agentleFS
yuchenmiao04yuchenmiao04/My_skills/skills/matlab/matlab-map-database-objects/SKILL.md

matlab-map-database-objects

Skill8 starsChanged 5 months agoMITView on GitHub

Generates MATLAB Object Relational Mapping (ORM) code using Database Toolbox. Use when mapping MATLAB classes to database tables, reading/writing objects with ormread/ormwrite/ormupdate, defining Mappable classes, converting classes to SQL with orm2sql, or using object-oriented database workflows.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 1 is distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use when mapping MATLAB classes to relational database tables using the Object Relational Mapping (ORM) layer in Database Toolbox.

    line 12
  2. Very specific:

    ALWAYS: use class-level attribute TableName to specify the database table name.

    line 36
  3. Very specific:

    Use ColumnName when the MATLAB property name differs from the database column name.

    line 321
  4. Very specific:

    Use ColumnType for types that need explicit mapping (e.g., "date" for datetime).

    line 322
  5. Very specific:

    Use ormread with RowFilter to import only the objects you need — pushes filter to database.

    line 323
  6. Very specific:

    Use orm2sql to verify your class mapping matches the expected database schema before writing.

    line 324
  7. Very specific:

    Solution: Ensure the class inherits from database.orm.mixin.Mappable.

    line 373
  8. Very specific:

    Solution: Verify the table exists in the database using sqlfind(conn, tableName).

    line 376
  9. Very specific:

    Solution: Use orm2sql(conn, "ClassName") to inspect the generated SQL.

    line 382

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

MATLAB Object Relational Mapper

Use when mapping MATLAB classes to relational database tables using the Object Relational Mapping (ORM) layer in Database Toolbox. Defines Mappable classes with property-to-column mappings and uses ormread, ormwrite, and ormupdate for CRUD operations on objects. Available since R2023b.

When to Use This Skill

  • Mapping MATLAB classes to database tables
  • Reading database rows as MATLAB objects
  • Writing MATLAB objects to database tables
  • Updating database rows from modified objects
  • Generating SQL CREATE TABLE from a class definition
  • Object-oriented database workflows
  • User mentions keywords: ORM, object relational mapping, Mappable, ormread, ormwrite, ormupdate, orm2sql, class mapping, property mapping, object database

When NOT to Use

  • Ad-hoc queries or data exploration — use sqlread/fetch with RowFilter instead
  • Bulk imports/exports of thousands of rows — ORM is slower than sqlwrite/sqlread for bulk operations
  • MongoDB, Cassandra, or Neo4j — ORM only works with relational databases (JDBC, ODBC, native MySQL/PostgreSQL/SQLite)
  • MATLAB releases before R2023b — ORM is not available

Critical Rules

ORM Requirements

  • ALWAYS inherit from database.orm.mixin.Mappable — this is required for ORM functionality.
  • ALWAYS define at least one PrimaryKey property — ORM requires it for identity.
  • ALWAYS use class-level attribute TableName to specify the database table name.
  • ALWAYS include a constructor that handles nargin == 0 (allows preallocation by ORM).
  • ORM requires R2023b or later.

Supported Connections

  • ORM works with: JDBC, ODBC, PostgreSQL native, MySQL native, SQLite native, DuckDB native.
  • ORM does NOT work with: MongoDB (mongoc), Cassandra (apacheCassandra), or Neo4j.

Decision Framework

ScenarioUse ORMUse sqlread/sqlwrite
Object identity and business logic neededYesNo
Class-based type safety requiredYesNo
Domain validation on read/writeYesNo
Ad-hoc queries or explorationNoYes
Bulk operations (thousands of rows)NoYes (faster)
No object mapping neededNoYes

Core Concepts

Defining a Mappable Class

A Mappable class maps MATLAB properties to database columns:

classdef (TableName = "employees") Employee < database.orm.mixin.Mappable

    properties (PrimaryKey, ColumnName = "EmployeeID")
        ID int32
    end

    properties
        Name string
        Department string
        Salary double
    end

    properties (ColumnName = "HireDate", ColumnType = "date")
        StartDate datetime
    end

    methods
        function obj = Employee(id, name, dept, salary, startDate)
            if nargin ~= 0
                obj.ID = id;
                obj.Name = name;
                obj.Department = dept;
                obj.Salary = salary;
                obj.StartDate = startDate;
            end
        end

        function obj = promote(obj, raise)
            obj.Salary = obj.Salary + raise;
        end
    end
end

