nicetable 0.7.0

A clean and elegant way to print text tables in Python with minimal boilerplate code

nicetable

• A clean and elegant way to print text tables in Python with minimal boilerplate code.
• Built with modern Python (including type annotations) and has an extensive test suite. Requires Python 3.6 and up.

Quickstart

'NiceTable' object is printable. In its simplest form, you just pass your data object to the constructor:

from nicetable.nicetable import NiceTable

input = [{"name": "Jones Green", "height_cm": 98.8, "shirt": "XL"},
         {"name": "Jill",        "height_cm": 175,   "birth_year": 1956}]
print(NiceTable(input))

Output:

+---------------+-------------+---------+--------------+
|  name         |  height_cm  |  shirt  |  birth_year  |
+---------------+-------------+---------+--------------+
|  Jones Green  |       98.8  |  XL     |        None  |
|  Jill         |      175.0  |  None   |        1956  |
+---------------+-------------+---------+--------------+

Note that:

1. The input is a list of dicts. A column was generated for each unique key in those dicts.
2. String columns are by default left adjusted, and their column width is set automatically by the longest value.
3. Numeric columns are nicely well-aligned by the digit to the right (see the height_cm column).

You can specify a different layout as the second parameter and pass other formatting options by name.
You can also use a dot notation to specify column-level options (by column name or column position).
For example, printing as a pipe-delimited CSV, or printing as a regular CSV, without an header line, when None values are printed as 'N/A' only for the 'shirt' column:

from nicetable.nicetable import NiceTable

input = [{"name": "Jones Green", "height_cm": 98.8, "shirt": "XL"},
         {"name": "Jill",        "height_cm": 175,   "birth_year": 1956}]

print(NiceTable(input, 'csv', sep_vertical='|'))
print(NiceTable(input, 'csv', header=False).set_col_options('shirt', none_string='N/A'))

Output:

name|height_cm|shirt|birth_year
Jones Green|167.8|XL|None
Jill|175|None|1956

Jones Green,167.8,XL,None
Jill,175,N/A,1956

Working with different input types and column names

List of lists / List of tuples

These inputs are interpreted as list of rows, each with a list / tuple of columns values.

• if you DO NOT specify column names, they will be assigned automatically, as 'C001', 'C002' etc:

from nicetable.nicetable import NiceTable

input = [[1], (1,2,3), [1,3,5,7,9]]
print(NiceTable(input))

Output:

+--------+--------+--------+--------+--------+
|  c001  |  c002  |  c003  |  c004  |  c005  |
+--------+--------+--------+--------+--------+
|     1  |  None  |  None  |  None  |  None  |
|     1  |     2  |     3  |  None  |  None  |
|     1  |     3  |     5  |     7  |     9  |
+--------+--------+--------+--------+--------+

• If you DO specify a list of column names, those will be used instead of the auto-generated names.
  The next example uses the function NiceTable.builtin_layouts() that returns a list of lists:

from nicetable.nicetable import NiceTable

print(NiceTable(NiceTable.builtin_layouts(), col_names=['Layout', 'Description']))

Output:

+-----------+------------------------------------------------------------------------------------------------------+
|  Layout   |  Description                                                                                         |
+-----------+------------------------------------------------------------------------------------------------------+
|  csv      |  comma-separated values with a one-line header.                                                      |
|  default  |  fixed-width table with data auto-alignment.                                                         |
|  grep     |  tab-separated values with no header. Great for CLI output, easily post-processed by cut, grep etc.  |
|  md       |  for tables inside Markdown(.md) files, using the GFM table extension. Ex: README.md on github.      |
|  tsv      |  tab-separated values with a one-line header.                                                        |
+-----------+------------------------------------------------------------------------------------------------------+

List of dicts

This input is interpreted as list of rows, each with a dict of {column name : column value} pairs.

