- warnings fixed
42 lines of code changed in 11 files:
- fixed illegal access to a value which has left the visible range
3 lines of code changed in 1 file:
- access to illegal position
2 lines of code changed in 1 file:
- what is the advantage of an array with one element?!
17 lines of code changed in 1 file:
- fixed warnings
3 lines of code changed in 2 files:
- updated prototypes
65 lines of code changed in 1 file:
- depends updated
14 lines of code changed in 1 file:
- added strupr / strlwr comments
3 lines of code changed in 1 file:
- cosmetics
1 lines of code changed in 1 file:
- moved DB repair to NT_dbrepair.cxx
- fixed illegal memory access
6 lines of code changed in 1 file:
- DB repair done here now
- added fix for dictionary compression bug
714 lines of code changed in 2 files:
- added NT_dbrepair.o
26 lines of code changed in 1 file:
- prototypes updated
15 lines of code changed in 3 files:
- declared anonymous struct DictData
1 lines of code changed in 1 file:
- added some comments
- cosmetics
- GB_read_pntr() now returns NULL on (decompression) error
61 lines of code changed in 1 file:
- gb_load_single_key_data() does no longer raise an internal error, if an
dictionary gets deleted
- added functions to access existing directories (GB_get_dictionary,
GB_set_dictionary, GB_free_dictionary)
64 lines of code changed in 1 file:
- added GBS_strforget()
5 lines of code changed in 1 file:
- if alignment has length -1, do not stop with error, just print a warning
- g_b_uncompress_single_sequence_by_master(): return decompressed size
- gb_uncompress_by_sequence():
. return decompressed size
. if error reading master, return NULL and set error (instead of
returning the error message as result (sic!))
16 lines of code changed in 1 file:
- cleaned up get_data_n_size()
- gb_convert_compression():
. separated copy and re-write
- gb_uncompress_by_dictionary_internal(): return decompressed size
- gb_uncompress_by_dictionary(): return NULL on error
99 lines of code changed in 1 file:
- added GB_UNCOMPRESSED_SIZE()
2 lines of code changed in 1 file:
- bugfix: calculate hash index according to case_sens defined in index
3 lines of code changed in 1 file:
- added GBS_hash_count_value()
18 lines of code changed in 1 file:
- several functions got static
- all compression functions return the new (uncompressed) size
- gb_uncompress_data():
. returns NULL in case of error
. check decompressed size vs. expected size. error on mismatch
- added GB_is_directory_compressed()
124 lines of code changed in 1 file:
- added -ggdb -ggdb3
2 lines of code changed in 1 file:
Fixed multiple nodes bug in sequence quality entries (only first one was shown).
20 lines of code changed in 1 file:
- replaced call to gbs_stricmp
- fixed calls to GBS_create_hash
7 lines of code changed in 1 file:
- replaced calls to GBS_stricmp etc.
4 lines of code changed in 2 files:
- fixed calls to GBS_create_hash
72 lines of code changed in 42 files:
- prototypes updated
1 lines of code changed in 2 files:
- uses GB_CALC_HASH_INDEX_CASE_IGNORED instead of own macro
3 lines of code changed in 1 file:
- defines hash-index macros (now used by adhash.c and adindex.c)
32 lines of code changed in 1 file:
- removed gbs_stricmp, GBS_stricmp and gbs_strnicmp (use strcasecmp and strncasecmp)
- fixed GBS_create_hash calls
4 lines of code changed in 1 file:
- hash uses GB_CASE instead of bool flag
- hash-index macros: moved to adlocal.h
- added find_hash_entry to avoid duplicated code
- added write_hash (used by GBS_write_hash and GBS_write_hash_no_strdup)
89 lines of code changed in 1 file:
- fixed use of GBS_strstruct functions
156 lines of code changed in 28 files:
- prototypes updated
11 lines of code changed in 2 files:
- changed void* to GBS_strstruct*
31 lines of code changed in 1 file:
- define incomplete type GBS_strstruct
1 lines of code changed in 1 file:
- bool/GB_BOOL/AW_BOOL-fixes
- replaced some casts
14 lines of code changed in 1 file:
- bool/GB_BOOL/AW_BOOL-fixes
43 lines of code changed in 15 files:
- gde_filter_weights: one return point
- replaced GBS_string_cmp by GBS_stricmp and GBS_string_matches
23 lines of code changed in 1 file:
- use GB_CASE param for client->server-search
9 lines of code changed in 1 file:
- prototypes updated
9 lines of code changed in 2 files:
- changed GBS_string_cmp
- convert GB_CASE to int for perl interface
2 lines of code changed in 1 file:
- use GB_CASE parameter
- changed GBS_string_[s]cmp
94 lines of code changed in 28 files:
- GB_create_index:
. specify case-sensitivity when creating
. single return point
- gb_index_find:
. specify case-sensitivity (much match with created index)
54 lines of code changed in 1 file:
- increased map file version to 5 (due to case-sensitivity flag
in indices)
1 lines of code changed in 1 file:
- all functions with case-sensitivity flag now use GB_CASE,
so it's obvious at caller whether case is ignored or not
24 lines of code changed in 1 file:
- GBS_string_cmp -> GBS_string_matches
. was working like strcmp, but this does not make sense, since
2nd parameter can contain wildcards, so it wasn't and won't
ever be used for sorting
. result changed (before: 0=="equal", now: GB_TRUE==matches)
- GBS_string_scmp -> GBS_string_matches_regexp
. result changed (before: 0=="equal", now: GB_TRUE==matches)
- added GBS_stricmp (case insensitive strcmp)
61 lines of code changed in 1 file:
- added GB_CASE
- GB_BOOL is enum now
14 lines of code changed in 1 file:
- added case_sensitivity to gb_index_files_struct
- new type GB_CASE
4 lines of code changed in 1 file:
- MG_transfer_fields_cb: due to wrong compare, sequences were not
realigned when they were transfered using Transfer fields.
2 lines of code changed in 1 file:
- simplified code
9 lines of code changed in 1 file:
- fixed a bug in GBS_hash_next_element (it stopped iteration at first
hash index conflict). Maybe fixes random strange behavior with cached
taxonomies and possible, unrecognized problems with add partial
sequences in ARB_PARS.
15 lines of code changed in 1 file:
- mostly const fixes
167 lines of code changed in 35 files:
- depends updated
2 lines of code changed in 1 file:
- avoid writing into const memory
7 lines of code changed in 1 file:
- fixed illegal free of DB-memory
1 lines of code changed in 1 file:
- call ./arb_valgrind not arb_valgrind
1 lines of code changed in 1 file:
- prototypes updated
33 lines of code changed in 3 files:
- rewrote GBS_eval_env(), now it does no longer modify const memory
and it does no rekursive expansion (if this is really used/needed
somewhere, it should be called several times instead)
- GBS_string_eval:
. do not write to const data
. removed very strange frees/dups
- several functions now return GB_CSTR:
. GBS_find_string
. GBS_regsearch
. gb_compile_regexpr
. GBS_regsearch
regexp-stuff is not test-compiled with undefined NO_REGEXPR;
have no such system.
52 lines of code changed in 1 file:
- GB_login:
. only one return-point
. changes due to constness
124 lines of code changed in 1 file:
- gb_findExtension gets unconst argument, cause its used
to modify writeable strings
- fixed some buffer-sizes
20 lines of code changed in 1 file:
- several functions now return GB_CSTR:
. GB_read_pntr
. GB_read_char_pntr
. GB_read_link_pntr
. GB_read_bits_pntr
. GB_read_bytes_pntr
. gb_read_pntr_ts
25 lines of code changed in 1 file:
- replaced strange typename GB_CPNTR (sounds const but isnt)
by GB_BUFFER and GB_CBUFFER
- added non-const wrapper for GBS_find_string (c++ only)
16 lines of code changed in 1 file:
- cosmetics
1 lines of code changed in 1 file:
- replacement was wrong if parent directory contained 'ARB.o' (e.g. 'ARB.openGL')
3 lines of code changed in 1 file:
- now not _only_ auto-generated
65 lines of code changed in 1 file:
- added target 'proto', put GDE_proto.h into CVS and updated
broken depends
9 lines of code changed in 1 file:
- removed GDE_proto.h
0 lines of code changed in 1 file:
- added ARB_GDE to proto
2 lines of code changed in 1 file:
- prototypes updated
16 lines of code changed in 3 files:
- depends updated
0 lines of code changed in 3 files:
- search for '@name' is now case-sensitive!
