profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/jhance/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.

HOLOGRAPHICpizza/GridToGo 5

An easy to use GUI for creating and managing ad-hoc OpenSim grids.

jhance/cereal-derive 3

Derive instances of Serialize

jhance/jmdict-parser 2

Parser for the JMdict XML File.

jhance/llcc 2

for-learning-how-to-use-llvm llvm-based not-quite-c compiler

jhance/aeson 1

A fast Haskell JSON library

jhance/binary-derive 1

derive instances of binary

jhance/conduit 1

A pull-based approach to streaming data.

jhance/data-object-yaml 1

Serialize data to and from Yaml files

issue commentdropbox/dbx_build_tools

Support test_suite for dbx_services_test

Mis-understanding of how test_suite works, so closing.

olschofkah

comment created time in 17 hours

issue closeddropbox/dbx_build_tools

Support test_suite for dbx_services_test

Would it be possible to add support for test suites for the 'test' invoked on a dbx_services_test? The use case is that a collection of tests that can run in parallel may share a set of defined daemon or task services that need to be performed. Right now the following error occurs:

Traceback (most recent call last): File "/root/.cache/bazel/_bazel_root/7ebbdddc3c30692c56fb633c56be95c3/external/dbx_build_tools/build_tools/services/svc.bzl", line 481, column 75, in services_bin_impl launcher_args.append("--svc.test-bin={}".format(ctx.executable.bin.short_path)) Error: 'NoneType' value has no field or method 'short_path'

closed time in 17 hours

olschofkah

issue closeddropbox/dbx_build_tools

How do I build a distribution archive (archive/master.tar.gz)?

I'd like to add support for OSX x86-64.

To start, I want to introduce a change and then package that change into the file archive/master.tar.gz. I would assume there is a target, but I haven't been able to find it so far.

How do I build the distribution archive?

closed time in 18 hours

jmyounker

issue commentdropbox/dbx_build_tools

How do I build a distribution archive (archive/master.tar.gz)?

That isn't built with bazel, its built with the command git archive presumably.

jmyounker

comment created time in 18 hours

push eventdropbox/dbx_build_tools

Jared Hance

commit sha 2de77ff90fc63ca8882d4a11fc0ac1b5efd5dc59

Remove internal-only code GitOrigin-RevId: 755ff99fe76c16ec8f80201458baf31834e825b1

view details

push time in 18 hours

fork jhance/git-archive-all.sh

A bash shell script wrapper for git-archive that archives a git superproject and its submodules, if it has any.

fork in 19 hours

push eventdropbox/dbx_build_tools

Jared Hance

commit sha 173ae8daa21f21759f0c5df6e946fc5afdaa2451

OpenSSL 1.1.1l GitOrigin-RevId: f5c3151054208f7c03d7f2f7fd4ec51e3ec447ef

view details

push time in 20 hours

push eventdropbox/dbx_build_tools

Ian Baker

commit sha f4d66013addfac9c63dea8a6ed733479767f7d94

Add dynamic library support to the go rules GitOrigin-RevId: fe94314b8c145bc6fea9c7fc9eb82517afe22821

view details

Athena Automated Reverts

commit sha 76dbf46447398221f2b2072547bd39bc18d2b2ab

Revert "Add support for the go rules" (Due to internal breakage) GitOrigin-RevId: 4bdc420f0a5241ba36c85ae111199cd42bcb3dbc

view details

michael schuldt

commit sha ea81dbb216b4decf469eb8f67d51f0181f33ac42

internal change GitOrigin-RevId: da1f9d2a77e03abc6a00cdb42143e64fafb8506e

view details

Ian Baker

commit sha d98a365a3a53266ea697ff4b25ebf0c268bc9cf5

Add support for dynamic linking to the go rules GitOrigin-RevId: 87377dc9028b38cb04481feadb87ec6f4079b9d3

view details

Ian Baker

commit sha ba3d6dfb790c847db52c4db22fbaf1f20b9d4e09

internal change GitOrigin-RevId: b4bf1f356afa251e2c62dc3fc4578bdf2ce183b1

view details

Ruslan Nigmatullin

commit sha 6d8f8a08ff47fba27a8b0df0f818da4503f010a7

[starlark] Update go.starlark.net to the latest master Summary: It adds proto and math modules, useful for configuration GitOrigin-RevId: 70a6eae5c5a7ef7400cc2dba9878a3812176393e

