style.py revision 8228:59d3bfa85f16 
1#! /usr/bin/env python
2# Copyright (c) 2006 The Regents of The University of Michigan
3# Copyright (c) 2007,2011 The Hewlett-Packard Development Company
4# All rights reserved.
5#
6# Redistribution and use in source and binary forms, with or without
7# modification, are permitted provided that the following conditions are
8# met: redistributions of source code must retain the above copyright
9# notice, this list of conditions and the following disclaimer;
10# redistributions in binary form must reproduce the above copyright
11# notice, this list of conditions and the following disclaimer in the
12# documentation and/or other materials provided with the distribution;
13# neither the name of the copyright holders nor the names of its
14# contributors may be used to endorse or promote products derived from
15# this software without specific prior written permission.
16#
17# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
18# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
19# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
20# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
21# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
22# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
23# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
24# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
25# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
26# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
27# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
28#
29# Authors: Nathan Binkert
30
31import heapq
32import os
33import re
34import sys
35
36from os.path import dirname, join as joinpath
37from itertools import count
38from mercurial import bdiff, mdiff
39
40current_dir = dirname(__file__)
41sys.path.insert(0, current_dir)
42sys.path.insert(1, joinpath(dirname(current_dir), 'src', 'python'))
43
44from m5.util import neg_inf, pos_inf, Region, Regions
45import sort_includes
46from file_types import lang_type
47
48all_regions = Region(neg_inf, pos_inf)
49
50tabsize = 8
51lead = re.compile(r'^([ \t]+)')
52trail = re.compile(r'([ \t]+)$')
53any_control = re.compile(r'\b(if|while|for)[ \t]*[(]')
54good_control = re.compile(r'\b(if|while|for) [(]')
55
56format_types = set(('C', 'C++'))
57
58def modified_regions(old_data, new_data):
59    regions = Regions()
60    beg = None
61    for pbeg, pend, fbeg, fend in bdiff.blocks(old_data, new_data):
62        if beg is not None and beg != fbeg:
63            regions.append(beg, fbeg)
64        beg = fend
65    return regions
66
67def modregions(wctx, fname):
68    fctx = wctx.filectx(fname)
69    pctx = fctx.parents()
70
71    file_data = fctx.data()
72    lines = mdiff.splitnewlines(file_data)
73    if len(pctx) in (1, 2):
74        mod_regions = modified_regions(pctx[0].data(), file_data)
75        if len(pctx) == 2:
76            m2 = modified_regions(pctx[1].data(), file_data)
77            # only the lines that are new in both
78            mod_regions &= m2
79    else:
80        mod_regions = Regions()
81        mod_regions.add(0, len(lines))
82
83    return mod_regions
84
85class UserInterface(object):
86    def __init__(self, verbose=False, auto=False):
87        self.auto = auto
88        self.verbose = verbose
89
90    def prompt(self, prompt, results, default):
91        if self.auto:
92            return self.auto
93
94        while True:
95            result = self.do_prompt(prompt, results, default)
96            if result in results:
97                return result
98
99class MercurialUI(UserInterface):
100    def __init__(self, ui, *args, **kwargs):
101        super(MercurialUI, self).__init__(*args, **kwargs)
102        self.ui = ui
103
104    def do_prompt(self, prompt, results, default):
105        return self.ui.prompt(prompt, default=default)
106
107    def write(self, string):
108        self.ui.write(string)
109
110class StdioUI(UserInterface):
111    def do_prompt(self, prompt, results, default):
112        return raw_input(prompt) or default
113
114    def write(self, string):
115        sys.stdout.write(string)
116
117class Region(object):
118    def __init__(self, asdf):
119        self.regions = Foo
120
121class Verifier(object):
122    def __init__(self, ui, repo=None):
123        self.ui = ui
124        self.repo = repo
125        if repo is None:
126            self.wctx = None
127
128    def __getattr__(self, attr):
129        if attr in ('prompt', 'write'):
130            return getattr(self.ui, attr)
131
132        if attr == 'wctx':
133            try:
134                wctx = repo.workingctx()
135            except:
136                from mercurial import context
137                wctx = context.workingctx(repo)
138            self.wctx = wctx
139            return wctx
140
141        raise AttributeError
142
143    def open(self, filename, mode):
