tree structure flag.

See references/script-report-examples.md for complete working examples.

Advanced Features

• •Charts: Return chart configuration as 4th element
• •Report Summary: Return summary metrics as 5th element
• •Custom Buttons: Add buttons in JavaScript onload function
• •Tree View: Enable hierarchical display with indent levels
• •Permissions: Use frappe.only_for() or frappe.has_permission()
• •Performance: Cache results, use proper indexing, limit data

See references/advanced-features.md for detailed documentation.

Reference Files

For detailed information on specific topics:

• •report-creation-workflow.md - Step-by-step guide to creating Script Reports
• •script-report-examples.md - Complete working examples of various report patterns
• •column-fieldtypes.md - All available column fieldtypes
• •column-fieldproperties.md - Column properties and formatting options
• •filter-fieldtypes.md - All available filter types
• •filter-fieldproperties.md - Filter properties and configuration options
• •advanced-features.md - Charts, summaries, custom buttons, tree reports, and optimization

Best Practices

Code Organization:

• •Use helper functions: get_columns(), get_data(), get_conditions()
• •Handle None filters: Check filters.get(key) not filters[key]

Performance:

• •Filter early in SQL WHERE clause
• •Use proper database indexes
• •Add LIMIT for large datasets
• •Cache expensive operations

Security:

• •Validate permissions with frappe.only_for() or frappe.has_permission()
• •Sanitize inputs in SQL queries
• •Use parameterized queries

User Experience:

• •Provide sensible default filter values
• •Use appropriate column widths
• •Add translations with _() or __()
• •Sort data logically

Troubleshooting

Report not showing up:

• •Check if report is disabled
• •Verify user has required role
• •Run bench migrate to sync

Data not displaying correctly:

• •Verify column count matches data row length
• •Check fieldtype matches data type

Filter not working:

• •Check fieldname matches in JS and Python
• •Verify filter value is passed correctly

See documentation for common issues and solutions.