cli: search: Do not output duplicate addresses

This filters out duplicate addresses from address outputs (sender,
receivers).

It also also adds tests for the new outputs.

The code here is an extended version of a patch from Jani Nikula.

diff --git a/doc/man1/notmuch-search.rst b/doc/man1/notmuch-search.rst
index b6607c922cc083c35b9add629c771d6d3927da1b..42f17e4aed66f5333bb39347afe5b5484472638a 100644 (file)
--- a/doc/man1/notmuch-search.rst
+++ b/doc/man1/notmuch-search.rst
@@ -85,6 +85,8 @@ Supported options for **search** include
             (--format=text0), as a JSON array (--format=json), or as
             an S-Expression list (--format=sexp).
 
+            Duplicate addresses are filtered out.
+
            Note: Searching for **sender** should be much faster than
            searching for **recipients**, because sender addresses are
            cached directly in the database whereas other addresses

diff --git a/notmuch-search.c b/notmuch-search.c
index 671fe4139981055e77e7f6c7c269a4bdb77ccb81..43d42c609756fb798a0ee678b259bce15818b1de 100644 (file)
--- a/notmuch-search.c
+++ b/notmuch-search.c
@@ -229,6 +229,27 @@ do_search_threads (search_options_t *opt)
     return 0;
 }
 
+/* Returns TRUE iff name and addr is duplicate. */
+static notmuch_bool_t
+is_duplicate (const search_options_t *opt, GHashTable *addrs, const char *name, const char *addr)
+{
+    notmuch_bool_t duplicate;
+    char *key;
+
+    key = talloc_asprintf (opt->format, "%s <%s>", name, addr);
+    if (! key)
+       return FALSE;
+
+    duplicate = g_hash_table_lookup_extended (addrs, key, NULL, NULL);
+
+    if (! duplicate)
+       g_hash_table_insert (addrs, key, NULL);
+    else
+       talloc_free (key);
+
+    return duplicate;
+}
+
 static void
 print_mailbox (const search_options_t *opt, const mailbox_t *mailbox)
 {
@@ -263,7 +284,8 @@ print_mailbox (const search_options_t *opt, const mailbox_t *mailbox)
 
 /* Print addresses from InternetAddressList.  */
 static void
-process_address_list (const search_options_t *opt, InternetAddressList *list)
+process_address_list (const search_options_t *opt, GHashTable *addrs,
+                     InternetAddressList *list)
 {
     InternetAddress *address;
     int i;
@@ -279,7 +301,7 @@ process_address_list (const search_options_t *opt, InternetAddressList *list)
            if (group_list == NULL)
                continue;
 
-           process_address_list (opt, group_list);
+           process_address_list (opt, addrs, group_list);
        } else {
            InternetAddressMailbox *mailbox = INTERNET_ADDRESS_MAILBOX (address);
            mailbox_t mbx = {
@@ -287,6 +309,9 @@ process_address_list (const search_options_t *opt, InternetAddressList *list)
                .addr = internet_address_mailbox_get_addr (mailbox),
            };
 
+           if (is_duplicate (opt, addrs, mbx.name, mbx.addr))
+               continue;
+
            print_mailbox (opt, &mbx);
        }
     }
@@ -294,7 +319,7 @@ process_address_list (const search_options_t *opt, InternetAddressList *list)
 
 /* Print addresses from a message header.  */
 static void
-process_address_header (const search_options_t *opt, const char *value)
+process_address_header (const search_options_t *opt, GHashTable *addrs, const char *value)
 {
     InternetAddressList *list;
 
@@ -305,11 +330,17 @@ process_address_header (const search_options_t *opt, const char *value)
     if (list == NULL)
        return;
 
-    process_address_list (opt, list);
+    process_address_list (opt, addrs, list);
 
     g_object_unref (list);
 }
 
+static void
+_my_talloc_free_for_g_hash (void *ptr)
+{
+    talloc_free (ptr);
+}
+
 static int
 do_search_messages (search_options_t *opt)
 {
@@ -317,8 +348,13 @@ do_search_messages (search_options_t *opt)
     notmuch_messages_t *messages;
     notmuch_filenames_t *filenames;
     sprinter_t *format = opt->format;
+    GHashTable *addresses = NULL;
     int i;
 
