style.py revision 8583:d182a475a668 
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        if len(lines) == 0:
237            return 0
238
239        language = lang_type(filename, lines[0])
240        sort_lines = list(self.sort_includes(lines, filename, language))
241        new = ''.join(line + '\n' for line in sort_lines)
242
243        mod = modified_regions(old, new)
244        modified = mod & regions
245
246        if modified:
247            self.write("invalid sorting of includes\n")
248            if self.ui.verbose:
249                for start, end in modified.regions:
250                    self.write("bad region [%d, %d)\n" % (start, end))
251            return 1
252
253        return 0
254
255    def fix(self, filename, regions=all_regions):
256        f = self.open(filename, 'r+')
257
258        old = f.readlines()
259        lines = [ l.rstrip('\n') for l in old ]
260        language = lang_type(filename, lines[0])
261        sort_lines = list(self.sort_includes(lines, filename, language))
262        new = ''.join(line + '\n' for line in sort_lines)
263
264        f.seek(0)
265        f.truncate()
266
267        for i,line in enumerate(sort_lines):
268            f.write(line)
269            f.write('\n')
270        f.close()
271
272def linelen(line):
273    tabs = line.count('\t')
274    if not tabs:
275        return len(line)
276
277    count = 0
278    for c in line:
279        if c == '\t':
280            count += tabsize - count % tabsize
281        else:
282            count += 1
283
284    return count
285
286class ValidationStats(object):
287    def __init__(self):
288        self.toolong = 0
289        self.toolong80 = 0
290        self.leadtabs = 0
291        self.trailwhite = 0
292        self.badcontrol = 0
293        self.cret = 0
294
295    def dump(self):
296        print '''\
297%d violations of lines over 79 chars. %d of which are 80 chars exactly.
298%d cases of whitespace at the end of a line.
299%d cases of tabs to indent.
300%d bad parens after if/while/for.
301%d carriage returns found.
302''' % (self.toolong, self.toolong80, self.trailwhite, self.leadtabs,
303       self.badcontrol, self.cret)
304
305    def __nonzero__(self):
306        return self.toolong or self.toolong80 or self.leadtabs or \
307               self.trailwhite or self.badcontrol or self.cret
308
309def validate(filename, stats, verbose, exit_code):
310    if lang_type(filename) not in format_types:
311        return
312
313    def msg(lineno, line, message):
314        print '%s:%d>' % (filename, lineno + 1), message
315        if verbose > 2:
316            print line
317
318    def bad():
319        if exit_code is not None:
320            sys.exit(exit_code)
321
322    try:
323        f = file(filename, 'r')
324    except OSError:
325        if verbose > 0:
326            print 'could not open file %s' % filename
327        bad()
328        return
329
330    for i,line in enumerate(f):
331        line = line.rstrip('\n')
332
333        # no carriage returns
334        if line.find('\r') != -1:
335            self.cret += 1
336            if verbose > 1:
337                msg(i, line, 'carriage return found')
338            bad()
339
340        # lines max out at 79 chars
341        llen = linelen(line)
342        if llen > 79:
343            stats.toolong += 1
344            if llen == 80:
345                stats.toolong80 += 1
346            if verbose > 1:
347                msg(i, line, 'line too long (%d chars)' % llen)
348            bad()
349
350        # no tabs used to indent
351        match = lead.search(line)
352        if match and match.group(1).find('\t') != -1:
353            stats.leadtabs += 1
354            if verbose > 1:
355                msg(i, line, 'using tabs to indent')
356            bad()
357
358        # no trailing whitespace
359        if trail.search(line):
360            stats.trailwhite +=1
361            if verbose > 1:
362                msg(i, line, 'trailing whitespace')
363            bad()
364
365        # for c++, exactly one space betwen if/while/for and (
366        if cpp:
367            match = any_control.search(line)
368            if match and not good_control.search(line):
369                stats.badcontrol += 1
370                if verbose > 1:
371                    msg(i, line, 'improper spacing after %s' % match.group(1))
372                bad()
373
374def do_check_style(hgui, repo, *files, **args):
375    """check files for proper m5 style guidelines"""
376    from mercurial import mdiff, util
377
378    auto = args.get('auto', False)
379    if auto:
380        auto = 'f'
381    ui = MercurialUI(hgui, hgui.verbose, auto)
