| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
to avoid keeping numbers in the makefile.
|
|
|
|
| |
C-stack-hungry than before, and faster than just using heap allocation.
|
| |
|
| |
|
| |
|
|
|
|
|
| |
Avoid checking for for the LF in a possible CRLF
sequence when EOF has already been found.
|
|
|
|
|
|
|
|
|
|
|
| |
Clear the TCL_ENCODING_END flag when end
bytes are written. This fix keep this method
from writing escape bytes for an encoding
like iso2022-jp multiple times when the
escape byte overlap with the end of the
IO buffer.
* tests/io.test: Add test for escape
byte overlap issue.
|
| |
|
| |
|
|
|
|
|
| |
flag succeeds before enabling 64bit build.
* unix/configure: autoconf-2.59
|
|
|
|
|
|
|
|
|
|
|
| |
* tests/load.test: unloading on Darwin (in addition to
* tests/unload.test: existing tests of .dylib loading).
* macosx/Tcl.xcodeproj/project.pbxproj: add building of dltest binaries
so that testsuite run from Xcode can use them; fix testsuite run script.
* unix/configure.in: add support for building dltest binaries as
* unix/dltest/Makefile.in: .bundle (in addition to .dylib) on Darwin.
* unix/Makefile.in: add stub lib dependency to dltest target.
* unix/configure: autoconf-2.59
|
| |
|
|
|
|
|
|
|
|
|
| |
are in use to support operator commands, might as well make them
the default for [expr] as well and avoid passing every parsed
expression through the inefficient Tcl_Token array format. This
addresses most issues in [RFE 1517602] Assuming no performance
disasters result from this, much dead code supporting the other
implementation might now be removed.
|
|
|
|
|
|
| |
* generic/tclCompExpr.c: of the operator commands through TEBC,
* generic/tclCompile.h: dropping all the routines in tclMathOp.c.
* generic/tclMathOp.c: Still needs Engineering Manual attention.
|
|
|
|
|
| |
* generic/tclCompExpr.c: now routing through TEBC via
* generic/tclCompile.h: TclSortingOpCmd().
|
|
|
|
|
|
|
|
| |
* generic/tclCompExpr.c: TclExecuteByteCode to implement the TIP 174
* generic/tclCompile.h: commands instead of using a mass of code
* generic/tclMathOp.c: duplication. Now all operator commands that
* tests/mathop.test: demand exactly one operation are implemented
via TclSingleOpCmd and a call to TEBC.
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclMathOp.c: to perform a bytecode compile / execute
sequence. This demonstrates a path toward avoiding mountains of
code duplication in tclMathOp.c and tclExecute.c.
* generic/tclCompile.h: Change TclExecuteByteCode() from static to
* generic/tclExecute.c: MODULE_SCOPE so all files including
tclCompile.h may call it.
|
|
|
|
| |
* tests/mathop.test:
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
* generic/tclExecute.c: failing tests illustrating bugs uncovered
* generic/tclMathOp.c: in [Path 1578137].
|
|
|
|
|
| |
the commands were not [namespace export]ed from the ::tcl::mathop
namespace. More bits from [Patch 1578137] correct that.
|
|
|
|
|
| |
found in [Patch 157837]. Many failures now demonstrate issues to
fix in the TIP 174 implementation.
|
| |
|
| |
|
|
|
|
| |
on windows.
|
|
|
|
|
| |
* tests/mathop.c: routines some of routines that compile
the new TIP 174 commands. This corrects some known bugs. More to come.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* doc/tclvars.n: pointerSize.
* win/Makefile.in: Added installation instructions for the platform
* win/makefile.vc: package. Added the platform package.
* win/makefile.bc:
* unix/Makefile.in:
* tests/platform.test:
* tests/safe.test:
* library/platform/platform.tcl:
* library/platform/shell.tcl:
* library/platform/pkgIndex.tcl:
|
|
|
|
|
|
|
|
|
|
|
|
| |
* tests/pkg.test: [package require $pkg] and [package unknown]
is invoked to find a satisfying package, pass the requirement argument
"0-" (which means all versions are acceptable). This permits a
registered [package unknown] command to call
[package vsatisfies $testVersion {*}$args] without any special
handling of the empty $args case. This fixes/avoids a bug in
[::tcl::tm::UnknownHandler] that was causing old TM versions to
be provided in preference to newer TM versions. Thanks to Julian
Noble for discovering the issue.
|
| |
|
|
|
|
|
|
|
| |
CompileExprTree() routine that can produce expression bytecode
directly from internal structures with no need to pass through
the Tcl_Token array representation. Still disabled by default.
#undef USE_EXPR_TOKENS to try it out.
|
|
|
|
|
|
|
|
|
|
| |
produce a different set of internal structures representing the parsed
expression, as well as routines that go on to convert those structures
into the traditional Tcl_Token array format. Use of these routines
is currently disabled. #undef PARSE_DIRECT_EXPR_TOKENS to enable
them. These routines will only become really useful when more
routines that compile directly from the new internal structures are
completed.
|
| |
|
| |
|
| |
|
|
|
|
| |
[Bug 1606860]
|
|
|
|
|
|
|
|
|
|
| |
* doc/chan.n: New subcommand [chan pending].
* generic/tclBasic.c: Thanks to Michael Cleverly for proposal
* generic/tclInt.h: and implementation.
* generic/tclIOCmd.c:
* library/init.tcl:
* tests/chan.test:
* tests/ioCmd.test:
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tcl.decls: Tcl_GetBignumAndClearObj -> Tcl_TakeBignumFromObj.
* generic/tclObj.c:
* generic/tclDecls.h: make genstubs
* generic/tclStubInit.c:
* generic/tclExecute.c: Update callers.
* generic/tclMathOp.c:
|
| |
|
| |
|
|
|
|
|
| |
Tcl_DecrRefCount() on the objPtr argument to plug memory leaks.
This makes the routine a consumer, which makes it easiest to use.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclCmdAH.c:
* generic/tclCmdIL.c:
* generic/tclCmdMZ.c:
* generic/tclCompCmds.c:
* generic/tclCompExpr.c:
* generic/tclCompile.c:
* generic/tclCompile.h:
* generic/tclExecute.c:
* generic/tclIOUtil.c:
* generic/tclInt.h:
* generic/tclInterp.c:
* generic/tclNamesp.c:
* generic/tclObj.c:
* generic/tclProc.c:
* tests/compile.test:
* tests/info.test:
* tests/platform.test:
* tests/safe.test:
|
|
|
|
|
|
|
| |
* tests/event.test (event-14.*): Corrected a bug where
TclUnixWaitForFile would present select() with the wrong mask
on an LP64 machine if a fd number exceeds 32. Thanks to
Jean-Luc Fontaine for reporting and diagnosing [Bug 1602208].
|
|
|
|
| |
floating-point increment values. Thanks to William Coleda [Bug 1602991]
|