Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | merge core-8-6-branch | jan.nijtmans | 2017-06-07 | 1 | -1/+1 |
|\ | |||||
| * | Tcl_GetWideIntFromObj() -> TclGetWideIntFromObj(), and minor other ↵ | jan.nijtmans | 2017-06-07 | 1 | -1/+1 |
| | | | | | | | | simplifications/optimizations. No functional change. | ||||
* | | merge core-8-6-branch | jan.nijtmans | 2017-06-02 | 1 | -1/+0 |
|\ \ | |/ | |||||
* | | Merge core-8-6-branch. This removes the work currently being done in ↵ | jan.nijtmans | 2017-06-02 | 1 | -345/+0 |
|\ \ | |/ | | | | | | | "sebres-8-6-clock-speedup-cr1" branch, but that will be merged again as soon as the work is done. All other changes in "trunk" since then (e.g. the INST_STR_CONCAT1 performance improvement, and the removal of SunOS-4) are retained. | ||||
* | | merge core-8-6-branch (except: the "timerate" command and the clock ↵ | jan.nijtmans | 2017-05-31 | 1 | -11/+11 |
|\ \ | | | | | | | | | | performance-tests are kept) | ||||
| * | | Remove "timerate" functionality: this definitely needs a TIP. Also undo ↵ | jan.nijtmans | 2017-05-31 | 1 | -347/+1 |
| | | | | | | | | | | | | changes in library/reg/pkgIndex.tcl, which are unrelated to clock functionality | ||||
| * | | back-ported branch sebres_trunk_timerate (new command "timerate" for 8.6) | sebres | 2017-05-09 | 1 | -1/+347 |
| |\ \ | |/ / |/| / | |/ | |||||
* | | timerate: don't calculate threshold by too few iterations, because sometimes ↵ | sebres | 2017-03-07 | 1 | -0/+7 |
| | | | | | | | | first iteration(s) can be too fast (cached, delayed clean up, etc). | ||||
* | | merge trunk | sebres | 2017-03-06 | 1 | -3/+27 |
|\ \ | |||||
| * \ | Merge trunk. Improve test-case. Bring back "source -nopkg", but only as ↵ | jan.nijtmans | 2017-01-05 | 1 | -3/+27 |
| |\ \ | | | | | | | | | | | | | undocumented internal helper for ::tcl::Pkg::source. | ||||
| * | | | In stead of "source -nopkg" use a ::tcl::Pkg::source utility function. | jan.nijtmans | 2017-01-04 | 1 | -25/+3 |
| | | | | |||||
| * | | | Merge trunk. First steps in eliminating "source -nopkg". Some test-cases ↵ | jan.nijtmans | 2017-01-04 | 1 | -1/+1 |
| |\ \ \ | | |/ / | | | | | | | | | still fail. | ||||
| * | | | Bring back stub table in original state. | jan.nijtmans | 2016-12-08 | 1 | -61/+4 |
| |\ \ \ | | | | | | | | | | | Merge trunk | ||||
| * | | | | Complete implementation, tests and documentation | jan.nijtmans | 2016-11-28 | 1 | -2/+25 |
| | | | | | |||||
| * | | | | Merge trunk. | jan.nijtmans | 2016-11-25 | 1 | -1/+0 |
| | | | | | | | | | | | | | | | Implementation looks complete. Still missing: test-cases and documentation. | ||||
* | | | | | [timerate] bug fix: missing scale conversion by Mac OSX on platform where ↵ | sebres | 2017-02-09 | 1 | -8/+16 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | high resolution clicks are not microseconds based; [win] use high resolution timer for the wide clicks and microseconds directly, prevent several forwards/backwards conversions; [win, unix, mac-osx] normalize some functions for common usage in different time units (clicks, micro- and nanoseconds) | ||||
* | | | | | resolve warning: enumeration value ‘TMRT_LAST’ not handled in switch ↵ | sebres | 2017-02-09 | 1 | -1/+1 |
| | | | | | | | | | | | | | | | | | | | | (impossible to handle in switch because of break); | ||||
* | | | | | New performance measurement routine "timerate" in opposition to "time" the ↵ | sebres | 2017-01-09 | 1 | -1/+332 |
| |_|/ / |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | execution limited by fixed time (in milliseconds) instead of repetition count (more precise results, to prevent very long execution time it is no more necessary to estimate repetition count) Syntax: timerate ?-direct? ?-calibrate? ?-overhead double? command ?time? | ||||
* | | | | Eliminate the internal macro/function TclNewIntObj: In all cases ↵ | jan.nijtmans | 2016-12-23 | 1 | -1/+1 |
| |/ / |/| | | | | | | | | TclNewLongObj is just as good. | ||||
* | | | Route all [string repeat] operations through a common implementation. | dgp | 2016-11-30 | 1 | -61/+4 |
|/ / | | | | | | | | | Code that to preserve bytearrays, eliminate unnecessary string rep generation, increase efficiency, and encapsulate access to internal rep details. | ||||
* | | typos | jan.nijtmans | 2016-11-24 | 1 | -1/+1 |
| | | |||||
* | | Route all [string last] operations through a common implementation. | dgp | 2016-11-08 | 1 | -56/+11 |
| | | |||||
* | | Implement direct eval [string first] with the refactored engine. | dgp | 2016-11-07 | 1 | -69/+9 |
