Python API

Tables

class RPA.Tables.Tables

Tables is a library for manipulating tabular data inside Robot Framework.

It can import data from various sources and apply different operations to it. Common use-cases are reading and writing CSV files, inspecting files in directories, or running tasks using existing Excel data.

Import types

The data a table can be created from can be of two main types:

An iterable of individual rows, like a list of lists, or list of dictionaries
A dictionary of columns, where each dictionary value is a list of values

For instance, these two input values:

data1 = [
    {"name": "Mark", "age": 58},
    {"name": "John", "age": 22},
    {"name": "Adam", "age": 67},
]

data2 = {
    "name": ["Mark", "John", "Adam"],
    "age":  [    58,     22,     67],
}

Would both result in the following table:

Index	Name	Age
0	Mark	58
1	John	22
2	Adam	67

Indexing columns and rows

Columns can be referred to in two ways: either with a unique string name or their position as an integer. Columns can be named either when the table is created, or they can be (re)named dynamically with keywords. The integer position can always be used, and it starts from zero.

For instance, a table with columns “Name”, “Age”, and “Address” would allow referring to the “Age” column with either the name “Age” or the number 1.

Rows do not have a name, but instead only have an integer index. This index also starts from zero. Keywords where rows are indexed also support negative values, which start counting backwards from the end.

For instance, in a table with five rows, the first row could be referred to with the number 0. The last row could be accessed with either 4 or -1.

Examples

Robot Framework

The Tables library can load tabular data from various other libraries and manipulate it inside Robot Framework.

*** Settings ***
Library    RPA.Tables

*** Keywords ***
Files to Table
    ${files}=    List files in directory    ${CURDIR}
    ${files}=    Create table    ${files}
    Filter table by column    ${files}    size  >=  ${1024}
    FOR    ${file}    IN    @{files}
        Log    ${file}[name]
    END
    Write table to CSV    ${files}    ${OUTPUT_DIR}${/}files.csv

Python

The library is also available directly through Python, where it is easier to handle multiple different tables or do more bespoke manipulation operations.

from RPA.Tables import Tables

library = Tables()
orders = library.read_table_from_csv(
    "orders.csv", columns=["name", "mail", "product"]
)

customers = library.group_table_by_column(rows, "mail")
for customer in customers:
    for order in customer:
        add_cart(order)
    make_order()

ROBOT_LIBRARY_DOC_FORMAT = 'REST'

ROBOT_LIBRARY_SCOPE = 'GLOBAL'

add_table_column(table: Table, name: str | None = None, values: Any | None = None)

Append a column to a table.

Parameters:

table – Table to modify
name – Name of new column
values – Value(s) for new column

The values can either be a list of values, one for each row, or one single value that is set for all rows.

Examples:

# Add empty column
Add table column    ${table}

# Add empty column with name
Add table column    ${table}    name=Home Address

# Add new column where every every row has the same value
Add table column    ${table}    name=TOS    values=${FALSE}

# Add new column where every row has a unique value
${is_first}=    Create list    ${TRUE}    ${FALSE}    ${FALSE}
Add table column    ${table}    name=IsFirst    values=${is_first}

add_table_row(table: Table, values: Any | None = None)

Append rows to a table.

Parameters:

table – Table to modify
values – Value(s) for new row

The values can either be a list of values, or a dictionary where the keys match current column names. Values for unknown keys are discarded.

It can also be a single value that is set for all columns, which is None by default.

Examples:

# Add empty row
Add table row    ${table}

# Add row where every column has the same value
Add table row    ${table}    Unknown

# Add values per column
${values}=    Create dictionary    Username=Mark    Mail=mark@robocorp.com
Add table row    ${table}    ${values}

clear_table(table: Table)

Clear table in-place, but keep columns.

Parameters:: table – Table to clear

Example:

Clear table    ${table}

copy_table(table: Table) → Table

Make a copy of a table object.

Parameters:: table – Table to copy
Returns:: Table object

${table_copy}= Copy table ${table}

Create Table object from data.

Data can be a combination of various iterable containers, e.g. list of lists, list of dicts, dict of lists.

Parameters:

data – Source data for table
trim – Remove all empty rows from the end of the worksheet, default False
columns – Names of columns (optional)

Returns:

Table object

See the main library documentation for more information about supported data types.

