def __init__(self, fname):
self.ntype = 'python'
name = os.path.basename(fname)
self.name = name.replace('.m', '')
self.nb = Notebook()
self.fname = fname
self._excercise_num = 1
self.excercises = []
def __init__(self, fname):
self.ntype = 'python'
name = os.path.basename(fname)
self.name = name.replace('.m', '')
self.nb = Notebook()
self.fname = fname
self._excercise_num = 1
self.excercises = []
def convert(fname):
"""Convert an m file into an IPython notebook
Takes most of the grunt work out of translating a numerical-tours
matlab file into an ipython notebook. It will auto-populate the
notebook with headings, markdown, code, and exercises as appropriate.
Ideally, all one then has to do is translate the code itself.
Some code transformations are in place (see CODE_REPLS and _parse_code
in this file).
"""
with open(fname) as fid:
lines = fid.readlines()
nb = Notebook()
nb.add_heading(lines[0][3:].rstrip())
state = 'markdown'
out_lines = []
for line in lines[1:]:
new_state, new_line = parse_line(line, state)
if not new_state == state:
get_section(nb, state, out_lines)
out_lines = [new_line]
state = new_state
else:
out_lines.append(new_line)
# handle the last section
get_section(nb, state, out_lines)
fname = fname.replace('.m', '.ipynb')
fname = os.path.basename(fname)
dname = os.path.dirname('__file__')
path = os.path.join(os.path.abspath(dname), fname)
nb.save(path)
class Converter(object):
"""Convert an m file into an IPython notebook
Takes most of the grunt work out of translating a numerical-tours
matlab file into an ipython notebook. It will auto-populate the
notebook with headings, markdown, code, and exercises as appropriate.
Ideally, all one then has to do is translate the code itself.
For python, some code transformations are in place
(see CODE_REPLS and `parse_code`).
"""
def __init__(self, fname):
self.ntype = 'python'
name = os.path.basename(fname)
self.name = name.replace('.m', '')
self.nb = Notebook()
self.fname = fname
self._excercise_num = 1
self.excercises = []
def convert(self, out_dir='.', ntype='python'):
self.ntype = ntype.lower()
with open(self.fname, 'rb') as fid:
text = fid.read().decode('utf-8', 'replace')
text = text.replace('\ufffd', ' ')
lines = text.splitlines()
header = lines[0][3:].rstrip()
header = [header, '=' * len(header)]
header += [lib.INTRO % ntype]
self.nb.add_markdown(header + lib.MATH_CMDS)
state = 'markdown'
out_lines = []
for line in lines[1:]:
new_state, new_line = self.parse_line(line, state)
if not new_state == state:
self.get_section(state, out_lines)
out_lines = [new_line]
state = new_state
else:
out_lines.append(new_line)
# handle the last section
self.get_section(state, out_lines)
fname = self.fname.replace('.m', '.ipynb')
basename = os.path.basename(fname)
path = os.path.join(out_dir, basename)
self.nb.save(path)
if self.ntype == 'python':
self._write_exercises(out_dir)
elif self.ntype == 'matlab':
convert_to_matlab_kernel(path)
def parse_line(self, line, state):
new_line = line
new_state = state
if state == 'excercise':
if new_line.startswith('%EXO'):
new_state = 'markdown'
new_line = ''
elif state == 'comment':
if new_line.startswith('%CMT'):
new_state = 'markdown'
new_line = ''
elif re.match(lib.SECTION_HEADING, new_line):
new_state = 'markdown'
new_line = new_line[3:]
new_line += '\n' + '-' * len(new_line)
elif new_line.startswith('perform_toolbox_installation('):
new_state = 'install'
# get the package names
new_line = re.split("'(\w*?)'", new_line)
new_line = ' '.join(new_line[1::2])
elif new_line.startswith('%EXO'):
new_state = 'excercise'
new_line = ''
elif new_line.startswith('%CMT'):
new_state = 'comment'
new_line = ''
elif new_line.startswith('%'):
new_state = 'markdown'
elif not new_line.strip():
new_state = 'markdown'
new_line = ''
else:
new_state = 'code'
if new_state == 'markdown':
pass
elif new_state == 'code' and self.ntype == 'python':
new_line = self.parse_code(new_line)
return new_state, new_line
def parse_code(self, line):
if line.rstrip().endswith(';'):
line = line.rstrip()[:-1]
for (pattern, repl) in lib.PY_REPLS:
line = re.sub(pattern, repl, line)
if line.lstrip().startswith('for '):
rest = line.lstrip().split('for ')[1]
var, _, rest = rest.partition('=')
rng, _, comment = rest.partition('%')
