1 Return-Path: <jani@nikula.org>
\r
2 X-Original-To: notmuch@notmuchmail.org
\r
3 Delivered-To: notmuch@notmuchmail.org
\r
4 Received: from localhost (localhost [127.0.0.1])
\r
5 by olra.theworths.org (Postfix) with ESMTP id 3AF1A431FAF
\r
6 for <notmuch@notmuchmail.org>; Tue, 8 May 2012 00:58:36 -0700 (PDT)
\r
7 X-Virus-Scanned: Debian amavisd-new at olra.theworths.org
\r
11 X-Spam-Status: No, score=-0.7 tagged_above=-999 required=5
\r
12 tests=[RCVD_IN_DNSWL_LOW=-0.7] autolearn=disabled
\r
13 Received: from olra.theworths.org ([127.0.0.1])
\r
14 by localhost (olra.theworths.org [127.0.0.1]) (amavisd-new, port 10024)
\r
15 with ESMTP id y4o8OzivJ0Rr for <notmuch@notmuchmail.org>;
\r
16 Tue, 8 May 2012 00:58:34 -0700 (PDT)
\r
17 Received: from mail-qc0-f181.google.com (mail-qc0-f181.google.com
\r
18 [209.85.216.181]) (using TLSv1 with cipher RC4-SHA (128/128 bits))
\r
19 (No client certificate requested)
\r
20 by olra.theworths.org (Postfix) with ESMTPS id C3993431FAE
\r
21 for <notmuch@notmuchmail.org>; Tue, 8 May 2012 00:58:34 -0700 (PDT)
\r
22 Received: by qcsk26 with SMTP id k26so152066qcs.26
\r
23 for <notmuch@notmuchmail.org>; Tue, 08 May 2012 00:58:33 -0700 (PDT)
\r
24 X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
\r
25 d=google.com; s=20120113;
\r
26 h=from:to:cc:subject:in-reply-to:references:user-agent:date
\r
27 :message-id:mime-version:content-type:x-gm-message-state;
\r
28 bh=AvnmJ/5kI/eNjOsRv0F9hhkYVq2Ln0xarqnCHxcmBXw=;
\r
29 b=i23hCtM6IMjfmjQDi7EQA7txBP9Zvs3E0FBPIto9SVU68UiIW5O5XHDLTzddbqWtGO
\r
30 0pZwJLoRsFxKAyeS6xEwENNWQYgJvefBRHdbuPWlNzGwDqD5dNKfy0zm1NIZCDhAPapk
\r
31 2rxfFidneztRtz3k6IUlXpC+yX8JaXYF/MOa0Zfs2GxfXkqExss47hm+ADc8Ouuf7fC6
\r
32 1/D5YF/yxzueCsLx1gSLTLUd8ABa97++EK4XFL6NfSutlF5ZtKf2/57Z0QzSIYSRxywt
\r
33 o3I2MfIBP4NkOk3l6u0oMlEPg90+lvBXu4PDjrDgBuWkVSYRKb8KWG1WzjyQqZiNokhI
\r
35 Received: by 10.224.205.6 with SMTP id fo6mr10210783qab.73.1336463912954;
\r
36 Tue, 08 May 2012 00:58:32 -0700 (PDT)
\r
37 Received: from localhost (nikula.org. [92.243.24.172])
\r
38 by mx.google.com with ESMTPS id bm15sm2680690qab.17.2012.05.08.00.58.30
\r
39 (version=SSLv3 cipher=OTHER); Tue, 08 May 2012 00:58:32 -0700 (PDT)
