Same as 221698c603dcb318c609b4d21cb2a9fada44a14c
We had a bunch of instances of
https://github.com/NixOS/nix/issues/2176,
where nix would exit with a “killed by signal 9” error.
According to Eelco in that issue, this is perfectly normal behaviour
of course, and appears if the last command in a loop closes `stdout`
or `stdin`, then the builder will SIGKILL it immediately. This is of
course also a perfectly fine error message for that case.
It turns out that mainly GNU coreutils exhibit this behaviour …
Let’s see if using a more sane tool suite fixes that.
Change-Id: Iaf9e542952ca36c02208a3f067f575ba978272b4
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2663
Reviewed-by: Profpatsch <mail@profpatsch.de>
Tested-by: BuildkiteCI
I'm looking at removing some of these because they can cause
unnecessary build steps during CI pipeline generation.
Change-Id: I84742968918090c050d2eedab8a1b42692632a42
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2655
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Setting meta.targets to include all derivations in the different package
sets in Profpatsch's user folder makes them checked by CI until they do
the readTree refactor as promised.
To reduce code duplication we handle this in a simple function which is
exposed from nix.utils which may be a good place for depot specific bits
and bops we accumulate over time.
To get around the issue of too nested sets we perform the following
renames:
* users.Profpatsch.tests gets moved into its own directory
* users.Profpatsch.arglib.netencode now lives in its own file instead of
the default.nix
* users.Profpatsch.netstring.tests gets moved into its own directory
Change-Id: Icd039c29d7760a711c1c53554504d6b0cd19e120
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2603
Tested-by: BuildkiteCI
Reviewed-by: Profpatsch <mail@profpatsch.de>
Go 1.16 makes "go list all" not work. "go list std" is what we should be
using instead anyway.
Change-Id: I3f867fde477030d2358085b3d64b5856fb9c421b
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2551
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
When a file is added to the depot tree that is picked up by read-tree,
but it’s not a function like ({...}: {}), `readTree` will fail on the
function application, leading to a bad error message.
We can do slightly better, by checking the type and throwing a nicer
trace message.
`assertMsg` is copied from `nixpkgs/lib/assert.nix`, since at this
point we don’t have a reference to the lib.
There is another evaluation failure that can happen, which is when the
function we try to call does not have dots; however, nix does not
provide any inflection capabilies for checking whether a function
attrset is open (`builtins.functionArgs` only tells us the attrs it
mentions explicitly). Maybe the locality of the error could be
improved somehow.
Change-Id: Ibe38ce78bb56902075f7c31f2eeeb93485b34be3
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2469
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Previously, for types defined using typedef (like all primitive types)
type.checkType would return a boolean. This is largely fine since in
most places `type.checkToBool (type.checkType x)` or similar is used.
However, some functions actually take type.checkType up on the promise
that it returns a set of the form:
{
ok = <bool>;
err = <option string>;
}
This is the case for restrict which has checkToBool = v: v.ok; and will
generate a proper set except if `t.checkToBool (t.checkType v) == false`
in which case it will return t.checkType v. If t was a primitive type or
defined using typedef, previously `t.checkType v` would be a boolean
which meant as soon as (restrict …).checkToBool was called on a restrict
checkType result in cases where the wrapped type didn't match, an
unrelated error would be thrown:
nix-repl> with nix.yants; restrict "foo" (_: true) int "lol"
error: value is a boolean while a set was expected, at /home/lukas/src/depot/nix/yants/default.nix:38:39
This is fixed by making typedef return a proper set from checkType and
adjusting its checkToBool accordingly.
Unfortunately I don't think we can easily add test cases for this except
by using recursive nix or VM tests as there is no way to introspect
error messages.
Change-Id: I96a7be065630f04ca33358f21809284911ec14fe
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2536
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Reviewed-by: Profpatsch <mail@profpatsch.de>
Nix unfortunately has terrible escaping syntax: If something is an
escape sequence it does not know, like \0, it just swallows the
backslash and returns the second character (byte) as is (yes,
"\0" == "0" is true). This created the following bug in nixFileName
which should have resulted in at least a parse error: "(.*)\.nix" is
the same as "(.*).nix" which meant that nixFileName matched anything
that is at least 4 characters long and ends in "nix". This lead to
readTree creating double attributes when directories are involved or
attributes for non-nix files.
