convert_nbformat.py

"""This conversion script should not be used. It is only there for reference.

This was used to convert the original notebooks in nbformat v3 (IPython 2.0)
to nbformat v4 (IPython 3.0). A difficulty was that the original notebooks
had invalid metadata so IPython failed at the conversion.

The fix_v3() function was used to fix this before converting to v4.

This script was run on 27/02/2015.

"""

import re
import json
import os
import sys
import os.path as op

from IPython.nbformat import read, convert, write, validate

def get_chapter_number(dir):
    return int(op.basename(dir)[7:9])

def get_chapter_name(number):
    return CHAPTER_NAMES[number-1]

def iter_chapters(root):
    for chapter in sorted([_ for _ in os.listdir(root)
                            if _.startswith('chapter')]):
        yield chapter

def iter_recipes(root, dir=''):
    files = sorted([_ for _ in os.listdir(op.join(root, dir))
        if re.match(r'\d{2}\_[^.]+\.ipynb', _)])
    for file in files:
        yield file

def fix_v3(path):
    with open(path, 'r') as f:
        nb = json.load(f)

    for cell in nb['worksheets'][0]['cells']:
        if isinstance(cell.get('metadata', {}), list):
            cell['metadata'] = {}

    with open(path, 'w') as f:
        json.dump(nb, f, indent=2)

def convert_to_v4(path):
    nb = read(path, 3)
    nb_new = convert(nb, 4)

    nb_new["metadata"] ={
         "kernelspec": {
          "display_name": "Python 3",
          "language": "python",
          "name": "python3"
         },
         "language_info": {
          "codemirror_mode": {
           "name": "ipython",
           "version": 3
          },
          "file_extension": ".py",
          "mimetype": "text/x-python",
          "name": "python",
          "nbconvert_exporter": "python",
          "pygments_lexer": "ipython3",
          "version": "3.4.2"
         }
        }

    validate(nb_new)
    write(nb_new, path)

curdir = op.realpath(op.dirname(os.path.abspath(__file__)))
root = op.realpath(op.join(curdir, '../notebooks'))
featured = op.realpath(op.join(curdir, '../featured'))

if __name__ == '__main__':

    # Chapter notebooks
    # for chapter in iter_chapters(root):
    #     for recipe in iter_recipes(root, chapter):
    #         file = op.join(root, chapter, recipe)

    #         if recipe in ('01_notebook.ipynb',
    #                       '02_pandas.ipynb',
    #                       '03_numpy.ipynb',):
    #             continue
    #         print("converting", file)

    #         # fix_v3(file)
    #         convert_to_v4(file)

    # Featured notebooks
    for recipe in iter_recipes(featured):
        file = op.join(featured, recipe)

        if recipe in ('01_notebook.ipynb',
                      '02_pandas.ipynb',
                      '03_numpy.ipynb',):
            continue
        print("converting", file)

        # fix_v3(file)
        convert_to_v4(file)