style.py revision 8227:f3aaa2470b5a 
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
45from file_types import lang_type
46
47all_regions = Region(neg_inf, pos_inf)
48
49tabsize = 8
50lead = re.compile(r'^([ \t]+)')
51trail = re.compile(r'([ \t]+)$')
52any_control = re.compile(r'\b(if|while|for)[ \t]*[(]')
53good_control = re.compile(r'\b(if|while|for) [(]')
54
55format_types = set(('C', 'C++'))
56
57def modified_regions(old_data, new_data):
58    regions = Regions()
59    beg = None
60    for pbeg, pend, fbeg, fend in bdiff.blocks(old_data, new_data):
61        if beg is not None and beg != fbeg:
62            regions.append(beg, fbeg)
63        beg = fend
64    return regions
65
66def modregions(wctx, fname):
67    fctx = wctx.filectx(fname)
68    pctx = fctx.parents()
69
70    file_data = fctx.data()
71    lines = mdiff.splitnewlines(file_data)
72    if len(pctx) in (1, 2):
73        mod_regions = modified_regions(pctx[0].data(), file_data)
74        if len(pctx) == 2:
75            m2 = modified_regions(pctx[1].data(), file_data)
76            # only the lines that are new in both
77            mod_regions &= m2
78    else:
79        mod_regions = Regions()
80        mod_regions.add(0, len(lines))
81
82    return mod_regions
83
84class UserInterface(object):
85    def __init__(self, verbose=False, auto=False):
86        self.auto = auto
87        self.verbose = verbose
88
89    def prompt(self, prompt, results, default):
90        if self.auto:
91            return self.auto
92
93        while True:
94            result = self.do_prompt(prompt, results, default)
95            if result in results:
96                return result
97
98class MercurialUI(UserInterface):
99    def __init__(self, ui, *args, **kwargs):
100        super(MercurialUI, self).__init__(*args, **kwargs)
101        self.ui = ui
102
103    def do_prompt(self, prompt, results, default):
104        return self.ui.prompt(prompt, default=default)
105
106    def write(self, string):
107        self.ui.write(string)
108
109class StdioUI(UserInterface):
110    def do_prompt(self, prompt, results, default):
111        return raw_input(prompt) or default
112
113    def write(self, string):
114        sys.stdout.write(string)
115
116class Region(object):
117    def __init__(self, asdf):
118        self.regions = Foo
119
120class Verifier(object):
121    def __init__(self, ui, repo=None):
122        self.ui = ui
123        self.repo = repo
124        if repo is None:
125            self.wctx = None
126
127    def __getattr__(self, attr):
128        if attr in ('prompt', 'write'):
129            return getattr(self.ui, attr)
130
131        if attr == 'wctx':
132            try:
133                wctx = repo.workingctx()
134            except:
135                from mercurial import context
136                wctx = context.workingctx(repo)
137            self.wctx = wctx
138            return wctx
139
140        raise AttributeError
141
142    def open(self, filename, mode):
143        if self.repo:
144            filename = self.repo.wjoin(filename)
145
146        try:
147            f = file(filename, mode)
148        except OSError, msg:
149            print 'could not open file %s: %s' % (filename, msg)
150            return None
151
152        return f
153
154    def skip(self, filename):
155        return lang_type(filename) not in self.languages
156
157    def check(self, filename, regions=all_regions):
158        f = self.open(filename, 'r')
159
160        errors = 0
161        for num,line in enumerate(f):
162            if num not in regions:
163                continue
164            if not self.check_line(line):
165                self.write("invalid %s in %s:%d\n" % \
166                               (self.test_name, filename, num + 1))
167                if self.ui.verbose:
168                    self.write(">>%s<<\n" % line[-1])
169                errors += 1
170        return errors
171
172    def fix(self, filename, regions=all_regions):
173        f = self.open(filename, 'r+')
174
175        lines = list(f)
176
177        f.seek(0)
178        f.truncate()
179
180        for i,line in enumerate(lines):
181            if i in regions:
182                line = self.fix_line(line)
183
184            f.write(line)
185        f.close()
186
187class Whitespace(Verifier):
188    languages = set(('C', 'C++', 'swig', 'python', 'asm', 'isa', 'scons'))
189    test_name = 'whitespace'
190    def check_line(self, line):