+    if (opt->output & OUTPUT_ADDRESS_FLAGS)
+       addresses = g_hash_table_new_full (g_str_hash, g_str_equal,
+                                          _my_talloc_free_for_g_hash, NULL);
+
     if (opt->offset < 0) {
        opt->offset += notmuch_query_count_messages (opt->query);
        if (opt->offset < 0)
@@ -366,7 +402,7 @@ do_search_messages (search_options_t *opt)
                const char *addrs;
 
                addrs = notmuch_message_get_header (message, "from");
-               process_address_header (opt, addrs);
+               process_address_header (opt, addresses, addrs);
            }
 
            if (opt->output & OUTPUT_RECIPIENTS) {
@@ -376,7 +412,7 @@ do_search_messages (search_options_t *opt)
 
                for (j = 0; j < ARRAY_SIZE (hdrs); j++) {
                    addrs = notmuch_message_get_header (message, hdrs[j]);
-                   process_address_header (opt, addrs);
+                   process_address_header (opt, addresses, addrs);
                }
            }
        }
@@ -384,6 +420,9 @@ do_search_messages (search_options_t *opt)
        notmuch_message_destroy (message);
     }
 
+    if (addresses)
+       g_hash_table_unref (addresses);
+
     notmuch_messages_destroy (messages);
 
     format->end (format);

diff --git a/test/T090-search-output.sh b/test/T090-search-output.sh
index 947d572ebeff71418ae282868500f6b50f195f7b..82380ac2af1472f6a4d632c578bb795061b88c06 100755 (executable)
--- a/test/T090-search-output.sh
+++ b/test/T090-search-output.sh
@@ -387,6 +387,93 @@ cat <<EOF >EXPECTED
 EOF
 test_expect_equal_file OUTPUT EXPECTED
 