view details

Yijia Li

commit sha ebb8de8a9f549d5831cb221a6121214d6eeaa6a5

Regenerate various build files GitOrigin-RevId: 6a6d21ff7a48cb77a36aa77a0e6578a3b6358204

view details

Zhen Wang

commit sha 1b99908bbba94916a87de6c125dd7c78289acca0

internal change GitOrigin-RevId: 58b745a3b6bbe6280339a9f23d412a1982b2fb56

view details

Jared Hance

commit sha 3513363a9d42cc91bc4fc5d8e8a6fab100d7bd59

Support extra healthchecks for service tasks GitOrigin-RevId: 56018be9a893aa2667a1b9cd70328e2e5cde1720

view details

Anup Chenthamarakshan

commit sha b5a3b5d5cdd4d1f311b4d820e57787f329b13f4f

Upgrade to Go 1.16.7 Summary: List of commits: https://github.com/golang/go/compare/go1.16.6...go1.16.7 GitOrigin-RevId: bba132848dc8e9d630b1043c194ad484421432c4

view details

Sai Teja Duthuluri

commit sha ebfcc94fd6092eded21e0d97734db0f285d7ad12

[go-v2-codegen] Generate code for hulk proto files, and fix issues in gen_build_proto Summary: In `gen_build_proto.py`, ignore `go_package` requirements for proto files in `configs/proto/google/protobuf` Use the script from (1) to add `go_package` option to hulk proto files, and all its dependencies. Modify `rpc_stubs.go` to change the package name for generated go code. Earlier, a proto package like `hulk.driver` would generate a go package with name `hulk_driver`. Instead, now we simply have the go package name `driver`. This was done to match `protoc-gen-godbx` behaviour with that of `protoc-gen-go` Mark `github.com/planetscale/vtprotobuf` as dirty in `.dbxvendor.json`, while we get a fix for https://github.com/planetscale/vtprotobuf/issues/18, and add a hack to the repo. Substitute using `github.com/gogo/protobuf/proto` with `github.com/golang/protobuf/proto` in `/go/src/dropbox/hulk/*` to remove usage of gogoproto GitOrigin-RevId: dddc8e25678549687004135369733d7ee4c75490

view details

Sai Teja Duthuluri

commit sha 88ad3872532ce68773c3c99c476cd5c1908cd729

[gogoproto] Limit visibility of gogoproto package Summary: This will ensure that newer usage of gogoproto is avoided since we are removing support for gogoproto. GitOrigin-RevId: db1c6139be87612376e0c81301d246e90e6dacd8

view details

michael schuldt

commit sha 2fef4ccac901831dda74d446461358a1feb674db

internal change GitOrigin-RevId: 9cd9e75b3511c04cfda0c33301ce3949706131a6

view details

push time in 20 hours

push eventjhance/mypy

Jared Hance

commit sha 6759d9eaf927acd9f07573aa227d18f2ae448cd4

flake8

view details

push time in 6 days

PR opened python/mypy

[mypyc] Coerce Register assignments

Fixes mypyc/mypyc#759 and probably some other issues. Including only the change to coerce the Register assignments fixes the missing type coercion for the specialized next-generator case, but causes some unusual box-then-unbox from short_ints.

These odd coercions are caused by the dict keys in dict literals being considered short_int, but dict_next_rtuple_single/pair return types bein annotated as int_primitive. Changing this to short_int_rprimitive removes the odd coercisions. The remainder of the diff is fixing the irbuild testcase output to have short_int.

+67 -17

0 comment

8 changed files

pr created time in 6 days

create barnchjhance/mypy

branch : next-coerce

created branch time in 6 days

PR opened python/mypy

[mypyc] Remangle names produced by allow-redefinition

Description

Fixes mypyc/mypyc#702 by replacing "'" with "redef", although this assumes nothing else will ever use demangling for now.

Test Plan

Only locals need any fix; a run testcase was added to verify that it works already for globals, and since the fix for locals is on the ir level only an ir testcase is added.

