Ditched '_find_SET()', since it was a no-value-added wrapper around
[python/dscho.git] / Tools / scripts / pindent.py
blob4c8e71de1379dd530d76cb7ee8c6c54b94545706
1 #! /usr/bin/env python
3 # This file contains a class and a main program that perform two
4 # related (though complimentary) formatting operations on Python
5 # programs. When called as "pindend -c", it takes a valid Python
6 # program as input and outputs a version augmented with block-closing
7 # comments. When called as "pindent -r" it assumes its input is a
8 # Python program with block-closing comments but with its indentation
9 # messed up, and outputs a properly indented version.
11 # A "block-closing comment" is a comment of the form '# end <keyword>'
12 # where <keyword> is the keyword that opened the block. If the
13 # opening keyword is 'def' or 'class', the function or class name may
14 # be repeated in the block-closing comment as well. Here is an
15 # example of a program fully augmented with block-closing comments:
17 # def foobar(a, b):
18 # if a == b:
19 # a = a+1
20 # elif a < b:
21 # b = b-1
22 # if b > a: a = a-1
23 # # end if
24 # else:
25 # print 'oops!'
26 # # end if
27 # # end def foobar
29 # Note that only the last part of an if...elif...else... block needs a
30 # block-closing comment; the same is true for other compound
31 # statements (e.g. try...except). Also note that "short-form" blocks
32 # like the second 'if' in the example must be closed as well;
33 # otherwise the 'else' in the example would be ambiguous (remember
34 # that indentation is not significant when interpreting block-closing
35 # comments).
37 # Both operations are idempotent (i.e. applied to their own output
38 # they yield an identical result). Running first "pindent -c" and
39 # then "pindent -r" on a valid Python program produces a program that
40 # is semantically identical to the input (though its indentation may
41 # be different).
43 # Other options:
44 # -s stepsize: set the indentation step size (default 8)
45 # -t tabsize : set the number of spaces a tab character is worth (default 8)
46 # file ... : input file(s) (default standard input)
47 # The results always go to standard output
49 # Caveats:
50 # - comments ending in a backslash will be mistaken for continued lines
51 # - continuations using backslash are always left unchanged
52 # - continuations inside parentheses are not extra indented by -r
53 # but must be indented for -c to work correctly (this breaks
54 # idempotency!)
55 # - continued lines inside triple-quoted strings are totally garbled
57 # Secret feature:
58 # - On input, a block may also be closed with an "end statement" --
59 # this is a block-closing comment without the '#' sign.
61 # Possible improvements:
62 # - check syntax based on transitions in 'next' table
63 # - better error reporting
64 # - better error recovery
65 # - check identifier after class/def
67 # The following wishes need a more complete tokenization of the source:
68 # - Don't get fooled by comments ending in backslash
69 # - reindent continuation lines indicated by backslash
70 # - handle continuation lines inside parentheses/braces/brackets
71 # - handle triple quoted strings spanning lines
72 # - realign comments
73 # - optionally do much more thorough reformatting, a la C indent
75 # Defaults
76 STEPSIZE = 8
77 TABSIZE = 8
79 import os
80 import re
81 import string
82 import sys
84 next = {}
85 next['if'] = next['elif'] = 'elif', 'else', 'end'
86 next['while'] = next['for'] = 'else', 'end'
87 next['try'] = 'except', 'finally'
88 next['except'] = 'except', 'else', 'end'
89 next['else'] = next['finally'] = next['def'] = next['class'] = 'end'
90 next['end'] = ()
91 start = 'if', 'while', 'for', 'try', 'def', 'class'
93 class PythonIndenter:
95 def __init__(self, fpi = sys.stdin, fpo = sys.stdout,
96 indentsize = STEPSIZE, tabsize = TABSIZE):
97 self.fpi = fpi
98 self.fpo = fpo
99 self.indentsize = indentsize
100 self.tabsize = tabsize
101 self.lineno = 0
102 self.write = fpo.write
103 self.kwprog = re.compile(
104 r'^\s*(?P<kw>[a-z]+)'
105 r'(\s+(?P<id>[a-zA-Z_]\w*))?'
106 r'[^\w]')
107 self.endprog = re.compile(
108 r'^\s*#?\s*end\s+(?P<kw>[a-z]+)'
109 r'(\s+(?P<id>[a-zA-Z_]\w*))?'