\r
40 From: Jani Nikula <jani@nikula.org>
\r
41 To: Austin Clements <amdragon@MIT.EDU>, notmuch@notmuchmail.org
\r
42 Subject: Re: [PATCH v2 2/2] new: Centralize file type stat-ing logic
\r
43 In-Reply-To: <1336429240-1114-3-git-send-email-amdragon@mit.edu>
\r
44 References: <1336414186-15293-1-git-send-email-amdragon@mit.edu>
\r
45 <1336429240-1114-1-git-send-email-amdragon@mit.edu>
\r
46 <1336429240-1114-3-git-send-email-amdragon@mit.edu>
\r
47 User-Agent: Notmuch/0.11.1+222~ga47a98c (http://notmuchmail.org) Emacs/23.1.1
\r
49 Date: Tue, 08 May 2012 07:58:28 +0000
\r
50 Message-ID: <87r4uvdryz.fsf@nikula.org>
\r
52 Content-Type: text/plain; charset=us-ascii
\r
54 ALoCoQl9ItXHqbHEak1VBDEU3KqrTo5GO5ue8xuA/6kGn/FOJzkzG/L7EUE8aMGQMSGTOZ6ZMPs7
\r
55 Cc: Vladimir Marek <vlmarek@volny.cz>
\r
56 X-BeenThere: notmuch@notmuchmail.org
\r
57 X-Mailman-Version: 2.1.13
\r
59 List-Id: "Use and development of the notmuch mail system."
\r
60 <notmuch.notmuchmail.org>
\r
61 List-Unsubscribe: <http://notmuchmail.org/mailman/options/notmuch>,
\r
62 <mailto:notmuch-request@notmuchmail.org?subject=unsubscribe>
\r
63 List-Archive: <http://notmuchmail.org/pipermail/notmuch>
\r
64 List-Post: <mailto:notmuch@notmuchmail.org>
\r
65 List-Help: <mailto:notmuch-request@notmuchmail.org?subject=help>
\r
66 List-Subscribe: <http://notmuchmail.org/mailman/listinfo/notmuch>,
\r
67 <mailto:notmuch-request@notmuchmail.org?subject=subscribe>
\r
68 X-List-Received-Date: Tue, 08 May 2012 07:58:36 -0000
\r
70 On Mon, 7 May 2012 18:20:40 -0400, Austin Clements <amdragon@MIT.EDU> wrote:
\r
71 > This moves our logic to get a file's type into one function. This has
\r
72 > several benefits: we can support OSes and file systems that do not
\r
73 > provide dirent.d_type or always return DT_UNKNOWN, complex
\r
74 > symlink-handling logic has been replaced by a simple stat fall-through
\r
75 > in one place, and the error message for un-stat-able file is more
\r
76 > accurate (previously, the error always mentioned directories, even
\r
77 > though a broken symlink is not a directory).
\r
82 > notmuch-new.c | 103 +++++++++++++++++++++++++++++++++++----------------------
\r
84 > 2 files changed, 64 insertions(+), 41 deletions(-)
\r
86 > diff --git a/notmuch-new.c b/notmuch-new.c
\r
87 > index cb720cc..8955677 100644
\r
88 > --- a/notmuch-new.c
\r
89 > +++ b/notmuch-new.c
\r
90 > @@ -154,6 +154,48 @@ dirent_sort_strcmp_name (const struct dirent **a, const struct dirent **b)
\r
91 > return strcmp ((*a)->d_name, (*b)->d_name);
\r
94 > +/* Return the type of a directory entry relative to path as a stat(2)
\r
95 > + * mode. Like stat, this follows symlinks. Returns -1 and sets errno
\r
96 > + * if the file's type cannot be determined (which includes dangling
\r
100 > +dirent_type (const char *path, const struct dirent *entry)
\r
102 > + struct stat statbuf;
\r
104 > + int err, saved_errno;
\r
106 > +#ifdef _DIRENT_HAVE_D_TYPE
\r
107 > + /* Mapping from d_type to stat mode_t. We omit DT_LNK so that
\r
108 > + * we'll fall through to stat and get the real file type. */
\r
109 > + static const mode_t modes[] = {
\r
110 > + [DT_BLK] = S_IFBLK,
\r
111 > + [DT_CHR] = S_IFCHR,
\r
112 > + [DT_DIR] = S_IFDIR,
\r
113 > + [DT_FIFO] = S_IFIFO,
\r
114 > + [DT_REG] = S_IFREG,
\r
115 > + [DT_SOCK] = S_IFSOCK
\r
117 > + if (entry->d_type < ARRAY_SIZE(modes) && modes[entry->d_type])
\r
118 > + return modes[entry->d_type];
\r
121 > + abspath = talloc_asprintf (NULL, "%s/%s", path, entry->d_name);
\r
122 > + if (!abspath) {
\r
123 > + errno = ENOMEM;
\r
126 > + err = stat(abspath, &statbuf);
\r
127 > + saved_errno = errno;
\r
128 > + talloc_free (abspath);
\r
130 > + errno = saved_errno;
\r
133 > + return statbuf.st_mode & S_IFMT;
\r
136 > /* Test if the directory looks like a Maildir directory.
\r
138 > * Search through the array of directory entries to see if we can find all
\r
139 > @@ -162,12 +204,12 @@ dirent_sort_strcmp_name (const struct dirent **a, const struct dirent **b)
\r
140 > * Return 1 if the directory looks like a Maildir and 0 otherwise.
