style.py revision 4745:33b409225928 
1#! /usr/bin/env python
2# Copyright (c) 2007 The Regents of The University of Michigan
3# All rights reserved.
4#
5# Redistribution and use in source and binary forms, with or without
6# modification, are permitted provided that the following conditions are
7# met: redistributions of source code must retain the above copyright
8# notice, this list of conditions and the following disclaimer;
9# redistributions in binary form must reproduce the above copyright
10# notice, this list of conditions and the following disclaimer in the
11# documentation and/or other materials provided with the distribution;
12# neither the name of the copyright holders nor the names of its
13# contributors may be used to endorse or promote products derived from
14# this software without specific prior written permission.
15#
16# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
17# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
18# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
19# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
20# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
21# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
22# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
26# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27#
28# Authors: Nathan Binkert
29
30import re
31import os
32import sys
33
34lead = re.compile(r'^([ \t]+)')
35trail = re.compile(r'([ \t]+)$')
36any_control = re.compile(r'\b(if|while|for)[ \t]*[(]')
37good_control = re.compile(r'\b(if|while|for) [(]')
38
39lang_types = { 'c'   : "C",
40               'h'   : "C",
41               'cc'  : "C++",
42               'hh'  : "C++",
43               'cxx' : "C++",
44               'hxx' : "C++",
45               'cpp' : "C++",
46               'hpp' : "C++",
47               'C'   : "C++",
48               'H'   : "C++",
49               'i'   : "swig",
50               'py'  : "python",
51               's'   : "asm",
52               'S'   : "asm",
53               'isa' : "isa" }
54whitespace_types = ('C', 'C++', 'swig', 'python', 'asm', 'isa')
55format_types = ( 'C', 'C++' )
56
57def file_type(filename):
58    extension = filename.split('.')
59    extension = len(extension) > 1 and extension[-1]
60    return lang_types.get(extension, None)
61
62def checkwhite_line(line):
63    match = lead.search(line)
64    if match and match.group(1).find('\t') != -1:
65        return False
66
67    match = trail.search(line)
68    if match:
69        return False
70
71    return True
72
73def checkwhite(filename):
74    if file_type(filename) not in whitespace_types:
75        return
76
77    try:
78        f = file(filename, 'r+')
79    except OSError, msg:
80        print 'could not open file %s: %s' % (filename, msg)
81        return
82
83    for num,line in enumerate(f):
84        if not checkwhite_line(line):
85            yield line,num + 1
86
87def fixwhite_line(line, tabsize):
88    if lead.search(line):
89        newline = ''
90        for i,c in enumerate(line):
91            if c == ' ':
92                newline += ' '
93            elif c == '\t':
94                newline += ' ' * (tabsize - len(newline) % tabsize)
95            else:
96                newline += line[i:]
97                break
98
99        line = newline
100
101    return line.rstrip() + '\n'
102
103def fixwhite(filename, tabsize, fixonly=None):
104    if file_type(filename) not in whitespace_types:
105        return
106
107    try:
108        f = file(filename, 'r+')
109    except OSError, msg:
110        print 'could not open file %s: %s' % (filename, msg)
111        return
112
113    lines = list(f)
114
115    f.seek(0)
116    f.truncate()
117
118    for i,line in enumerate(lines):
119        if fixonly is None or i in fixonly:
120            line = fixwhite_line(line, tabsize)
121
122        print >>f, line,
123
124def linelen(line):
125    tabs = line.count('\t')
126    if not tabs:
127        return len(line)
128
129    count = 0
130    for c in line:
131        if c == '\t':
132            count += tabsize - count % tabsize
133        else:
134            count += 1
135
136    return count
137
138class ValidationStats(object):
139    def __init__(self):
140        self.toolong = 0
141        self.toolong80 = 0
142        self.leadtabs = 0
143        self.trailwhite = 0
144        self.badcontrol = 0
145        self.cret = 0
146
147    def dump(self):
148        print '''\
149%d violations of lines over 79 chars. %d of which are 80 chars exactly.
150%d cases of whitespace at the end of a line.
151%d cases of tabs to indent.
152%d bad parens after if/while/for.
153%d carriage returns found.
