3 # This file contains a class and a main program that perform three
4 # related (though complimentary) formatting operations on Python
5 # programs. When called as "pindent -c", it takes a valid Python
6 # program as input and outputs a version augmented with block-closing
7 # comments. When called as "pindent -d", it assumes its input is a
8 # Python program with block-closing comments and outputs a commentless
9 # version. When called as "pindent -r" it assumes its input is a
10 # Python program with block-closing comments but with its indentation
11 # messed up, and outputs a properly indented version.
13 # A "block-closing comment" is a comment of the form '# end <keyword>'
14 # where <keyword> is the keyword that opened the block. If the
15 # opening keyword is 'def' or 'class', the function or class name may
16 # be repeated in the block-closing comment as well. Here is an
17 # example of a program fully augmented with block-closing comments:
31 # Note that only the last part of an if...elif...else... block needs a
32 # block-closing comment; the same is true for other compound
33 # statements (e.g. try...except). Also note that "short-form" blocks
34 # like the second 'if' in the example must be closed as well;
35 # otherwise the 'else' in the example would be ambiguous (remember
36 # that indentation is not significant when interpreting block-closing
39 # The operations are idempotent (i.e. applied to their own output
40 # they yield an identical result). Running first "pindent -c" and
41 # then "pindent -r" on a valid Python program produces a program that
42 # is semantically identical to the input (though its indentation may
43 # be different). Running "pindent -e" on that output produces a
44 # program that only differs from the original in indentation.
47 # -s stepsize: set the indentation step size (default 8)
48 # -t tabsize : set the number of spaces a tab character is worth (default 8)
49 # -e : expand TABs into spaces
50 # file ... : input file(s) (default standard input)
51 # The results always go to standard output
54 # - comments ending in a backslash will be mistaken for continued lines
55 # - continuations using backslash are always left unchanged
56 # - continuations inside parentheses are not extra indented by -r
57 # but must be indented for -c to work correctly (this breaks
59 # - continued lines inside triple-quoted strings are totally garbled
62 # - On input, a block may also be closed with an "end statement" --
63 # this is a block-closing comment without the '#' sign.
65 # Possible improvements:
66 # - check syntax based on transitions in 'next' table
67 # - better error reporting
68 # - better error recovery
69 # - check identifier after class/def
71 # The following wishes need a more complete tokenization of the source:
72 # - Don't get fooled by comments ending in backslash
73 # - reindent continuation lines indicated by backslash
74 # - handle continuation lines inside parentheses/braces/brackets
75 # - handle triple quoted strings spanning lines
77 # - optionally do much more thorough reformatting, a la C indent
88 next['if'] = next['elif'] = 'elif', 'else', 'end'
89 next['while'] = next['for'] = 'else', 'end'
90 next['try'] = 'except', 'finally'
91 next['except'] = 'except', 'else', 'finally', 'end'
92 next['else'] = next['finally'] = next['def'] = next['class'] = 'end'
94 start = 'if', 'while', 'for', 'try', 'with', 'def', 'class'
98 def __init__(self, fpi = sys.stdin, fpo = sys.stdout,
99 indentsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
102 self.indentsize = indentsize
103 self.tabsize = tabsize
105 self.expandtabs = expandtabs
106 self._write = fpo.write
107 self.kwprog = re.compile(
108 r'^\s*(?P<kw>[a-z]+)'
109 r'(\s+(?P<id>[a-zA-Z_]\w*))?'
111 self.endprog = re.compile(
112 r'^\s*#?\s*end\s+(?P<kw>[a-z]+)'
113 r'(\s+(?P<id>[a-zA-Z_]\w*))?'
115 self.wsprog = re.compile(r'^[ \t]*')
118 def write(self, line):
120 self._write(line.expandtabs(self.tabsize))
127 line = self.fpi.readline()
128 if line: self.lineno = self.lineno + 1
133 def error(self, fmt, *args):
134 if args: fmt = fmt % args
136 sys.stderr.write('Error at line %d: %s\n' % (self.lineno, fmt))
137 self.write('### %s ###\n' % fmt)
141 line = self.readline()
142 while line[-2:] == '\\\n':
143 line2 = self.readline()
151 def putline(self, line, indent = None):
156 tabs, spaces = divmod(indent*self.indentsize, self.tabsize)
158 m = self.wsprog.match(line)
161 self.write('\t'*tabs + ' '*spaces + line[i:])
167 line = self.getline()
168 if not line: break # EOF
170 m = self.endprog.match(line)
175 self.error('unexpected end')
176 elif stack[-1][0] != kw2:
177 self.error('unmatched end')
180 self.putline(line, len(stack))
183 m = self.kwprog.match(line)
187 self.putline(line, len(stack))
188 stack.append((kw, kw))
191 if next.has_key(kw) and stack:
192 self.putline(line, len(stack)-1)
198 self.putline(line, len(stack))
201 self.error('unterminated keywords')
202 for kwa, kwb in stack:
203 self.write('\t%s\n' % kwa)
212 line = self.getline()
213 if not line: break # EOF
215 m = self.endprog.match(line)
217 end_counter = end_counter + 1
220 m = self.kwprog.match(line)
224 begin_counter = begin_counter + 1
229 if begin_counter - end_counter < 0:
230 sys.stderr.write('Warning: input contained more end tags than expected\n')
231 elif begin_counter - end_counter > 0:
232 sys.stderr.write('Warning: input contained less end tags than expected\n')
241 current, firstkw, lastkw, topid = 0, '', '', ''
243 line = self.getline()
245 m = self.wsprog.match(line)
248 m = self.endprog.match(line)
251 endkw = m.group('kw')
252 thisid = m.group('id')
254 m = self.kwprog.match(line)