(fixes a serious DB bug, where directories were overwritten during
DB optimization, resulting in data corruption)
9 lines of code changed in 1 file:
- shorten displayed name for db in DB viewer
28 lines of code changed in 1 file:
- replaced calls to GB_find with new find-functions
1092 lines of code changed in 111 files:
- TEST_SOME now allows to specify the keys
15 lines of code changed in 1 file:
- show key index
9 lines of code changed in 1 file:
- fixed non-working 'skip newline'
3 lines of code changed in 1 file:
- added POST_COMPILE
- explicit rules (avoids deleting objects after compile)
- compile arb_test + arb_perf_test only in DEBUG mode
5 lines of code changed in 1 file:
- test program using different case for field names
167 lines of code changed in 1 file:
- pass DEBUG flag to sub-Makefiles
2 lines of code changed in 1 file:
- changed predefined prototype for P2A_find
2 lines of code changed in 1 file:
- added GB_has_key()
16 lines of code changed in 1 file:
- added GB_expect_error
8 lines of code changed in 1 file:
- gbcms_talking_find/GBCMC_find use GB_NONE to search unspecific child
15 lines of code changed in 1 file:
- added case-sensitivity parameter to functions searching for childs
with specific content
- added several replacement functions for GB_find:
GB_find_string,
GB_child, GB_nextChild,
GB_entry, GB_nextEntry,
GB_brother.
GB_find should now only be used is special cases!
80 lines of code changed in 1 file:
- fix dependency paths
2 lines of code changed in 1 file:
- removed test code
15 lines of code changed in 1 file:
- removed adseqcompr.h
0 lines of code changed in 1 file:
- prototypes updated
3 lines of code changed in 2 files:
- depends updated
0 lines of code changed in 3 files:
- fixed DEBUG+RELEASE compile
4 lines of code changed in 3 files:
- align tree selection box to longest tree name
7 lines of code changed in 1 file:
- type fix
4 lines of code changed in 1 file:
- comment
2 lines of code changed in 1 file:
- renamed 2 globals (size,low)
16 lines of code changed in 1 file:
- added comments for GB_warning, GB_information, GB_status, GB_status2
- added GBS_readable_size (show size info in human readable form)
42 lines of code changed in 1 file:
- GBT_link_tree_using_species_hash skips generation of "seen species"-hash
if duplicate-counting and status-display is off.
Else, it detects size of "seen species"-hash by counting nodes in tree,
not by counting existing species (fixes slow link on huge "zombie"-trees
in DBs with much less species)
- new parameter maxTreeNameLen to GBT_tree_info_string (used for proper text-alignment)
- comments on GBT_message
58 lines of code changed in 1 file:
- Status trace now has 2 levels of detail
- Renamed 'Kill' into 'Abort' in status window
- After timeout w/o aborting, the popup-window shows
a new option 'Continue' -> Encourage user to mail to devel.
- in aw_status_timer_listen_event, one read did not update 'gaugeValue'.
Therefore some status updates where skipped, which was no problem in
most cases, but if a zero-gaugeValue (which is interpreted as "Reset
starting time") was ignored, estimated time went completely wrong.
- avoid sending int 0 to status process for low (but non-zero) gauge values
87 lines of code changed in 1 file:
- nni_rek() only opens and updates status if flag is set;
before it updated always.
abort now also works if flag is cleared.
23 lines of code changed in 1 file:
- added status-flag to struct isits.
when set, update status for each single species
when cleared, update status for progress over multiple species
- global optimization now abortable
40 lines of code changed in 1 file:
- added functions to mark deep and degenerated branches
105 lines of code changed in 2 files:
- global optimize: single status windows for optimizing sequence and
non-sequence data
- added functions to mark deep and degenerated branches
43 lines of code changed in 1 file:
- GB_CTREE moved into adseqcompr.c
0 lines of code changed in 1 file:
- sequence compression
. allow up to 50 sequences per master
. removed header adseqcompr.h (types were local)
. can be aborted now
. much better master distribution, to support compression trees with
degenerated branches. I had one DB where more than 300 decompression steps
where necessary to get the data of one sequence - impossible to work with.
. added test code to save tree with master-locations as groups
. much code now static
. warns if used tree is very degenerated
. displays correct status for different phases of optimization
. localized variable-scopes
. fixed compression statistic
. fixed missing frees
462 lines of code changed in 1 file:
removed (types only used locally in adseqcompr.c)
0 lines of code changed in 1 file:
- comment on GB_STRING_SHRT
14 lines of code changed in 1 file:
- added debug code to gb_read_bin_rek_V2()
- open files in binary mode
21 lines of code changed in 1 file:
- gb_write_bin_rek:
. auto-change db-type back to GB_STRING (when size is too big)
. test GB_STRING_SHRT for correct size and final zero-byte.
if test fails -> fallback to GB_STRING
Before corrupt databases where saved to disk, whenever
the string data contained a zero-byte, resulting in
unloadable databases ("Unknown DB type 0")
. added assertion to catch unknown DB types
50 lines of code changed in 1 file: