summaryrefslogtreecommitdiffstats
path: root/generic
Commit message (Collapse)AuthorAgeFilesLines
* Allow [array names -regexp] to use backreferences. This capability was ↵andy2016-11-251-3/+10
| | | | broken by [71270e9141]. See also bug [1366683].
* Remove spurious article in commentsandy2016-11-251-2/+2
|
* Partially backout [ef4da65408] because AppendPrintfToObjVA() (used ↵andy2016-11-241-2/+2
| | | | indirectly by [tcl::unsupported::disassemble]) does not yet support wide and bignum arguments. This fixes a SIGSEGV in compile.test and makes [disassemble] work again.
* Make compileEpoch "unsigned int", and start counting at 1.jan.nijtmans2016-11-243-3/+3
|
* typosjan.nijtmans2016-11-243-3/+3
|
* Corrections to misleading comments.dkf2016-11-221-5/+6
|
* More internal use of size_t in stead of int.jan.nijtmans2016-11-218-23/+19
|
* Minor: fix the indentation of comments.dkf2016-11-181-24/+26
|
* Fix mp_cnt_lsb() signature, so it matches the signature used in Tcljan.nijtmans2016-11-181-2/+1
|\
| * Fix mp_cnt_lsb() signature, so it matches the signature used in Tcl. Add two ↵jan.nijtmans2016-11-181-1/+1
| | | | | | | | missing "static" keywords in tclUnixNotfy.c. Remove some more files in libtommath directory which are not useful for Tcl.
* | Fix [e6f27aa56fa51bfc1752ce768bf0d301c60bfd2c|e6f27aa56f]: Update libtommath ↵jan.nijtmans2016-11-185-157/+142
|\ \ | | | | | | | | | to 1.0
| * | Restore bn_mp_radix_size.c to exact copy of libtommath-1.0 version: Since ↵libtommath_1_0jan.nijtmans2016-11-172-7/+29
| | | | | | | | | | | | the radix_size of "9" should return 2, not 3. Add test-case to prove that.
| * | Add test-case for mp_iseven(). If mp_iseven(0) ever returns 0, we will be ↵jan.nijtmans2016-11-172-3/+24
| | | | | | | | | | | | warned that we are using the wrong mp_iseven() function.
| * | Merge trunk. Re-generate tclTomMath.h. Use faster exponentiation-function ↵jan.nijtmans2016-11-172-149/+120
| |\ \ | | | | | | | | | | | | from libtommath 1.0 (in tclExecute.c).
| * | | (experimental) Upgrade to libtommath 1.0 (actually by merging all changes ↵jan.nijtmans2016-11-163-0/+13
| |/ / | | | | | | | | | between libtommath 0.42.0 and 1.0). Still to be tested thourougly, before doing anything with it.
* | | Fix libtommath's mp_radix_size() function such that it returns 2 for ↵jan.nijtmans2016-11-172-7/+50
|\ \ \ | |/ / |/| / | |/ single-digit numbers. Add testcases for mp_radix_size() and mp_iseven(). Undo useless change in bn_mp_add_d.c (bring back libtommath's version).
| * Fix libtommath's mp_radix_size() function such that it returns 2 for ↵jan.nijtmans2016-11-172-7/+50
| |\ | | | | | | | | | single-digit numbers. Add testcases for mp_radix_size() and mp_iseven(). Undo useless change in bn_mp_add_d.c (bring back libtommath's version).
| | * Fix libtommath's mp_radix_size() function such that it returns 2 for ↵jan.nijtmans2016-11-172-7/+50
| | | | | | | | | | | | single-digit numbers. Add testcases for mp_radix_size() and mp_iseven(). Undo useless change in bn_mp_add_d.c (bring back libtommath's version).
* | | Use more "size_t" in stead of "int" internall. Also eliminate a lot of ↵jan.nijtmans2016-11-1614-94/+93
| | | | | | | | | | | | type-casts which are not necessary any more.
* | | See Itcl Bug [1b28657c0e]. Make methodNameType honor the Tcl_ObjType contract.dkf2016-11-151-0/+1
|\ \ \ | |/ /
| * | See Itcl Bug [1b28657c0e]. Make methodNameType honor the Tcl_ObjType contract.dkf2016-11-151-0/+1
| |\ \
| | * | Slightly more efficient version as this is Hot Path code.bug_itcl_1b2865dkf2016-11-151-1/+1
| | | |
| | * | See Itcl Bug [1b28657c0e]. The methodNameType fails to honor the Tcl_ObjTypedgp2016-11-141-0/+1
| | | | | | | | | | | | contract. Need dkf review on suitability of this fix.