• If you DO NOT specify column names, they will be collected from the input, as in the first example:

from nicetable.nicetable import NiceTable

input = [{"name": "Jones Green", "height_cm": 98.8, "shirt": "XL"},
        {"name": "Jill",        "height_cm": 175,   "birth_year": 1956}]
print(NiceTable(input))

Output:

+---------------+-------------+---------+--------------+
|  name         |  height_cm  |  shirt  |  birth_year  |
+---------------+-------------+---------+--------------+
|  Jones Green  |       98.8  |  XL     |        None  |
|  Jill         |      175.0  |  None   |        1956  |
+---------------+-------------+---------+--------------+

• If you DO specify a list of column names, ONLY THOSE COLUMNS WILL BE COLLECTED.
  For example, collecting only three columns, and setting a specific column order:

from nicetable.nicetable import NiceTable

input = [{"name": "Jones Green", "height_cm": 98.8, "shirt": "XL"},
        {"name": "Jill",        "height_cm": 175,   "birth_year": 1956}]
print(NiceTable(input, col_names=['name', 'birth_year', 'height_cm']))

Output:

+---------------+--------------+-------------+
|  name         |  birth_year  |  height_cm  |
+---------------+--------------+-------------+
|  Jones Green  |        None  |       98.8  |
|  Jill         |        1956  |      175.0  |
+---------------+--------------+-------------+

• If you want to collect all columns, but provide them a new name, use the rename_columns() function.

from nicetable.nicetable import NiceTable

input = [{"name": "Jones Green", "height_cm": 98.8, "shirt": "XL"},
        {"name": "Jill",        "height_cm": 175,   "birth_year": 1956}]
print(NiceTable(input).rename_columns(['Name', 'Height(cm)', 'Shirt Size', 'Year of Birth']))

Output:

+---------------+--------------+--------------+-----------------+
|  Name         |  Height(cm)  |  Shirt Size  |  Year of Birth  |
+---------------+--------------+--------------+-----------------+
|  Jones Green  |        98.8  |  XL          |           None  |
|  Jill         |       175.0  |  None        |           1956  |
+---------------+--------------+--------------+-----------------+

Fine-grained NiceTable control

Instead of creating a NiceTable object inside a print() statement, you can alternatively:

1. Create a standalone NiceTable object, specifying a list of column names.
2. Populate it iteratively with the append() function, passing a list, a tuple or a dict, representing a new row.
3. Print it multiple times with different formatting.

This example uses the string NiceTable.SAMPLE_JSON, parses it as JSON, and chery-pick four columns:

import json
from nicetable.nicetable import NiceTable

out = NiceTable(col_names=['Name', 'Type', 'Height(cm)', 'Weight(kg)'])
for pokemon in json.loads(NiceTable.SAMPLE_JSON):
    out.append([pokemon['name'], pokemon['type'], pokemon['height'], pokemon['weight']])

print(out)
out.layout = 'md'
print(out)

Output:

+-------------+----------------+--------------+--------------+
|  Name       |  Type          |  Height(cm)  |  Weight(kg)  |
+-------------+----------------+--------------+--------------+
|  Bulbasaur  |  Grass/Poison  |          70  |       6.901  |
|  Pikachu    |  Electric      |          40  |       6.100  |
|  Mewtwo     |  Psychic       |         200  |     122.000  |
+-------------+----------------+--------------+--------------+

|  Name       |  Type          |  Height(cm)  |  Weight(kg)  |
|-------------|----------------|--------------|--------------|
|  Bulbasaur  |  Grass/Poison  |          70  |       6.901  |
|  Pikachu    |  Electric      |          40  |       6.100  |
|  Mewtwo     |  Psychic       |         200  |     122.000  |

Table-level settings

Below is the list of the table-level settings, which you can use in the constructor, or set on an existing NiceTable object:

Setting	Type	Default	Description
header	bool	1	whether the table header will be printed
header_sepline	bool	1	if the header is printed, whether a sepline will be printed after it
header_adjust	str	left	adjust of the column names, one of: ['left', 'center', 'right', 'compact']
sep_vertical	str	|	a vertical separator string
sep_horizontal	str	-	a horizontal separator string
sep_cross	str	+	a crossing separator string (where vertical and horizontal separators meet)
border_top	bool	1	whether the table top border will be printed
border_bottom	bool	1	whether the table bottom border will be printed
border_left	bool	1	whether the table left border will be printed
border_right	bool	1	whether the table right border will be printed
cell_adjust	str	auto	adjust of the values, one of: ['auto', 'left', 'center', 'right', 'compact', 'strict_left', 'strict_center', 'strict_right']
cell_spacing	int	2	number of spaces to add to each side of a value
value_min_len	int	1	minimal string length of a value. Shorter values will be space-padded
value_max_len	int	9999	maximum string length of a value
value_too_long_policy	str	wrap	handling of a string longer than value_max_len, one of: ['truncate', 'wrap']
value_newline_replace	str	None	if set, replace newlines in string value with this
value_none_string	str	None	string representation of the None value
value_escape_type	str	ignore	handling of sep_vertical inside a value, one of: ['remove', 'replace', 'prefix', 'ignore']
value_escape_char	str	\	a string to replace or prefix sep_vertical, based on value_escape_type
value_func	function	None	a function to pre-process the value before any other settings apply