Example:

# Create a new table using a Dictionary of Lists
# Because of the dictionary keys the column names will be automatically set
@{Table_Data_name}=    Create List    Mark    John    Amy
@{Table_Data_age}=    Create List    ${58}    ${22}    ${67}
&{Table_Data}=    Create Dictionary
...    name=${Table_Data_name}
...    age=${Table_Data_age}
${table}=    Create Table    ${Table_Data}

export_table(table: Table, with_index: bool = False, as_list: bool = True) → List | Dict

Convert a table object into standard Python containers.

Parameters:

table – Table to convert to dict
with_index – Include index in values
as_list – Export data as list instead of dict

Returns:

A List or Dictionary that represents the table

Example:

${orders}=       Read worksheet as table    orders.xlsx
Sort table by column    ${orders}    CustomerId
${export}=       Export table    ${orders}
# The following keyword expects a dictionary:
Write as JSON    ${export}

filter_empty_rows(table: Table)

Remove all rows from a table which have only None values.

Parameters:: table – Table to filter

The filtering will be done in-place.

Example:

Filter empty rows    ${table}

filter_table_by_column(table: Table, column: int | str, operator: str, value: Any)

Remove all rows where column values don’t match the given condition.

Parameters:

table – Table to filter
column – Column to filter with
operator – Filtering operator, e.g. >, <, ==, contains
value – Value to compare column to (using operator)

See the keyword Find table rows for all supported operators and their descriptions.

The filtering will be done in-place.

Examples:

# Only accept prices that are non-zero
Filter table by column    ${table}   price  !=  ${0}

# Remove uwnanted product types
@{types}=    Create list    Unknown    Removed
Filter table by column    ${table}   product_type  not in  ${types}

filter_table_with_keyword(table: Table, name: str, *args)

Run a keyword for each row of a table, then remove all rows where the called keyword returns a falsy value.

Can be used to create custom RF keyword based filters.

Parameters:

table – Table to modify.
name – Keyword name used as filter.
args – Additional keyword arguments to be passed. (optional)

The row object will be given as the first argument to the filtering keyword.

find_table_rows(table: Table, column: int | str, operator: str, value: Any)

Find all the rows in a table which match a condition for a given column.

Parameters:

table – Table to search into.
column – Name or position of the column to compare with.
operator – Comparison operator used with every cell value on the specified column.
value – Value to compare against.

Returns:

New Table object containing all the rows matching the condition.

Supported operators:

Operator	Description
>	Cell value is larger than
<	Cell value is smaller than
>=	Cell value is larger or equal than
<=	Cell value is smaller or equal than
==	Cell value is equal to
!=	Cell value is not equal to
is	Cell value is the same object
not is	Cell value is not the same object
contains	Cell value contains given value
not contains	Cell value does not contain given value
in	Cell value is in given value
not in	Cell value is not in given value

Returns the matches as a new Table instance.

Examples:

# Find all rows where price is over 200
@{rows} =    Find table rows    ${table}    Price  >  ${200}

# Find all rows where the status does not contain "removed"
@{rows} =    Find table rows    ${table}    Status  not contains  removed

get_table_cell(table: Table, row: int | str, column: int | str) → Any

Get a cell value from a table.

Parameters:

table – Table to read from
row – Row of cell
column – Column of cell

Returns:

Cell value

Examples:

# Get the value in the first row and first column
Get table cell    ${table}    0    0

# Get the value in the last row and first column
Get table cell    ${table}   -1    0

# Get the value in the last row and last column
Get table cell    ${table}   -1    -1

# Get the value in the third row and column "Name"
Get table cell    ${table}    2    Name

get_table_column(table: Table, column: int | str) → List

Get all values for a single column in a table.

Parameters:

table – Table to read
column – Column to read

Returns:

List of the rows in the selected column

Example:

${emails}=    Get table column    ${users}    E-Mail Address

get_table_dimensions(table: Table) → Tuple[int, int]

Return table dimensions, as (rows, columns).

Parameters:: table – Table to inspect
Returns:: Two integer values that represent the number of rows and columns

Example:

${rows}  ${columns}=    Get table dimensions    ${table}
Log    Table has ${rows} rows and ${columns} columns.

get_table_row(table: Table, row: int | str, as_list: bool = False) → Dict | List

Get a single row from a table.

Parameters:

table – Table to read
row – Row to read
as_list – Return list instead of dictionary

Returns:

Dictionary or List of table row

Examples:

# returns the first row in the table
${first}=    Get table row    ${orders}

# returns the last row in the table
${last}=      Get table row    ${orders}    -1    as_list=${TRUE}

Return a new Table from a range of given Table rows.

Parameters:

table – Table to read from
start – Start index (inclusive)
start – End index (exclusive)

Returns:

Table object of the selected rows

If start is not defined, starts from the first row. If end is not defined, stops at the last row.

