Adds a common project that allows sharing code between the `runtime` and `std-bits`.
Due to classpath separation and the way it is compiled, the classes will be duplicated - we will have one copy for the `runtime` classpath and another copy as a small JAR for `Standard.Base` library.
This is still much better than having the code duplicated - now at least we have a single source of truth for the shared implementations.
Due to the copying we should not expand this project too much, but I encourage to put here any methods that would otherwise require us to copy the code itself.
This may be a good place to put parts of the hashing logic to then allow sharing the logic between the `runtime` and the `MultiValueKey` in the `Table` library (cc: @Akirathan).
close#5070
Changelog:
- Include the original exception to log expressions
- Enable logging of Akka Actors' lifecycle events on debug logging level
- Decrease the severity of interruption log messages because interruptions are part of the workflow. The computation can be interrupted at any time, and still be recomputed after. Warnings are just misleading in this case.
This change downgrades hashing algorithm used in caching IR and library bindings to SHA-1. It is sufficient and significantly faster for the purpose of simple checksum we use it for.
Additionally, don't calculate the digest for serialized bytes - if we get the expected object type then we are confident about the integrity.
Don't initialize Jackson's ObjectMapper for every metadata serialization/de-serialization. Initialization is very costly.
Avoid unnecessary conversions between Scala and Java. Those back-and-forth `asScala` and `asJava` are pretty expensive.
Finally fix an SBT warning when generating library cache.
Closes https://github.com/enso-org/enso/issues/5763
# Important Notes
The change cuts roughly 0.8-1s from the overall startup.
This change will certainly lead to invalidation of existing caches. It is advised to simply start with a clean slate.
This change adds serialization and deserialization of library bindings.
In order to be functional, one needs to first generate IR and
serialize bindings using `--compiled <path-to-library>` command. The bindings
will be stored under the library with `.bindings` suffix.
Bindings are being generated during `buildEngineDistribution` task, thus not
requiring any extra steps.
When resolving import/exports the compiler will first try to load
module's bindings from cache. If successful, it will not schedule its
imports/exports for immediate compilation, as we always did, but use the
bindings info to infer the dependent modules.
The current change does not make any optimizations when it comes to
compiling the modules, yet. It only delays the actual
compilation/loading IR from cache so that it can be done in bulk.
Further optimizations will come from this opportunity such as parallel
loading of caches or lazily inferring only the necessary modules.
Part of https://github.com/enso-org/enso/issues/5568 work.
In cloud we want to allow users to create new project from the template. List of templates is a bit outdated and doesn't contain all from the https://github.com/enso-org/project-templates. This PR simply adds missing ones
A combination of commands triggered separate compilation that only recompiled part of builtins. A mechanism that workaround annotation processor's problems with separate compilation was updated to include changes from https://github.com/enso-org/enso/pull/4111. This was pretty tough to find given the rather unusual circumstances in CI.
# Important Notes
This eliminates the _random_ failures in CI related to separate compilation.
To reproduce a specific set of steps has to be executed:
```
sbt> all buildEngineDistribution engine-runner/assembly runtime/Benchmark/compile language-server/Benchmark/compile searcher/Benchmark/compile
(exit sbt)
sbt> test
```
Expressions returning polyglot values were not reporting the type of the result because we have to do additional magic that infers the correct Enso type. Since this is exactly what `TypeOfNode` does, I re-used the logic.
Straightforward solution failed in tests because of assertions:
```
[enso] WARNING: Execution of function main failed (Invalid library usage. Cached library must be adopted by a RootNode before it is executed.).
java.lang.AssertionError: Invalid library usage. Cached library must be adopted by a RootNode before it is executed.
```
That is why this PR replaces `ExecutionEventListener` with `ExecutionEventNodeFactory`.
# Important Notes
Usage of `TypeOfNode` for programs that **do not** import stdlib means that we report types that do not involve stdlib e.g.