110 r'[^\w]')
111 self.wsprog = re.compile(r'^[ \t]*')
112 # end def __init__
114 def readline(self):
115 line = self.fpi.readline()
116 if line: self.lineno = self.lineno + 1
117 # end if
118 return line
119 # end def readline
121 def error(self, fmt, *args):
122 if args: fmt = fmt % args
123 # end if
124 sys.stderr.write('Error at line %d: %s\n' % (self.lineno, fmt))
125 self.write('### %s ###\n' % fmt)
126 # end def error
128 def getline(self):
129 line = self.readline()
130 while line[-2:] == '\\\n':
131 line2 = self.readline()
132 if not line2: break
133 # end if
134 line = line + line2
135 # end while
136 return line
137 # end def getline
139 def putline(self, line, indent = None):
140 if indent is None:
141 self.write(line)
142 return
143 # end if
144 tabs, spaces = divmod(indent*self.indentsize, self.tabsize)
145 i = 0
146 m = self.wsprog.match(line)
147 if m: i = m.end()
148 # end if
149 self.write('\t'*tabs + ' '*spaces + line[i:])
150 # end def putline
152 def reformat(self):
153 stack = []
154 while 1:
155 line = self.getline()
156 if not line: break # EOF
157 # end if
158 m = self.endprog.match(line)
159 if m:
160 kw = 'end'
161 kw2 = m.group('kw')
162 if not stack:
163 self.error('unexpected end')
164 elif stack[-1][0] != kw2:
165 self.error('unmatched end')
166 # end if
167 del stack[-1:]
168 self.putline(line, len(stack))
169 continue
170 # end if
171 m = self.kwprog.match(line)
172 if m:
173 kw = m.group('kw')
174 if kw in start:
175 self.putline(line, len(stack))
176 stack.append((kw, kw))
177 continue
178 # end if
179 if next.has_key(kw) and stack:
180 self.putline(line, len(stack)-1)
181 kwa, kwb = stack[-1]
182 stack[-1] = kwa, kw
183 continue
184 # end if
185 # end if
186 self.putline(line, len(stack))
187 # end while
188 if stack:
189 self.error('unterminated keywords')
190 for kwa, kwb in stack:
191 self.write('\t%s\n' % kwa)
192 # end for
193 # end if
194 # end def reformat
196 def complete(self):
197 self.indentsize = 1
198 stack = []
199 todo = []
200 current, firstkw, lastkw, topid = 0, '', '', ''
201 while 1:
202 line = self.getline()
203 i = 0
204 m = self.wsprog.match(line)
205 if m: i = m.end()
206 # end if
207 m = self.endprog.match(line)
208 if m:
209 thiskw = 'end'
210 endkw = m.group('kw')
211 thisid = m.group('id')
212 else:
213 m = self.kwprog.match(line)
214 if m:
215 thiskw = m.group('kw')
216 if not next.has_key(thiskw):
217 thiskw = ''
218 # end if
219 if thiskw in ('def', 'class'):
220 thisid = m.group('id')
221 else:
222 thisid = ''
223 # end if
224 elif line[i:i+1] in ('\n', '#'):
225 todo.append(line)
226 continue
227 else:
228 thiskw = ''
229 # end if
230 # end if
231 indent = len(string.expandtabs(line[:i], self.tabsize))
232 while indent < current:
233 if firstkw:
234 if topid:
235 s = '# end %s %s\n' % (
236 firstkw, topid)
237 else:
238 s = '# end %s\n' % firstkw
239 # end if
240 self.putline(s, current)
241 firstkw = lastkw = ''
242 # end if
243 current, firstkw, lastkw, topid = stack[-1]
244 del stack[-1]
245 # end while
246 if indent == current and firstkw:
247 if thiskw == 'end':
248 if endkw != firstkw:
249 self.error('mismatched end')
250 # end if
251 firstkw = lastkw = ''
252 elif not thiskw or thiskw in start:
253 if topid:
254 s = '# end %s %s\n' % (
255 firstkw, topid)
256 else:
257 s = '# end %s\n' % firstkw
258 # end if
259 self.putline(s, current)
260 firstkw = lastkw = topid = ''
261 # end if
262 # end if
263 if indent > current:
264 stack.append((current, firstkw, lastkw, topid))
265 if thiskw and thiskw not in start:
266 # error
267 thiskw = ''
268 # end if
269 current, firstkw, lastkw, topid = \
270 indent, thiskw, thiskw, thisid
271 # end if
272 if thiskw:
273 if thiskw in start:
274 firstkw = lastkw = thiskw
275 topid = thisid
276 else:
277 lastkw = thiskw
278 # end if
279 # end if
280 for l in todo: self.write(l)
281 # end for
282 todo = []
283 if not line: break
284 # end if
285 self.write(line)
286 # end while
287 # end def complete
289 # end class PythonIndenter
291 # Simplified user interface
292 # - xxx_filter(input, output): read and write file objects
293 # - xxx_string(s): take and return string object
294 # - xxx_file(filename): process file in place, return true iff changed
296 def complete_filter(input= sys.stdin, output = sys.stdout,