+test_begin_subtest "--output=sender"
+notmuch search --output=sender '*' >OUTPUT
+cat <<EOF >EXPECTED
+François Boulogne <boulogne.f@gmail.com>
+Olivier Berger <olivier.berger@it-sudparis.eu>
+Chris Wilson <chris@chris-wilson.co.uk>
+Carl Worth <cworth@cworth.org>
+Alexander Botero-Lowry <alex.boterolowry@gmail.com>
+Keith Packard <keithp@keithp.com>
+Jjgod Jiang <gzjjgod@gmail.com>
+Rolland Santimano <rollandsantimano@yahoo.com>
+Jan Janak <jan@ryngle.com>
+Stewart Smith <stewart@flamingspork.com>
+Lars Kellogg-Stedman <lars@seas.harvard.edu>
+Alex Botero-Lowry <alex.boterolowry@gmail.com>
+Ingmar Vanhassel <ingmar@exherbo.org>
+Aron Griffis <agriffis@n01se.net>
+Adrian Perez de Castro <aperez@igalia.com>
+Israel Herraiz <isra@herraiz.org>
+Mikhail Gusarov <dottedmag@dottedmag.net>
+EOF
+test_expect_equal_file OUTPUT EXPECTED
+
+test_begin_subtest "--output=sender --format=json"
+notmuch search --output=sender --format=json '*' >OUTPUT
+cat <<EOF >EXPECTED
+[{"name": "François Boulogne", "address": "boulogne.f@gmail.com", "name-addr": "François Boulogne <boulogne.f@gmail.com>"},
+{"name": "Olivier Berger", "address": "olivier.berger@it-sudparis.eu", "name-addr": "Olivier Berger <olivier.berger@it-sudparis.eu>"},
+{"name": "Chris Wilson", "address": "chris@chris-wilson.co.uk", "name-addr": "Chris Wilson <chris@chris-wilson.co.uk>"},
+{"name": "Carl Worth", "address": "cworth@cworth.org", "name-addr": "Carl Worth <cworth@cworth.org>"},
+{"name": "Alexander Botero-Lowry", "address": "alex.boterolowry@gmail.com", "name-addr": "Alexander Botero-Lowry <alex.boterolowry@gmail.com>"},
+{"name": "Keith Packard", "address": "keithp@keithp.com", "name-addr": "Keith Packard <keithp@keithp.com>"},
+{"name": "Jjgod Jiang", "address": "gzjjgod@gmail.com", "name-addr": "Jjgod Jiang <gzjjgod@gmail.com>"},
+{"name": "Rolland Santimano", "address": "rollandsantimano@yahoo.com", "name-addr": "Rolland Santimano <rollandsantimano@yahoo.com>"},
+{"name": "Jan Janak", "address": "jan@ryngle.com", "name-addr": "Jan Janak <jan@ryngle.com>"},
+{"name": "Stewart Smith", "address": "stewart@flamingspork.com", "name-addr": "Stewart Smith <stewart@flamingspork.com>"},
+{"name": "Lars Kellogg-Stedman", "address": "lars@seas.harvard.edu", "name-addr": "Lars Kellogg-Stedman <lars@seas.harvard.edu>"},
+{"name": "Alex Botero-Lowry", "address": "alex.boterolowry@gmail.com", "name-addr": "Alex Botero-Lowry <alex.boterolowry@gmail.com>"},
+{"name": "Ingmar Vanhassel", "address": "ingmar@exherbo.org", "name-addr": "Ingmar Vanhassel <ingmar@exherbo.org>"},
+{"name": "Aron Griffis", "address": "agriffis@n01se.net", "name-addr": "Aron Griffis <agriffis@n01se.net>"},
+{"name": "Adrian Perez de Castro", "address": "aperez@igalia.com", "name-addr": "Adrian Perez de Castro <aperez@igalia.com>"},
+{"name": "Israel Herraiz", "address": "isra@herraiz.org", "name-addr": "Israel Herraiz <isra@herraiz.org>"},
+{"name": "Mikhail Gusarov", "address": "dottedmag@dottedmag.net", "name-addr": "Mikhail Gusarov <dottedmag@dottedmag.net>"}]
+EOF
+test_expect_equal_file OUTPUT EXPECTED
+
+test_begin_subtest "--output=recipients"
+notmuch search --output=recipients '*' >OUTPUT
+cat <<EOF >EXPECTED
+Allan McRae <allan@archlinux.org>
+"Discussion about the Arch User Repository (AUR)" <aur-general@archlinux.org>
+olivier.berger@it-sudparis.eu
+notmuch@notmuchmail.org
+notmuch <notmuch@notmuchmail.org>
+Keith Packard <keithp@keithp.com>
+Mikhail Gusarov <dottedmag@dottedmag.net>
+EOF
+test_expect_equal_file OUTPUT EXPECTED
+
+test_begin_subtest "--output=sender --output=recipients"
+notmuch search --output=sender --output=recipients '*' >OUTPUT
+cat <<EOF >EXPECTED
+François Boulogne <boulogne.f@gmail.com>
+Allan McRae <allan@archlinux.org>
+"Discussion about the Arch User Repository (AUR)" <aur-general@archlinux.org>
+Olivier Berger <olivier.berger@it-sudparis.eu>
+olivier.berger@it-sudparis.eu
+Chris Wilson <chris@chris-wilson.co.uk>
+notmuch@notmuchmail.org
+Carl Worth <cworth@cworth.org>
+Alexander Botero-Lowry <alex.boterolowry@gmail.com>
+Keith Packard <keithp@keithp.com>
+Jjgod Jiang <gzjjgod@gmail.com>
+Rolland Santimano <rollandsantimano@yahoo.com>
+Jan Janak <jan@ryngle.com>
+Stewart Smith <stewart@flamingspork.com>
+Lars Kellogg-Stedman <lars@seas.harvard.edu>
+notmuch <notmuch@notmuchmail.org>
+Alex Botero-Lowry <alex.boterolowry@gmail.com>
+Ingmar Vanhassel <ingmar@exherbo.org>
+Aron Griffis <agriffis@n01se.net>
+Adrian Perez de Castro <aperez@igalia.com>
+Israel Herraiz <isra@herraiz.org>
+Mikhail Gusarov <dottedmag@dottedmag.net>
+EOF
+test_expect_equal_file OUTPUT EXPECTED
+
 test_begin_subtest "sanitize output for quoted-printable line-breaks in author and subject"
 add_message "[subject]='two =?ISO-8859-1?Q?line=0A_subject?=
        headers'"