+47 -1

0 comment

3 changed files

pr created time in 6 days

create barnchjhance/mypy

branch : mangling

created branch time in 6 days

issue commentmypyc/mypyc

GCC on Linux fails to compile tuple assignment with somewhat complex type annotation with "array subscript 1 is above array bounds"

This seems like a GCC version added new checks that are somehow getting tripped up.

ichard26

comment created time in 6 days

PR opened python/mypy

[mypyc] Remove decorator_helper_name

Description

Per mypyc/mypyc#718, renaming the original function name in the module mapping causes the name of decorated functions to have be something like mypyc*decorator_helper. When checking if just using the original name as the module name (and then later the function will be overriden by the decorated value) no tests fail. And there are indeed tests that exhibit rather complex decorator behavior so it seems there may not actually be anything wrong with using the original name.

(Explain how this PR changes mypy.)

Test Plan

Added a runtime test that exhibits better name behavior, existing tests to prevent regression.

+24 -23

0 comment

5 changed files

pr created time in a month

create barnchjhance/mypy

branch : decorator-names

created branch time in a month

push eventjhance/mypy

Michael J. Sullivan

commit sha 869f7b447d850f8b9607ea93e6f891954a781237

Add an xfailed test for issue #9852 (#10932) As part of doing this I upgraded the pep561 tests to support incremental tests, and removed at least a little of the duplication between different incremental test suites.

view details

Ethan Smith

commit sha e73432162b86f901c54e87d62858b0bdde062d24

Try removing skipped test on macOS (#10934) This test was original skipped because nesting a virtualenv inside of a venv on macOS was a problem, but now that we use a newer virtualenv, hopefully that issue has gone away.

view details

jhance

commit sha 97a1b3f81f90ebc2739773cd949b92d891f47c4d

Add primitive for bytes join() method (#10929)

view details

97littleleaf11

commit sha baea65a338e67a09a23158b4067b75e86de009a7

[mypyc] Faster len(bytes) (#10936) Similar to list and tuple, we can directly get length of bytes from PyVarObject->ob_size. Implements part of mypyc/mypyc#880.

view details

97littleleaf11

commit sha 68a67aedb072dc811424636140a87aa76f1b59bb

[mypyc] Add tests for chr(), ord(), encode() and decode() (#10914) Add several run tests: * chr() * ord() * decode() * encode()

view details

Michael J. Sullivan

commit sha daed963d1f849fa88875d2d6ea2775a7a9b16c2f

Pin tomli to 1.1.0 (#10938) While we figure out what the plan is in coordination with black and pip (see #10893), just pin to the previous version of tomli that isn't causing our CI to fail.

view details

97littleleaf11

commit sha e7161ac06d6b05f6b59669d2d0a6b9406bb112bc

[mypyc] Add bytes concat op (#10926)

view details

97littleleaf11

commit sha 58c0a051b72731f093385cc7001b1f5c2abd2c18

[mypyc] Introduce FormatOp and add a tokenizer for .format() call (#10935) This PR adds a tokenizer that convert a str.format() format string into literals and specifiers. By doing so, the code structure of `translate_str_format` is clearer. This PR also introduces `FormatOp`. Compare to `ConversionSpecifier`, `FormatOp` has fewer attributes and indicates compile time optimizations. For example, to mark a conversion from any object to string, `ConversionSpecifier` may have several representations, like '%s', '{}' or '{:{}}'. However, there would only exist one corresponding `FormatOp`. Currently `FormatOp` is just an Enum for convenience. We might add several attributes later and upgrade it to a class if we need to support more conversions. To help for the future optimization, these parts of code are extracted into new functions: * `generate_format_ops` that shrink `ConversionSpecifier` into `FormatOp` * `convert_expr` that can help convert the expressions into desired results.

view details

Michael J. Sullivan

commit sha c90026bd867635a4efe70cdacf0caaf00a986914

Fix caching behavior of PEP561-installed namespace packages (#10937) Since PEP561-installed namespace packages only show up in FindModuleCache as a side-effect, the submodules need to be listed first in the dependencies for things to work. This was handled already in the Import case but not the ImportFrom case. Also fix the cache handling of namespace packages so it doesn't always get reported stale. Fixes #9852.

view details

Tim Gallant

commit sha ed2b4c76cb8662942096de1cfbf08aa422a3b6f2

[dataclasses plugin] Support kw_only=True (#10867) Fixes #10865

view details

Michael J. Sullivan

commit sha 3c935bdd1332672f5daeae7f3f9a858a453333d4

Switch to using Id functions everywhere in lib-rt and dropping format calls (#10941)

view details

97littleleaf11

commit sha 6e48f7e18b39e5fdab594e111da133965081c728

[mypyc] Speed up string formatting using FormatOp (#10942) This PR speeds up FStrings using FormatOp and convert_expr.

