Add update_column and add_column to SingleTableMetadata

.github/workflows/integration.yml

@@ -10,7 +10,7 @@ jobs:
     strategy:
       matrix:
         python-version: [3.6, 3.7, 3.8, 3.9]
-        os: [ubuntu-latest, macos-10.15, windows-latest]
+        os: [ubuntu-latest, macos-latest, windows-latest]
     steps:
     - uses: actions/checkout@v1
     - name: Set up Python ${{ matrix.python-version }}

.github/workflows/minimum.yml

@@ -10,7 +10,7 @@ jobs:
     strategy:
       matrix:
         python-version: [3.6, 3.7, 3.8, 3.9]
-        os: [ubuntu-latest, macos-10.15, windows-latest]
+        os: [ubuntu-latest, macos-latest, windows-latest]
     steps:
     - uses: actions/checkout@v1
     - name: Set up Python ${{ matrix.python-version }}

.github/workflows/readme.yml

@@ -10,7 +10,7 @@ jobs:
     strategy:
       matrix:
         python-version: [3.6, 3.7, 3.8, 3.9]
-        os: [ubuntu-latest, macos-10.15]   # skip windows bc rundoc fails
+        os: [ubuntu-latest, macos-latest]   # skip windows bc rundoc fails
     steps:
     - uses: actions/checkout@v1
     - name: Set up Python ${{ matrix.python-version }}

.github/workflows/tutorials.yml

@@ -10,7 +10,7 @@ jobs:
     strategy:
       matrix:
         python-version: [3.6, 3.7, 3.8, 3.9]
-        os: [ubuntu-latest, macos-10.15, windows-latest]
+        os: [ubuntu-latest, macos-latest, windows-latest]
     steps:
     - uses: actions/checkout@v1
     - name: Set up Python ${{ matrix.python-version }}

.github/workflows/unit.yml

@@ -10,7 +10,7 @@ jobs:
     strategy:
       matrix:
         python-version: [3.6, 3.7, 3.8, 3.9]
-        os: [ubuntu-latest, macos-10.15, windows-latest]
+        os: [ubuntu-latest, macos-latest, windows-latest]
     steps:
     - uses: actions/checkout@v1
     - name: Set up Python ${{ matrix.python-version }}

sdv/metadata/single_table.py

@@ -1,7 +1,9 @@
 """Single Table Metadata."""
 
-import copy
 import json
+import re
+from copy import deepcopy
+from datetime import datetime
 from pathlib import Path
 
 import pandas as pd
@@ -12,16 +14,92 @@
 class SingleTableMetadata:
     """Single Table Metadata class."""
 
+    _EXPECTED_KWARGS = {
+        'numerical': frozenset(['representation']),
+        'datetime': frozenset(['datetime_format']),
+        'categorical': frozenset(['order', 'order_by']),
+        'boolean': frozenset([]),
+        'text': frozenset(['regex_format']),
+    }
+
     _DTYPES_TO_SDTYPES = {
         'i': 'numerical',
         'f': 'numerical',
         'O': 'categorical',
         'b': 'boolean',
         'M': 'datetime',
     }
-    KEYS = ['columns', 'primary_key', 'alternate_keys', 'constraints', 'SCHEMA_VERSION']
+
+    _NUMERICAL_REPRESENTATIONS = frozenset([
+        'int', 'int64', 'int32', 'int16', 'int8',
+        'uint', 'uint64', 'uint32', 'uint16', 'uint8',
+        'float', 'float64', 'float32', 'float16', 'float8',
+    ])
+    _KEYS = frozenset([
+        'columns',
+        'primary_key',
+        'alternate_keys',
+        'constraints',
+        'SCHEMA_VERSION'
+    ])
     SCHEMA_VERSION = 'SINGLE_TABLE_V1'
 
+    def _validate_numerical(self, column_name, **kwargs):
+        representation = kwargs.get('representation')
+        if representation and representation not in self._NUMERICAL_REPRESENTATIONS:
+            raise ValueError(
+                f"Invalid value for 'representation' {representation} for column '{column_name}'.")
+
+    @staticmethod
+    def _validate_datetime(column_name, **kwargs):
+        datetime_format = kwargs.get('datetime_format')
+        if datetime_format:
+            try:
+                formated_date = datetime.now().strftime(datetime_format)
+            except Exception as exception:
+                raise ValueError(
+                    f"Invalid datetime format string '{datetime_format}' "
+                    f"for datetime column '{column_name}'."
+                ) from exception
+
+            matches = re.findall('(%.)|(%)', formated_date)
+            if matches:
+                raise ValueError(
+                    f"Invalid datetime format string '{datetime_format}' "
+                    f"for datetime column '{column_name}'."
+                )
+
+    @staticmethod
+    def _validate_categorical(column_name, **kwargs):
+        order = kwargs.get('order')
+        order_by = kwargs.get('order_by')
+        if order and order_by:
+            raise ValueError(
+                f"Categorical column '{column_name}' has both an 'order' and 'order_by' "
+                'attribute. Only 1 is allowed.'
+            )
+        if order_by and order_by not in ('numerical_value', 'alphabetical'):
+            raise ValueError(
+                f"Unknown ordering method '{order_by}' provided for categorical column "
+                f"'{column_name}'. Ordering method must be 'numerical_value' or 'alphabetical'."
+            )
+        if (isinstance(order, list) and not len(order)) or\
+           (not isinstance(order, list) and order is not None):
+            raise ValueError(
+                f"Invalid order value provided for categorical column '{column_name}'. "
+                "The 'order' must be a list with 1 or more elements."
+            )
+
+    @staticmethod
+    def _validate_text(column_name, **kwargs):
+        regex = kwargs.get('regex_format')
+        try:
+            re.compile(regex)
+        except Exception as exception:
+            raise ValueError(
+                f"Invalid regex format string '{regex}' for text column '{column_name}'."
+            ) from exception
+
     def __init__(self):
         self._columns = {}
         self._primary_key = None
@@ -36,6 +114,88 @@ def __init__(self):
             'SCHEMA_VERSION': self.SCHEMA_VERSION
         }
 