297 stepsize = STEPSIZE, tabsize = TABSIZE):
298 pi = PythonIndenter(input, output, stepsize, tabsize)
299 pi.complete()
300 # end def complete_filter
302 def reformat_filter(input = sys.stdin, output = sys.stdout,
303 stepsize = STEPSIZE, tabsize = TABSIZE):
304 pi = PythonIndenter(input, output, stepsize, tabsize)
305 pi.reformat()
306 # end def reformat
308 class StringReader:
309 def __init__(self, buf):
310 self.buf = buf
311 self.pos = 0
312 self.len = len(self.buf)
313 # end def __init__
314 def read(self, n = 0):
315 if n <= 0:
316 n = self.len - self.pos
317 else:
318 n = min(n, self.len - self.pos)
319 # end if
320 r = self.buf[self.pos : self.pos + n]
321 self.pos = self.pos + n
322 return r
323 # end def read
324 def readline(self):
325 i = string.find(self.buf, '\n', self.pos)
326 return self.read(i + 1 - self.pos)
327 # end def readline
328 def readlines(self):
329 lines = []
330 line = self.readline()
331 while line:
332 lines.append(line)
333 line = self.readline()
334 # end while
335 return lines
336 # end def readlines
337 # seek/tell etc. are left as an exercise for the reader
338 # end class StringReader
340 class StringWriter:
341 def __init__(self):
342 self.buf = ''
343 # end def __init__
344 def write(self, s):
345 self.buf = self.buf + s
346 # end def write
347 def getvalue(self):
348 return self.buf
349 # end def getvalue
350 # end class StringWriter
352 def complete_string(source, stepsize = STEPSIZE, tabsize = TABSIZE):
353 input = StringReader(source)
354 output = StringWriter()
355 pi = PythonIndenter(input, output, stepsize, tabsize)
356 pi.complete()
357 return output.getvalue()
358 # end def complete_string
360 def reformat_string(source, stepsize = STEPSIZE, tabsize = TABSIZE):
361 input = StringReader(source)
362 output = StringWriter()
363 pi = PythonIndenter(input, output, stepsize, tabsize)
364 pi.reformat()
365 return output.getvalue()
366 # end def reformat_string
368 def complete_file(filename, stepsize = STEPSIZE, tabsize = TABSIZE):
369 source = open(filename, 'r').read()
370 result = complete_string(source, stepsize, tabsize)
371 if source == result: return 0
372 # end if
373 import os
374 try: os.rename(filename, filename + '~')
375 except os.error: pass
376 # end try
377 f = open(filename, 'w')
378 f.write(result)
379 f.close()
380 return 1
381 # end def complete_file
383 def reformat_file(filename, stepsize = STEPSIZE, tabsize = TABSIZE):
384 source = open(filename, 'r').read()
385 result = reformat_string(source, stepsize, tabsize)
386 if source == result: return 0
387 # end if
388 import os
389 os.rename(filename, filename + '~')
390 f = open(filename, 'w')
391 f.write(result)
392 f.close()
393 return 1
394 # end def reformat_file
396 # Test program when called as a script
398 usage = """
399 usage: pindent (-c|-r) [-s stepsize] [-t tabsize] [file] ...
400 -c : complete a correctly indented program (add #end directives)
401 -r : reformat a completed program (use #end directives)
402 -s stepsize: indentation step (default %(STEPSIZE)d)
403 -t tabsize : the worth in spaces of a tab (default %(TABSIZE)d)
404 [file] ... : files are changed in place, with backups in file~
405 If no files are specified or a single - is given,
406 the program acts as a filter (reads stdin, writes stdout).
407 """ % vars()
409 def test():
410 import getopt
411 try:
412 opts, args = getopt.getopt(sys.argv[1:], 'crs:t:')
413 except getopt.error, msg:
414 sys.stderr.write('Error: %s\n' % msg)
415 sys.stderr.write(usage)
416 sys.exit(2)
417 # end try
418 action = None
419 stepsize = STEPSIZE
420 tabsize = TABSIZE
421 for o, a in opts:
422 if o == '-c':
423 action = 'complete'
424 elif o == '-r':
425 action = 'reformat'
426 elif o == '-s':
427 stepsize = string.atoi(a)
428 elif o == '-t':
429 tabsize = string.atoi(a)
430 # end if
431 # end for
432 if not action:
433 sys.stderr.write(
434 'You must specify -c(omplete) or -r(eformat)\n')
435 sys.stderr.write(usage)
436 sys.exit(2)
437 # end if
438 if not args or args == ['-']:
439 action = eval(action + '_filter')
440 action(sys.stdin, sys.stdout, stepsize, tabsize)
441 else:
442 action = eval(action + '_file')
443 for file in args:
444 action(file, stepsize, tabsize)
445 # end for
446 # end if
447 # end def test
449 if __name__ == '__main__':
450 test()
451 # end if