in a lot of places now. It _returns_ if it fails to get the necessary memory,
so any function using it should be returning a dns_result_t.
chain_name moved toward start of file, for inlining by findnode.
move_chain_to_last broke out some functionality of dns_rbtnodechain_last
(which now calls it), also for the benefit of findnode.
dns_rbtnodechain_init on the chain used by addnode.
All of the new_foundname concatenation gook was removed from findnode,
which now just sets foundname by using chain_name.
The chain set up by findnode points to the predecessor when the searched
for name is not found.
nodechain_current was changed to take name, origin and node parameters while
returning a dns_result_t. This allows the chain returned by findnode to
be named.
The nodechain_{first,last,prev} use nodechain_current to set their
names and origins. nodechain_next does not need it because it will never
have to do the "set the origin to '.' and remove '.' from names in the
root level" bit. The root level will never be more than one name, and
everything in the megatree is a successor to that node, so 'next' never
reaches the root level.
what happens at ".". For most origins, they do NOT include themself in
a zone file ... that is, the record for "rc" appears in the $ORIGIN for
".vix.com."
I added support for the full search algorithm to the RBT database
code. The database has some special needs in searching, in particular
certain nodes need to be examined before the DOWN pointer is followed.
I solved this problem by adding a 'callback' bit to the node structure.
When findnode is called, the caller can supply a callback. Before we go
DOWN at a node, we call the callback if the callback bit is set at the
node. If the callback returns DNS_R_CONTINUE, we proceed. Otherwise, we
will stop right there and return what the caller said to return. I
added an 'empty_data_ok' flag to findnode as well, so that nodes with
empty data can be made candidates for partial match status.
I also wanted to make dns_rbtnodechain_t public, so that a chain could
be allocated on the stack by the caller. I added two routines,
dns_rbtnodechain_init() and dns_rbtnodechain_reset() to work with them.
The latter routine frees any dynamically allocated memory associated with
the chain. The chain structure now contains the memory context. I also
moved get_ancestor_mem() and put_ancestor_mem() up in the file, so that
inlining could be more effective.
In the nits department, you wrote
*node = result == DNS_R_SUCCESS ? current : NULL;
In the future, please write this instead (the patch has this change in it):
if (result == DNS_R_SUCCESS)
*node = current;
else
*node = NULL;
is currently fine, because there are only two attribute flags a name
can have, one of which the node definitely needs to have set, and the
other of which is functionally irrelevant (because it happens to be set
by the NODENAME macro, namely the read-only flag). This might not
be the right thing to do in the future when more name attributes get
added, but we'll cross that bridge if we get to it. (_Some_ change was
necessary because the node's attributes was not properly initialized when
created, and under certain circumstances this ended up triggering an
exception in the dns_name_* functions.
Watch for level overflow during dns_rbt_addnode(). Returns DNS_R_NOSPACE,
which is probably not the right thing.
Comment cleanups.
returns a dns_result_t.
dns_rbt_created requires deleter_arg NULL if deleter is NULL.
dns_rbt_addnode always returns DNS_R_EXISTS if the node is found, regardless
of if it has data.
dns_rbt_addname will add the data and return SUCCESS if addnode says
EXISTS but there is no data at the node.
make sure memory is freed even when get_ancestor_mem to reallocate fails.
partialmatch returned by findnode must have data associated with it.
fixed up deletename; prematurely exited without freeing chain memory if
the node was not found. made deletename responsible for identifying whether
a node could be removed rather than zapnode.
turned a few of the print support functions static.
name attributes, name length, offset length, and number of bytes
of "padding" that are left over when a node is split.
adjusted node splitting algorithm to preserve the original location
of the lower portion of the split node (eg, when a node has "a.b" as
its name and is split into "a" and "b", "a" retains the same address
space as "a.b".)
and use it within dns_rbt_findnode.
moved the guts of dns_rbt_deletename into its own function to clean up
the freeing of ancestor memory into just one location.
deletefromlevel required that ancestor_count be > 1, which would abort
trying to delete the root of the tree when the root had no children.
ancestor/level history no longer kept in rbt structure; node_chain structure
was created for this purpose, and each dns_rbt_deletename call will pass
the chain structure to dns_rbt_findnode.
color enum removed from rbtnode structure; single bit used instead
length of name in rbtnode no longer kept as int in rbtnode structure;
stored as single byte just past end of rbtnode, and the actual name data starts
two bytes past the end.
dns_rbt_create takes additional argument that is a pointer to a function
which is responsible for deleting the space pointed to by an rbtnode's data
pointer when the node is removed.
dns_rbt_addnode renamed to dns_rbt_addonlevel so that dns_rbt_addnode could
be a public function that can provide the address of the new node back
to the caller.
removed several #if 0 blocks that were held over from when parent pointers
were used.
dns_rbt_node_t changed to dns_rbtnode_t to match ISC conventions.
main() test routine and support functions removed, to be put in
bin/tests/rbt_test.c
adding a node and rotating left/right taught to not use parent pointers.
deletion disabled because it currently does not know how to not use
parent pointers, and since they are not being maintained by insertion,
deleting has no prayer of working.
several isc_result_t returns changed to dns_result_t.
