| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| |
| | |
[d5dec2b0514639f44929cbbba9c4130798a8a54a|d5dec2b051])
|
| | |
|
| |
| |
| |
| |
| | |
a package was defined in was, or does that happen at too high a level?"
Answer: Even though the name of the pkgIndex file is available earlier, it is very well possible to remember it and store it with the other files. This commit does exactly that.
|
| | |
|
| |
| |
| |
| |
| | |
TCL_ERROR is returned: The caller doesn't do anything with this.
Don't access (non-const) variable tclEmptyStringRep any more, use its value (&tclEmptyString) directly. Only keep it in tclPkg.c, for error checking.
|
| |
| |
| |
| | |
case that a filename contains spaces.
|
| | |
|
| | |
|
|\ \
| | |
| | | |
Implementation looks complete. Still missing: test-cases and documentation.
|
|/ / |
|
| |
| |
| |
| | |
in many places where possible.
|
|/
|
|
| |
Tcl_InitStubs() arguments. Useful for debugging and detection of stub incompatibilities (e.g. for Tcl9)
|
|\ |
|
|\ \
| |/
| |
| | |
space, is (marginally) faster, while still being fully up/down compatible. It makes pkgb.so loadable in "novem" without the need to change the Tcl_PkgProvide() call to Tcl_PkgProvideEx().
|
| |\
| | |
| | |
| | | |
space, is (marginally) faster, while still being fully up/down compatible. It makes pkgb.so loadable in "novem" without the need to change the Tcl_PkgProvide() call to Tcl_PkgProvideEx().
|
| | |
| | |
| | |
| | | |
space, is (marginally) faster, while still being fully up/down compatible. It makes pkgb.so loadable in "novem" without the need to change the Tcl_PkgProvide() call to Tcl_PkgProvideEx().
|
| | |
| | |
| | | |
fix compilation with -DNO_CONST
|
| | | |
|
|\ \ \
| |/ / |
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | |
| | |
| | |
| | | |
API.
|
| | |
| | |
| | | |
pipeline creation, package handling, procedures, [scan] formats)
|
| | |
| | |
| | | |
rest of Tcl source code. No ABI change. API change *should* be harmless.
|
|\ \ \
| |/ /
| | | |
cause more harm than good. Purged them (except in zlib files).
|
| |\ \
| | |/
| | | |
more harm than good. Purged them.
|
| | |
| | |
| | | |
more harm than good. Purged them.
|
| | |
| | |
| | |
| | | |
* tests/pkg.test: TCL_TIP268 sections.
|
| | |
| | |
| | |
| | |
| | |
| | | |
* generic/tclPkg.c (CheckVersion): Tcl_Alloc() with calls to
* unix/tclUnixTime.c (SetTZIfNecessary): ckalloc(), which better
* win/tclAppInit.c (setargv): supports memory debugging.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When no requirements are supplied to a [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.
|
| | |
| | |
| | |
| | |
| | | |
argument version==NULL passed in. Backport of the fix for the
same problem in 8.5.
|
| | |
| | |
| | |
| | |
| | |
| | | |
return information from 'Tcl_PkgRequireProc'. Keep the
interpreter result empty. Backport of fix for problem found
while testing #268 under 8.5. More details in the comments.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* generic/tclBasic.c: Define TCL_TIP268 to activate the new
* generic/tclDecls.h: features.
* generic/tclInt.h:
* generic/tclPkg.c:
* generic/tclStubInit.c:
* generic/tclTest.c:
* library/init.tcl
* library/package.tcl:
* tests/pkg.test:
* tests/platform.test:
* tests/safe.test:
* doc/PkgRequire.3:
|
| | |
| | |
| | |
| | |
| | | |
* generic/tclPkg.c (Tcl_PkgRequireEx): replace Tcl_GlobalEval(Obj)
with more efficient Tcl_Eval(Obj)Ex
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
to be even more forgiving of package version mismatch errors in
[package ifneeded] commands, not even logging any warning messages.
This further reduces the ***POTENTIAL INCOMPATIBILITY*** noted for
that change.
|
| | |
| | |
| | |
| | |
| | |
| | | |
* tests/pkg.test: to be more forgiving of package version
mismatch errors in [package ifneeded] commands. This reduces the
***POTENTIAL INCOMPATIBILITY*** noted for that change.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* tests/pkg.test: by Tcl_PkgRequire(Ex) so that the returned
values will always agree with what is stored in the package database.
This way repeated calls to Tcl_PkgRequire(Ex) have the same results.
Thanks to Hemang Lavana. [Bug 1162286].
* tests/namespace.test (25.7,8): Backport test of knownBug.
|
| | |
| | |
| | |
| | | |
backported from trunk.
|
| | |
| | |
| | |
| | | |
* generic/tclPkg.c (Tcl_PkgInitStubsCheck):
|
| | |
| | |
| | |
| | | |
Typo in rules.vc
|
| | | |
|
| | | |
|
| | |
| | |
| | |
| | |
| | | |
warning: array subscript has type 'char'
win/makefile.vc Revert to version 1.203 [Bug #2912773]
|
| | |
| | |
| | |
| | |
| | | |
warning: array subscript has type 'char'
win/makefile.vc Revert to version 1.203 [Bug #2912773]
|
| | |
| | |
| | |
| | |
| | | |
- some internal const decorations
- spacing
|
| | |
| | |
| | |
| | |
| | | |
const tables. No functional
or API change.
|
| | | |
|
| | |
| | |
| | |
| | | |
?options? to the form ?-option value ...?
|