On Thu, Oct 18, 2012 at 01:11:00PM +0200, Pavel Březina wrote:
>
https://fedorahosted.org/sssd/ticket/1482
> From b40a0e4d5a05e78adce3c2eff0e981333a1a6e40 Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?Pavel=20B=C5=99ezina?= <pbrezina(a)redhat.com>
> Date: Thu, 18 Oct 2012 12:42:50 +0200
> Subject: [PATCH] do not default fullname to gecos when schema = ad
>
>
https://fedorahosted.org/sssd/ticket/1482
> ---
> src/db/sysdb.h | 3 +++
> src/db/sysdb_ops.c | 29 ++++++++++++++++++++---------
> src/db/sysdb_subdomains.c | 5 +++--
> src/providers/ldap/sdap_async_users.c | 21 +++++++++++++++++++--
> src/providers/proxy/proxy_id.c | 1 +
> src/responder/pac/pacsrv_cmd.c | 2 +-
> src/tests/sysdb-tests.c | 18 +++++++++++-------
> src/tools/sss_seed.c | 2 +-
> src/tools/sss_sync_ops.c | 2 +-
> 9 files changed, 60 insertions(+), 23 deletions(-)
>
> diff --git a/src/db/sysdb.h b/src/db/sysdb.h
> index
0ce6beaa361bcb5930e64c31fc3763a230881e9a..6b36db0fa3567ed1d6b3d0deb6313154b57265b1 100644
> --- a/src/db/sysdb.h
> +++ b/src/db/sysdb.h
> @@ -609,6 +609,7 @@ int sysdb_get_new_id(struct sysdb_ctx *sysdb,
> /* Add user (only basic attrs and w/o checks) */
> int sysdb_add_basic_user(struct sysdb_ctx *sysdb,
> const char *name,
> + const char *fullname,
> uid_t uid, gid_t gid,
> const char *gecos,
> const char *homedir,
> @@ -617,6 +618,7 @@ int sysdb_add_basic_user(struct sysdb_ctx *sysdb,
> /* Add user (all checks) */
> int sysdb_add_user(struct sysdb_ctx *sysdb,
> const char *name,
> + const char *fullname,
> uid_t uid, gid_t gid,
> const char *gecos,
> const char *homedir,
> @@ -662,6 +664,7 @@ int sysdb_mod_group_member(struct sysdb_ctx *sysdb,
>
> int sysdb_store_user(struct sysdb_ctx *sysdb,
> const char *name,
> + const char *fullname,
> const char *pwd,
> uid_t uid, gid_t gid,
> const char *gecos,
> diff --git a/src/db/sysdb_ops.c b/src/db/sysdb_ops.c
> index
52f0cef54b9819b1f5d754dcbd56abd1d7cb0e96..859d1d58b9959da778fac08b6f5cf426d2433d25 100644
> --- a/src/db/sysdb_ops.c
> +++ b/src/db/sysdb_ops.c
> @@ -776,6 +776,7 @@ done:
>
> int sysdb_add_basic_user(struct sysdb_ctx *sysdb,
> const char *name,
> + const char *fullname,
> uid_t uid, gid_t gid,
> const char *gecos,
> const char *homedir,
I would prefer not to add another argument to sysdb_add_user(),
sysdb_store_user() and especially not to sysdb_add_basic_user(). I think
the same result can be reached with fewer changes if ...
>
> errno_t sysdb_delete_domuser(struct sss_domain_info *domain,
> diff --git a/src/providers/ldap/sdap_async_users.c
b/src/providers/ldap/sdap_async_users.c
> index
8974e6a249f1c7db94a7dda63762d84ddd4f0347..f392b594a2fbf043a7dca963161715197639b36b 100644
> --- a/src/providers/ldap/sdap_async_users.c
> +++ b/src/providers/ldap/sdap_async_users.c
> @@ -42,6 +42,7 @@ int sdap_save_user(TALLOC_CTX *memctx,
> struct ldb_message_element *el;
> int ret;
> const char *name = NULL;
> + const char *fullname = NULL;
> const char *pwd;
> const char *gecos;
> const char *homedir;
> @@ -82,6 +83,22 @@ int sdap_save_user(TALLOC_CTX *memctx,
> goto fail;
> }
>
> + if (opts->schema_type == SDAP_SCHEMA_AD) {
> + ret = sysdb_attrs_get_el(attrs,
> + opts->user_map[SDAP_AT_USER_FULLNAME].sys_name,
&el);
> + if (ret == EOK) {
> + if (el->num_values == 0) {
> + fullname = NULL;
> + } else {
> + fullname = (const char *)el->values[0].data;
> + }
> + } else if (ret == ENOENT) {
> + fullname = NULL;
> + } else {
> + goto fail;
> + }
> + }
> +
> ret = sysdb_attrs_get_el(attrs,
> opts->user_map[SDAP_AT_USER_PWD].sys_name,
&el);
> if (ret) goto fail;
> @@ -343,8 +360,8 @@ int sdap_save_user(TALLOC_CTX *memctx,
>
> DEBUG(6, ("Storing info for user %s\n", name));
>
> - ret = sysdb_store_user(ctx, name, pwd, uid, gid, gecos, homedir, shell,
> - user_attrs, missing, cache_timeout, now);
> + ret = sysdb_store_user(ctx, name, fullname, pwd, uid, gid, gecos, homedir,
> + shell, user_attrs, missing, cache_timeout, now);
> if (ret) goto fail;
>
> if (_usn_value) {
... you add fullname to the user_attrs. Then sysdb_add_basic_user() will
set fullname==gecos when it initially creates the user object in the
cache, but it will be overwritten in the same transaction when
sysdb_store_user() add all the user_attrs.
Hi,
that's a good idea. I'm sending the patch. I did not test it though,
because I need to leave now. But it is pretty straightforward so I hope
it's OK.
Thanks,
Pavel.