Examples:

# Get all rows except first five
${slice}=    Get table slice    ${table}    start=5

# Get rows at indexes 5, 6, 7, 8, and 9
${slice}=    Get table slice    ${table}    start=5    end=10

# Get all rows except last five
${slice}=    Get table slice    ${table}    end=-5

group_table_by_column(table: Table, column: int | str) → List[Table]

Group a table by column and return a list of grouped Tables.

Parameters:

table – Table to use for grouping
column – Column which is used as grouping criteria

Returns:

List of Table objects

Example:

# Groups rows of matching customers from the `customer` column
# and returns the groups or rows as Tables
@{groups}=    Group table by column    ${orders}    customer
# An example of how to use the List of Tables once returned
FOR    ${group}    IN    @{groups}
    # Process all orders for the customer at once
    Process order    ${group}
END

map_column_values(table: Table, column: int | str, name: str, *args)

Run a keyword for each cell in a given column, and replace its content with the return value.

Can be used to easily convert column types or values in-place.

Parameters:

table – Table to modify.
column – Column to modify.
name – Mapping keyword name.
args – Additional keyword arguments. (optional)

The cell value will be given as the first argument to the mapping keyword.

Examples:

# Convert all columns values to a different type
Map column values    ${table}    Price    Convert to integer

# Look up values with a custom keyword
Map column values    ${table}    User     Map user ID to name

merge_tables(*tables: Table, index: str | None = None) → Table

Create a union of two tables and their contents.

Parameters:

tables – Tables to merge
index – Column name to use as index for merge

Returns:

Table object

By default rows from all tables are appended one after the other. Optionally a column name can be given with index, which is used to merge rows together.

Example:

For instance, a name column could be used to identify unique rows and the merge operation should overwrite values instead of appending multiple copies of the same name.

Name	Price
Egg	10.0
Cheese	15.0
Ham	20.0

Name	Stock
Egg	12.0
Cheese	99.0
Ham	0.0

${products}=    Merge tables    ${prices}    ${stock}    index=Name
FOR    ${product}    IN    @{products}
    Log many
    ...    Product: ${product}[Name]
    ...    Price: ${product}[Price]
    ...    Stock: ${product}[Stock]
END

pop_table_column(table: Table, column: int | str | None = None) → Dict | List

Remove column from table and return it.

Parameters:

table – Table to modify
column – Column to remove

Returns:

Dictionary or List of the removed, popped, column

Examples:

# Remove column from table and discard it
Pop table column    ${users}   userId

# Remove column from table and iterate over it
${ids}=    Pop table column    ${users}    userId
FOR    ${id}    IN    @{ids}
    Log    User id: ${id}
END

pop_table_row(table: Table, row: int | str | None = None, as_list: bool = False) → Dict | List

Remove row from table and return it.

Parameters:

table – Table to modify
row – Row index, pops first row if none given
as_list – Return list instead of dictionary

Returns:

Dictionary or List of the removed, popped, row

Examples:

# Pop the firt row in the table and discard it
Pop table row    ${orders}

# Pop the last row in the table and store it
${row}=      Pop table row    ${data}    -1    as_list=${TRUE}

Read a CSV file as a table.

Parameters:

path – Path to CSV file
header – CSV file includes header
columns – Names of columns in resulting table
dialect – Format of CSV file
delimiters – String of possible delimiters
column_unknown – Column name for unknown fields
encoding – Text encoding for input file, uses system encoding by default

Returns:

Table object

By default attempts to deduce the CSV format and headers from a sample of the input file. If it’s unable to determine the format automatically, the dialect and header will have to be defined manually.

Builtin dialect values are excel, excel-tab, and unix, and header is boolean argument (True/False). Optionally a set of valid delimiters can be given as a string.

The columns argument can be used to override the names of columns in the resulting table. The amount of columns must match the input data.

If the source data has a header and rows have more fields than the header defines, the remaining values are put into the column given by column_unknown. By default it has the value “Unknown”.

Examples:

# Source dialect is deduced automatically
${table}=    Read table from CSV    export.csv
Log   Found columns: ${table.columns}

# Source dialect is known and given explicitly
${table}=    Read table from CSV    export-excel.csv    dialect=excel
Log   Found columns: ${table.columns}

rename_table_columns(table: Table, names: List[str | None], strict: bool = False)

Renames columns in the Table with given values. Columns with name as None will use the previous value.