line = 'for %s in %s:' % (var, rng.rstrip())
if comment:
line += ' # %s' % comment
# remove right side spacing
for op in ['\\(', '\[', '{']:
line = re.sub('%s\s+' % op, op[-1], line)
# remove left side spacing
for op in ['\\)', '\]', ':', '}']:
line = re.sub('\s+%s' % op, op[-1], line)
# add a space on the left
for op in [r'\+', '<', '>', '=']:
line = re.sub(r'(\S)%s' % op,
lambda m: '%s %s' % (m.groups()[0], op[-1]),
line)
# add a space on the right
for op in ['/', r'\+', '=', ':', ',', ';']:
line = re.sub(r'%s(\S)' % op,
lambda m: '%s %s' % (op[-1], m.groups()[0]),
line)
line = re.sub('< =', '<=', line)
line = re.sub('> =', '>=', line)
return line
def get_section(self, state, out_lines):
nb = self.nb
if state == 'excercise':
header = '__Exercise %s__' % self._excercise_num
lines = [header, '']
code_lines = []
for line in out_lines:
if line.startswith('%'):
lines.append(line[3:])
else:
code_lines.append(line)
lines = self._parse_markdown(lines)
self.excercises.append((lines[2:], code_lines))
nb.add_markdown(lines)
if self.ntype == 'python':
nb.add_code('solutions.exo%s()' % self._excercise_num)
else:
nb.add_code('%%%%matlab\nexo%s()' % self._excercise_num)
self._excercise_num += 1
if self.ntype == 'matlab':
nb.add_code("%%matlab\n%% Insert your code here.")
else:
nb.add_code("## Insert your code here.")
elif state == 'markdown':
nb.add_markdown(self._parse_markdown(out_lines))
elif state == 'code':
if not self.ntype == 'python':
out_lines.insert(0, '%%{0}'.format(self.ntype))
nb.add_code(out_lines)
elif state == 'install':
func = getattr(self, 'get_%s_intro' % self.ntype)
func(out_lines[0].split())
def get_python_intro(self, toolboxes):
setup = r"""
from __future__ import division
import nt_toolbox as nt
from nt_solutions import {0} as solutions
%matplotlib inline
%load_ext autoreload
%autoreload 2
""".format(self.name)
self.nb.add_code(self._reformat(setup))
def get_matlab_intro(self, toolboxes):
setup = ['%load_ext pymatbridge']
self.nb.add_code(setup)
setup = ['%%matlab']
for toolbox in toolboxes:
setup += ["addpath('toolbox_%s')" % toolbox]
setup += ["addpath('solutions/%s')" % self.name]
self.nb.add_code(setup)
def get_scilab_intro(self, toolboxes):
setup = ['%load_ext scilab2py.ipython']
self.nb.add_code(setup)
setup = ['%%scilab']
for toolbox in toolboxes:
setup += ["addpath('toolbox_%s')" % toolbox]
setup += ["addpath('solutions/%s')" % self.name]
self.nb.add_code(setup)
def _parse_markdown(self, lines):
lines = self._handle_links(lines)
return self._handle_latex(lines)
@staticmethod
def _handle_links(lines):
links = []
new_lines = []
for line in lines:
matches = re.findall(lib.LINK, line)
for match in matches:
if not isinstance(match, tuple):
continue
link, title = match
links.append(link[1:-1])
line = line.replace(link, '')
new_link = '[%s][%s]' % (title[1:-2], len(links))
line = line.replace(title, new_link)
biblio_links = re.findall(lib.BIBLIO_LINK, line)
for link in biblio_links:
line = line.replace('<#biblio %s>' % link,
'%s(#biblio)' % link)
new_lines.append(line)
if links:
new_lines.append('')
for (ind, link) in enumerate(links):
new_lines.append('[%s]:%s' % (ind + 1, link))
return new_lines
@staticmethod
def _handle_latex(lines):
output = []
for line in lines:
while line.startswith('%'):
line = line[1:]
if line.startswith(' '):
line = line[1:]
if '\\' in line:
for (pattern, repl) in lib.MATH_REPLS:
line = re.sub(pattern, repl, line)
output.append(line.rstrip())
return output
@staticmethod
def _reformat(text):
lines = text.splitlines()
return '\n'.join([l.strip() for l in lines])
def _write_exercises(self, out_dir):
sfile = 'solutions/%s.py' % self.name
sfile = os.path.join(out_dir, sfile)
with open(sfile, 'w') as fid:
for (ind, (comments, lines)) in enumerate(self.excercises):
fid.write('def exo%s():\n """\n' % (ind + 1))
for comment in comments:
fid.write(' %s\n' % comment)
fid.write(' """\n')
for line in lines:
line = self.parse_code(line)
if line.strip():
fid.write(' %s\n' % line)
fid.write('\n\n')
class Converter(object):
"""Convert an m file into an IPython notebook
Takes most of the grunt work out of translating a numerical-tours
matlab file into an ipython notebook. It will auto-populate the
notebook with headings, markdown, code, and exercises as appropriate.
