Skip to content

Instantly share code, notes, and snippets.

@ColtonProvias
Created January 3, 2014 04:18
Show Gist options
  • Star 5 You must be signed in to star a gist
  • Fork 1 You must be signed in to fork a gist
  • Save ColtonProvias/8232624 to your computer and use it in GitHub Desktop.
Save ColtonProvias/8232624 to your computer and use it in GitHub Desktop.
Fountain parser for Python 3.
"""
fountain.py
Ported to Python 3 by Colton J. Provias - cj@coltonprovias.com
Based on Fountain by Nima Yousefi & John August
Original code for Objective-C at https://github.com/nyousefi/Fountain
"""
COMMON_TRANSITIONS = {'FADE OUT.', 'CUT TO BLACK.', 'FADE TO BLACK.'}
class FountainElement:
def __init__(self, element_type, element_text, section_depth=0, scene_number='', is_centered=False, is_dual_dialogue=False):
self.element_type = element_type
self.element_text = element_text
self.section_depth = section_depth
self.scene_number = scene_number
self.is_centered = is_centered
self.is_dual_dialogue = is_dual_dialogue
def __repr__(self):
return self.element_type + ': ' + self.element_text
class Fountain:
def __init__(self, string=None, path=None):
if path:
with open(path) as fp:
self.contents = fp.read()
else:
self.contents = string
self.parse()
def parse(self):
self.metadata = dict()
contents = self.contents.strip().replace('\r', '')
if ':' in contents[:20]:
script_head, script_body = contents.split('\n\n', 1)
self._parse_head(script_head.splitlines())
self._parse_body(script_body.splitlines())
else:
self._parse_body(contents.splitlines())
def _parse_head(self, script_head):
open_key = None
for line in script_head:
line = line.rstrip()
if line[0].isspace():
self.metadata[open_key].append(line.strip())
elif line[-1] == ':':
open_key = line[0:-1].lower()
self.metadata[open_key] = list()
else:
key, value = line.split(':', 1)
self.metadata[key.strip().lower()] = [value.strip()]
def _parse_body(self, script_body):
is_comment_block = False
is_inside_dialogue_block = False
newlines_before = 0
index = -1
self.elements = list()
comment_text = list()
for line in script_body:
assert type(line) is str
index += 1
line = line.lstrip()
full_strip = line.strip()
if (not line or line.isspace()) and not is_comment_block:
is_inside_dialogue_block = False
newlines_before += 1
continue
if line.startswith('/*'):
line = line.rstrip()
if line.endswith('*/'):
text = line.replace('/*', '').replace('*/', '')
self.elements.append(FountainElement('Boneyard', text))
is_comment_block = False
newlines_before = 0
else:
is_comment_block = True
comment_text.append('')
continue
if line.rstrip().endswith('*/'):
text = line.replace('*/', '')
comment_text.append(text.strip())
self.elements.append(FountainElement('Boneyard',
'\n'.join(comment_text)))
is_comment_block = False
comment_text = list()
newlines_before = 0
continue
if is_comment_block:
comment_text.append(line)
continue
if line.startswith('==='):
self.elements.append(FountainElement('Page Break', line))
newlines_before = 0
continue
if len(full_strip) > 0 and full_strip[0] == '=':
self.elements.append(FountainElement('Synopsis',
full_strip[1:].strip()))
continue
if newlines_before > 0 and full_strip.startswith('[[') and full_strip.endswith(']]'):
self.elements.append(FountainElement('Comment',
full_strip.strip('[] \t')))
continue
if len(full_strip) > 0 and full_strip[0] == '#':
newlines_before = 0
depth = full_strip.split()[0].count('#')
self.elements.append(FountainElement('Section Heading',
full_strip[depth:],
section_depth=depth))
continue
if len(line) > 1 and line[0] == '.' and line[1] != '.':
newlines_before = 0
if full_strip[-1] == '#' and full_strip.count('#') > 1:
scene_number_start = len(full_strip) - full_strip[::-1].find('#', 1) - 1
self.elements.append(FountainElement('Scene Heading', full_strip[1:scene_number_start].strip(), scene_number=full_strip[scene_number_start:].strip('#').strip()))
else:
self.elements.append(FountainElement('Scene Heading', full_strip[1:].strip()))
continue
if line[0:4] in ['INT ', 'INT.', 'EXT ', 'EXT.', 'EST ', 'EST.', 'I/E ', 'I/E.'] or\
line[0:8] in ['INT/EXT ', 'INT/EXT.'] or\
line[0:9] in ['INT./EXT ', 'INT./EXT.']:
newlines_before = 0
if full_strip[-1] == '#' and full_strip.count('#') > 1:
scene_number_start = len(full_strip) - full_strip[::-1].find('#', 1) - 1
self.elements.append(FountainElement('Scene Heading', full_strip[:scene_number_start].strip(), scene_number=full_strip[scene_number_start:].strip('#').strip()))
else:
self.elements.append(FountainElement('Scene Heading', full_strip.strip()))
continue
if full_strip.endswith(' TO:'):
newlines_before = 0
self.elements.append(FountainElement('Transition', full_strip))
continue
if full_strip in COMMON_TRANSITIONS:
newlines_before = 0
self.elements.append(FountainElement('Transition', full_strip))
continue
if full_strip[0] == '>':
newlines_before = 0
if len(full_strip) > 1 and full_strip[-1]:
self.elements.append(FountainElement('Action', full_strip[1:-1].strip(), is_centered=True))
else:
self.elements.append(FountainElement('Transition', full_strip[1:].strip()))
continue
if newlines_before > 0 and index + 1 < len(script_body) and script_body[index + 1]:
newlines_before = 0
if full_strip[-1] == '^':
for element in reversed(self.elements):
if element.element_type == 'Character':
element.is_dual_dialogue = True
break
self.elements.append(FountainElement('Character', full_strip.rstrip('^').strip(), is_dual_dialogue=True))
else:
self.elements.append(FountainElement('Character', full_strip))
is_inside_dialogue_block = True
continue
if is_inside_dialogue_block:
if newlines_before == 0 and full_strip[0] == '(':
self.elements.append(FountainElement('Parenthetical', full_strip))
else:
if self.elements[-1].element_type == 'Dialogue':
self.elements[-1].element_text = '\n'.join([self.elements[-1].element_text, full_strip])
else:
self.elements.append(FountainElement('Dialogue', full_strip))
continue
if newlines_before == 0 and len(self.elements) > 0:
self.elements[-1].element_text = '\n'.join([self.elements[-1].element_text, full_strip])
newlines_before = 0
else:
self.elements.append(FountainElement('Action', full_strip))
newlines_before = 0
@Tagirijus
Copy link

Tagirijus commented Jul 29, 2016

Quite cool, but the parser seems to have a problem with dual dialogue. The "second" character (the one, which ends with "^") is element_type == 'Character", but the following dialogue text has no element_type. I will try to fix this on my own, but I wanted you to know. Maybe you will fix it as well and will be faster than me. (-; ... moreover: I am still not that familiar with the github-system and how to upload a fix for an existing code. :D

Edit:
Haha, nevermind. I already got the fix. After line 170 there is missing: is_inside_dialogue_block = True

I uploaded a fix on my own account here: https://gist.github.com/Tagirijus/d4fafe1b922cd65622e0d1491efb7f14

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment