from everywhere - now RENAME, SHOW FULL TABLES, and TRUNCATE work with discovery.
improve error messages in truncate

SELECT * FROM `t-1` will no longer read "t-1.frm" file, use `#mysql50#t-1` for this.

"unknown" tables. Duplicates are possible - deal with them.

may not exist (the table exists only in the engine).

fixes for need_full_discover_for_existence mode

replace enum read_frm_op with a bitmap flags.
remove always-unused 'error' argument of get_table_share

Pass db and table_name into a function instead of the table_list,
when only db and table name are needed.

* rename check_if_table_exists() -> table_exists() and remove unneeded arguments

"
  revision-id: sanja@askmonty.org-20110511110948-4kdevwzomvk56y1w
  committer: sanja@askmonty.org
  branch nick: work-maria-5.1-CREATE-merge
  timestamp: Wed 2011-05-11 14:09:48 +0300
    Bugfix: New table creation/renaming block added if old encoded table present
"
the old behavior was less inconsistent than the new one.
In the new one the error message was sometimes different (under LOCK TABLES e.g.),
and there were race conditions (if this CREATE happened when a concurrent ALTER
has renamed the old table away but haven't put the new table in place)

The old one was like "(when using old table names) for DML #mysql50# prefix
is optional, for DDL it's required".

simplify open_table_def()

* print "table doesn't exist in engine" when a table doesn't exist in the engine,
  instead of "file not found" (if no file was involved)
* print a complete filename that cannot be found ('t1.MYI', not 't1')
* it's not an error for a DROP if a table doesn't exist in the engine (or some table
  files cannot be found) - if the DROP succeeded regardless

have a specially defined enum with clearly named values

and now-unused read_string() function

and let open_binary_frm to parse it from the buffer, not a file.
this avoids jumping back in forth in the frm file, and doing
intermediate buffer mallocs.

also:
* move few tests from archive/archive.test to archive/discover.test
* fix (unintentionally) bug#34104, updated result files

after all engines have discovered their tables
  
side effect: correct alphabetical sorting as in ORDER BY ... COLLATE utf8_bin,
information_schema is no longer the first after find_files(),
tables like #mysql50#zzz are sorted first (as per table name),
not last (as per file name zzz).

* make find_files static

there is no "upstream" OQGraph version that needs to stay compatible
with different MariaDB releases.

* provide a default bas_ext value of the empty list

and handlerton::table_exists_in_engine (new implementation is coming),
and handlerton::license (redundant)

when the function has a dedicated flag for this very use case.

(PLUGIN_VAR_MEMALLOC ix 0x8000 and cannot be saved in a char as such)

* replace pointer acrobatics with a struct
* make sorting explicit: MY_DONT_SORT -> MY_WANT_SORT
(if you want something to be done - say it. fixes all places where
my_dir() was used without thinking)
* typo s/number_off_files/number_of_files/
* directory_file_name() doesn't need to be extern
* remove #ifdef __BORLANDC__
* ignore '.' and '..' entries