Change-Id: Ibf3be2bd189d48881c82ee795c50151bfb365627
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2535
Reviewed-by: tazjin <mail@tazj.in>
Tested-by: BuildkiteCI
This adds the star of NixCon 2017 from vuizvui, slightly reformatted and
now using yants. While it has some flaws, I realized that it is ideal to
run the tests of rustSimple{Lib,Bin} where the normal and the -tests
variant would have to be rebuilt if either the tests or the library /
executable itself changes.
Change-Id: Ie8f84f98c51c9fafc046eff916c8f0df7e8f224b
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2528
Tested-by: BuildkiteCI
Reviewed-by: Profpatsch <mail@profpatsch.de>
By using an extremely trivial derivation we can ensure that it will not
throw if evaluated using deepSeq. When using stdenv.mkDerivation or
similar at some point something will most likely throw or generate some
kind of error which is alright in the context of nixpkgs, but makes
testing yants harder than you'd think it should be.
Change-Id: I61ff7dc01a00a4815ef39066e4e223123356ddd5
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2507
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Port existing tests to runTestsuite and add some obvious additional
tests that wouldn't be possible before (using assertThrows and
assertEq).
Change-Id: Ibe950a7a0cda3e23ebb226bdff35f52cdfec5ddf
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2479
Tested-by: BuildkiteCI
Reviewed-by: Profpatsch <mail@profpatsch.de>
Reviewed-by: tazjin <mail@tazj.in>
assertDoesNotThrow is like assertThrows, but fails if the expression
throws. In that case the new unexpected-throw branch of AssertErrorContext
is returned.
Change-Id: I7195eb5df8965456e9ab9b69e35ec96b33f00a35
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2476
Tested-by: BuildkiteCI
Reviewed-by: Profpatsch <mail@profpatsch.de>
Refactor assertEq and assertThrows to be implemented in terms of a more
generic assertBool to reduce code duplication and ease adding new assert
types.
To preserve meaningful error messages AssertResult is changed slightly:
nope-eq and nope-throw have been replaced by a single nope branch which
contains an AssertErrorContext which contains error information. To
implement an assert assertBoolContext (which is not exposed) can be
used: It takes an AssertErrorContext which is returned in case of an
error and a boolean determining whether the assert was successful.
The currently possible AssertErrorContext are:
* should-throw: error result of assertThrows, formerly nope-throw
* not-equal: error result of assertEq, formerly nope-eq
Change-Id: Ifd6b3aa4187c90c3add2df63fa7c906c8f03fd2d
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2473
Tested-by: BuildkiteCI
Reviewed-by: Profpatsch <mail@profpatsch.de>
Introduces the concept of a “tag”, a single-keyed attrset which
annotates a nix value with a name.
This can be used to implement tagged unions (by implying the list of
possible tags is well-known), which has some overlap with how
`nix.yants` does it.
However, the more fascinating use-case is in concert with a
so-called discriminator, `match` and hylomorphisms.
The discriminator can take a nix value, and add tags to it based on
some predicate.
With `match`, we can then use that information to convert the
discriminated values again.
With `hylo`, we can combine both the “constructive” discriminator step
with the “destructive” match step to recursively walk over a nix data
structure (based on a description of how to recurse, e.g. through attrset
values or list values), and then apply a transformation in one go.
Change-Id: Ia335ca8b0881447fbbcb6bcd80f49feb835f1715
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2434
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
This should closely match the documented behaviour. It might still be
missing some edge cases of course.
Change-Id: I5c75fa045d5f3be8cf5eab787a02644500c14522
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2466
Reviewed-by: tazjin <mail@tazj.in>
Tested-by: BuildkiteCI
First step to slowly giving readTree some coverage, so we can do
refactoring without breaking functionality.
Change-Id: If25a8c0fa9c4ac7472c0473372f10a9326cccaf7
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2465
Reviewed-by: tazjin <mail@tazj.in>
Tested-by: BuildkiteCI
We are going to export some tests under `nix.readTree.tests`, so in
order to do that and still have `nix.readTree` be a function, let’s
move it to `__functor`.
This requires wiring the `args` and `initPath` arguments through
explicitly.
Change-Id: Ife7956b85d35e59c22174b42dcb7cca83ed868ea
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2464
Reviewed-by: tazjin <mail@tazj.in>
Tested-by: BuildkiteCI
Uses `builtins.tryEval` to check that the expression throws when
`deepSeq`-ed.
