json-excel-converter 1.2.0

JSON to excel convertor

JSON to excel converter

A package that converts json to CSV, excel or other table formats

Sample output

Simple json

[
  {
    "col1": "val1",
    "col2": "val2" 
  }
]

the generated CSV/excel is:

col1          col2
==================
val1          val2

Nested json

[
  {
    "col1": "val1",
    "col2": {
      "col21": "val21",
      "col22": "val22"
    }
  }
]

the generated CSV/excel is (in excel, col2 spans two cells horizontally):

col1          col2
              col21         col22
=================================
val1          val21         val22

json with array property

[
  {
    "col1": "val1",
    "col2": [
      {
        "col21": "val21"
      },
      {
        "col21": "val22"
      }
    ]
  }
]

the generated CSV/excel is (in excel, col2 spans two cells horizontally):

col1          col2         
              col21         col21
=================================
val1          val21         val22

Installation

pip install json-excel-converter[extra]

where extra is:

• xlsxwriter to use the xlsxwriter

Usage

Simple usage

from json_excel_converter import Converter 
from json_excel_converter.xlsx import Writer

data = [
    {'a': [1], 'b': 'hello'},
    {'a': [1, 2, 3], 'b': 'world'}
]

conv = Converter()
conv.convert(data, Writer(file='/tmp/test.xlsx'))

Streaming usage with restarts

from json_excel_converter import Converter, LinearizationError 
from json_excel_converter.csv import Writer

conv = Converter()
writer = Writer(file='/tmp/test.csv')
while True:
    try:
        data = get_streaming_data()     # custom function to get iterator of data
        conv.convert_streaming(data, writer)
        break
    except LinearizationError:
        pass

Arrays

When the first row is processed, the library guesses the columns layout. In case of arrays, a column (or more columns if the array contains json objects) is created for each of the items in the array, as shown in the example above.

On subsequent rows the array might contain more items. The library reacts by adjusting the number of columns in the layout and raising LinearizationError as previous rows might be already output.

Converter.convert_streaming just raises this exception - it is the responsibility of caller to take the right action.

Converter.convert captures this error and restarts the processing. In case of CSV this means truncating the output file to 0 bytes and processing the data again. XLSX writer caches all the data before writing them to excel so the restart just means discarding the cache.

If you know the size of the array in advance, you should pass it in options. Then no processing restarts are required and LinearizationError is not raised.

from json_excel_converter import Converter, Options
from json_excel_converter.xlsx import Writer

data = [
   {'a': [1]},
   {'a': [1, 2, 3]}
]
options = Options()
options['a'].cardinality = 3

conv = Converter(options=options)
writer = Writer(file='/tmp/test.xlsx')
conv.convert(data, writer)
# or
conv.convert_streaming(data, writer)    # no exception occurs here

XLSX Formatting

Cell format

XLSX writer enables you to format the header and data by passing an array of header_formatters or data_formatters. Take these from json_excel_converter.xlsx.formats package or create your own.

from json_excel_converter import Converter

from json_excel_converter.xlsx import Writer
from json_excel_converter.xlsx.formats import LastUnderlined, Bold, \
    Centered, Format

data = [
    {'a': 'Hello'},
    {'a': 'World'}
]

w = Writer('/tmp/test3.xlsx',
           header_formats=(
               Centered, Bold, LastUnderlined,
               Format({
                   'font_color': 'red'
               })),
           data_formats=(
               Format({
                   'font_color': 'green'
               }),)
           )

conv = Converter()
conv.convert(data, w)

See https://xlsxwriter.readthedocs.io/format.html for details on formats in xlsxwriter

Column widths

Pass the required column widths to writer:

w = Writer('/tmp/test3.xlsx', column_widths={
    'a': 20
})

Width of nested data can be specified as well:

data = [
    {'a': {'b': 1, 'c': 2}}
]

w = Writer('/tmp/test3.xlsx', column_widths={
    'a.b': 20,
    'a.c': 30,
})

To set the default column width, pass it as DEFAULT_COLUMN_WIDTH property:

w = Writer('/tmp/test3.xlsx', column_widths={
    DEFAULT_COLUMN_WIDTH: 20
})

Row heights

Row heights can be specified via the row_heights writer option:

w = Writer('/tmp/test3.xlsx', row_heights={
    DEFAULT_ROW_HEIGHT: 20,     # a bit taller rows
    1: 40                       # extra tall header
})

Urls

To render url, pass a function that gets data of a row and returns url to options

data = [
   {'a': 'https://google.com'},
]

options = Options()
options['a'].url = lambda data: data['a']

conv = Converter(options)
conv.convert(data, w)

Note: this will only be rendered in XLSX output, CSV output will silently ignore the link.

Custom cell rendering

Override the write_cell method. The method receives cell_data (instance of json_excel_converter.Value) and data (the original data being written to this row). Note that this method is used both for writing header and rows - for header the data parameter is None.

class UrlWriter(Writer):
    def write_cell(self, row, col, cell_data, cell_format, data):
        if cell_data.path == 'a' and data:
            self.sheet.write_url(row, col,
                                 'https://test.org/' + data['b'],
                                 string=cell_data.value)
        else:
            super().write_cell(row, col, cell_data, cell_format, data)