+    def _validate_unexpected_kwargs(self, column_name, sdtype, **kwargs):
+        expected_kwargs = self._EXPECTED_KWARGS.get(sdtype, ['pii'])
+        unexpected_kwargs = set(list(kwargs)) - set(expected_kwargs)
+        if unexpected_kwargs:
+            unexpected_kwargs = list(unexpected_kwargs)
+            unexpected_kwargs.sort()
+            unexpected_kwargs = ', '.join(unexpected_kwargs)
+            raise ValueError(
+                f"Invalid values '({unexpected_kwargs})' for {sdtype} column '{column_name}'.")
+
+    def _validate_column_exists(self, column_name):
+        if column_name not in self._columns:
+            raise ValueError(
+                f"Column name ('{column_name}') does not exist in the table. "
+                "Use 'add_column' to add new column."
+            )
+
+    def _validate_column(self, column_name, sdtype, **kwargs):
+        self._validate_unexpected_kwargs(column_name, sdtype, **kwargs)
+        if sdtype == 'categorical':
+            self._validate_categorical(column_name, **kwargs)
+        elif sdtype == 'numerical':
+            self._validate_numerical(column_name, **kwargs)
+        elif sdtype == 'datetime':
+            self._validate_datetime(column_name, **kwargs)
+        elif sdtype == 'text':
+            self._validate_text(column_name, **kwargs)
+
+    def add_column(self, column_name, **kwargs):
+        """Add a column to the ``SingleTableMetadata``.
+
+        Args:
+            column_name (str):
+                The column name to be added.
+
+            kwargs (type):
+                Any additional key word arguments for the column, where ``sdtype`` is required.
+
+        Raises:
+            - ``ValueError`` if the column already exists.
+            - ``ValueError`` if the ``kwargs`` do not contain ``sdtype``.
+            - ``ValueError`` if the column has unexpected values or ``kwargs`` for the given
+              ``sdtype``.
+        """
+        if column_name in self._columns:
+            raise ValueError(
+                f"Column name '{column_name}' already exists. Use 'update_column' "
+                'to update an existing column.'
+            )
+
+        sdtype = kwargs.get('sdtype')
+        if sdtype is None:
+            raise ValueError(f"Please provide a 'sdtype' for column '{column_name}'.")
+
+        self._validate_column(column_name, **kwargs)
+        self._columns[column_name] = deepcopy(kwargs)
+
+    def update_column(self, column_name, **kwargs):
+        """Update an existing column in the ``SingleTableMetadata``.
+
+        Args:
+            column_name (str):
+                The column name to be updated.
+            **kwargs (type):
+                Any key word arguments that describe metadata for the column.
+
+        Raises:
+            - ``ValueError`` if the column doesn't already exist in the ``SingleTableMetadata``.
+            - ``ValueError`` if the column has unexpected values or ``kwargs`` for the current
+              ``sdtype``.
+        """
+        self._validate_column_exists(column_name)
+        _kwargs = deepcopy(kwargs)
+        if 'sdtype' in kwargs:
+            sdtype = kwargs.pop('sdtype')
+        else:
+            sdtype = self._columns[column_name]['sdtype']
+            _kwargs['sdtype'] = sdtype
+
+        self._validate_column(column_name, sdtype, **kwargs)
+        self._columns[column_name] = _kwargs
+
     def detect_from_dataframe(self, data):
         """Detect the metadata from a ``pd.DataFrame`` object.
 
@@ -99,7 +259,7 @@ def to_dict(self):
             elif value:
                 metadata[key] = value
 
-        return copy.deepcopy(metadata)
+        return deepcopy(metadata)
 
     def _set_metadata_dict(self, metadata):
         """Set a ``metadata`` dictionary to the current instance.
@@ -109,8 +269,8 @@ def _set_metadata_dict(self, metadata):
                 Python dictionary representing a ``SingleTableMetadata`` object.
         """
         self._metadata = {}
-        for key in self.KEYS:
-            value = copy.deepcopy(metadata.get(key))
+        for key in self._KEYS:
+            value = deepcopy(metadata.get(key))
             if key == 'constraints' and value:
                 value = [Constraint.from_dict(constraint_dict) for constraint_dict in value]