`Standard.Builtins.Main.Integer` instead of `Standard.Base.Data.Numbers.Integer`. While surprising, this is correct and I would say desirable. While reviewing the code, notice the difference in expectations in our runtime tests.
In order to investigate `engine/language-server` project, I need to be able to open its sources in IGV and NetBeans.
# Important Notes
By adding same Java source (this time `package-info.java`) and compiling with our Frgaal compiler the necessary `.enso-sources*` files are generated for `engine/language-server` and then the `enso4igv` plugin can open them and properly understand their compile settings.
In addition to that this PR enhances the _"logical view"_ presentation of the project by including all source roots found under `src/*/*`.
https://github.com/enso-org/enso/pull/3764 introduced static wrappers for instance methods. Except it had a limitation to only be allowed for types with at least a single constructor.
That excluded builtin types as well which, by default, don't have them. This limitation is problematic for Array/Vector consolidation and makes builtin types somehow second-citizens.
This change lifts the limitation for builtin types only. Note that we do want to share the implementation of the generated builtin methods. At the same time due to the additional argument we have to adjust the starting index of the arguments.
This change avoids messing with the existing dispatch logic, to avoid unnecessary complexity.
As a result it is now possible to call builtin types' instance methods, statically:
```
arr = Array.new_1 42
Array.length arr
```
That would previously lead to missing method exception in runtime.
# Important Notes
The only exception is `Nothing`. Primarily because it requires `Nothing` to have a proper eigentype (`Nothing.type`) which would messed up a lot of existing logic for no obvious benefit (no more calling of `foo=Nothing` in parameters being one example).
Introduces unboxed (and arity-specialized) storage schemes for Atoms. It results in improvements both in memory consumption and runtime.
Memory wise: instead of using an array, we now use object fields. We also enable unboxing. This cuts a good few pointers in an unboxed object. E.g. a quadruple of integers is now 64 bytes (4x8 bytes for long fields + 16 bytes for layout and constructor pointers + 16 bytes for a class header). It used to be 168 bytes (4x24 bytes for boxed Longs + 16 bytes for array header + 32 bytes for array contents + 8 bytes for constructor ptr + 16 bytes for class header), so we're saving 104 bytes a piece. In the least impressive scenarios (all-boxed fields) we're saving 8 bytes per object (saving 16 bytes for array header, using 8 bytes for the new layout field). In the most-benchmarked case (list of longs), we save 32 bytes per cons-cell.
Time wise:
All list-summing benchmarks observe a ~2x speedup. List generation benchmarks get ~25x speedups, probably both due to less GC activity and better allocation characteristics (only allocating one object per Cons, rather than Cons + Object[] for fields). The "map-reverse" family gets a neat 10x speedup (part of the work is reading, which is 2x faster, the other is allocating, which is now 25x faster, we end up with 10x when combined).
While doing regular node manipulation in editior, noticed a number of situations when
```
java.lang.IndexOutOfBoundsException: None
at java.base/java.lang.Character.offsetByCodePoints(Character.java:8699)
at java.base/java.lang.String.offsetByCodePoints(String.java:820)
at org.enso.text.buffer.CodePointView$Ops$.drop(CodePointView.scala:98)
at org.enso.text.buffer.CodePointView$Ops$.drop(CodePointView.scala:57)
at org.enso.text.buffer.Node.drop(Tree.scala:218)
at org.enso.text.buffer.Rope.dropWith(Rope.scala:86)
at org.enso.text.buffer.CodePointView.drop(CodePointView.scala:30)
at org.enso.text.editing.RopeTextEditor$.cutOutTail(RopeTextEditor.scala:42)
...
```
would be thrown. Further text edits would simply be rejected requiring a complete restart. I doubt we should propagate
`IndexOutOfBoundsException`. Instead it is safer to just apply the edit to the end of the rope.