| | | |||||
* | | WIP | dgp | 2016-10-28 | 1 | -1/+2 |
| | | |||||
* | | Start bringing all `string cat` operations into one place so it can be coded | dgp | 2016-10-27 | 1 | -9/+8 |
| | | | | | | correctly one time instead of badly multiple times. | ||||
* | | Use memcmp where possible. | dgp | 2016-10-18 | 1 | -2/+2 |
| | | |||||
* | | Update [string map] tests to get illustration of need for pure dict | dgp | 2016-10-07 | 1 | -1/+1 |
|/ | | | restriction up to date with ordered dicts. | ||||
* | Simplify all the Tcl_NRPostProc declarations. | dgp | 2016-07-06 | 1 | -8/+4 |
| | |||||
* | Align comment. Remove some end-of-line spacing. No functional changes. | jan.nijtmans | 2015-09-28 | 1 | -2/+2 |
|\ | |||||
| * | Fix more corner-cases like ↵ | jan.nijtmans | 2014-05-01 | 1 | -6/+6 |
| | | | | | | | | [0e92c404f19ede5b2eb06e6db27647d3138cc56|0e92c404f1]: The only place where a type of &tclByteArrayType can be trusted is when determining its length, because the character length of a (UTF-8) string is always equal to the byte length of the byte array. | ||||
| * | Better (safer) fix for [0e92c404f1] | jan.nijtmans | 2014-04-30 | 1 | -1/+1 |
| | | |||||
* | | More optimized non-BC [string cat] using Tcl_AppendObjToObj() - thx Donal | ferrieux | 2014-08-05 | 1 | -31/+14 |
| | | |||||
* | | TIP 429 Implementation: [string cat] | ferrieux | 2014-08-02 | 1 | -0/+71 |
| | | |||||
* | | [f0f876c141] Improve consistency in error messages. | dkf | 2014-06-17 | 1 | -7/+9 |
| | | |||||
* | | add compilation of [string is] | dkf | 2014-02-02 | 1 | -1/+1 |
|\ \ | |||||
| * \ | merge main working branch | dkf | 2014-01-02 | 1 | -34/+1 |
| |\ \ | |||||
| * \ \ | merge main working branch | dkf | 2013-12-30 | 1 | -16/+19 |
| |\ \ \ | |||||
| * \ \ \ | merge main dev branch | dkf | 2012-11-05 | 1 | -5/+28 |
| |\ \ \ \ | |||||
| * | | | | | Work on compilation of [string is]. | dkf | 2012-11-02 | 1 | -1/+1 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Hit some problem edge cases with differences in strictness of edge cases that will force a rethink ([string is boolean] is significantly more strict than Tcl_GetBooleanFromObj). | ||||
* | | | | | | Make DEFAULT_TRIM_SET a MODULE_SCOPE string constant, so its value can be ↵ | jan.nijtmans | 2014-01-08 | 1 | -6/+39 |
| |_|_|/ / |/| | | | | | | | | | | | | | | | | | | | shared in tclCmdMZ.o and TclCompCmdsSZ.o and it no longer pollutes the tclStringTrim.h header file. | ||||
* | | | | | Factor out the definition of the default string trim set; define it once only. | dkf | 2013-12-30 | 1 | -34/+1 |
| |_|/ / |/| | | | |||||
* | | | | simple compilation of [string replace] | dkf | 2013-12-12 | 1 | -1/+1 |
| | | | | |||||
* | | | | Now do [string toupper], [string tolower] and [string totitle]. Only handles ↵ | dkf | 2013-10-29 | 1 | -3/+3 |
| | | | | | | | | | | | | | | | | the no-indices case; that's the only case anyone actually commonly uses. | ||||
* | | | | First attempt at [string trim] compilation. | dkf | 2013-09-30 | 1 | -3/+5 |
| | | | | |||||
* | | | | [3613609]: Replace strcasecmp() with UTF-8-aware version. | dkf | 2013-05-22 | 1 | -1/+1 |
|\ \ \ \ | | |_|/ | |/| | | |||||
| * | | | Proposed solution for 3613609: lsort -nocase does not sort non-ASCII correctly | jan.nijtmans | 2013-05-21 | 1 | -1/+1 |
| | | | | |||||
* | | | | Implement Tcl_NewBooleanObj, Tcl_DbNewBooleanObj and Tcl_SetBooleanObj as ↵ | jan.nijtmans | 2013-04-23 | 1 | -2/+2 |
|\ \ \ \ | |/ / / | | | | | | | | | | | | | | | | | macros using Tcl_NewIntObj, Tcl_DbNewLongObj and Tcl_SetIntObj. Starting with Tcl 8.5, this is exactly the same, it only eliminates code duplication. Eliminate use of NO_WIDE_TYPE everywhere: It's exactly the same as TCL_WIDE_INT_IS_LONG | ||||
* | | | | Eliminate all Tcl_ConvertToType calls and all direct calls to ↵ | jan.nijtmans | 2013-02-04 | 1 | -1/+2 |
| | | | | | | | | | | | | | | | | typePtr->setFromAnyProc (except the call from inside the Tcl_ConvertToType function) from the Tcl core. | ||||
* | | | | Add super-simple compiler to many ensemble subcommands to allow better code | dkf | 2013-01-04 | 1 | -11/+11 |
| |_|/ |/| | | | | | | | | | | | generation where we can detect that we're not in the WrongNumArgs case. The compiler just checks that the argument count is in the right range and issues a standard dispatch; that's enough to do an efficient job. | ||||
* | | | Added bytecode compilation of many Tcl commands, merged from development branch. | dkf | 2012-11-05 | 1 | -4/+4 |
|\ \ \ |