Property Attributes Reference

AttributePurposeExample
PrimaryKeyMarks property as the primary key (required)properties (PrimaryKey)
ColumnNameMaps property to a differently-named columnproperties (ColumnName = "EmpID")
ColumnTypeSpecifies the database column typeproperties (ColumnType = "date")
TableNameClass-level attribute for the target table nameclassdef (TableName = "employees")

Common Mistakes

% INCORRECT — class without PrimaryKey (ormwrite will error)
classdef (TableName = "employees") Employee < database.orm.mixin.Mappable
    properties
        Name string
        Dept string
    end
end
% ormwrite(conn, emp) → Error: No PrimaryKey defined

% CORRECT — PrimaryKey is required for ORM operations
classdef (TableName = "employees") Employee < database.orm.mixin.Mappable
    properties(PrimaryKey)
        EmployeeID int32
    end
    properties
        Name string
        Dept string
    end
end

% INCORRECT — class not inheriting Mappable
classdef Employee
    properties(PrimaryKey)
        EmployeeID int32
    end
end
% ormread(conn, "Employee") → Error! Not a Mappable class.

% CORRECT — must inherit from database.orm.mixin.Mappable
classdef Employee < database.orm.mixin.Mappable
    properties(PrimaryKey)
        EmployeeID int32
    end
end

% INCORRECT — constructor without nargin==0 guard
classdef (TableName = "emp") Employee < database.orm.mixin.Mappable
    methods
        function obj = Employee(id, name)
            obj.EmployeeID = id;
            obj.Name = name;
        end
    end
end
% ormread will fail because ORM needs to construct empty objects

% CORRECT — nargin==0 guard allows ORM to construct empty objects
classdef (TableName = "emp") Employee < database.orm.mixin.Mappable
    methods
        function obj = Employee(id, name)
            if nargin == 0
                return;
            end
            obj.EmployeeID = id;
            obj.Name = name;
        end
    end
end

Writing Objects with ormwrite

% Create and insert a single object
emp = Employee(1, "Alice", "Engineering", 95000, datetime(2023,3,15));
ormwrite(conn, emp);

% Create and insert an array of objects
emps = [Employee(2, "Bob", "Sales", 72000, datetime(2023,6,1)), ...
        Employee(3, "Carol", "Engineering", 105000, datetime(2022,1,10))];
ormwrite(conn, emps);

Reading Objects with ormread

% Read all objects
allEmployees = ormread(conn, "Employee");

% Read with a row filter
rf = rowfilter("Salary");
highEarners = ormread(conn, "Employee", RowFilter=rf.Salary > 90000);

% Refresh an existing object from database
emp = ormread(conn, emp);

Updating Objects with ormupdate

% Modify object in MATLAB
emp = promote(emp, 5000);

% Push changes to database
ormupdate(conn, emp);

Deleting Records

ORM does not provide an ormdelete function. Use execute for deletion:

execute(conn, "DELETE FROM employees WHERE EmployeeID = 42");

Error Handling

try
    ormwrite(conn, employeeObj);
catch ME
    if contains(ME.message, "UNIQUE") || contains(ME.message, "primary key")
        warning("Duplicate primary key. Use ormupdate instead.");
        ormupdate(conn, employeeObj);
    else
        rethrow(ME);
    end
end

Generating SQL from Class Definition

% View the CREATE TABLE SQL that corresponds to the class
sql = orm2sql(conn, "Employee");
disp(sql);
% Output: "CREATE TABLE employees (EmployeeID integer, Name text, ...)"

Complete Examples

Example 1: Full ORM Workflow — Define, Write, Read, Update

Step 1: Define the class (save as Product.m):

classdef (TableName = "products") Product < database.orm.mixin.Mappable

    properties (PrimaryKey, ColumnName = "ProductNumber")
        ID int32
    end

    properties
        Name string
        Description string
        Quantity int32
    end

    properties (ColumnName = "UnitCost")
        CostPerItem double
    end

    properties (ColumnType = "date")
        InventoryDate datetime
    end

    methods
        function obj = Product(id, name, desc, cost, qty, invDate)
            if nargin ~= 0
                obj.ID = id;
                obj.Name = name;
                obj.Description = desc;
                obj.CostPerItem = cost;
                obj.Quantity = qty;
                obj.InventoryDate = invDate;
            end
        end

        function obj = restock(obj, amount)
            obj.Quantity = obj.Quantity + amount;
            obj.InventoryDate = datetime("today");
        end
    end
end

Step 2: Use ORM operations (save as ormWorkflow.m):

(Uses the Product class defined above.)

