Skip to content

Instantly share code, notes, and snippets.

@necolas
necolas / _todo.md
Created June 30, 2012 18:07
Grunt tasks to process HTML files and produce a deploy directory of optimized files
  • Avoid reprocessing the same block in different HTML files.
  • Throw warning when processing a different block to an existing destination file. Hashing will avoid collisions, but introduce confusion.
  • Add file versioning for inline media and CSS images.
  • Avoid need for 'usemin' task - get the replacement element pattern from the first/last HTML element in actual block being replaced. Added benefit of preserving other attributes that may exist (e.g. title, media).

Acknowledgements: This is an adaption of some of Mickael Daniel's work on h5bp/node-build-script

@MattFaus
MattFaus / ContentRevisionsInputReader.py
Last active August 21, 2021 08:11
A custom appengine-mapreduce input_reader and output_writer. The ContentRevisionsInputReader iterates over a complex data structure we use to keep track of modifications to each of the content nodes within the Khan Academy content management system. This is a utility class that is used to intelligently rollout schema changes to these nodes. The …
class ContentRevisionsInputReader(
third_party.mapreduce.input_readers.InputReader):
"""Mapper that loads the latest commit's revisions from the datastore."""
def __init__(self, kinds, start, step, current=0):
# List of content kinds to filter on
self._kinds = kinds
# Start index in snapshot array
self._start = start
# Stride through the snapshot array
self._step = step
@ZipCPU
ZipCPU / axi_regs_top.sby
Created October 16, 2020 22:37
Verified AXI-lite slave design
## Here's the SymbiYosys script we used for this task
##
[tasks]
prf
cvr
[options]
prf: mode prove
prf: depth 4
cvr: mode cover