256 thiskw = m.group('kw')
257 if not next.has_key(thiskw):
260 if thiskw in ('def', 'class'):
261 thisid = m.group('id')
265 elif line[i:i+1] in ('\n', '#'):
272 indent = len(line[:i].expandtabs(self.tabsize))
273 while indent < current:
276 s = '# end %s %s\n' % (
279 s = '# end %s\n' % firstkw
281 self.putline(s, current)
282 firstkw = lastkw = ''
284 current, firstkw, lastkw, topid = stack[-1]
287 if indent == current and firstkw:
290 self.error('mismatched end')
292 firstkw = lastkw = ''
293 elif not thiskw or thiskw in start:
295 s = '# end %s %s\n' % (
298 s = '# end %s\n' % firstkw
300 self.putline(s, current)
301 firstkw = lastkw = topid = ''
305 stack.append((current, firstkw, lastkw, topid))
306 if thiskw and thiskw not in start:
310 current, firstkw, lastkw, topid = \
311 indent, thiskw, thiskw, thisid
315 firstkw = lastkw = thiskw
321 for l in todo: self.write(l)
330 # end class PythonIndenter
332 # Simplified user interface
333 # - xxx_filter(input, output): read and write file objects
334 # - xxx_string(s): take and return string object
335 # - xxx_file(filename): process file in place, return true iff changed
337 def complete_filter(input = sys.stdin, output = sys.stdout,
338 stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
339 pi = PythonIndenter(input, output, stepsize, tabsize, expandtabs)
341 # end def complete_filter
343 def delete_filter(input= sys.stdin, output = sys.stdout,
344 stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
345 pi = PythonIndenter(input, output, stepsize, tabsize, expandtabs)
347 # end def delete_filter
349 def reformat_filter(input = sys.stdin, output = sys.stdout,
350 stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
351 pi = PythonIndenter(input, output, stepsize, tabsize, expandtabs)
353 # end def reformat_filter
356 def __init__(self, buf):
359 self.len = len(self.buf)
361 def read(self, n = 0):
363 n = self.len - self.pos
365 n = min(n, self.len - self.pos)
367 r = self.buf[self.pos : self.pos + n]
368 self.pos = self.pos + n
372 i = self.buf.find('\n', self.pos)
373 return self.read(i + 1 - self.pos)
377 line = self.readline()
380 line = self.readline()
384 # seek/tell etc. are left as an exercise for the reader
385 # end class StringReader
392 self.buf = self.buf + s
397 # end class StringWriter
399 def complete_string(source, stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
400 input = StringReader(source)
401 output = StringWriter()
402 pi = PythonIndenter(input, output, stepsize, tabsize, expandtabs)
404 return output.getvalue()
405 # end def complete_string
407 def delete_string(source, stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
408 input = StringReader(source)
409 output = StringWriter()
410 pi = PythonIndenter(input, output, stepsize, tabsize, expandtabs)
412 return output.getvalue()
413 # end def delete_string
415 def reformat_string(source, stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
416 input = StringReader(source)
417 output = StringWriter()
418 pi = PythonIndenter(input, output, stepsize, tabsize, expandtabs)
420 return output.getvalue()
421 # end def reformat_string
423 def complete_file(filename, stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
424 source = open(filename, 'r').read()
425 result = complete_string(source, stepsize, tabsize, expandtabs)
426 if source == result: return 0
429 try: os.rename(filename, filename + '~')
430 except os.error: pass
432 f = open(filename, 'w')
436 # end def complete_file
438 def delete_file(filename, stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
439 source = open(filename, 'r').read()
440 result = delete_string(source, stepsize, tabsize, expandtabs)
441 if source == result: return 0
444 try: os.rename(filename, filename + '~')
445 except os.error: pass
447 f = open(filename, 'w')
451 # end def delete_file
453 def reformat_file(filename, stepsize = STEPSIZE, tabsize = TABSIZE, expandtabs = EXPANDTABS):
454 source = open(filename, 'r').read()
455 result = reformat_string(source, stepsize, tabsize, expandtabs)
456 if source == result: return 0
459 try: os.rename(filename, filename + '~')
460 except os.error: pass
462 f = open(filename, 'w')
466 # end def reformat_file
468 # Test program when called as a script
471 usage: pindent (-c|-d|-r) [-s stepsize] [-t tabsize] [-e] [file] ...
472 -c : complete a correctly indented program (add #end directives)
473 -d : delete #end directives
474 -r : reformat a completed program (use #end directives)
475 -s stepsize: indentation step (default %(STEPSIZE)d)
476 -t tabsize : the worth in spaces of a tab (default %(TABSIZE)d)
477 -e : expand TABs into spaces (defailt OFF)
478 [file] ... : files are changed in place, with backups in file~
479 If no files are specified or a single - is given,
480 the program acts as a filter (reads stdin, writes stdout).
483 def error_both(op1, op2):
484 sys.stderr.write('Error: You can not specify both '+op1+' and -'+op2[0]+' at the same time\n')
485 sys.stderr.write(usage)
492 opts, args = getopt.getopt(sys.argv[1:], 'cdrs:t:e')
493 except getopt.error, msg:
494 sys.stderr.write('Error: %s\n' % msg)
495 sys.stderr.write(usage)
501 expandtabs = EXPANDTABS
504 if action: error_both(o, action)
508 if action: error_both(o, action)
512 if action: error_both(o, action)
525 'You must specify -c(omplete), -d(elete) or -r(eformat)\n')
526 sys.stderr.write(usage)
529 if not args or args == ['-']:
530 action = eval(action + '_filter')
531 action(sys.stdin, sys.stdout, stepsize, tabsize, expandtabs)
533 action = eval(action + '_file')
534 for filename in args:
535 action(filename, stepsize, tabsize, expandtabs)
540 if __name__ == '__main__':