The table above was generated from NiceTable.FORMATTING_SETTINGS, using the md layout:

from nicetable.nicetable import NiceTable

print(NiceTable(NiceTable.FORMATTING_SETTINGS,
                'md', 
                ['Setting', 'Type', 'Default', 'Description']))

Column-level settings

The set_col_options() function sets allows you to set the following settings at the column-level:

Parameter	Meaning
adjust	overrides the table-wide cell_adjust
max_len	overrides the table-wide value_max_len
newline_replace	overrides the table-wide value_newline_replace
none_string	overrides the table-wide value_none_string
func	overrides the table-wide value_func

This function accepts either a column name or a column position for the first parameter. For example:

import json
from nicetable.nicetable import NiceTable

out = NiceTable(json.loads(NiceTable.SAMPLE_JSON))
out.rename_columns(['ID','Name', 'Type', 'Height(cm)', ' Weight(kg)'])
# set the second column options by position (column positions starts from zero)
out.set_col_options(1, adjust='center')
# set the third column options by column name
out.set_col_options('Type',
                    func=lambda x: x.lower() if x != 'Electric' else None,
                    none_string='N/A')
print(out)

Output:

+-------+-------------+----------------+--------------+---------------+
|  ID   |  Name       |  Type          |  Height(cm)  |   Weight(kg)  |
+-------+-------------+----------------+--------------+---------------+
|  001  |  Bulbasaur  |  grass/poison  |          70  |        6.901  |
|  025  |   Pikachu   |  N/A           |          40  |        6.100  |
|  150  |    Mewtwo   |  psychic       |         200  |      122.000  |
+-------+-------------+----------------+--------------+---------------+

Cell adjustment

• Cell contents can be adjusted left, center or right, and are space-padded to the width of the longest value in the column (see also next section on wrapping).
  Alternatively, cell contents can be kept as-is with compact adjustment, though it means that the table vertical lines will not align (this is used in some layouts such as csv).
• The default adjustment is auto, meaning that numeric columns (those with only numbers or None values) are adjusted right, and non-numeric columns are adjusted left.
• Numeric columns automatically well-aligned, meaning all their ones digit are printed in the same position.
  To print them as strings, add a strict_ prefix to the adjust, like strict_left. For example:

+-----------------+-------------------+------------------+---------------+-----------------+----------------+
|  standard left  |  standard center  |  standard right  |  strict_left  |  strict_center  |  strict_right  |
+-----------------+-------------------+------------------+---------------+-----------------+----------------+
|    6.901        |        6.901      |           6.901  |  6.901        |      6.901      |         6.901  |
|    6.000        |        6.000      |           6.000  |  6            |        6        |             6  |
|    1.000        |        1.000      |           1.000  |  1            |        1        |             1  |
|  122.000        |      122.000      |         122.000  |  122          |       122       |           122  |
+-----------------+-------------------+------------------+---------------+-----------------+----------------+

The example above uses long column names on purpose, otherwise left, center and right would look the same, as all the numbers in each column have the same fixed width (based on their longest column value).

Text wrapping and newlines

NiceTable supports handling long values and newlines in both column names and cell values.

Text wrapping

When a value is longer than value_max_len, it handled by a value_too_long_policy policy.
The default policy is wrap, which means the value will be broken to multiple lines every value_max_len characters.
Alternatively, specify the truncate policy to have to values truncated.
The following examples demonstrates the two policies:

from nicetable.nicetable import NiceTable

out = NiceTable(col_names=['Code', 'Product Description(Long)'])
out.append([1, 'Boeing 777. Batteries not included. May contain nuts.'])
out.append([2, 'Sack of sand'])
print(out)
out.value_max_len = 19
print(out)
out.value_too_long_policy = 'truncate'
print(out)

Output:

+--------+---------------------------------------------------------+
|  Code  |  Product Description(Long)                              |
+--------+---------------------------------------------------------+
|     1  |  Boeing 777. Batteries not included. May contain nuts.  |
|     2  |  Sack of sand                                           |
+--------+---------------------------------------------------------+

+--------+-----------------------+
|  Code  |  Product Description  |
|        |  (Long)               |
+--------+-----------------------+
|     1  |  Boeing 777. Batteri  |
|        |  es not included. Ma  |
|        |  y contain nuts.      |
|     2  |  Sack of sand         |
+--------+-----------------------+