154''' % (self.toolong, self.toolong80, self.trailwhite, self.leadtabs,
155       self.badcontrol, self.cret)
156
157    def __nonzero__(self):
158        return self.toolong or self.toolong80 or self.leadtabs or \
159               self.trailwhite or self.badcontrol or self.cret
160
161def validate(filename, stats, verbose, exit_code):
162    if file_type(filename) not in format_types:
163        return
164
165    def msg(lineno, line, message):
166        print '%s:%d>' % (filename, lineno + 1), message
167        if verbose > 2:
168            print line
169
170    def bad():
171        if exit_code is not None:
172            sys.exit(exit_code)
173
174    cpp = filename.endswith('.cc') or filename.endswith('.hh')
175    py = filename.endswith('.py')
176
177    if py + cpp != 1:
178        raise AttributeError, \
179              "I don't know how to deal with the file %s" % filename
180
181    try:
182        f = file(filename, 'r')
183    except OSError:
184        if verbose > 0:
185            print 'could not open file %s' % filename
186        bad()
187        return
188
189    for i,line in enumerate(f):
190        line = line.rstrip('\n')
191
192        # no carriage returns
193        if line.find('\r') != -1:
194            self.cret += 1
195            if verbose > 1:
196                msg(i, line, 'carriage return found')
197            bad()
198
199        # lines max out at 79 chars
200        llen = linelen(line)
201        if llen > 79:
202            stats.toolong += 1
203            if llen == 80:
204                stats.toolong80 += 1
205            if verbose > 1:
206                msg(i, line, 'line too long (%d chars)' % llen)
207            bad()
208
209        # no tabs used to indent
210        match = lead.search(line)
211        if match and match.group(1).find('\t') != -1:
212            stats.leadtabs += 1
213            if verbose > 1:
214                msg(i, line, 'using tabs to indent')
215            bad()
216
217        # no trailing whitespace
218        if trail.search(line):
219            stats.trailwhite +=1
220            if verbose > 1:
221                msg(i, line, 'trailing whitespace')
222            bad()
223
224        # for c++, exactly one space betwen if/while/for and (
225        if cpp:
226            match = any_control.search(line)
227            if match and not good_control.search(line):
228                stats.badcontrol += 1
229                if verbose > 1:
230                    msg(i, line, 'improper spacing after %s' % match.group(1))
231                bad()
232
233def modified_lines(old_data, new_data):
234    from itertools import count
235    from mercurial import bdiff, mdiff
236
237    modified = set()
238    counter = count()
239    for pbeg, pend, fbeg, fend in bdiff.blocks(old_data, new_data):
240        for i in counter:
241            if i < fbeg:
242                modified.add(i)
243            elif i + 1 >= fend:
244                break
245    return modified
246
247def check_whitespace(ui, repo, hooktype, node, parent1, parent2):
248    from mercurial import mdiff
249
250    if hooktype != 'pretxncommit':
251        raise AttributeError, \
252              "This hook is only meant for pretxncommit, not %s" % hooktype
253
254    tabsize = 8
255    verbose = ui.configbool('style', 'verbose', False)
256    def prompt(name, fixonly=None):
257        result = ui.prompt("(a)bort, (i)gnore, or (f)ix?", "^[aif]$", "a")
258        if result == 'a':
259            return True
260        elif result == 'i':
261            pass
262        elif result == 'f':
263            fixwhite(name, tabsize, fixonly)
264        else:
265            raise RepoError, "Invalid response: '%s'" % result
266
267        return False
268
269    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
270
271    for fname in added:
272        ok = True
273        for line,num in checkwhite(fname):
274            ui.write("invalid whitespace in %s:%d\n" % (fname, num))
275            if verbose:
276                ui.write(">>%s<<\n" % line[-1])
277            ok = False
278
279        if not ok:
280            if prompt(fname):
281                return True
282
283    wctx = repo.workingctx()
284    for fname in modified:
285        fctx = wctx.filectx(fname)
286        pctx = fctx.parents()
287        assert len(pctx) in (1, 2)
288
289        file_data = fctx.data()
290        mod_lines = modified_lines(pctx[0].data(), file_data)
291        if len(pctx) == 2:
292            m2 = modified_lines(pctx[1].data(), file_data)
293            mod_lines = mod_lines & m2 # only the lines that are new in both
294
295        fixonly = set()
296        for i,line in enumerate(mdiff.splitnewlines(file_data)):
297            if i not in mod_lines:
298                continue
299
300            if checkwhite_line(line):
301                continue
302
303            ui.write("invalid whitespace: %s:%d\n" % (fname, i+1))
304            if verbose:
305                ui.write(">>%s<<\n" % line[:-1])
306            fixonly.add(i)
307
308        if fixonly:
309            if prompt(fname, fixonly):
310                return True
311
312def check_format(ui, repo, hooktype, node, parent1, parent2):
313    if hooktype != 'pretxncommit':
314        raise AttributeError, \
315              "This hook is only meant for pretxncommit, not %s" % hooktype
316
317    modified, added, removed, deleted, unknown, ignore, clean = repo.status()
318
319    verbose = 0
320    stats = ValidationStats()
321    for f in modified + added:
322        validate(f, stats, verbose, None)
323
324    if stats:
325        stats.dump()
326        result = ui.prompt("invalid formatting\n(i)gnore or (a)bort?",
327                           "^[ia]$", "a")
328        if result.startswith('i'):
329            pass
330        elif result.startswith('a'):
331            return True
332        else:
333            raise RepoError, "Invalid response: '%s'" % result
334
335    return False
336
337if __name__ == '__main__':
338    import getopt
339
340    progname = sys.argv[0]
341    if len(sys.argv) < 2:
342        sys.exit('usage: %s <command> [<command args>]' % progname)
343
344    fixwhite_usage = '%s fixwhite [-t <tabsize> ] <path> [...] \n' % progname
345    chkformat_usage = '%s chkformat <path> [...] \n' % progname
346    chkwhite_usage = '%s chkwhite <path> [...] \n' % progname
347
348    command = sys.argv[1]
349    if command == 'fixwhite':
350        flags = 't:'
351        usage = fixwhite_usage
352    elif command == 'chkwhite':
353        flags = 'nv'
354        usage = chkwhite_usage
355    elif command == 'chkformat':
356        flags = 'nv'
357        usage = chkformat_usage
358    else:
359        sys.exit(fixwhite_usage + chkwhite_usage + chkformat_usage)
360
361    opts, args = getopt.getopt(sys.argv[2:], flags)
362
363    code = 1
364    verbose = 1
365    tabsize = 8
366    for opt,arg in opts:
367        if opt == '-n':
368            code = None
369        if opt == '-t':
370            tabsize = int(arg)
371        if opt == '-v':
372            verbose += 1
373
374    if command == 'fixwhite':
375        for filename in args:
376            fixwhite(filename, tabsize)
377    elif command == 'chkwhite':
378        for filename in args:
379            line = checkwhite(filename)
380            if line:
381                print 'invalid whitespace at %s:%d' % (filename, line)
382    elif command == 'chkformat':
383        stats = ValidationStats()
384        for filename in files:
385            validate(filename, stats=stats, verbose=verbose, exit_code=code)
386
387        if verbose > 0:
388            stats.dump()
389    else:
390        sys.exit("command '%s' not found" % command)
391