* Hash codes prototype
* Remove Any.hash_code
* Improve caching of hashcode in atoms
* [WIP] Add Hash_Map type
* Implement Any.hash_code builtin for primitives and vectors
* Add some values to ValuesGenerator
* Fix example docs on Time_Zone.new
* [WIP] QuickFix for HashCodeTest before PR #3956 is merged
* Fix hash code contract in HashCodeTest
* Add times and dates values to HashCodeTest
* Fix docs
* Remove hashCodeForMetaInterop specialization
* Introduce snapshoting of HashMapBuilder
* Add unit tests for EnsoHashMap
* Remove duplicate test in Map_Spec.enso
* Hash_Map.to_vector caches result
* Hash_Map_Spec is a copy of Map_Spec
* Implement some methods in Hash_Map
* Add equalsHashMaps specialization to EqualsAnyNode
* get and insert operations are able to work with polyglot values
* Implement rest of Hash_Map API
* Add test that inserts elements with keys with same hash code
* EnsoHashMap.toDisplayString use builder storage directly
* Add separate specialization for host objects in EqualsAnyNode
* Fix specialization for host objects in EqualsAnyNode
* Add polyglot hash map tests
* EconomicMap keeps reference to EqualsNode and HashCodeNode.
Rather than passing these nodes to `get` and `insert` methods.
* HashMapTest run in polyglot context
* Fix containsKey index handling in snapshots
* Remove snapshots field from EnsoHashMapBuilder
* Prepare polyglot hash map handling.
- Hash_Map builtin methods are separate nodes
* Some bug fixes
* Remove ForeignMapWrapper.
We would have to wrap foreign maps in assignments for this to be efficient.
* Improve performance of Hash_Map.get_builtin
Also, if_nothing parameter is suspended
* Remove to_flat_vector.
Interop API requires nested vector (our previous to_vector implementation). Seems that I have misunderstood the docs the first time I read it.
- to_vector does not sort the vector by keys by default
* Fix polyglot hash maps method dispatch
* Add tests that effectively test hash code implementation.
Via hash map that behaves like a hash set.
* Remove Hashcode_Spec
* Add some polyglot tests
* Add Text.== tests for NFD normalization
* Fix NFD normalization bug in Text.java
* Improve performance of EqualsAnyNode.equalsTexts specialization
* Properly compute hash code for Atom and cache it
* Fix Text specialization in HashCodeAnyNode
* Add Hash_Map_Spec as part of all tests
* Remove HashMapTest.java
Providing all the infrastructure for all the needed Truffle nodes is no longer manageable.
* Remove rest of identityHashCode message implementations
* Replace old Map with Hash_Map
* Add some docs
* Add TruffleBoundaries
* Formatting
* Fix some tests to accept unsorted vector from Map.to_vector
* Delete Map.first and Map.last methods
* Add specialization for big integer hash
* Introduce proper HashCodeTest and EqualsTest.
- Use jUnit theories.
- Call nodes directly
* Fix some specializations for primitives in HashCodeAnyNode
* Fix host object specialization
* Remove Any.hash_code
* Fix import in Map.enso
* Update changelog
* Reformat
* Add truffle boundary to BigInteger.hashCode
* Fix performance of HashCodeTest - initialize DataPoints just once
* Fix MetaIsATest
* Fix ValuesGenerator.textual - Java's char is not Text
* Fix indent in Map_Spec.enso
* Add maps to datapoints in HashCodeTest
* Add specialization for maps in HashCodeAnyNode
* Add multiLevelAtoms to ValuesGenerator
* Provide a workaround for non-linear key inserts
* Fix specializations for double and BigInteger
* Cosmetics
* Add truffle boundaries
* Add allowInlining=true to some truffle boundaries.
Increases performance a lot.
* Increase the size of vectors, and warmup time for Vector.Distinct benchmark
* Various small performance fixes.