144        if self.repo:
145            filename = self.repo.wjoin(filename)
146
147        try:
148            f = file(filename, mode)
149        except OSError, msg:
150            print 'could not open file %s: %s' % (filename, msg)
151            return None
152
153        return f
154
155    def skip(self, filename):
156        return lang_type(filename) not in self.languages
157
158    def check(self, filename, regions=all_regions):
159        f = self.open(filename, 'r')
160
161        errors = 0
162        for num,line in enumerate(f):
163            if num not in regions:
164                continue
165            if not self.check_line(line):
166                self.write("invalid %s in %s:%d\n" % \
167                               (self.test_name, filename, num + 1))
168                if self.ui.verbose:
169                    self.write(">>%s<<\n" % line[-1])
170                errors += 1
171        return errors
172
173    def fix(self, filename, regions=all_regions):
174        f = self.open(filename, 'r+')
175
176        lines = list(f)
177
178        f.seek(0)
179        f.truncate()
180
181        for i,line in enumerate(lines):
182            if i in regions:
183                line = self.fix_line(line)
184
185            f.write(line)
186        f.close()
187
188    def apply(self, filename, prompt, regions=all_regions):
189        if not self.skip(filename):
190            errors = self.check(filename, regions)
191            if errors:
192                if prompt(filename, self.fix, regions):
193                    return True
194        return False
195
196
197class Whitespace(Verifier):
198    languages = set(('C', 'C++', 'swig', 'python', 'asm', 'isa', 'scons'))
199    test_name = 'whitespace'
200    def check_line(self, line):
201        match = lead.search(line)
202        if match and match.group(1).find('\t') != -1:
203            return False
204
205        match = trail.search(line)
206        if match:
207            return False
208
209        return True
210
211    def fix_line(self, line):
212        if lead.search(line):
213            newline = ''
214            for i,c in enumerate(line):
215                if c == ' ':
216                    newline += ' '
217                elif c == '\t':
218                    newline += ' ' * (tabsize - len(newline) % tabsize)
219                else:
220                    newline += line[i:]
221                    break
222
223            line = newline
224
225        return line.rstrip() + '\n'
226
227class SortedIncludes(Verifier):
228    languages = sort_includes.default_languages
229    def __init__(self, *args, **kwargs):
230        super(SortedIncludes, self).__init__(*args, **kwargs)
231        self.sort_includes = sort_includes.SortIncludes()
232
233    def check(self, filename, regions=all_regions):
234        f = self.open(filename, 'r')
235
236        lines = [ l.rstrip('\n') for l in f.xreadlines() ]
237        old = ''.join(line + '\n' for line in lines)
238        f.close()
239
240        language = lang_type(filename, lines[0])
241        sort_lines = list(self.sort_includes(lines, filename, language))
242        new = ''.join(line + '\n' for line in sort_lines)
243
244        mod = modified_regions(old, new)
245        modified = mod & regions
246        print mod, regions, modified
247
248        if modified:
249            self.write("invalid sorting of includes\n")
250            if self.ui.verbose:
251                for start, end in modified.regions:
252                    self.write("bad region [%d, %d)\n" % (start, end))
253            return 1
254
255        return 0
256
257    def fix(self, filename, regions=all_regions):
258        f = self.open(filename, 'r+')
259
260        old = f.readlines()
261        lines = [ l.rstrip('\n') for l in old ]
262        language = lang_type(filename, lines[0])
263        sort_lines = list(self.sort_includes(lines, filename, language))
264        new = ''.join(line + '\n' for line in sort_lines)
265
266        f.seek(0)
267        f.truncate()
268
269        for i,line in enumerate(sort_lines):
270            f.write(line)
271            f.write('\n')
272        f.close()
273
274def linelen(line):
275    tabs = line.count('\t')
276    if not tabs:
277        return len(line)
278
279    count = 0
280    for c in line:
281        if c == '\t':
282            count += tabsize - count % tabsize
283        else:
284            count += 1
285
286    return count
287
288class ValidationStats(object):
289    def __init__(self):
290        self.toolong = 0
291        self.toolong80 = 0
292        self.leadtabs = 0
293        self.trailwhite = 0
294        self.badcontrol = 0
295        self.cret = 0
296
297    def dump(self):
298        print '''\
299%d violations of lines over 79 chars. %d of which are 80 chars exactly.