Change-Id: I0d57cc37f473bb733f57a1b1c0d889084152fd2f
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2463
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
Previously we would throw or return `{}`, which doesn’t integrate
nicely into our CI; thus, let’s wrap it into a derivation which either
fails the build or doesn’t.
Change-Id: I65880d86b8393094661e57a0b32aafe748bf1dd5
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2462
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Nobody has actually done any experimentation with typed Go, so we're
getting rid of it for now - it's causing annoying IFD during build
graph generation.
Change-Id: Ibac3dea98ebed1b3ee08acda184d24c500cf695d
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2458
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
Reviewed-by: lukegb <lukegb@tvl.fyi>
Reviewed-by: Profpatsch <mail@profpatsch.de>
`restrict` uses a predicate function to restrict a type, giving the
restricting a descriptive name in the process.
First, the wrapped type definition is checked (e.g. int) and then the
value is checked with the predicate, so the predicate can already
depend on the value being of the wrapped type.
Change-Id: Ic3edde45a8f34c31bc164414580d0a1aa5a821d5
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2312
Tested-by: BuildkiteCI
Reviewed-by: Profpatsch <mail@profpatsch.de>
Reviewed-by: tazjin <mail@tazj.in>
Similar to runCommandLocal, this turns off substitutions and prefers
building locally.
Change-Id: I823b34c7fc54990b54a82324172c299aeffdbf41
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2309
Tested-by: BuildkiteCI
Reviewed-by: Profpatsch <mail@profpatsch.de>
Reviewed-by: lukegb <lukegb@tvl.fyi>
I can review all nix-related changes.
Change-Id: I13e5bb7b523d4b9c79dbe2083d9e23c217466651
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2308
Tested-by: BuildkiteCI
Reviewed-by: lukegb <lukegb@tvl.fyi>
Machines on which LANG is misconfigured have trouble with SBCL loading
files that contain characters in certain encodings. This overrides
whichever local LANG (if any) is set.
Change-Id: Ic4341a01c4393e7f697de6cecc58dea4f2d85987
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2076
Tested-by: BuildkiteCI
Reviewed-by: glittershark <grfn@gws.fyi>
This is required to automatically walk the tree (see subsequent
commits).
Note: Lisp packages are removed from the CI builds in this commit
because the attrValues of third_party.lisp will contain an element
that is simply `true`, which causes a type error.
These packages are re-added when CI refactoring is complete.
Change-Id: I21e2b719e6c7161c23d2867a216f4daa1c6c8394
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1848
Tested-by: BuildkiteCI
Reviewed-by: glittershark <grfn@gws.fyi>
Expose an `sbcl` attribute on packages and programs, to allow for easier
development either with SLY or on a REPL.
Change-Id: Ide4d087a5223561e1fe192ef32dc593c54b5a20e
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1834
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
depot-scanner is a tool that runs Nix and parses the --trace-file-access output to deduce what files are necessary to evaluate a derivation.
Take DEPOT_ROOT from the environment. If depotRoot doesn't exist, print an error early.
Fix the build of the protobuf library. Switch to the GRPC build rule, as a service is in this proto file.
Create the PathType enum and parse it from cmdline flags.
Change-Id: I537b5c6bceecf76ca510f7ac04ab9dad7785feb1
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1769
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Reviewed-by: lukegb <lukegb@tvl.fyi>
This accepts a function for args, and passes it an attrset containing
the locatedAt parameter.
Change-Id: I3c0f8ca00605e02a787fda88f32b06f5ef9998f2
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1639
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
This was based off fetchzip, but unlike fetchzip we don't need to
support arbitrary archive formats, so we don't actually need the unpack
hooks.
Change-Id: I5655c031f420422f13e72bb79f4b4b95f6f1204e
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1640
Tested-by: BuildkiteCI
Reviewed-by: kanepyork <rikingcoding@gmail.com>
Reviewed-by: tazjin <mail@tazj.in>
This removes the need for `map (x: x.gopkg)` in dependency lists.
Change-Id: Ia19e51f14110bf22f5cdbb64fa7ce182e1b34241
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1624
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
This makes the derivations more uniform, since both third-party
packages generated by buildGo.external and native buildGo.package
expose their libraries as gopkg attributes now.
Change-Id: I547f9860082f36e3300139bf67613eb4fc600d24
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1623
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Add support for explicitly specifying tests as part of a buildLisp
program or library.