* Fix Geo_Spec tests to accept unsorted Map.to_vector
* Implement Map.remove
* FIx Visualization tests to accept unsorted Map.to_vector
* Treat java.util.Properties as Map
* Add truffle boundaries
* Invoke polyglot methods on java.util.Properties
* Ignore python tests if python lang is missing
Fixes an error when the logger processes NPE:
```
[internal-logger-error] One of the printers failed to write a message: java.lang.NullPointerException
```
Implements `getMetaObject` and related messages from Truffle interop for Enso values and types. Turns `Meta.is_a` into builtin and re-uses the same functionality.
# Important Notes
Adds `ValueGenerator` testing infrastructure to provide unified access to special Enso values and builtin types that can be reused by other tests, not just `MetaIsATest` and `MetaObjectTest`.
This removes the special handling of polyglot exceptions and allows matching on Java exceptions in the same way as for any other types.
`Polyglot_Error`, `Panic.catch_java` and `Panic.catch_primitive` are gone
The change mostly deals with the backslash of removing `Polyglot_Error` and two `Panic` methods.
`Panic.catch` was implemented as a builtin instead of delegating to `Panic.catch_primitive` builtin that is now gone.
This fixes https://www.pivotaltracker.com/story/show/182844611
* Sequence literal (Vector) should preserve warnings
When Vector was created via a sequence literal, we simply dropped any
associated any warnings associated with it.
This change propagates Warnings during the creation of the Vector.
Ideally, it would be sufficient to propagate warnings from the
individual elements to the underlying storage but doesn't go well with
`Vector.fromArray`.
* update changelog
* Array-like structures preserver warnings
Added a WarningsLibrary that exposes `hasWarnings` and `getWarnings`
messages. That way we can have a single storage that defines how to
extract warnings from an Array and the others just delegate to it.
This simplifies logic added to sequence literals to handle warnings.
* Ensure polyglot method calls are warning-free
Since warnings are no longer automatically extracted from Array-like
structures, we delay the operation until an actual polyglot method call
is performed.
Discovered a bug in `Warning.detach_selected_warnings` which was missing
any usage or tests.
* nits
* Support multi-dimensional Vectors with warnings
* Propagate warnings from case branches
* nit
* Propagate all vector warnings when reading element
Previously, accessing an element of an Array-like structure would only
return warnings of that element or of the structure itself.
Now, accessing an element also returns warnings from all its elements as
well.
It appears that we were always adding builtin methods to the scope of the module and the builtin type that shared the same name.
This resulted in some methods being accidentally available even though they shouldn't.
This change treats differently builtins of types and modules and introduces auto-registration feature for builtins.
By default all builtin methods are registered with a type, unless explicitly defined in the annotation property.
Builtin methods that are auto-registered do not have to be explicitly defined and are registered with the underlying type.
Registration correctly infers the right type, depending whether we deal with static or instance methods.
Builtin methods that are not auto-registered have to be explicitly defined **always**. Modules' builtin methods are the prime example.
# Important Notes
Builtins now carry information whether they are static or not (inferred from the lack of `self` parameter).
They also carry a `autoRegister` property to determine if a builtin method should be automatically registered with the type.
This change adds support for Version Controlled projects in language server.
Version Control supports operations:
- `init` - initialize VCS for a project
- `save` - commit all changes to the project in VCS
- `restore` - ability to restore project to some past `save`
- `status` - show the status of the project from VCS' perspective
- `list` - show a list of requested saves
# Important Notes
Behind the scenes, Enso's VCS uses git (or rather [jGit](https://www.eclipse.org/jgit/)) but nothing stops us from using a different implementation as long as it conforms to the establish API.
Computing length of a text takes time. Let's cache it after first computation.
# Important Notes
Wrote `StringBenchmarks` that sums lengths of (the same) `Text` present many time in a `Vector`. Initially it took `383.673 ms` per operation. Then it took `0.031 ms/op`. Looks like the `length` calls are returning instantly as they get cached.
