style.py revision 8238:d8ec0a7b3f0c 
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 = 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 Verifier(object):
118    def __init__(self, ui, repo=None):
119        self.ui = ui
120        self.repo = repo
121        if repo is None:
122            self.wctx = None
123
124    def __getattr__(self, attr):
125        if attr in ('prompt', 'write'):
126            return getattr(self.ui, attr)
127
128        if attr == 'wctx':
129            try:
130                wctx = repo.workingctx()
131            except:
132                from mercurial import context
133                wctx = context.workingctx(repo)
134            self.wctx = wctx
135            return wctx
136
137        raise AttributeError
138
139    def open(self, filename, mode):
140        if self.repo:
141            filename = self.repo.wjoin(filename)
142
143        try:
144            f = file(filename, mode)
145        except OSError, msg:
146            print 'could not open file %s: %s' % (filename, msg)
147            return None
148
149        return f
150
151    def skip(self, filename):
152        return lang_type(filename) not in self.languages
153
154    def check(self, filename, regions=all_regions):
155        f = self.open(filename, 'r')
156
157        errors = 0
158        for num,line in enumerate(f):
159            if num not in regions:
160                continue
161            if not self.check_line(line):
162                self.write("invalid %s in %s:%d\n" % \
163                               (self.test_name, filename, num + 1))
164                if self.ui.verbose:
165                    self.write(">>%s<<\n" % line[-1])
166                errors += 1
167        return errors
168
169    def fix(self, filename, regions=all_regions):
170        f = self.open(filename, 'r+')
171
172        lines = list(f)
173
174        f.seek(0)
175        f.truncate()
176
177        for i,line in enumerate(lines):
178            if i in regions:
179                line = self.fix_line(line)
180
181            f.write(line)
182        f.close()
183
184    def apply(self, filename, prompt, regions=all_regions):
185        if not self.skip(filename):
186            errors = self.check(filename, regions)
187            if errors:
188                if prompt(filename, self.fix, regions):
189                    return True
190        return False
191
192
193class Whitespace(Verifier):
194    languages = set(('C', 'C++', 'swig', 'python', 'asm', 'isa', 'scons'))
195    test_name = 'whitespace'
196    def check_line(self, line):
197        match = lead.search(line)
198        if match and match.group(1).find('\t') != -1:
199            return False
200
201        match = trail.search(line)
202        if match:
203            return False
204
205        return True
206
207    def fix_line(self, line):
208        if lead.search(line):
209            newline = ''
210            for i,c in enumerate(line):
211                if c == ' ':
212                    newline += ' '
213                elif c == '\t':
214                    newline += ' ' * (tabsize - len(newline) % tabsize)
215                else:
216                    newline += line[i:]
217                    break
218
219            line = newline
220
221        return line.rstrip() + '\n'
222
223class SortedIncludes(Verifier):
224    languages = sort_includes.default_languages
225    def __init__(self, *args, **kwargs):
226        super(SortedIncludes, self).__init__(*args, **kwargs)
227        self.sort_includes = sort_includes.SortIncludes()
228
229    def check(self, filename, regions=all_regions):
230        f = self.open(filename, 'r')
231
232        lines = [ l.rstrip('\n') for l in f.xreadlines() ]
233        old = ''.join(line + '\n' for line in lines)
234        f.close()
235
236        language = lang_type(filename, lines[0])
237        sort_lines = list(self.sort_includes(lines, filename, language))
238        new = ''.join(line + '\n' for line in sort_lines)
239
240        mod = modified_regions(old, new)
241        modified = mod & regions
242
243        if modified:
244            self.write("invalid sorting of includes\n")
245            if self.ui.verbose:
246                for start, end in modified.regions:
247                    self.write("bad region [%d, %d)\n" % (start, end))
248            return 1
249
250        return 0
251
252    def fix(self, filename, regions=all_regions):
253        f = self.open(filename, 'r+')
254
255        old = f.readlines()
256        lines = [ l.rstrip('\n') for l in old ]
257        language = lang_type(filename, lines[0])
258        sort_lines = list(self.sort_includes(lines, filename, language))
259        new = ''.join(line + '\n' for line in sort_lines)
260
261        f.seek(0)
262        f.truncate()
263
264        for i,line in enumerate(sort_lines):
265            f.write(line)
266            f.write('\n')
267        f.close()
268
269def linelen(line):
270    tabs = line.count('\t')
271    if not tabs:
272        return len(line)
273
274    count = 0
275    for c in line:
276        if c == '\t':
277            count += tabsize - count % tabsize
278        else:
279            count += 1
280
281    return count
282
283class ValidationStats(object):
284    def __init__(self):
285        self.toolong = 0
286        self.toolong80 = 0
287        self.leadtabs = 0
288        self.trailwhite = 0
289        self.badcontrol = 0
290        self.cret = 0
291
292    def dump(self):
293        print '''\
294%d violations of lines over 79 chars. %d of which are 80 chars exactly.
295%d cases of whitespace at the end of a line.
296%d cases of tabs to indent.
297%d bad parens after if/while/for.
298%d carriage returns found.
299''' % (self.toolong, self.toolong80, self.trailwhite, self.leadtabs,
300       self.badcontrol, self.cret)
301
302    def __nonzero__(self):
303        return self.toolong or self.toolong80 or self.leadtabs or \
304               self.trailwhite or self.badcontrol or self.cret
305
306def validate(filename, stats, verbose, exit_code):
307    if lang_type(filename) not in format_types:
308        return
309
310    def msg(lineno, line, message):
311        print '%s:%d>' % (filename, lineno + 1), message
312        if verbose > 2:
313            print line
314
315    def bad():
316        if exit_code is not None:
317            sys.exit(exit_code)
318
319    try:
320        f = file(filename, 'r')
321    except OSError:
322        if verbose > 0:
323            print 'could not open file %s' % filename
324        bad()
325        return
326
327    for i,line in enumerate(f):
328        line = line.rstrip('\n')
329
330        # no carriage returns
331        if line.find('\r') != -1:
332            self.cret += 1
333            if verbose > 1:
334                msg(i, line, 'carriage return found')
335            bad()
336
337        # lines max out at 79 chars
338        llen = linelen(line)
339        if llen > 79:
340            stats.toolong += 1
341            if llen == 80:
342                stats.toolong80 += 1
343            if verbose > 1:
344                msg(i, line, 'line too long (%d chars)' % llen)
345            bad()
346
347        # no tabs used to indent
348        match = lead.search(line)
349        if match and match.group(1).find('\t') != -1:
350            stats.leadtabs += 1
351            if verbose > 1:
352                msg(i, line, 'using tabs to indent')
353            bad()
354
355        # no trailing whitespace
356        if trail.search(line):
357            stats.trailwhite +=1
358            if verbose > 1:
359                msg(i, line, 'trailing whitespace')
360            bad()
361
362        # for c++, exactly one space betwen if/while/for and (
363        if cpp:
364            match = any_control.search(line)
365            if match and not good_control.search(line):
366                stats.badcontrol += 1
367                if verbose > 1:
368                    msg(i, line, 'improper spacing after %s' % match.group(1))
369                bad()
370
371def do_check_style(hgui, repo, *files, **args):
372    """check files for proper m5 style guidelines"""
373    from mercurial import mdiff, util
374
375    auto = args.get('auto', False)
376    if auto:
377        auto = 'f'
378    ui = MercurialUI(hgui, hgui.verbose, auto)
379
380    if files:
381        files = frozenset(files)
382
383    def skip(name):
384        return files and name in files
385
386    def prompt(name, func, regions=all_regions):
387        result = ui.prompt("(a)bort, (i)gnore, or (f)ix?", 'aif', 'a')
388        if result == 'a':
389            return True
390        elif result == 'f':
391            func(repo.wjoin(name), regions)
392
393        return False
394
395    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
396
397    whitespace = Whitespace(ui)
398    sorted_includes = SortedIncludes(ui)
399    for fname in added:
400        if skip(fname):
401            continue
402
403        if whitespace.apply(fname, prompt):
404            return True
405
406        if sorted_includes.apply(fname, prompt):
407            return True
408
409    try:
410        wctx = repo.workingctx()
411    except:
412        from mercurial import context
413        wctx = context.workingctx(repo)
414
415    for fname in modified:
416        if skip(fname):
417            continue
418
419        regions = modregions(wctx, fname)
420
421        if whitespace.apply(fname, prompt, regions):
422            return True
423
424        if sorted_includes.apply(fname, prompt, regions):
425            return True
426
427    return False
428
429def do_check_format(hgui, repo, **args):
430    ui = MercurialUI(hgui, hgui.verbose, auto)
431
432    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
433
434    verbose = 0
435    stats = ValidationStats()
436    for f in modified + added:
437        validate(f, stats, verbose, None)
438
439    if stats:
440        stats.dump()
441        result = ui.prompt("invalid formatting\n(i)gnore or (a)bort?",
442                           'ai', 'a')
443        if result == 'a':
444            return True
445
446    return False
447
448def check_hook(hooktype):
449    if hooktype not in ('pretxncommit', 'pre-qrefresh'):
450        raise AttributeError, \
451              "This hook is not meant for %s" % hooktype
452
453def check_style(ui, repo, hooktype, **kwargs):
454    check_hook(hooktype)
455    args = {}
456
457    try:
458        return do_check_style(ui, repo, **args)
459    except Exception, e:
460        import traceback
461        traceback.print_exc()
462        return True
463
464def check_format(ui, repo, hooktype, **kwargs):
465    check_hook(hooktype)
466    args = {}
467
468    try:
469        return do_check_format(ui, repo, **args)
470    except Exception, e:
471        import traceback
472        traceback.print_exc()
473        return True
474
475try:
476    from mercurial.i18n import _
477except ImportError:
478    def _(arg):
479        return arg
480
481cmdtable = {
482    '^m5style' :
483    ( do_check_style,
484      [ ('a', 'auto', False, _("automatically fix whitespace")) ],
485      _('hg m5style [-a] [FILE]...')),
486    '^m5format' :
487    ( do_check_format,
488      [ ],
489      _('hg m5format [FILE]...')),
490}
491
492if __name__ == '__main__':
493    import getopt
494
495    progname = sys.argv[0]
496    if len(sys.argv) < 2:
497        sys.exit('usage: %s <command> [<command args>]' % progname)
498
499    fixwhite_usage = '%s fixwhite [-t <tabsize> ] <path> [...] \n' % progname
500    chkformat_usage = '%s chkformat <path> [...] \n' % progname
501    chkwhite_usage = '%s chkwhite <path> [...] \n' % progname
502
503    command = sys.argv[1]
504    if command == 'fixwhite':
505        flags = 't:'
506        usage = fixwhite_usage
507    elif command == 'chkwhite':
508        flags = 'nv'
509        usage = chkwhite_usage
510    elif command == 'chkformat':
511        flags = 'nv'
512        usage = chkformat_usage
513    else:
514        sys.exit(fixwhite_usage + chkwhite_usage + chkformat_usage)
515
516    opts, args = getopt.getopt(sys.argv[2:], flags)
517
518    code = 1
519    verbose = 1
520    for opt,arg in opts:
521        if opt == '-n':
522            code = None
523        if opt == '-t':
524            tabsize = int(arg)
525        if opt == '-v':
526            verbose += 1
527
528    if command == 'fixwhite':
529        for filename in args:
530            fixwhite(filename, tabsize)
531    elif command == 'chkwhite':
532        for filename in args:
533            for line,num in checkwhite(filename):
534                print 'invalid whitespace: %s:%d' % (filename, num)
535                if verbose:
536                    print '>>%s<<' % line[:-1]
537    elif command == 'chkformat':
538        stats = ValidationStats()
539        for filename in args:
540            validate(filename, stats=stats, verbose=verbose, exit_code=code)
541
542        if verbose > 0:
543            stats.dump()
544    else:
545        sys.exit("command '%s' not found" % command)
546