reports/ipynb_generator.py

   1 import sys, os, re, logging
   2
   3 # Languages mapping as used by markdown/pandoc
   4 shortname2language = dict(
   5     c='C',
   6     cpp='Cpp',
   7     f='Fortran',
   8     html='HTML',
   9     js='JavaScript',
  10     r='R',
  11     rb='Ruby',
  12     pl='Perl',
  13     py='Python',
  14     sh='Bash',
  15     tex='Tex',
  16     )
  17
  18 def read(text, argv=sys.argv[2:]):
  19     lines = text.splitlines()
  20     # First read all include statements
  21     for i in range(len(lines)):
  22         if lines[i].startswith('#include "'):
  23             filename = lines[i].split('"')[1]
  24             with open(filename, 'r') as f:
  25                 include_text = f.read()
  26             lines[i] = include_text
  27     text = '\n'.join(lines)
  28     logging.info('******* text after include:\n{}'.format(text))
  29
  30     # Run Mako
  31     mako_kwargs = {}
  32     for arg in argv:
  33         key, value = arg.split('=')
  34         mako_kwargs[key] = value
  35
  36     try:
  37         import mako
  38         has_mako = True
  39     except ImportError:
  40         print('Cannot import mako - mako is not run')
  41         has_mako = False
  42
  43     if has_mako:
  44         from mako.template import Template
  45         from mako.lookup import TemplateLookup
  46         lookup = TemplateLookup(directories=[os.curdir])
  47         text = text.encode('utf-8')
  48         temp = Template(text=text, lookup=lookup, strict_undefined=True)
  49         logging.info('******* mako_kwargs: {}'.format(str(mako_kwargs)))
  50         text = temp.render(**mako_kwargs)
  51
  52     logging.info('******* text after mako:\n{}'.format(text))
  53
  54     # Parse the cells
  55     lines = text.splitlines()
  56     cells = []
  57     inside = None    # indicates which type of cell we are inside
  58     fullname = None  # full language name in code cells
  59     for line in lines:
  60         if line.startswith('-----'):
  61             # New cell, what type?
  62             m = re.search(r'-----([a-z0-9-]+)?', line)
  63             if m:
  64                 shortname = m.group(1)
  65                 if shortname:
  66                     # Check if code is to be typeset as static
  67                     # Markdown code (e.g., shortname=py-t)
  68                     logging.info('******* found shortname {}'
  69                         .format(shortname))
  70                     astext = shortname[-2:] == '-t'
  71                     logging.info('******* cell: astext={} shortname={}'
  72                         .format(astext, shortname))
  73                     if astext:
  74                         # Markdown
  75                         shortname = shortname[:-2]
  76                         inside = 'markdown'
  77                         cells.append(['markdown', 'code', ['\n']])
  78                         cells[-1][2].append('```%s\n' % fullname)
  79                     else:
  80                         # Code cell
  81                         if shortname in shortname2language:
  82                             fullname = shortname2language[shortname]
  83                         inside = 'codecell'
  84                         cells.append(['codecell', fullname, []])
  85                 else:
  86                     logging.info('******* cell: markdown')
  87                     # Markdown cell
  88                     inside = 'markdown'
  89                     cells.append(['markdown', 'text', ['\n']])
  90             else:
  91                 raise SyntaxError('Wrong syntax of cell delimiter:\n{}'
  92                     .format(repr(line)))
  93         else:
  94             # Ordinary line in a cell
  95             if inside in ('markdown', 'codecell'):
  96                 cells[-1][2].append(line)
  97             else:
  98                 raise SyntaxError('line\n {}\nhas no beginning cell delimiter'
  99                     .format(line))
 100     # Merge the lines in each cell to a string
 101     for i in range(len(cells)):
 102         if cells[i][0] == 'markdown' and cells[i][1] == 'code':
 103             # Add an ending ``` of code
 104             cells[i][2].append('```\n')
 105         cells[i][2] = '\n'.join(cells[i][2])
 106     # TODO: optional logging
 107     import pprint
 108     logging.info('******* cell data structure:\b%s' % pprint.pformat(cells))
 109     return cells
 110
 111 def write(cells):
 112     """Turn cells list into valid IPython notebook code."""
 113     # Use Jupyter nbformat functionality for writing the notebook
 114
 115     from nbformat.v4 import (
 116         new_code_cell, new_markdown_cell, new_notebook, writes)
 117     nb_cells = []
 118
 119     for cell_tp, language, block in cells:
 120         if cell_tp == 'markdown':
 121             nb_cells.append(
 122                 new_markdown_cell(source=block))
 123         elif cell_tp == 'codecell':
 124             nb_cells.append(new_code_cell(source=block))
 125
 126     nb = new_notebook(cells=nb_cells)
 127     filestr = writes(nb)
 128     return filestr
 129
 130 def driver():
 131     """Compile a document and its variables."""
 132     try:
 133         filename = sys.argv[1]
 134         with open(filename, 'r') as f:
 135             text = f.read()
 136     except (IndexError, IOError) as e:
 137         print('Usage: %s filename' % (sys.argv[0]))
 138         print(e)
 139         sys.exit(1)
 140     cells = read(text, argv=sys.argv[2:])
 141     filestr = write(cells)
 142     # Assuming file extension .gj (generate Jupyter); TODO: less strict
 143     filename = filename[:-3] + '.ipynb'
 144     with open(filename, 'w') as f:
 145         f.write(filestr)
 146
 147 if __name__ == '__main__':
 148     logfile = 'tmp.log'
 149     if os.path.isfile:
 150         os.remove(logfile)
 151     logging.basicConfig(format='%(message)s', level=logging.DEBUG,
 152                         filename=logfile)
 153     driver()