191        match = lead.search(line)
192        if match and match.group(1).find('\t') != -1:
193            return False
194
195        match = trail.search(line)
196        if match:
197            return False
198
199        return True
200
201    def fix_line(self, line):
202        if lead.search(line):
203            newline = ''
204            for i,c in enumerate(line):
205                if c == ' ':
206                    newline += ' '
207                elif c == '\t':
208                    newline += ' ' * (tabsize - len(newline) % tabsize)
209                else:
210                    newline += line[i:]
211                    break
212
213            line = newline
214
215        return line.rstrip() + '\n'
216
217def linelen(line):
218    tabs = line.count('\t')
219    if not tabs:
220        return len(line)
221
222    count = 0
223    for c in line:
224        if c == '\t':
225            count += tabsize - count % tabsize
226        else:
227            count += 1
228
229    return count
230
231class ValidationStats(object):
232    def __init__(self):
233        self.toolong = 0
234        self.toolong80 = 0
235        self.leadtabs = 0
236        self.trailwhite = 0
237        self.badcontrol = 0
238        self.cret = 0
239
240    def dump(self):
241        print '''\
242%d violations of lines over 79 chars. %d of which are 80 chars exactly.
243%d cases of whitespace at the end of a line.
244%d cases of tabs to indent.
245%d bad parens after if/while/for.
246%d carriage returns found.
247''' % (self.toolong, self.toolong80, self.trailwhite, self.leadtabs,
248       self.badcontrol, self.cret)
249
250    def __nonzero__(self):
251        return self.toolong or self.toolong80 or self.leadtabs or \
252               self.trailwhite or self.badcontrol or self.cret
253
254def validate(filename, stats, verbose, exit_code):
255    if lang_type(filename) not in format_types:
256        return
257
258    def msg(lineno, line, message):
259        print '%s:%d>' % (filename, lineno + 1), message
260        if verbose > 2:
261            print line
262
263    def bad():
264        if exit_code is not None:
265            sys.exit(exit_code)
266
267    try:
268        f = file(filename, 'r')
269    except OSError:
270        if verbose > 0:
271            print 'could not open file %s' % filename
272        bad()
273        return
274
275    for i,line in enumerate(f):
276        line = line.rstrip('\n')
277
278        # no carriage returns
279        if line.find('\r') != -1:
280            self.cret += 1
281            if verbose > 1:
282                msg(i, line, 'carriage return found')
283            bad()
284
285        # lines max out at 79 chars
286        llen = linelen(line)
287        if llen > 79:
288            stats.toolong += 1
289            if llen == 80:
290                stats.toolong80 += 1
291            if verbose > 1:
292                msg(i, line, 'line too long (%d chars)' % llen)
293            bad()
294
295        # no tabs used to indent
296        match = lead.search(line)
297        if match and match.group(1).find('\t') != -1:
298            stats.leadtabs += 1
299            if verbose > 1:
300                msg(i, line, 'using tabs to indent')
301            bad()
302
303        # no trailing whitespace
304        if trail.search(line):
305            stats.trailwhite +=1
306            if verbose > 1:
307                msg(i, line, 'trailing whitespace')
308            bad()
309
310        # for c++, exactly one space betwen if/while/for and (
311        if cpp:
312            match = any_control.search(line)
313            if match and not good_control.search(line):
314                stats.badcontrol += 1
315                if verbose > 1:
316                    msg(i, line, 'improper spacing after %s' % match.group(1))
317                bad()
318
319def do_check_style(hgui, repo, *files, **args):
320    """check files for proper m5 style guidelines"""
321    from mercurial import mdiff, util
322
323    auto = args.get('auto', False)
324    if auto:
325        auto = 'f'
326    ui = MercurialUI(hgui, hgui.verbose, auto)
327
328    if files:
329        files = frozenset(files)
330
331    def skip(name):
332        return files and name in files
333
334    def prompt(name, func, regions=all_regions):
335        result = ui.prompt("(a)bort, (i)gnore, or (f)ix?", 'aif', 'a')
336        if result == 'a':