1. Changes how we do monadic state – rather than a haskelly solution, we now have an implicit env with mutable data inside. It's better for the JVM. It also opens the possibility to have state ratained on exceptions (previously not possible) – both can now be implemented.
2. Introduces permission check system for IO actions.
We've had an old attempt at integrating a Rust parser with our Scala/Java projects. It seems to have been abandoned and is not used anywhere - it is also superseded by the new integration of the Rust parser. I think it was used as an experiment to see how to approach such an integration.
Since it is not used anymore - it make sense to remove it, because it only adds some (slight, but non-zero) maintenance effort. We can always bring it back from git history if necessary.
- Reimplement the `Duration` type to a built-in type.
- `Duration` is an interop type.
- Allow Enso method dispatch on `Duration` interop coming from different languages.
# Important Notes
- The older `Duration` type should now be split into new `Duration` builtin type and a `Period` type.
- This PR does not implement `Period` type, so all the `Period`-related functionality is currently not working, e.g., `Date - Period`.
- This PR removes `Integer.milliseconds`, `Integer.seconds`, ..., `Integer.years` extension methods.
Use an `ArraySlice` to slice `Vector`.
Avoids memory copying for the slice function.
# Important Notes
| Test | Ref | New |
| --- | --- | --- |
| New Vector | 71.9 | 71.0 |
| Append Single | 26.0 | 27.7 |
| Append Large | 15.1 | 14.9 |
| Sum | 156.4 | 165.8 |
| Drop First 20 and Sum | 171.2 | 165.3 |
| Drop Last 20 and Sum | 170.7 | 163.0 |
| Filter | 76.9 | 76.9 |
| Filter With Index | 166.3 | 168.3 |
| Partition | 278.5 | 273.8 |
| Partition With Index | 392.0 | 393.7 |
| Each | 101.9 | 102.7 |
- Note: the performance of New and Append has got slower from previous tests.
This PR adds a possibility to generate native-image for engine-runner.
Note that due to on-demand loading of stdlib, programs that make use of it are currently not yet supported
(that will be resolved at a later point).
The purpose of this PR is only to make sure that we can generate a bare minimum runner because due to lack TruffleBoundaries or misconfiguration in reflection config, this can get broken very easily.
To generate a native image simply execute:
```
sbt> engine-runner-native/buildNativeImage
... (wait a few minutes)
```
The executable is called `runner` and can be tested via a simple test that is in the resources. To illustrate the benefits
see the timings difference between the non-native and native one:
```
>time built-distribution/enso-engine-0.0.0-dev-linux-amd64/enso-0.0.0-dev/bin/enso --no-ir-caches --in-project test/Tests/ --run engine/runner-native/src/test/resources/Factorial.enso 6
720
real 0m4.503s
user 0m9.248s
sys 0m1.494s
> time ./runner --run engine/runner-native/src/test/resources/Factorial.enso 6
720
real 0m0.176s
user 0m0.042s
sys 0m0.038s
```
# Important Notes
Notice that due to a [bug in GraalVM](https://github.com/oracle/graal/issues/4200), which is already fixed in 22.x, and us still being on 21.x for the time being, I had to add a workaround to our sbt build to build a different fat jar for native image. To workaround it I had to exclude sqlite jar. Hence native image task is on `engine-runner-native` and not on `engine-runner`.
Will need to add the above command to CI.
`Vector` type is now a builtin type. This requires a bunch of additional builtin methods for its creation:
- Use `Vector.from_array` to convert any array-like structure into a `Vector` [by copy](f628b28f5f)
- Use (already existing) `Vector.from_polyglot_array` to convert any array-like structure into a `Vector` **without** copying
- Use (already existing) `Vector.fill 1 item` to create a singleton `Vector`
Additional, for pattern matching purposes, we had to implement a `VectorBranchNode`. Use following to match on `x` being an instance of `Vector` type:
```
import Standard.Base.Data.Vector
size = case x of
Vector.Vector -> x.length
_ -> 0
```
Finally, `VectorLiterals` pass that transforms `[1,2,3]` to (roughly)
```
a1 = 1
a2 = 2
a3 = 3
Vector (Array (a1,a2, a3))
```
had to be modified to generate
```
a1 = 1
a2 = 2
a3 = 3
Vector.from_array (Array (a1, a2, a3))
```
instead to accomodate to the API changes. As of 025acaa676 all the known CI checks passes. Let's start the review.