view details

pranavrajpal

commit sha 0bde4b9e82e337f1b63d1e7dcc6922ed648e86f7

[mypyc] Use MRO of argument for dispatch in singledispatch functions (#10930) This changes the implementation of singledispatch to look up the correct registered implementation to use at runtime using the MRO of the passed argument, instead of using a chain of isinstance checks for each registered implementation, for the reasons mentioned in mypyc/mypyc#802 (comment). This doesn't support dynamically registering implementations yet, so 2 of the existing singledispatch tests that rely on that are marked as xfails for now.

view details

pranavrajpal

commit sha 0dcd2c636e28d27473c084e5148e7dfb86886831

[mypyc] Add support for dynamically registering singledispatch functions (#10968) Instead of generating a regular native function for singledispatch functions, generate a callable class. That gives us a place to put the registry dict (instead of storing the registry dict in a static variable) and also will allow us to support dynamically registering functions by adding a register method to that callable class.

view details

97littleleaf11

commit sha ba7b50c07886f744268d664adb5a6f026ce7be37

[mypyc] Add bytes index primitive (#10950) Implements part of mypyc/mypyc#880.

view details

97littleleaf11

commit sha 5adb0a05c157d9f5afbdb048a6d06a3234586ab6

[mypyc] Add primitive for bytes.decode (#10951) Implements part of mypyc/mypyc#880.

view details

97littleleaf11

commit sha 3970cc3baf98dd456927accb550659beb14ea0bd

Add primitives for str.encode() (#10974)

view details

97littleleaf11

commit sha 644496828841035d097d7afd80b96081e0ea5c7f

[mypyc] Add primitive for bytes slicing (#10966) Add a C helper function for bytes and bytearray slicing.

view details

pranavrajpal

commit sha 4c90aa42e6a74f127956124aec94f55dbc46a281

[mypyc] Use a cache to speed up singledispatch calls (#10972) This adds a cache to the dispatching code for singledispatch functions, which makes calling singledispatch functions much faster.

view details

pranavrajpal

commit sha cd013d13e600539a6c2d7a667c95b6c4602e62ee

[mypyc] Use native calls to singledispatch functions (#10981) * Wrap singledispatch callable class in glue function mypyc doesn't have support for calling callable classes without going through the python API, so instead of having to special case the singledispatch callable class when we want to generate a native call to it, we create a glue function that loads the callable class from the globals dict and then makes a native call to that callable class's __call__ method. * Use native calls for singledispatch functions Use a native call to the glue function that we generate for singledispatch functions instead of calling those singledispatch functions through the python API. * Add irbuild test for native calls to singledispatch functions Add an irbuild test to make sure we use native calls to the glue functions for singledispatch functions, and that the glue functions make native calls to the __call__ methods on the callable classes. * Fix irbuild test

view details

push time in a month

push eventjhance/mypy

Jared Hance

commit sha 256840e564d443d791abb49c0747245fb9e357af

Fix header

view details

push time in 2 months

PR opened python/mypy

[mypyc] Add primitive for bytes join() method

Description

This is mostly adapted from how it is done for str. We call undocumented API function _CPyBytes_Join which behaves similarly to the publicly available unicode one. Care is taken to deal with the case where we happen to have bytearray instead. As a result of needing to do the CheckExact anyway, we also support wonky subclasses of bytes.

Test Plan

Added both ir and runtime tests. Benchmark yielded reasonable gains but the benchmark also included concat which is likely slower than join.

+61 -1

0 comment

5 changed files

pr created time in 2 months

PR opened python/mypy

[mypyc] Implement bytes equality optimizations

Description

Implements part of mypyc/mypyc#880.

Implementation for the bytes comparison itself is stolen from the implementation of bytes_compare from cpython and inlined. Later we can switch unicode to use a similar API

Test Plan

Added both IR tests and runtime tests. Made a custom benchmark that yielded performance gain of around 4x (from 2.3x faster to 9x faster than uncompiled vs master).

+74 -3

0 comment

5 changed files

pr created time in 2 months

push eventjhance/mypy

Jared Hance

commit sha d777ac4a1c5ef9c0b6ff23011f22bf50a53cfec6

Add comment

view details

push time in 2 months

create barnchjhance/mypy

branch : bytes-join

created branch time in 2 months

create barnchjhance/mypy

branch : bytes-compare

created branch time in 2 months