Change-Id: I733213c1618f0fa60f645465560bce0522641efd
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1481
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
This was already happening for libraries, but not for programs - as a
result, programs built with libraries that contained unicode (eg
cl-unicode, uax-15, ...) would fail to build due to character encoding
issues when loading the FASLs.
Change-Id: I66149b585e85b213d0c026153140a1925536bd29
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1469
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Without this, we would end up with "double" store paths like this:
/nix/store/848js1fvbjniv5n00hifmhgzzszl97vv--nix-store-ad6piq18wdkxnfzsbyn88ixvv7gfb1dp-main.go.drv
We really only care about the `main.go` bit though.
Change-Id: Ib0644781a0e232a45e1cae3dd05b9b828c9087ee
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1321
Tested-by: BuildkiteCI
Reviewed-by: ericvolp12 <ericvolp12@gmail.com>
Breaking change detection will run but not enforce.
Emoji of water buffalo was chosen by @pedge fiat in the bufbuild slack.
Change-Id: Ie292f2bfddc0e3bc512e4a138c0b5d0fa2603bad
Reviewed-on: https://cl.tvl.fyi/c/depot/+/1247
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Reviewed-by: glittershark <grfn@gws.fyi>
If you compiled dirA/test.lisp and dirB/test.lisp in the same library,
they'd both get written to /test.fasl and the second would overwrite the
first. Instead, use the whole store path (with / swapped for -) as the fasl
filename.
Change-Id: I4eb88b5d33757751e1f67e72ed328bd58079b1b9
Reviewed-on: https://cl.tvl.fyi/c/depot/+/944
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
We have this nice `runExecline` now, so we don’t need to use
`runCommand` (which spawns bash) just to write a simple script.
Change-Id: I2941ed8c1448fa1d7cc02dc18b24a8a945b2c38b
Reviewed-on: https://cl.tvl.fyi/c/depot/+/704
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Reviewed-by: BuildkiteCI
runExecline is a primitive that just does not care.
It’s similar to `runCommand`, but instead of concatenating bash
scripts left and right, it actually *uses* the features of
`derivation`, passing things to `args` and making it possible to
overwrite the `builder` in a sensible manner.
Additionally, it provides a way to pass a nix string to `stdin` of the
build script.
Similar to `writeExecline`, the passed script is not a string, but a
nested list of nix lists representing execline blocks. Escaping is
done by the implementation, the user can just use normal nix strings.
Change-Id: I890d9e5d921207751cdc8cc4309381395d92742f
Reviewed-on: https://cl.tvl.fyi/c/depot/+/701
Reviewed-by: BuildkiteCI
Reviewed-by: isomer <isomer@tvl.fyi>
Reviewed-by: tazjin <mail@tazj.in>
Tested-by: BuildkiteCI
The escaping functions are going to be used by both `writeExecline`
and `runExecline`, so let’s move them to their own namespace.
Change-Id: Iccf69eaeca3062573e0751a17c548b7def86196d
Reviewed-on: https://cl.tvl.fyi/c/depot/+/706
Reviewed-by: tazjin <mail@tazj.in>
Reviewed-by: Kane York <rikingcoding@gmail.com>
The most trivial of all derivations. It is more useful than it looks.
Can be used to bind nix expressions (e.g. test suites) to a
derivation, so that `nix-build` does not crap itself.
Change-Id: I61c24d8c129c9505733161207f3c30e820f5b15e
Reviewed-on: https://cl.tvl.fyi/c/depot/+/665
Reviewed-by: tazjin <mail@tazj.in>
This is a simple-stupid “unix import system” for nix, for referencing
binaries in `/bin/` by their name and lifting them to a Nix attrset.
Allows for simple aliasing of executable names.
Change-Id: Ifa23cb377201c3b08050c5026e9751e736afaf56
Reviewed-on: https://cl.tvl.fyi/c/depot/+/664
Reviewed-by: tazjin <mail@tazj.in>
This is a very simple test suite for nix expressions.
It should help us set up a good suite of unit tests for our nix-based
stuff.
Since we allow import from derivation, these tests can also depend on
derivations and e.g. use `builtins.readFile` to check outputs.
This is a first PoC to get us going, we can always replace it by
something different in the future if we don’t like it.
Change-Id: I206c7b624db2b1dabd9c73ffce4f87e658919958
Reviewed-on: https://cl.tvl.fyi/c/depot/+/662
Reviewed-by: tazjin <mail@tazj.in>
Tested-by: tazjin <mail@tazj.in>