Ideally, all one then has to do is translate the code itself.
For python, some code transformations are in place
(see CODE_REPLS and `parse_code`).
"""
def __init__(self, fname):
self.ntype = 'python'
name = os.path.basename(fname)
self.name = name.replace('.m', '')
self.nb = Notebook()
self.fname = fname
self._excercise_num = 1
self.excercises = []
def convert(self, out_dir='.', ntype='python'):
self.ntype = ntype.lower()
with open(self.fname, 'rb') as fid:
text = fid.read().decode('utf-8', 'replace')
text = text.replace('\ufffd', ' ')
lines = text.splitlines()
header = lines[0][3:].rstrip()
header = [header, '=' * len(header)]
header += [lib.INTRO % ntype]
self.nb.add_markdown(header + lib.MATH_CMDS)
state = 'markdown'
out_lines = []
for line in lines[1:]:
new_state, new_line = self.parse_line(line, state)
if not new_state == state:
self.get_section(state, out_lines)
out_lines = [new_line]
state = new_state
else:
out_lines.append(new_line)
# handle the last section
self.get_section(state, out_lines)
fname = self.fname.replace('.m', '.ipynb')
basename = os.path.basename(fname)
path = os.path.join(out_dir, basename)
self.nb.save(path)
if self.ntype == 'python':
self._write_exercises(out_dir)
def parse_line(self, line, state):
new_line = line
new_state = state
if state == 'excercise':
if new_line.startswith('%EXO'):
new_state = 'markdown'
new_line = ''
elif state == 'comment':
if new_line.startswith('%CMT'):
new_state = 'markdown'
new_line = ''
elif re.match(lib.SECTION_HEADING, new_line):
new_state = 'markdown'
new_line = new_line[3:]
new_line += '\n' + '-' * len(new_line)
elif new_line.startswith('perform_toolbox_installation('):
new_state = 'install'
# get the package names
new_line = re.split("'(\w*?)'", new_line)
new_line = ' '.join(new_line[1::2])
elif new_line.startswith('%EXO'):
new_state = 'excercise'
new_line = ''
elif new_line.startswith('%CMT'):
new_state = 'comment'
new_line = ''
elif new_line.startswith('%'):
new_state = 'markdown'
elif not new_line.strip():
new_state = 'markdown'
new_line = ''
else:
new_state = 'code'
if new_state == 'markdown':
pass
elif new_state == 'code' and self.ntype == 'python':
new_line = self.parse_code(new_line)
return new_state, new_line
def parse_code(self, line):
if line.rstrip().endswith(';'):
line = line.rstrip()[:-1]
for (pattern, repl) in lib.PY_REPLS:
line = re.sub(pattern, repl, line)
if line.lstrip().startswith('for '):
rest = line.lstrip().split('for ')[1]
var, _, rest = rest.partition('=')
rng, _, comment = rest.partition('%')
line = 'for %s in %s:' % (var, rng.rstrip())
if comment:
line += ' # %s' % comment
# remove right side spacing
for op in ['\\(', '\[', '{']:
line = re.sub('%s\s+' % op, op[-1], line)
# remove left side spacing
for op in ['\\)', '\]', ':', '}']:
line = re.sub('\s+%s' % op, op[-1], line)
# add a space on the left
for op in [r'\+', '<', '>', '=']:
line = re.sub(r'(\S)%s' % op,
lambda m: '%s %s' % (m.groups()[0], op[-1]),
line)
# add a space on the right
for op in ['/', r'\+', '=', ':', ',', ';']:
line = re.sub(r'%s(\S)' % op,
lambda m: '%s %s' % (op[-1], m.groups()[0]),
line)
line = re.sub('< =', '<=', line)
line = re.sub('> =', '>=', line)
return line
def get_section(self, state, out_lines):
nb = self.nb
if state == 'excercise':
header = '__Exercise %s__' % self._excercise_num
lines = [header, '']
code_lines = []
for line in out_lines:
if line.startswith('%'):
lines.append(line[3:])
else:
code_lines.append(line)
lines = self._parse_markdown(lines)
self.excercises.append((lines[2:], code_lines))
nb.add_markdown(lines)
if self.ntype == 'python':
nb.add_code('solutions.exo%s()' % self._excercise_num)
else:
nb.add_code('%%%%matlab\nexo%s()' % self._excercise_num)
self._excercise_num += 1
if self.ntype == 'matlab':
nb.add_code("%%matlab\n%% Insert your code here.")
