Method needed for skipping lines

[Bruno Desthuilliers]

Gustaf a écrit :
> Hi all,
> 
> Just for fun, I'm working on a script to count the number of lines in 
> source files. Some lines are auto-generated (by the IDE) and shouldn't 
> be counted. The auto-generated part of files start with "Begin VB.Form" 
> and end with "End" (first thing on the line). The "End" keyword may 
> appear inside the auto-generated part, but not at the beginning of the 
> line.
> 
> I imagine having a flag variable to tell whether you're inside the 
> auto-generated part, but I wasn't able to figure out exactly how. Here's 
> the function, without the ability to skip auto-generated code:
> 
> # Count the lines of source code in the file
> def count_lines(f):
>  file = open(f, 'r')

1/ The param name is not very explicit.
2/ You're shadowing the builtin file type.
3/ It migh be better to pass an opened file object instead - this would 
make your function more generic (ok, perhaps a bit overkill here, but 
still a better practice IMHO).

>  rows = 0

Shouldn't that be something like 'line_count' ?

>  for line in file:
>    rows = rows + 1

Use augmented assignment instead:
      rows += 1

>  return rows

You forgot to close the file.

> How would you modify this to exclude lines between "Begin VB.Form" and 
> "End" as described above?

Here's a straightforward solution:

def count_loc(path):
   loc_count = 0
   in_form = False
   opened_file = open(path)
   try:
     # striping lines, and skipping blank lines
     for line in opened_file:
       line = line.strip()
       # skipping blank lines
       if not line:
         continue
       # skipping VB comments
       # XXX: comment mark should not be hardcoded
       if line.startswith(';'):
         continue
       # skipping autogenerated code
       if line.startswith("Begin VB.Form"):
         in_form = True
         continue
       elif in_form:
         if line.startswith("End"):
             in_form = False
         continue
       # Still here ? ok, we count this one
       loc_count += 1
   finally:
     opened_file.close()
   return loc_count

HTH

PS : If you prefer a more functional approach
(warning: the following code may permanently damage innocent minds):

def chain(*predicates):
     def _chained(arg):
         for p in predicates:
             if not p(arg):
                 return False
         return True
     return _chained

def not_(predicate):
     def _not_(arg):
         return not predicate(arg)
     return _not_

class InGroupPredicate(object):
     def __init__(self, begin_group, end_group):
         self.in_group = False
         self.begin_group = begin_group
         self.end_group = end_group

     def __call__(self, line):
         if self.begin_group(line):
             self.in_group = True
             return True
         elif self.in_group and self.end_group(line):
             self.in_group = False
             return True # this one too is part of the group
         return self.in_group

def count_locs(lines, count_line):
     return len(filter(
                chain(lambda line: bool(line), count_line),
                map(str.strip,lines)
            ))

def count_vb_locs(lines):
     return count_locs(lines, chain(
                 not_(InGroupPredicate(
                     lambda line: line.startswith('Begin VB.Form'),
                     lambda line: line.startswith('End')
                 )),
             lambda line: not line.startswith(';')
           ))

# and finally our count_lines function, greatly simplified !-)
def count_lines(path):
     f = open(path)
     try:
         return count_vb_locs(f)
     finally:
         f.close()

(anyone on doing it with itertools ?-)