Lymask

Bug fixes and tutorial

Local dataprep and DRC imports
Write your dbStep functions in a file in your dataprep directory, then import them with the YML.

python
dataprep/mySteps.py
from lymask.dataprep_steps import dpStep

dpStep
def do_something(cell):
pass


yaml
dataprep/default.yml
---
- add_library: {filename: mySteps.py}
- do_something
...


Layer refresh hotkeyed
Ctrl+L

No more SOEN
Everything is technology generic; however, the dpSteps currently there were developed for this platform. But now you can write your own (see above).

Acceleration
Sizing and spacing are faster. They can also be more generic using the new `turbo` function. See lymask/library.py. Parallel tiles and cores can now be specified in yaml.

Including the macros
They did not show up using the `package_data` technique, so not it uses the `MANIFEST.in` technique.

DRC
DRC engine based on klayout Regions. This is an alternative to lydrc. Advantages are that it is specified by YAML files, so it can be modified without reloading the klayout application. Also, it can be run from command line or API. Also, it is declarative, which means it can be read by anything else that can read YAML – useful if you want to get DRC parameters into your layout script.

Also new ways to specify yml
Dictionary based because the list based one was really syntactically weird

- nanowire_sleeve: {Delta: 1.5, delta: 0.2, do_photo: true}

instead of

- - nanowire_sleeve:
- Delta: 1.5
delta: 0.2
do_photo: true

- new protocol for specifying dataprep layers with tabs
- cleaner tab loading
- more robust file handling
- more tests

Currently, this package only works with the NIST SOEN-PDK project.

Linkage dependencies specified

lygadgets_link lymask

User doesn't have to first link lygadgets. User doesn't even have to know what lygadgets is

GUI functionality
Call the scripts from the GUI to have them work with the active layout and the active technology. Find the menu under "Tools>Mask Dataprep".

Reloading layer properties is sort of useful. It will be better when the dataprep layers can be deleted from the main tab, and likewise with mask layers