337            return True
338        elif result == 'f':
339            func(repo.wjoin(name), regions)
340
341        return False
342
343    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
344
345    whitespace = Whitespace(ui)
346    for fname in added:
347        if skip(fname) or whitespace.skip(fname):
348            continue
349
350        errors = whitespace.check(fname)
351        if errors:
352            print errors
353            if prompt(fname, whitespace.fix):
354                return True
355
356    try:
357        wctx = repo.workingctx()
358    except:
359        from mercurial import context
360        wctx = context.workingctx(repo)
361
362    for fname in modified:
363        if skip(fname) or whitespace.skip(fname):
364            continue
365
366        regions = modregions(wctx, fname)
367
368        errors = whitespace.check(fname, regions)
369        if errors:
370            if prompt(fname, whitespace.fix, regions):
371                return True
372
373def do_check_format(hgui, repo, **args):
374    ui = MercurialUI(hgui, hgui.verbose, auto)
375
376    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
377
378    verbose = 0
379    stats = ValidationStats()
380    for f in modified + added:
381        validate(f, stats, verbose, None)
382
383    if stats:
384        stats.dump()
385        result = ui.prompt("invalid formatting\n(i)gnore or (a)bort?",
386                           'ai', 'a')
387        if result == 'a':
388            return True
389
390    return False
391
392def check_hook(hooktype):
393    if hooktype not in ('pretxncommit', 'pre-qrefresh'):
394        raise AttributeError, \
395              "This hook is not meant for %s" % hooktype
396
397def check_style(ui, repo, hooktype, **kwargs):
398    check_hook(hooktype)
399    args = {}
400
401    try:
402        return do_check_style(ui, repo, **args)
403    except Exception, e:
404        import traceback
405        traceback.print_exc()
406        return True
407
408def check_format(ui, repo, hooktype, **kwargs):
409    check_hook(hooktype)
410    args = {}
411
412    try:
413        return do_check_format(ui, repo, **args)
414    except Exception, e:
415        import traceback
416        traceback.print_exc()
417        return True
418
419try:
420    from mercurial.i18n import _
421except ImportError:
422    def _(arg):
423        return arg
424
425cmdtable = {
426    '^m5style' :
427    ( do_check_style,
428      [ ('a', 'auto', False, _("automatically fix whitespace")) ],
429      _('hg m5style [-a] [FILE]...')),
430    '^m5format' :
431    ( do_check_format,
432      [ ],
433      _('hg m5format [FILE]...')),
434}
435
436if __name__ == '__main__':
437    import getopt
438
439    progname = sys.argv[0]
440    if len(sys.argv) < 2:
441        sys.exit('usage: %s <command> [<command args>]' % progname)
442
443    fixwhite_usage = '%s fixwhite [-t <tabsize> ] <path> [...] \n' % progname
444    chkformat_usage = '%s chkformat <path> [...] \n' % progname
445    chkwhite_usage = '%s chkwhite <path> [...] \n' % progname
446
447    command = sys.argv[1]
448    if command == 'fixwhite':
449        flags = 't:'
450        usage = fixwhite_usage
451    elif command == 'chkwhite':
452        flags = 'nv'
453        usage = chkwhite_usage
454    elif command == 'chkformat':
455        flags = 'nv'
456        usage = chkformat_usage
457    else:
458        sys.exit(fixwhite_usage + chkwhite_usage + chkformat_usage)
459
460    opts, args = getopt.getopt(sys.argv[2:], flags)
461
462    code = 1
463    verbose = 1
464    for opt,arg in opts:
465        if opt == '-n':
466            code = None
467        if opt == '-t':
468            tabsize = int(arg)
469        if opt == '-v':
470            verbose += 1
471
472    if command == 'fixwhite':
473        for filename in args:
474            fixwhite(filename, tabsize)
475    elif command == 'chkwhite':
476        for filename in args:
477            for line,num in checkwhite(filename):
478                print 'invalid whitespace: %s:%d' % (filename, num)
479                if verbose:
480                    print '>>%s<<' % line[:-1]
481    elif command == 'chkformat':
482        stats = ValidationStats()
483        for filename in args:
484            validate(filename, stats=stats, verbose=verbose, exit_code=code)
485
486        if verbose > 0:
487            stats.dump()
488    else:
489        sys.exit("command '%s' not found" % command)
490