Fix problems with initial connections and name-setting in conferences

[zugz]

This change is 

[iphydf]

Reviewed 1 of 6 files at r1.
Reviewable status: 0 of 1 LGTMs obtained

---

auto_tests/conference_test.c, line 105 at r1 (raw file):

    for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
        for (uint16_t j = 0; j < NUM_GROUP_TOX; ++j) {
            const int len = tox_conference_peer_get_name_size(toxes[i], 0, j, nullptr);

size_t or uint32_t

---

auto_tests/conference_test.c, line 154 at r1 (raw file):

        char name[NAMELEN+1];
        snprintf(name,NAMELEN+1,"Tox #%4d",tox_index[i]);

Run format-source.

---

toxcore/group.c, line 2043 at r1 (raw file):

static bool check_message_info(uint32_t message_number, uint8_t message_id, Group_Peer *peer)
{
    bool ignore_older = (message_id == GROUP_MESSAGE_NAME_ID || message_id == GROUP_MESSAGE_TITLE_ID);

const bool

---

toxcore/group.c, line 2046 at r1 (raw file):

    Message_Info *i;
    for (i = peer->last_message_infos; i < peer->last_message_infos + peer->num_last_message_infos; ++i) {

Can you factor out this for-loop into a separate function returning Message_Info* or nullptr?

---

toxcore/group.c, line 2067 at r1 (raw file):

    for (Message_Info *j = peer->last_message_infos + peer->num_last_message_infos - 1; j > i; --j) {
        *j = *(j-1);

This looks like memmove. Use memmove?

[iphydf]

Reviewable status: 0 of 1 LGTMs obtained

---

auto_tests/conference_test.c, line 155 at r1 (raw file):

        char name[NAMELEN+1];
        snprintf(name,NAMELEN+1,"Tox #%4d",tox_index[i]);
        tox_self_set_name(toxes[i], name, NAMELEN, nullptr);

name needs a cast to 'const uint8_t *`.

[iphydf]

Reviewable status: 0 of 1 LGTMs obtained

---

auto_tests/conference_test.c, line 105 at r1 (raw file):

Previously, iphydf wrote…

size_t or uint32_t

Still (this comment is on line 105, regarding len).

---

auto_tests/conference_test.c, line 114 at r2 (raw file):

            snprintf(expected_name, NAMELEN + 1, NAME_FORMAT_STR, tox_index[j]);
            ck_assert_msg(memcmp(name, expected_name, NAMELEN) == 0,
                          "name of #%d according to #%d is \"" NAME_FORMAT "\"; expected \"%s\"",

%u since tox_indexes are unsigned.

---

auto_tests/conference_test.c, line 164 at r2 (raw file):

        char name[NAMELEN + 1];
        snprintf(name, NAMELEN + 1, NAME_FORMAT_STR, tox_index[i]);
        tox_self_set_name(toxes[i], (const uint8_t *) name, NAMELEN, nullptr);

No space after cast.

---

toxcore/group.c, line 2042 at r2 (raw file):

}

Message_Info *find_message_slot_or_reject(uint32_t message_number, uint8_t message_id, Group_Peer *peer)

Make this internal linkage.

---

toxcore/group.c, line 2048 at r2 (raw file):

    Message_Info *i;

    for (i = peer->last_message_infos; i < peer->last_message_infos + peer->num_last_message_infos; ++i) {

You can write for (Message_Info *i = ...) { and instead of break, write return i and instead of return nullptr, write break, and at the bottom return nullptr.

---

toxcore/group.c, line 2072 at r2 (raw file):

static bool check_message_info(uint32_t message_number, uint8_t message_id, Group_Peer *peer)
{
    Message_Info *i = find_message_slot_or_reject(message_number, message_id, peer);

Message_Info *const i = ...

---

toxcore/group.c, line 2121 at r2 (raw file):

    if (g->number_joined != -1 && count_close_connected(g) >= DESIRED_CLOSE_CONNECTIONS) {
        int fr_close_index = friend_in_close(g, g->number_joined);

const

[zugz]

* Sunday, 2018-07-29 at 17:13 +0000 - iphydf <notifications@github.com>:

You can write `for (Message_Info *i = ...) {` and instead of `break`, write `return i` and instead of `return nullptr`, write `break`, and at the bottom `return nullptr`.

That doesn't work, sadly - we're also interested in the case that the condition in the 'for' fails. Done the rest, except for fixing the other conference tests.

[iphydf]

Reviewable status: 0 of 1 LGTMs obtained

---

auto_tests/conference_test.c, line 21 at r3 (raw file):

#define GROUP_MESSAGE "Install Gentoo"

#define NAME_FORMAT_STR "Tox #%4d"

%4u

---

toxcore/group.c, line 2143 at r3 (raw file):

    uint32_t message_number;
    memcpy(&message_number, data + sizeof(uint16_t), sizeof(message_number));

Can you use net_unpack_u32 here?

[iphydf]

Reviewable status: 0 of 1 LGTMs obtained

---

auto_tests/conference_test.c, line 108 at r3 (raw file):

        for (uint16_t j = 0; j < NUM_GROUP_TOX; ++j) {
            const size_t len = tox_conference_peer_get_name_size(toxes[i], 0, j, nullptr);
            ck_assert_msg(len == NAMELEN, "name of #%u according to #%u has incorrect length %u", tox_index[j], tox_index[i], len);

Now you need to cast len to (unsigned int) to fit into %u.

[zugz]

* Sunday, 2018-07-29 at 21:34 +0000 - iphydf <notifications@github.com>:

*[toxcore/group.c, line 2143 at r3](https://reviewable.io/reviews/toktok/c-toxcore/1035#-LIbxWo1AMSHkR82qOPX:-LIbxWo1AMSHkR82qOPY:bllger9) ([raw file](https://github.com/toktok/c-toxcore/blob/ed89dc79aaac37132a21dd5eb8a4e2b681366235/toxcore/group.c#L2143)):* > ```C > > uint32_t message_number; > memcpy(&message_number, data + sizeof(uint16_t), sizeof(message_number)); > ``` Can you use `net_unpack_u32` here?

This was existing code, just moved. I can change it, but maybe it would make more sense to convert the whole codebase to use net_unpack_* in one go, in a separate PR?

[CLAassistant]

All committers have signed the CLA.

[iphydf]

Reviewed 1 of 3 files at r2, 6 of 6 files at r4.
Reviewable status: complete! 1 of 1 LGTMs obtained

[robinlinden]

Reviewed 1 of 3 files at r2, 6 of 6 files at r4.
Reviewable status: complete! 2 of 1 LGTMs obtained

---

auto_tests/conference_test.c, line 132 at r4 (raw file):

            const size_t len = tox_conference_peer_get_name_size(toxes[i], 0, j, nullptr);
            ck_assert_msg(len == NAMELEN, "name of #%u according to #%u has incorrect length %u", state[j].id, state[i].id,
                          (unsigned int) len);

Nit: no space between cast and variable.

[zugz]

* Wednesday, 2018-08-01 at 19:35 +0000 - Robin Lindén <notifications@github.com>:

Nit: no space between cast and variable.

Done. Also improved connectedness detetection, and signed the commit.

[robinlinden]

Reviewed 2 of 2 files at r5.
Reviewable status: complete! 2 of 1 LGTMs obtained