300%d cases of whitespace at the end of a line.
301%d cases of tabs to indent.
302%d bad parens after if/while/for.
303%d carriage returns found.
304''' % (self.toolong, self.toolong80, self.trailwhite, self.leadtabs,
305       self.badcontrol, self.cret)
306
307    def __nonzero__(self):
308        return self.toolong or self.toolong80 or self.leadtabs or \
309               self.trailwhite or self.badcontrol or self.cret
310
311def validate(filename, stats, verbose, exit_code):
312    if lang_type(filename) not in format_types:
313        return
314
315    def msg(lineno, line, message):
316        print '%s:%d>' % (filename, lineno + 1), message
317        if verbose > 2:
318            print line
319
320    def bad():
321        if exit_code is not None:
322            sys.exit(exit_code)
323
324    try:
325        f = file(filename, 'r')
326    except OSError:
327        if verbose > 0:
328            print 'could not open file %s' % filename
329        bad()
330        return
331
332    for i,line in enumerate(f):
333        line = line.rstrip('\n')
334
335        # no carriage returns
336        if line.find('\r') != -1:
337            self.cret += 1
338            if verbose > 1:
339                msg(i, line, 'carriage return found')
340            bad()
341
342        # lines max out at 79 chars
343        llen = linelen(line)
344        if llen > 79:
345            stats.toolong += 1
346            if llen == 80:
347                stats.toolong80 += 1
348            if verbose > 1:
349                msg(i, line, 'line too long (%d chars)' % llen)
350            bad()
351
352        # no tabs used to indent
353        match = lead.search(line)
354        if match and match.group(1).find('\t') != -1:
355            stats.leadtabs += 1
356            if verbose > 1:
357                msg(i, line, 'using tabs to indent')
358            bad()
359
360        # no trailing whitespace
361        if trail.search(line):
362            stats.trailwhite +=1
363            if verbose > 1:
364                msg(i, line, 'trailing whitespace')
365            bad()
366
367        # for c++, exactly one space betwen if/while/for and (
368        if cpp:
369            match = any_control.search(line)
370            if match and not good_control.search(line):
371                stats.badcontrol += 1
372                if verbose > 1:
373                    msg(i, line, 'improper spacing after %s' % match.group(1))
374                bad()
375
376def do_check_style(hgui, repo, *files, **args):
377    """check files for proper m5 style guidelines"""
378    from mercurial import mdiff, util
379
380    auto = args.get('auto', False)
381    if auto:
382        auto = 'f'
383    ui = MercurialUI(hgui, hgui.verbose, auto)
384
385    if files:
386        files = frozenset(files)
387
388    def skip(name):
389        return files and name in files
390
391    def prompt(name, func, regions=all_regions):
392        result = ui.prompt("(a)bort, (i)gnore, or (f)ix?", 'aif', 'a')
393        if result == 'a':
394            return True
395        elif result == 'f':
396            func(repo.wjoin(name), regions)
397
398        return False
399
400    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
401
402    whitespace = Whitespace(ui)
403    sorted_includes = SortedIncludes(ui)
404    for fname in added:
405        if skip(fname):
406            continue
407
408        if whitespace.apply(fname, prompt):
409            return True
410
411        if sorted_includes.apply(fname, prompt):
412            return True
413
414    try:
415        wctx = repo.workingctx()
416    except:
417        from mercurial import context
418        wctx = context.workingctx(repo)
419
420    for fname in modified:
421        if skip(fname):