# Important Notes
Matching in `case` statement is currently done via `Vector_Data`. Use:
```
case x of
Vector.Vector_Data -> True
```
until a better alternative is found.
This is a step towards the new language spec. The `type` keyword now means something. So we now have
```
type Maybe a
Some (from_some : a)
None
```
as a thing one may write. Also `Some` and `None` are not standalone types now – only `Maybe` is.
This halfway to static methods – we still allow for things like `Number + Number` for backwards compatibility. It will disappear in the next PR.
The concept of a type is now used for method dispatch – with great impact on interpreter code density.
Some APIs in the STDLIB may require re-thinking. I take this is going to be up to the libraries team – some choices are not as good with a semantically different language. I've strived to update stdlib with minimal changes – to make sure it still works as it did.
It is worth mentioning the conflicting constructor name convention I've used: if `Foo` only has one constructor, previously named `Foo`, we now have:
```
type Foo
Foo_Data f1 f2 f3
```
This is now necessary, because we still don't have proper statics. When they arrive, this can be changed (quite easily, with SED) to use them, and figure out the actual convention then.
I have also reworked large parts of the builtins system, because it did not work at all with the new concepts.
It also exposes the type variants in SuggestionBuilder, that was the original tiny PR this was based on.
PS I'm so sorry for the size of this. No idea how this could have been smaller. It's a breaking language change after all.
- Added `Zone`, `Date_Time` and `Time_Of_Day` to `Standard.Base`.
- Renamed `Zone` to `Time_Zone`.
- Added `century`.
- Added `is_leap_year`.
- Added `length_of_year`.
- Added `length_of_month`.
- Added `quarter`.
- Added `day_of_year`.
- Added `Day_Of_Week` type and `day_of_week` function.
- Updated `week_of_year` to support ISO.
# Important Notes
- Had to pass locale to formatter for date/time tests to work on my PC.
- Changed default of `week_of_year` to use ISO.
* Builtin Date_Time, Time_Of_Day, Zone
Improved polyglot support for Date_Time (formerly Time), Time_Of_Day and
Zone. This follows the pattern introduced for Enso Date.
Minor caveat - in tests for Date, had to bend a lot for JS Date to pass.
This is because JS Date is not really only a Date, but also a Time and
Timezone, previously we just didn't consider the latter.
Also, JS Date does not deal well with setting timezones so the trick I
used is to first call foreign function returning a polyglot JS Date,
which is converted to ZonedDateTime and only then set the correct
timezone. That way none of the existing tests had to be changes or
special cased.
Additionally, JS deals with milliseconds rather than nanoseconds so
there is loss in precision, as noted in Time_Spec.
* Add tests for Java's LocalTime
* changelog
* Make date formatters in table happy
* PR review, add more tests for zone
* More tests and fixed a bug in column reader
Column reader didn't take into account timezone but that was a mistake
since then it wouldn't map to Enso's Date_Time.
Added tests that check it now.
* remove redundant conversion
* Update distribution/lib/Standard/Base/0.0.0-dev/src/Data/Time.enso
Co-authored-by: Radosław Waśko <radoslaw.wasko@enso.org>
* First round of addressing PR review
* don't leak java exceptions in Zone
* Move Date_Time to top-level module
* PR review
Co-authored-by: Radosław Waśko <radoslaw.wasko@enso.org>
Co-authored-by: Jaroslav Tulach <jaroslav.tulach@enso.org>