* | | | Merge core-8-6-branchjan.nijtmans2016-11-141-1/+1
|\ \ \ \ | |/ / /
| * | | Better fix for the usability of TCL_NO_DEPRECATED on MacOSXjan.nijtmans2016-11-141-1/+1
| |\ \ \ | | |/ / | |/| / | | |/
| | * Better fix for the usability of TCL_NO_DEPRECATED on MacOSX.jan.nijtmans2016-11-141-0/+2
| | |
* | | Fix harmless gcc warning message: using "unsigned" doesn't serve any ↵jan.nijtmans2016-11-113-10/+10
| | | | | | | | | | | | purpose, actually.
* | | Fix [79614fb8b61983ac8ef30ea8752c310465798fc7|79614fb8b6]: "glob", "encoding ↵jan.nijtmans2016-11-112-3/+4
|\ \ \ | |/ / | | | | | | system" and encoding-free filesystems
| * | Fix [79614fb8b61983ac8ef30ea8752c310465798fc7|79614fb8b6]: "glob", "encoding ↵jan.nijtmans2016-11-112-3/+4
| |\ \ | | |/ | | | | | | system" and encoding-free filesystems
| | * Fix [79614fb8b61983ac8ef30ea8752c310465798fc7|79614fb8b6]: "glob", "encoding ↵jan.nijtmans2016-11-112-3/+4
| | | | | | | | | | | | system" and encoding-free filesystems
* | | On OSX, there is a conflict with the "define panic" and definitions in "mach.h".jan.nijtmans2016-11-101-0/+2
|\ \ \ | |/ /
| * | On OSX, there is a conflict with the "define panic" and definitions in "mach.h".jan.nijtmans2016-11-101-0/+2
| | |
* | | Route all [string last] operations through a common implementation.dgp2016-11-084-72/+120
| | |
* | | Purge disabled code.dgp2016-11-071-17/+0
| | |
* | | Optimize case of all single-byte chars.dgp_string_finddgp2016-11-071-2/+20
| | |
* | | Consolidate the "find empty string" cases.dgp2016-11-071-16/+11
| | |
* | | Implement direct eval [string first] with the refactored engine.dgp2016-11-071-69/+9
| | |
* | | First draft refactoring the [string first] functionality.dgp2016-11-043-0/+90
| | |
* | | [824752f10e] Use LIST_MAX instead of INT_MAX for unknown number of elements ↵dgp2016-11-041-1/+1
|\ \ \ | |/ / | | | | | | to the end of the list.
| * | [8245752f10e] Use LIST_MAX instead of INT_MAX for unknown number of elementsdgp2016-11-041-1/+1
| | | | | | | | | to the end of the list.
| * | [824752f10e] More robust, portable check for integer overflow.dgp2016-11-041-5/+2
| |\ \ | | |/
| | * [824752f10e] More robust, portable check for integer overflow.dgp2016-11-041-5/+2
| | |
| * | merge core-8-5-branch ([824752f10e] Avoid calling Tcl_SetObjResult if interp ↵gahr2016-11-041-2/+5
| |\ \ | | |/ | | | | | | is NULL)
* | | [824752f10e] More robust, portable check for integer overflow.dgp2016-11-041-5/+2
| | |
* | | close forkdgp2016-11-041-2/+5
|\ \ \
| * \ \ merge core-8-5-branch ([824752f10e] Avoid calling Tcl_SetObjResult if interp ↵gahr2016-11-041-2/+5
| |\ \ \ | | | |/ | | |/| | | | | is NULL)
| | * | [824752f10e] Avoid calling Tcl_SetObjResult if interp is NULLgahr2016-11-041-2/+5
| | | |
* | | | Make packages msgcat and tcltest work unmodified in "novem". Make more ↵jan.nijtmans2016-11-041-2/+2
|\ \ \ \ | |/ / / |/| | / | | |/ | |/| test-cases work unmodified in "novem".
* | | Replace indexing with pointer increments.dgp_string_catdgp2016-11-011-13/+18
| | |
* | | Trim away obsolete code.dgp2016-11-011-145/+1
| | |