+--------+-----------------------+
|  Code  |  Product Description  |
+--------+-----------------------+
|     1  |  Boeing 777. Batteri  |
|     2  |  Sack of sand         |
+--------+-----------------------+

Newlines

When newlines are encountered in a column name or a value, they by default cause the text to wrap. Alternatively, you can ask that newlines will be replaced, by setting value_newline_replace to an alternative string (default is None).
The following example first shows the default behavior, and than shows replacing newlines with the string \n:

from nicetable.nicetable import NiceTable

out = NiceTable(col_names=['Code', 'Product Description\n(Long)']) \
    .append([1, 'Boeing 777\nBatteries not included.\nMay contain nuts.']) \
    .append([2, 'Sack of sand'])
print(out)
out.value_newline_replace = '\\n'
print(out)

Output:

+--------+---------------------------+
|  Code  |  Product Description      |
|        |  (Long)                   |
+--------+---------------------------+
|     1  |  Boeing 777               |
|        |  Batteries not included.  |
|        |  May contain nuts.        |
|     2  |  Sack of sand             |
+--------+---------------------------+

+--------+----------------------------------------------------------+
|  Code  |  Product Description\n(Long)                             |
+--------+----------------------------------------------------------+
|     1  |  Boeing 777\nBatteries not included.\nMay contain nuts.  |
|     2  |  Sack of sand                                            |
+--------+----------------------------------------------------------+

Escaping

The values in different columns of the same row are separated by the vertical separator string (default is |, set by the sep_vertical property).
What happens if the content of a cell contains that string? It might be irrelevant if the output is just viewed by a person, but it might matter if the string output will be processed by another program (for example, for the CSV layout).
There are four supported behaviors you can choose from, if the one set by the layout you picked is not appropriate:

1. ignore: no special handling of the vertical separator in a a cell, it is printed as is. This is the default escaping behavior.
2. remove: the vertical separator is removed.
   This is set by the csv layout and its derivatives (tsv and grep layouts).
3. prefix: the vertical separator is prefixed by another string, controlled by value_escape_char.
   This is set by the md layout, which uses \ as a prefix.
4. replace: the vertical separator is prefixed by another string, controlled by value_escape_char.

Others

get_column(col)
returns a List of the column values.

Adding a custom layout

To add a custom layout based on the existing options, you can inherit from NiceTable and define your own layout function.
The description of your function will be incorporated in the builtin_layouts() output

from nicetable.nicetable import NiceTable

class MyNiceTable(NiceTable):
    def _layout_as_winter_columns(self) -> None:
        """Table with a winter-themed separator. Quite Ugly."""
        self.sep_vertical = '❄☂🌧☂❄'
        self.sep_cross = '❄☂🌧☂❄'
        self.sep_horizontal = 'ˣ'

print(MyNiceTable(MyNiceTable.builtin_layouts(),
                  'winter_columns',
                  ['Layout', 'Description']))

Output:

❄☂🌧☂❄ˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣ❄☂🌧☂❄ˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣ❄☂🌧☂❄
❄☂🌧☂❄  Layout          ❄☂🌧☂❄  Description                                                                                         ❄☂🌧☂❄
❄☂🌧☂❄ˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣ❄☂🌧☂❄ˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣ❄☂🌧☂❄
❄☂🌧☂❄  csv             ❄☂🌧☂❄  comma-separated values with a one-line header.                                                      ❄☂🌧☂❄
❄☂🌧☂❄  default         ❄☂🌧☂❄  fixed-width table with data auto-alignment.                                                         ❄☂🌧☂❄
❄☂🌧☂❄  grep            ❄☂🌧☂❄  tab-separated values with no header. Great for CLI output, easily post-processed by cut, grep etc.  ❄☂🌧☂❄
❄☂🌧☂❄  md              ❄☂🌧☂❄  for tables inside Markdown(.md) files, using the GFM table extension. Ex: README.md on github.      ❄☂🌧☂❄
❄☂🌧☂❄  tsv             ❄☂🌧☂❄  tab-separated values with a one-line header.                                                        ❄☂🌧☂❄
❄☂🌧☂❄  winter_columns  ❄☂🌧☂❄  Table with a winter-themed separator. Quite Ugly.                                                   ❄☂🌧☂❄
❄☂🌧☂❄ˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣ❄☂🌧☂❄ˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣˣ❄☂🌧☂❄

Note that the new layout and its description were added the output of builtin_layouts() of the new class.