\r
143 > -_entries_resemble_maildir (struct dirent **entries, int count)
\r
144 > +_entries_resemble_maildir (const char *path, struct dirent **entries, int count)
\r
146 > int i, found = 0;
\r
148 > for (i = 0; i < count; i++) {
\r
149 > - if (entries[i]->d_type != DT_DIR && entries[i]->d_type != DT_UNKNOWN)
\r
150 > + if (dirent_type (path, entries[i]) != S_IFDIR)
\r
153 > if (strcmp(entries[i]->d_name, "new") == 0 ||
\r
154 > @@ -250,7 +292,7 @@ add_files_recursive (notmuch_database_t *notmuch,
\r
155 > notmuch_status_t status, ret = NOTMUCH_STATUS_SUCCESS;
\r
156 > notmuch_message_t *message = NULL;
\r
157 > struct dirent **fs_entries = NULL;
\r
158 > - int i, num_fs_entries;
\r
159 > + int i, num_fs_entries, entry_type;
\r
160 > notmuch_directory_t *directory;
\r
161 > notmuch_filenames_t *db_files = NULL;
\r
162 > notmuch_filenames_t *db_subdirs = NULL;
\r
163 > @@ -317,7 +359,7 @@ add_files_recursive (notmuch_database_t *notmuch,
\r
166 > /* Pass 1: Recurse into all sub-directories. */
\r
167 > - is_maildir = _entries_resemble_maildir (fs_entries, num_fs_entries);
\r
168 > + is_maildir = _entries_resemble_maildir (path, fs_entries, num_fs_entries);
\r
170 > for (i = 0; i < num_fs_entries; i++) {
\r
172 > @@ -325,17 +367,16 @@ add_files_recursive (notmuch_database_t *notmuch,
\r
174 > entry = fs_entries[i];
\r
176 > - /* We only want to descend into directories.
\r
177 > - * But symlinks can be to directories too, of course.
\r
179 > - * And if the filesystem doesn't tell us the file type in the
\r
180 > - * scandir results, then it might be a directory (and if not,
\r
181 > - * then we'll stat and return immediately in the next level of
\r
182 > - * recursion). */
\r
183 > - if (entry->d_type != DT_DIR &&
\r
184 > - entry->d_type != DT_LNK &&
\r
185 > - entry->d_type != DT_UNKNOWN)
\r
187 > + /* We only want to descend into directories (and symlinks to
\r
188 > + * directories). */
\r
189 > + entry_type = dirent_type (path, entry);
\r
190 > + if (entry_type == -1) {
\r
191 > + /* Be pessimistic, e.g. so we don't lose lots of mail just
\r
192 > + * because a user broke a symlink. */
\r
193 > + fprintf (stderr, "Error reading file %s/%s: %s\n",
\r
194 > + path, entry->d_name, strerror (errno));
\r
195 > + return NOTMUCH_STATUS_FILE_ERROR;
\r
196 > + } else if (entry_type != S_IFDIR) {
\r
200 > @@ -425,31 +466,13 @@ add_files_recursive (notmuch_database_t *notmuch,
\r
201 > notmuch_filenames_move_to_next (db_subdirs);
\r
204 > - /* If we're looking at a symlink, we only want to add it if it
\r
205 > - * links to a regular file, (and not to a directory, say).
\r
207 > - * Similarly, if the file is of unknown type (due to filesystem
\r
208 > - * limitations), then we also need to look closer.
\r
210 > - * In either case, a stat does the trick.
\r
212 > - if (entry->d_type == DT_LNK || entry->d_type == DT_UNKNOWN) {
\r
215 > - next = talloc_asprintf (notmuch, "%s/%s", path, entry->d_name);
\r
216 > - err = stat (next, &st);
\r
217 > - talloc_free (next);
\r
220 > - /* Don't emit an error for a link pointing nowhere, since
\r
221 > - * the directory-traversal pass will have already done
\r
226 > - if (! S_ISREG (st.st_mode))
\r
228 > - } else if (entry->d_type != DT_REG) {
\r
229 > + /* Only add regular files (and symlinks to regular files). */
\r
230 > + entry_type = dirent_type (path, entry);
\r
231 > + if (entry_type == -1) {
\r
232 > + fprintf (stderr, "Error reading file %s/%s: %s\n",
\r
233 > + path, entry->d_name, strerror (errno));
\r
234 > + return NOTMUCH_STATUS_FILE_ERROR;
\r
235 > + } else if (entry_type != S_IFREG) {
\r
239 > diff --git a/test/new b/test/new
\r
240 > index 26253db..e3900f5 100755
\r
243 > @@ -140,7 +140,7 @@ test_begin_subtest "Broken symlink aborts"
\r
244 > ln -s does-not-exist "${MAIL_DIR}/broken"
\r
245 > output=$(NOTMUCH_NEW 2>&1)
\r
246 > test_expect_equal "$output" \
\r
247 > -"Error reading directory /run/shm/nm/tmp.new/mail/broken: No such file or directory
\r
248 > +"Error reading file /run/shm/nm/tmp.new/mail/broken: No such file or directory
\r
249 > Note: A fatal error was encountered: Something went wrong trying to read or write a file
\r
251 > rm "${MAIL_DIR}/broken"
\r
255 > _______________________________________________
\r
256 > notmuch mailing list
\r
257 > notmuch@notmuchmail.org
\r
258 > http://notmuchmail.org/mailman/listinfo/notmuch
\r