422            continue
423
424        regions = modregions(wctx, fname)
425
426        if whitespace.apply(fname, prompt, regions):
427            return True
428
429        if sorted_includes.apply(fname, prompt, regions):
430            return True
431
432    return False
433
434def do_check_format(hgui, repo, **args):
435    ui = MercurialUI(hgui, hgui.verbose, auto)
436
437    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
438
439    verbose = 0
440    stats = ValidationStats()
441    for f in modified + added:
442        validate(f, stats, verbose, None)
443
444    if stats:
445        stats.dump()
446        result = ui.prompt("invalid formatting\n(i)gnore or (a)bort?",
447                           'ai', 'a')
448        if result == 'a':
449            return True
450
451    return False
452
453def check_hook(hooktype):
454    if hooktype not in ('pretxncommit', 'pre-qrefresh'):
455        raise AttributeError, \
456              "This hook is not meant for %s" % hooktype
457
458def check_style(ui, repo, hooktype, **kwargs):
459    check_hook(hooktype)
460    args = {}
461
462    try:
463        return do_check_style(ui, repo, **args)
464    except Exception, e:
465        import traceback
466        traceback.print_exc()
467        return True
468
469def check_format(ui, repo, hooktype, **kwargs):
470    check_hook(hooktype)
471    args = {}
472
473    try:
474        return do_check_format(ui, repo, **args)
475    except Exception, e:
476        import traceback
477        traceback.print_exc()
478        return True
479
480try:
481    from mercurial.i18n import _
482except ImportError:
483    def _(arg):
484        return arg
485
486cmdtable = {
487    '^m5style' :
488    ( do_check_style,
489      [ ('a', 'auto', False, _("automatically fix whitespace")) ],
490      _('hg m5style [-a] [FILE]...')),
491    '^m5format' :
492    ( do_check_format,
493      [ ],
494      _('hg m5format [FILE]...')),
495}
496
497if __name__ == '__main__':
498    import getopt
499
500    progname = sys.argv[0]
501    if len(sys.argv) < 2:
502        sys.exit('usage: %s <command> [<command args>]' % progname)
503
504    fixwhite_usage = '%s fixwhite [-t <tabsize> ] <path> [...] \n' % progname
505    chkformat_usage = '%s chkformat <path> [...] \n' % progname
506    chkwhite_usage = '%s chkwhite <path> [...] \n' % progname
507
508    command = sys.argv[1]
509    if command == 'fixwhite':
510        flags = 't:'
511        usage = fixwhite_usage
512    elif command == 'chkwhite':
513        flags = 'nv'
514        usage = chkwhite_usage
515    elif command == 'chkformat':
516        flags = 'nv'
517        usage = chkformat_usage
518    else:
519        sys.exit(fixwhite_usage + chkwhite_usage + chkformat_usage)
520
521    opts, args = getopt.getopt(sys.argv[2:], flags)
522
523    code = 1
524    verbose = 1
525    for opt,arg in opts:
526        if opt == '-n':
527            code = None
528        if opt == '-t':
529            tabsize = int(arg)
530        if opt == '-v':
531            verbose += 1
532
533    if command == 'fixwhite':
534        for filename in args:
535            fixwhite(filename, tabsize)
536    elif command == 'chkwhite':
537        for filename in args:
538            for line,num in checkwhite(filename):
539                print 'invalid whitespace: %s:%d' % (filename, num)
540                if verbose:
541                    print '>>%s<<' % line[:-1]
542    elif command == 'chkformat':
543        stats = ValidationStats()
544        for filename in args:
545            validate(filename, stats=stats, verbose=verbose, exit_code=code)
546
547        if verbose > 0:
548            stats.dump()
549    else:
550        sys.exit("command '%s' not found" % command)
551