Parameters:

table – Table to modify
names – List of new column names
strict – If True, raises ValueError if column lengths do not match

The renaming will be done in-place.

Examples:

# Initially set the column names
${columns}=    Create list   First  Second  Third
Rename table columns    ${table}    ${columns}
# First, Second, Third

# Update the first and second column names to Uno and Dos
${columns}=    Create list   Uno  Dos
Rename table columns    ${table}    ${columns}
# Uno, Dos, Third

set_row_as_column_names(table: Table, row: int | str)

Set existing row as names for columns.

Parameters:

table – Table to modify
row – Row to use as column names

Example:

# Set the column names based on the first row
Set row as column names    ${table}    0

set_table_cell(table: Table, row: int | str, column: int | str, value: Any)

Set a cell value in a table.

Parameters:

table – Table to modify to
row – Row of cell
column – Column of cell
value – Value to set

Examples:

# Set the value in the first row and first column to "First"
Set table cell    ${table}    0    0       First

# Set the value in the last row and first column to "Last"
Set table cell    ${table}   -1    0       Last

# Set the value in the last row and last column to "Corner"
Set table cell    ${table}   -1    -1       Corner

# Set the value in the third row and column "Name" to "Unknown"
Set table cell    ${table}    2    Name    Unknown

set_table_column(table: Table, column: int | str, values: Any)

Assign values to a column in the table.

Parameters:

table – Table to modify
column – Column to modify
values – Value(s) to set

The values can either be a list of values, one for each row, or one single value that is set for all rows.

Examples:

# Set different value for each row (sizes must match)
${ids}=    Create list    1  2  3  4  5
Set table column    ${users}    userId    ${ids}

# Set the same value for all rows
Set table column    ${users}    email     ${NONE}

set_table_row(table: Table, row: int | str, values: Any)

Assign values to a row in the table.

Parameters:

table – Table to modify
row – Row to modify
values – Value(s) to set

The values can either be a list of values, or a dictionary where the keys match current column names. Values for unknown keys are discarded.

It can also be a single value that is set for all columns.

Examples:

${columns}=  Create list     One  Two  Three
${table}=    Create table    columns=${columns}

${values}=   Create list     1  2  3
Set table row    ${table}    0    ${values}

${values}=   Create dictionary    One=1  Two=2  Three=3
Set table row    ${table}    1    ${values}

Set table row    ${table}    2    ${NONE}

sort_table_by_column(table: Table, column: int | str, ascending: bool = True)

Sort a table in-place according to column.

Parameters:

table – Table to sort
column – Column to sort with
ascending – Table sort order

Examples:

# Sorts the `order_date` column ascending
Sort table by column    ${orders}    order_date

# Sorts the `order_date` column descending
Sort table by column    ${orders}    order_date    ascending=${FALSE}

table_head(table: Table, count: int = 5, as_list: bool = False) → Table | List[List]

Return first count rows from a table.

Parameters:

table – Table to read from
count – Number of lines to read
as_list – Return list instead of Table

Returns:

Return Table object or List of the selected rows

Example:

# Get the first 10 employees
${first}=    Table head    ${employees}    10

table_tail(table: Table, count: int = 5, as_list: bool = False) → Table | List[List]

Return last count rows from a table.

Parameters:

table – Table to read from
count – Number of lines to read
as_list – Return list instead of Table

Returns:

Return Table object or List of the selected rows

Example:

# Get the last 10 orders
${latest}=    Table tail    ${orders}    10

trim_column_names(table: Table)

Remove all extraneous whitespace from column names.

Parameters:: table – Table to filter

The filtering will be done in-place.

Example:

# This example will take colums such as:
# "One", "Two ", "  Three "
# and trim them to become the below:
# "One", "Two", "Three"
Trim column names     ${table}

trim_empty_rows(table: Table)

Remove all rows from the end of a table which have only None as values.

Parameters:: table – Table to filter

The filtering will be done in-place.

Example:

Trim empty rows    ${table}

write_table_to_csv(table: Table, path: str, header: bool = True, dialect: str | Dialect = Dialect.Excel, encoding: str | None = None, delimiter: str | None = ',')

Write a table as a CSV file.

Parameters:

table – Table to write
path – Path to write to
header – Write columns as header to CSV file
dialect – The format of output CSV
encoding – Text encoding for output file, uses system encoding by default
delimiter – Delimiter character between columns

Builtin dialect values are excel, excel-tab, and unix.

Example:

${sheet}=    Read worksheet as table    orders.xlsx    header=${TRUE}
Write table to CSV    ${sheet}    output.csv