summaryrefslogtreecommitdiffstats
path: root/generic/tclStrIdxTree.c
Commit message (Collapse)AuthorAgeFilesLines
* clock option re-orderingjan.nijtmans2024-05-221-4/+3
|\
| * clock option re-orderingjan.nijtmans2024-05-221-4/+3
| |
* | Tcl_DuplicateObj can't return NULLdkf2024-04-181-22/+30
| |
* | Merge 8.7jan.nijtmans2024-04-141-1/+1
|\ \ | |/
| * Some int -> Tcl_Size changesjan.nijtmans2024-04-141-4/+4
| |
* | Merge 8.7dkf2024-04-121-1/+1
|\ \ | |/
| * More cleaning up, notably no more symbols starting with _ as they're a ↵dkf2024-04-121-1/+1
| | | | | | | | reserved 'namespace'
* | Merge 8.7jan.nijtmans2024-04-111-107/+129
|\ \ | |/
| * No need for TCL_OBJTYPE_V0 in Tcl 8.7. Make StrIdxTreeObjType static const. ↵jan.nijtmans2024-04-111-3/+2
| | | | | | | | Define TCL_OBJTYPE_V1/TCL_OBJTYPE_V2 for 8.7
| * A few more small changesdkf2024-04-111-5/+13
| |
| * Style cleanup, plus added comments on memory managementdkf2024-04-111-108/+122
| |
* | Merge 8.7. encodingPtr -> encodingNamePtr (for consistency)jan.nijtmans2024-04-041-1/+1
|\ \ | |/
| * Merge 8.6jan.nijtmans2024-04-041-1/+1
| |
* | Oops, ckalloc -> Tcl_Allocjan.nijtmans2024-04-011-5/+5
|/
* TIP 688: clock command revision and speedupjan.nijtmans2024-04-011-5/+5
|
* Review: use Tcl_Size for "length" in tclStrIdxTree.h. Macro tweaks.jan.nijtmans2024-03-201-14/+12
|
* Fix [1acd172c424b57c9] (by just reverting the change causing this). Not ↵jan.nijtmans2024-03-111-4/+5
| | | | crucial for TIP #688. Also, make it compile/run using -DTCL_NO_DEPRECATED=1
* (cherry-pick) More tweaksjan.nijtmans2024-02-261-1/+1
|
* Rebase to latest 9.0jan.nijtmans2024-02-261-6/+7
|
* Merge 8.7. Use more TCL_UNUSED()jan.nijtmans2021-02-191-4/+4
|
* Merge 8.6. More warning fixesjan.nijtmans2020-10-151-3/+3
|
* Fixes sebres/tclclockmod#18 (Fails to parse short month name for June); sebres2019-07-151-1/+1
| | | | | The reason for that was the wrong length calculation by scanning through my string index tree, so the ambiguity check `j->(jan,ju->(jun,jul))` failed for 2nd element Ju(ne) with length 2. Simple fix and test-cases covering that, but it looks like this has a good potential for speedup (todo: move length calculation from search to build)
* generic\tclStrIdxTree.c: bug fix (lost abbreviation), if tree will be built ↵sebres2018-05-291-1/+2
| | | | | | | from 2 lists, and the short form differentiate from long form ("jan." vs "january", note the dot-char), don't use longest form - should be split in 2 entries with common parent "jan". Thus use simple case (don't split) only if found item is covered in full (e. g. "jan" vs "january", note without dot-char). Test covered now (clock-29.181x, locale "fr") - `clock scan [clock format 0 -format {%a %d-%m-%Y} -locale fr] -format {%a %d-%m-%Y} -locale fr`
* Initialize prevf to fix (used before set) warning. sebres2018-05-291-1/+1
| | | | * Prevf doesn't get used at line 145 unless `prevItem != NULL` and prevf also gets set in the same block, so this is safe.
* Fix LookupLastTransition() for behavior when tick < compVal, undo ↵jan.nijtmans2017-06-071-6/+6
| | | | Tcl_EvalObjEx -> TclEvalObjEx change, eliminate two inline macro's which are only used once. Eliminate many unnecessary MODULE_SCOPE declarations (duplicated with header files)
* missing static keyword for inline declarationssebres2017-06-021-2/+2
|
* More code review, e.g. use Tcl_SetObjResult in stead of Tcl_SetResult, ↵jan.nijtmans2017-05-311-18/+18
| | | | preventing a (char *) type case. No functional changes.
* small code review: resolves several warning on some compilerssebres2017-05-301-2/+2
|
* [clock] tclStrIdxTree extended with possibility to hold client data; also ↵sebres2017-05-111-26/+33
| | | | changed in clock - indices starts with 1 instead of 0, and 0(NULL) instead of -1 used as sign of ambiguous keys.
* Added files missing after merge/back-port (rebase with merge point)sebres2017-05-101-0/+520
generated by cgit v0.12 at 2025-12-05 05:43:54 (GMT)