382
383    if files:
384        files = frozenset(files)
385
386    def skip(name):
387        return files and name in files
388
389    def prompt(name, func, regions=all_regions):
390        result = ui.prompt("(a)bort, (i)gnore, or (f)ix?", 'aif', 'a')
391        if result == 'a':
392            return True
393        elif result == 'f':
394            func(repo.wjoin(name), regions)
395
396        return False
397
398    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
399
400    whitespace = Whitespace(ui)
401    sorted_includes = SortedIncludes(ui)
402    for fname in added:
403        if skip(fname):
404            continue
405
406        fpath = joinpath(repo.root, fname)
407
408        if whitespace.apply(fpath, prompt):
409            return True
410
411        if sorted_includes.apply(fpath, 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        fpath = joinpath(repo.root, fname)
425        regions = modregions(wctx, fname)
426
427        if whitespace.apply(fpath, prompt, regions):
428            return True
429
430        if sorted_includes.apply(fpath, prompt, regions):
431            return True
432
433    return False
434
435def do_check_format(hgui, repo, **args):
436    ui = MercurialUI(hgui, hgui.verbose, auto)
437
438    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
439
440    verbose = 0
441    stats = ValidationStats()
442    for f in modified + added:
443        validate(joinpath(repo.root, f), stats, verbose, None)
444
445    if stats:
446        stats.dump()
447        result = ui.prompt("invalid formatting\n(i)gnore or (a)bort?",
448                           'ai', 'a')
449        if result == 'a':
450            return True
451
452    return False
453
454def check_hook(hooktype):
455    if hooktype not in ('pretxncommit', 'pre-qrefresh'):
456        raise AttributeError, \
457              "This hook is not meant for %s" % hooktype
458
459def check_style(ui, repo, hooktype, **kwargs):
460    check_hook(hooktype)
461    args = {}
462
463    try:
464        return do_check_style(ui, repo, **args)
465    except Exception, e:
466        import traceback
467        traceback.print_exc()
468        return True
469
470def check_format(ui, repo, hooktype, **kwargs):
471    check_hook(hooktype)
472    args = {}
473
474    try:
475        return do_check_format(ui, repo, **args)
476    except Exception, e:
477        import traceback
478        traceback.print_exc()
479        return True
480
481try:
482    from mercurial.i18n import _
483except ImportError:
484    def _(arg):
485        return arg
486
487cmdtable = {
488    '^m5style' :
489    ( do_check_style,
490      [ ('a', 'auto', False, _("automatically fix whitespace")) ],
491      _('hg m5style [-a] [FILE]...')),
492    '^m5format' :
493    ( do_check_format,
494      [ ],
495      _('hg m5format [FILE]...')),
496}
497
498if __name__ == '__main__':
499    import getopt
500
501    progname = sys.argv[0]
502    if len(sys.argv) < 2:
503        sys.exit('usage: %s <command> [<command args>]' % progname)
504
505    fixwhite_usage = '%s fixwhite [-t <tabsize> ] <path> [...] \n' % progname
506    chkformat_usage = '%s chkformat <path> [...] \n' % progname
507    chkwhite_usage = '%s chkwhite <path> [...] \n' % progname
508
509    command = sys.argv[1]
510    if command == 'fixwhite':
511        flags = 't:'
512        usage = fixwhite_usage
513    elif command == 'chkwhite':
514        flags = 'nv'
515        usage = chkwhite_usage
516    elif command == 'chkformat':
517        flags = 'nv'
518        usage = chkformat_usage
519    else:
520        sys.exit(fixwhite_usage + chkwhite_usage + chkformat_usage)
521
522    opts, args = getopt.getopt(sys.argv[2:], flags)
523
524    code = 1
525    verbose = 1
526    for opt,arg in opts:
527        if opt == '-n':
528            code = None
529        if opt == '-t':
530            tabsize = int(arg)
531        if opt == '-v':
532            verbose += 1
533
534    if command == 'fixwhite':
535        for filename in args:
536            fixwhite(filename, tabsize)
537    elif command == 'chkwhite':
538        for filename in args:
539            for line,num in checkwhite(filename):
540                print 'invalid whitespace: %s:%d' % (filename, num)
541                if verbose:
542                    print '>>%s<<' % line[:-1]
543    elif command == 'chkformat':
544        stats = ValidationStats()
545        for filename in args:
546            validate(filename, stats=stats, verbose=verbose, exit_code=code)
547
548        if verbose > 0:
549            stats.dump()
550    else:
551        sys.exit("command '%s' not found" % command)
552