This protects against rogue template metaprograms that don't terminate.
Change-Id: I9558ceaaed17baddbed84aac67a3c72397b62b64
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>
Value.ERROR as its value
This avoids infinite recursion in metaprograms whose termination depends
on the inputs being known.
Change-Id: Ic6972188f69bf324431ce1331ad079fbf2c9039e
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>
patterns used by some commonly-used libraries
In particular, Boost.MPL supports compile-time computations with type
lists of up to 20 elements, which require a depth of 40 to support
Change-Id: I89b33205837f0acdb84e2d79cef53ec636a6bc8a
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>
When file language mapping is reset back to inherit, empty (without
configuration, language and path attributes) file-mapping element is
created in .cproject file. This leads to project corruption.
Add a check for empty entry value in
LanguageMappingStore#addFileMappings.
475344: Broken .cproject file after file language mapping is reset to
inherit
https://bugs.eclipse.org/bugs/show_bug.cgi?id=475344
Change-Id: I6e7d0b8d1199501e087ce42b75f0d8e83fca77cd
Signed-off-by: Ivan Furnadjiev <ivan@eclipsesource.com>
(cherry picked from commit a48852f21a)
There is an assumption in
FileLanguageMappingPropertyPage#computeInheritedMapping that there is
only one overridden file mapping. If both project and workspace file
mappings have been overridden, a CoreException is thrown when they both
are reset to inherit together.
Change-Id: I5c845831ac446eaf65c782b4428b3b72a274fb23
Signed-off-by: Ivan Furnadjiev <ivan@eclipsesource.com>
(cherry picked from commit 5cc4336072)
Added further check to isStatementWithMixedLocation return false if all
macro expansions are further enclosed in child statements.
Change-Id: Ida6994adbbf76674c37a4c1c214aeea79b4963f2
Signed-off-by: Thomas Corbat <tcorbat@hsr.ch>
member, include the class name in the search view
Change-Id: Idbf86a22c9000328aecbab50f3f562f721bb64f6
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>
from top to bottom, a chance to handle a click
Change-Id: I294b5511e65dfe593dfbb58beda6888da1870779
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>
SemanticQueries.getPureVirtualMethods() so it can be reused
Change-Id: I5a48bdc3111ac75f2c1ca4f5595fe86defd686b8
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>
expressions in some of the parser's heuristic checks
Change-Id: Ia91a00b91e0050838e990d1daad81b524816f8d0
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>
It does not make sense for findTypeRelevantDeclarator to check whether
the innermost declarator (result) is of type IASTArrayModifier:
- There is no common subtype of IASTDeclarator and IASTArrayModifier.
- IASTArrayModifier are attached to IASTArrayDeclarators but they are
not resolved as nested declarators.
- Similar to the existence of pointer operators, the resolution of the
innermost type relevant declarator should stop at array declarators.
Change-Id: I9c3673f60a48245eb51ccc8cb94757b87abf0265
Signed-off-by: Thomas Corbat <tcorbat@hsr.ch>
prefer files closer to the origin file in the directory structure
Change-Id: I82a3c1dc3f09cecb69e07511dd5b8bed62676b6a
Signed-off-by: Nathan Ridge <zeratul976@hotmail.com>