TODO
* add LZXPRESSED compressed string to compression test
* add leaf value (data definition) cache?
* use leaf page previous/next values?

* fix XOR checksum in esedb_test_checksum_calculate_little_endian_ecc32 test
* libfdata next search keep going on a key match?

* fix flaky test for System_Search_AutoSummary (one byte off?)
* pyesedb
  - null value support
  - long value: add data size support
* refactor libesedb_record_get_value_binary_data_size and
  libesedb_record_get_value_binary_data
* complete libfdata btree refactor
  - libesedb_values_tree.c replace by fdata functions?
  - fix is_flexible_match for xp_search.edb indexes
* record flags
  - add indication of corrupt records

* format notes:
  - value of un-initialized page checksum: 0xdeadbeefbaadf00d
    "ESE Super ECCXORChecksum"

* create tool to analyze cache usage, hits and misses
* move read of data definition into leaf value callback
  - create separate callback for long values?
  - this will clean up data definition
* refactor:
  - libesedb_multi_value_get_value what about value_type ?
  - clean up multi value calls to libfvalue
  - libesedb_long_value_get_segment_data
* deprecate:
  - libesedb_record_get_value
  - libesedb_multi_value_get_value
* complete libfvalue update integration - multi value

* windows search add support for windows 8 SystemIndex_PropertyStore
* libesedb check esedbtools_libX header files for _LIBESEDB_LIBX_ defs
* handle compressed values inside the library
  - compressed binary data
* esedbexport -t /root creates /root.export ?
* esedbexport skip table on error
* check Win2K ntds.dit support
* clean up and refactor info handle
* export handle: fix creation of indexes directory

API:
* add function to retrieve database signature
  - can be used to match up EDB/STM pairs
* add long value functions (string and binary data?)
* remove
  - libesedb_record_get_value
  - libesedb_long_value_get_segment_data
  - libesedb_multi_value_get_value

* windows search
  - multi long value segment decryption/compression support
* export error in more test files: nibble count value exceeds maximum

* code clean up
  - change page value index and number of values into int
  - clean up TODO's
  - refactor info handle
* combine libesedb_io_handle_read_page with libesedb_page_read
  - need libesedb_page_read for recovery ?

* page tree on demand
  - check sanity of page value 0 
* clean up
  - libesedb_data_type_definition remove offset
* handle default value
  - need database which uses default value
* parse variable types even when they are not part of the table catalog definition
* add support for build in table number 1
* implement cloning of items for improved asynchronous and multi threaded usage

libesedb:
* add get_table_by_object_identifier function
* ECC checksum
  - non 4 byte modulus buffer support
  - add alignment optimization
  - error correction
  - Windows 7 extended checksum ?
* template table
  - what about non-tagged data types
* multi value
  - what about multi value 0x18
* database header
  - compare primary and secondary
* catalog
  - compare primary and secondary
  - limit the size of the catalog ?
  - support for type 0x0006
  - support for type 0x0007 (line 3457 in libesedb_page_tree.c)
* long values
  - support for non-linear stored long value segments - current solution does a sanity check
* default values
   - apply to data definition

Error tolerance
* probe test for the page size using the backup database header
* use backup catalog

Debug
* is function libesedb_debug_print_table_group_of_bits needed?
* print value flags
* print part of keys as strings

Tools
* esedbinfo
  - page summary (scan for uninitialized, empty, etc. pages)
  - file header summary (check backup file header)
  - summary of catalog (check backup catalog)
  - print database state
* esedbexport
  - add export mode

Windows search:
* improve windows_search_decompress_byte_indexed_compressed_data

File format
* where is the root object defined? or is it implied to be object 1?
* compressed table, column, index names ?
* page flags
  - 0x0400, 0x0800, 0x8000
* tagged values
  - flag 0x4000
* multi value offsets
  - flag 0x8000
* page flags
* page keys
* (revision 0x11) extended page header
* what about the first byte in 7-bit ascii compressed data
* codepage 1200 string in Exchange EDB starts with 0xff 0xff
* streaming file
  - only consists of ESEDB header with paged streamed data,
    how to reconstruct streams ?

Deployment:
* update configure
  - improve libfdata detection
  - improve libfguid detection
  - improve libfwnt detection
* update manuals
  - add esedbexport.1

Performance
* create data definition tree in page tree for faster lookup of keys
  keep list for sequential access (B+-tree)
* create a UTF-8 version of run-length compressed UTF-16 string decompression function
* use register alignment for key comparison
* optimize libesedb_checksum_calculate_little_endian_ecc32

Test suite:
* windows search: work on index-based compression
  - create test program to validate output

Notes
* create a root page table instead?
  - no
* read catalog data type definitions and convert them into column definitions?
  - no

20090913
* initial version based on liblnk
* see `git log' for more recent change log