Goal is to support "*!" which is a variant of multiplication
without the commutative nature.
The refactoring yields a more consistent handling of
Python specializations. Now this happens when the methods
are collected, rather than later. This way, specific behavior
for identically named synonyms can be implemented for example.
This is the case for the "*" operator in Trans which is partially
commutative and partially not.
* Fixed issue-1164 (internal error on delete cells on archlinux)
* Found another location where take_cell would spoil cell_name
Co-authored-by: matthias <matthias@klayout.de>
* [CONSIDER MERGING] Clarification and typo fixed in DRC doc.
* Enhanced Shapes::is_valid so it's more reliable in detecting invalid shape objects.
* ObjectInstPath now has a validity attribute
* First implementation of solution for edtService (partial service to follow)
* Validating selection also for Partial Edit mode
* [CONSIDER MERGING] Clarification and typo fixed in DRC doc.
* Avoids a segfault during drawing: make sure the cell bbox update is not triggered in the drawing thread before it is stopped
* Fixed#1144 (wrong layer taken on copy layer)
* Added python stubs with type hinting and documentation.
The documentation was extracted by inspecting the docstrings within each class and methods.
This should enable type hinting and checking by IDEs like VSCode.
The stubs were automatically generated, and have not been manually curated. (TODO)
* created tl ArgType to python type translator
* Fixing a pyaModule.cc bug
* almost finished stubgen
* fixing little bug in classes defined within classes
* Release version of the stubgen script. Good enough for release.
* Adding notes
* Including stubs when installing from source
* typechecking bugfix: missed checking is_iter()
* Cleaning up unused code for merge review
* Including stubs when installing from source (part deux)
* Putting the GenericDeviceExtractor name setting into the right place.
* Updating python stubs
Co-authored-by: Matthias Koefferlein <matthias@koefferlein.de>
* Fixed issue-1135 (LVS mismatch on parallel devices)
The fix consists of a more elaborate device identity analysis
following the topological matching. In this step, the devices
are identified according to their connections and parameters.
It is important to properly identify devices taking their
parameters into account as well as their connections.
* Second part of issue fixed (inverter chain ambiguity)
* Added test
* Updated tests
* Updated golden test results
* Updated golden test data for Windows
Co-authored-by: klayoutmatthias <matthias@klayout.org>
* First attempt to fix. Rather experimental.
* Debugging and bug fixing
The basic issue was a missing break
However, correct computation of cache results for instance-to-instance cluster
interaction is implemented
Plus: identical and overlapping instances are no longer ignored except in the
case of exact duplicates. Otherwise these instance generate dead nets which
are not connected elsewhere.
* Added tests, fixed duplicate cells test, added missing files.
* Code simplification (removed invariant from transformation in cluster-to-cluster interaction cache)
* Skipping cell instance duplicates as some real-world testcases mandate so
* Updated test data
* Two patches to enable building on MSYS with gcc 12.1
1. Complete explicit instantiations in dbHierProcessor
2. Strict weak ordering is not present for std::type_info
(using "before" on gcc 12.1). We need to work with plain
pointers and use a translation table to associate pointers
with classes.
* Fixed another place where std::type_info got important
Co-authored-by: Matthias Koefferlein <matthias@klayout.de>
This was a problem with universal DRC which does not specifically
set the merged status. Furthermore, the default for DeepRegion and
DeepEdges was "false" already and now is aligned.
* WIP: some convenience (self.param=() for Ruby PCell impl), a bug fix (crash on call of pcell_parameter on wrong cell), update of tests
* Implemented solution for #1079 (refresh calls coerce_parameters)
* Basic.CIRCLE modified such that coerce_parameters_impl also catches the case of a programmatic change of actual_radius and this function will update radius and the handle accordingly
* Modified Basic PCells such that they work with coerce_parameters also if generated programmatically
* Added tests
* Fixed first issue (UI problem with GDS2Text options)
* Fixed problem with 'save' when the file extension does not indicate one of the known formats
In this case, and when plain 'save' is used, the original format is delivered.
The session files also store the original format now.
The statistics page will now indicate the format of the file that was loaded.
* Implementated patch as suggested by issue-1047
* Implementing blend-mode for LEF/DEF for a first solution for issue-1048
* Added testcase, fixed other tests
* Issue-1050 fixed - now REGION, BLOCKAGE and DIEAREA are relevant from mapping file, nothing else
* Bug fixes
* Experimental: DEF multipart paths
* Experimental: ignore LEF wire extension, ignore PATTERNNAME
* Implemented PATTERNNAME in DEF instead of LEF ..
* Implemented nondefaultrule vias
* LEF/DEF: non-defaultrule vias need a name suffix.
* Refined solution for NDR via names, updated test data
* Updated test data
* Multiple map files and 'REGIONS FENCE' + 'REGIONS GUIDE' layer mapping entries.
* Added test data, updated doc.
* Some enhancements and feedback
REGIONS in map files -> REGION
REGION NONE for map files (maps REGIONs which are neither GUIDE or FENCE)
Bugfix: VIA with ORIGIN
No via cell name prefix by default for buddy tools
* Simplified buddy tool options for LEF/DEF property names
Dropped --lefdef-dont-produce-instance-names,
--lefdef-dont-produce-pin-names and --lefdef-dont-produce-net-names
options. Instead, these options have to be enabled by giving
the respective property names.
Writing the property names takes some space and these
options are simply confusing and redundant.
* Issue a warning about missing cut geometry for a DEF via with PATTERNNAME (which is ignored)
* Fixed issue-1065
* Opportunity fix: library selection box listed libraries with technology even though no technology is selected.
* Opportunity: show PCell errors on new error layer. Previously, errors were only shown when the PCell had at least a single layer parameter
The error text is drawn together with guiding shapes. Disabling guiding shapes will
turn off error display.
* Opportunity: fixed an issue when reloading a library immediately after loading a layout: in this case, the library connection got lost and cells were substituted by static ones
* Opportunity: added a layout cleanup for the following case: a layout is loaded without the library present - giving defunct cells. Then the library is registered and defunct cells are reconnected.
* Some doc fixes as preparation for PDF generation
* Some doc bug fixes and enhancements for PDF manual generation
* Fixed a number of invalid references found during PDF generation
* Enable PCellDeclarationHelper doc for help system
* Fixed doc text
* Fixed some doc issues with invalid links etc.
* First step for solution:
Problem was: the ambiguity resolver was making decisions which later resulted in
a name conflict. Later on, another branch of the backtracking algorithm came
across the same situation but decided based on names, creating an conflict.
First part of the solution is to establish the backtracking information
during ambiguity resolution and using that rather than an alternative branch
later on. This avoids this conflict, but does not favor names as mandated
by the "use_names" flag. This will be the second step of the solution.
* Bugfixed solution (partially)
* Introducing third pass in netlist compare
The second pass is "ambiguity resolution by name" and
is skipped if "consider_net_names" is false. The third
pass then is ignoring the names.
* Bugfix
* Comment fixed, test updates
* Added tests
* Added test data variant for CentOS 8
* Fixed issue-1011 by using single-point polygons for texts in deep mode. Tests need fixing.
* Updated tests (text become a single point)
* Added test for issue-1011
* PyPI deployment: corrected branch filter
* skip redundant points in partial edit, but not colinear ones
* Polygon transformation now preserves raw points
Matthias Koefferlein
Thomas Ferreira de Lima
klayoutmatthias