In the verbal description of the algorithm, "should be less than" implies "not equal", which would not be correct. A wording such as "must be at most" would be totally clear.

Your code has a bug because you are not quite following rule 1. The rule says "at any point...", but you only check the rule after a smiley. You miss the case when a lone ) breaks the rule.

Instead of canceling += 1 with -= 1 here...

 left += 1
 if i > 0 and source[i-1] == ':': # :(
     left -= 1
     frown += 1

... it would be clearer to use else:

 if i > 0 and source[i-1] == ':': # :(
     frown += 1
 else:
     left += 1

Instead of if i > 0 and source[i-1] == ':' it would be simpler to remember the previous character in a variable:

  previous_char = None 
  for char in source:
  if char == 'c':
      if previous_char == ':':
  ...
  previous_char = char

In the verbal description of the algorithm, "should be less than" implies "not equal", which would not be correct. A wording such as "must be at most" would be totally clear.

Your code has a bug because you are not quite following rule 1. The rule says "at any point...", but you only check the rule after a smiley. You miss the case when a lone ) breaks the rule.

Instead of canceling += 1 with -= 1 here...

 left += 1
 if i > 0 and source[i-1] == ':': # :(
     left -= 1
     frown += 1

... it would be clearer to use else:

 if i > 0 and source[i-1] == ':': # :(
     frown += 1
 else:
     left += 1

Instead of if i > 0 and source[i-1] == ':' it would be simpler to remember the previous character in a variable:

  previous_char = None 
  for char in source:
      if char == 'c':
          if previous_char == ':':
      ...
      previous_char = char

In the verbal description of the algorithm, "should be less than" implies "not equal", which would not be correct. A wording such as "must be at most" would be totally clear.

Your code has a bug because you are not quite following rule 1. The rule says "at any point...", but you only check the rule after a smiley. You miss the case when a lone ) breaks the rule.

Instead of canceling += 1 with -= 1 here...

 left += 1
 if i > 0 and source[i-1] == ':': # :(
     left -= 1
     frown += 1

... it would be clearer to use else:

 if i > 0 and source[i-1] == ':': # :(
     frown += 1
 else:
     left += 1

Instead of if i > 0 and source[i-1] == ':' it would be simpler to remember the previous character in a variable:

  previous_char = None 
  for char in source:
  if char == 'c':
      if previous_char == ':':

In the verbal description of the algorithm, "should be less than" implies "not equal", which would not be correct. A wording such as "must be at most" would be totally clear.

Your code has a bug because you are not quite following rule 1. The rule says "at any point...", but you only check the rule after a smiley. You miss the case when a lone ) breaks the rule.

Instead of canceling += 1 with -= 1 here...

 left += 1
 if i > 0 and source[i-1] == ':': # :(
     left -= 1
     frown += 1

... it would be clearer to use else:

 if i > 0 and source[i-1] == ':': # :(
     frown += 1
 else:
     left += 1

Instead of if i > 0 and source[i-1] == ':' it would be simpler to remember the previous character in a variable:

  previous_char = None 
  for char in source:
  if char == 'c':
      if previous_char == ':':
  ...
  previous_char = char