Table Detection

Detect and classify tables in spreadsheets.

Advanced table detection and classification utilities.

class sheetwise.smart_tables.TableType(value)

Types of tables that can be detected.

DATA_TABLE = 'data_table'

PIVOT_TABLE = 'pivot_table'

MATRIX = 'matrix'

FORM = 'form'

MIXED = 'mixed'

SPARSE = 'sparse'

class sheetwise.smart_tables.EnhancedTableRegion(top_left, bottom_right, rows, cols, confidence=1.0, table_type=TableType.DATA_TABLE, has_headers=False, header_rows=None, header_cols=None)

Extended table region with additional metadata.

table_type: TableType = 'data_table'

has_headers: bool = False

header_rows: List[int] = None

header_cols: List[int] = None

confidence: float = 1.0

property start_row: int

property end_row: int

property start_col: int

property end_col: int

__init__(top_left, bottom_right, rows, cols, confidence=1.0, table_type=TableType.DATA_TABLE, has_headers=False, header_rows=None, header_cols=None)

top_left: str

bottom_right: str

rows: range

cols: range

class sheetwise.smart_tables.SmartTableDetector(min_table_size=2, max_empty_ratio=0.7, header_detection=True)

Advanced table detection with enhanced capabilities.

This class provides utilities to: 1. Detect multiple tables in spreadsheets 2. Identify table headers and structures 3. Classify tables by type 4. Handle complex table layouts

__init__(min_table_size=2, max_empty_ratio=0.7, header_detection=True)

Initialize the detector.

Parameters:

• min_table_size (int) – Minimum number of rows/columns to consider a table

• max_empty_ratio (float) – Maximum ratio of empty cells allowed in a table

• header_detection (bool) – Whether to detect headers

detect_tables(df)

Detect multiple tables in a spreadsheet.

Parameters:

df (DataFrame) – Input DataFrame

Return type:

List[EnhancedTableRegion]

Returns:

List of detected enhanced table regions

extract_tables_to_dataframes(df)

Extract all tables from a spreadsheet into separate dataframes.

Parameters:

df (DataFrame) – Input DataFrame

Return type:

Dict[str, DataFrame]

Returns:

Dictionary mapping table names to extracted DataFrames