else:
nb.add_code("## Insert your code here.")
elif state == 'markdown':
nb.add_markdown(self._parse_markdown(out_lines))
elif state == 'code':
if not self.ntype == 'python':
out_lines.insert(0, '%%{0}'.format(self.ntype))
nb.add_code(out_lines)
elif state == 'install':
func = getattr(self, 'get_%s_intro' % self.ntype)
func(out_lines[0].split())
def get_python_intro(self, toolboxes):
setup = r"""
from __future__ import division
import nt_toolbox as nt
from nt_solutions import {0} as solutions
%matplotlib inline
%load_ext autoreload
%autoreload 2
""".format(self.name)
self.nb.add_code(self._reformat(setup))
def get_matlab_intro(self, toolboxes):
setup = ['%load_ext pymatbridge']
self.nb.add_code(setup)
setup = ['%%matlab']
for toolbox in toolboxes:
setup += ["addpath('toolbox_%s')" % toolbox]
setup += ["addpath('solutions/%s')" % self.name]
self.nb.add_code(setup)
def get_scilab_intro(self, toolboxes):
setup = ['%load_ext scilab2py.ipython']
self.nb.add_code(setup)
setup = ['%%scilab']
for toolbox in toolboxes:
setup += ["addpath('toolbox_%s')" % toolbox]
setup += ["addpath('solutions/%s')" % self.name]
self.nb.add_code(setup)
def _parse_markdown(self, lines):
lines = self._handle_links(lines)
return self._handle_latex(lines)
@staticmethod
def _handle_links(lines):
links = []
new_lines = []
for line in lines:
matches = re.findall(lib.LINK, line)
for match in matches:
if not isinstance(match, tuple):
continue
link, title = match
links.append(link[1:-1])
line = line.replace(link, '')
new_link = '[%s][%s]' % (title[1:-2], len(links))
line = line.replace(title, new_link)
biblio_links = re.findall(lib.BIBLIO_LINK, line)
for link in biblio_links:
line = line.replace('<#biblio %s>' % link,
'%s(#biblio)' % link)
new_lines.append(line)
if links:
new_lines.append('')
for (ind, link) in enumerate(links):
new_lines.append('[%s]:%s' % (ind + 1, link))
return new_lines
@staticmethod
def _handle_latex(lines):
output = []
for line in lines:
while line.startswith('%'):
line = line[1:]
if line.startswith(' '):
line = line[1:]
if '\\' in line:
for (pattern, repl) in lib.MATH_REPLS:
line = re.sub(pattern, repl, line)
output.append(line.rstrip())
return output
@staticmethod
def _reformat(text):
lines = text.splitlines()
return '\n'.join([l.strip() for l in lines])
def _write_exercises(self, out_dir):
sfile = 'solutions/%s.py' % self.name
sfile = os.path.join(out_dir, sfile)
with open(sfile, 'w') as fid:
for (ind, (comments, lines)) in enumerate(self.excercises):
fid.write('def exo%s():\n """\n' % (ind + 1))
for comment in comments:
fid.write(' %s\n' % comment)
fid.write(' """\n')
for line in lines:
line = self.parse_code(line)
if line.strip():
fid.write(' %s\n' % line)
fid.write('\n\n')