% Connect to SQLite (no driver needed)
conn = sqlite("inventory.db", "create");

% Create and insert products
p1 = Product(1, "Widget", "Small widget", 9.99, 100, datetime(2024,1,1));
p2 = Product(2, "Gadget", "Large gadget", 29.99, 50, datetime(2024,1,1));
ormwrite(conn, [p1, p2]);

% Read all products back as objects
allProducts = ormread(conn, "Product");
disp(allProducts(1));

% Filter: find products under $15
rf = rowfilter("CostPerItem");
cheapProducts = ormread(conn, "Product", RowFilter=rf.CostPerItem < 15);

% Update: restock a product
cheapProducts(1) = restock(cheapProducts(1), 200);
ormupdate(conn, cheapProducts(1));

% Verify
refreshed = ormread(conn, cheapProducts(1));
disp(refreshed.Quantity); % Should be 300

close(conn);

Best Practices

  • ALWAYS define Mappable classes in their own .m file — one class per file (MATLAB requirement).
  • ALWAYS include a constructor that handles nargin == 0 (allows preallocation).
  • ALWAYS mark exactly one property block with PrimaryKey.
  • Use ColumnName when the MATLAB property name differs from the database column name.
  • Use ColumnType for types that need explicit mapping (e.g., "date" for datetime).
  • Use ormread with RowFilter to import only the objects you need — pushes filter to database.
  • Use orm2sql to verify your class mapping matches the expected database schema before writing.
  • Modify objects in MATLAB using class methods, then push changes with ormupdate.

Common Patterns

Pattern 1: Define-Write-Read-Update

% Define class → Product.m
p = Product(1, "Item", "Desc", 9.99, 10, datetime("today"));
ormwrite(conn, p);
p = ormread(conn, p);        % Refresh from DB
p = restock(p, 50);          % Modify in MATLAB
ormupdate(conn, p);          % Push to DB

Pattern 2: Filtered Read with Business Logic

rf = rowfilter("Quantity");
lowStock = ormread(conn, "Product", RowFilter=rf.Quantity < 10);
for i = 1:numel(lowStock)
    lowStock(i) = restock(lowStock(i), 100);
    ormupdate(conn, lowStock(i));
end

Pattern 3: Schema Verification

sql = orm2sql(conn, "MyClass");
disp(sql); % Verify CREATE TABLE matches expectations

Checklist

Before finalizing, verify:

  • Mappable class inherits from database.orm.mixin.Mappable
  • Class has TableName attribute on classdef
  • At least one property block has PrimaryKey attribute
  • Constructor handles nargin == 0 for preallocation
  • Class saved in its own .m file
  • Connection established and verified (isopen(conn))
  • Connection type is supported (JDBC, ODBC, MySQL/PostgreSQL/SQLite/DuckDB native)
  • R2023b or later (ORM not available in earlier releases)

Troubleshooting

Issue: ormwrite fails with "class is not Mappable"

  • Solution: Ensure the class inherits from database.orm.mixin.Mappable. The class definition must include < database.orm.mixin.Mappable.

Issue: ormread returns empty array

  • Solution: Verify the table exists in the database using sqlfind(conn, tableName). Verify the class TableName attribute matches the actual table name.

Issue: ormupdate doesn't change database values

  • Solution: ormupdate matches rows by primary key. Verify the object's primary key property value exists in the database. Use ormread to refresh and check.

Issue: Property-to-column mapping is incorrect

  • Solution: Use orm2sql(conn, "ClassName") to inspect the generated SQL. Verify ColumnName and ColumnType attributes match the database schema.

Issue: ORM functions not found ("Undefined function")

  • Solution: ORM requires R2023b or later. Check your release with ver('database'). For older releases, use sqlread/sqlwrite instead.

Copyright 2026 The MathWorks, Inc.


Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.