fd9716fa95
Various semantic checks on formal parameters.
...
R=danrubel@google.com
Review-Url: https://codereview.chromium.org/2926953004 .
2017-06-10 10:10:07 +02:00
6b65fc22f7
Annotate targets of implicit this property gets/sets in inference tests.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2933663002 .
2017-06-09 17:41:13 -07:00
8f5bcddc31
Implement type inference for assignments to a static variable.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2933643002 .
2017-06-09 17:36:20 -07:00
ef94f715a9
Add type inference for assignment to local variables.
...
Also fix some upwards inference cases for assignment to index
expressions.
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2928033005 .
2017-06-09 15:13:14 -07:00
c2f18fc930
Implement promised watch behaviour.
...
1. There are no reasons to use different watch function between
computeDelta() invocations, so I moved it to the constructor.
2. We need to call it, and await, before reading files, not after.
3. Added tests.
R=ahe@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2931183003 .
2017-06-09 15:04:30 -07:00
6f4bba5f03
Compute API signatures of files.
...
Note, that in non-incemental case this change has negative performance
implications - compilation is 10% slower. Computing MD5 hashes is very
expensive.
In real life scenarios we will cache API signatures and URIs as
part of unlinked information for each file.
R=ahe@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2926883003 .
2017-06-09 11:41:29 -07:00
4abff2291b
Rework type inference of assignments to index expressions.
...
The approach from the las CL (try to figure out the structure of the
assignment expression from the desugared result) turned out to be
unsustainably complex. In this version we keep track of the structure
while doing the desugaring, and then store the result in a wrapper
object the kernel AST. The wrapper object defers visit methods to the
object it wraps, so when the kernel objects are serialized to disk,
the wrapper disappears.
I also took the liberty of removing the code that inserts types
into the temporary variables and conditional expressions
introduced by desugaring. I will add this in a later CL if it
proves to be necessary.
R=ahe@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2927013004 .
2017-06-09 10:48:07 -07:00
d9f14d6a2b
Add --sdk option.
...
R=sigmund@google.com
Review-Url: https://codereview.chromium.org/2929783002 .
2017-06-09 15:52:14 +02:00
2042128817
Include uriToSource from dill-based libraries in output.
...
R=ahe@google.com
Review-Url: https://codereview.chromium.org/2929103002 .
2017-06-09 14:11:43 +02:00
aa8ca9244a
Add getTarget method to CompilerCommandLine
...
The method is used to create an instance of a backend target class.
CompilerCommandLine already has the necessary information for this ("target"
name string and "strongMode" flag). Additionally, a backed target instance is
now not created in fasta.dart/parseScriptInFileSystem, but should be passed at
the call site.
R=ahe@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2932513003 .
2017-06-09 11:04:44 +02:00
cf59b1a0f9
Remove ClassHierarchy factory constructor.
...
KernelTarget already uses IncrementalClassHierarchy, and other clients
either need ClosedWorldClassHierarchy, so should explicitly create it,
or don't care, and can be switched to IncrementalClassHierarchy.
R=ahe@google.com , kmillikin@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2930973002 .
2017-06-08 18:42:22 -07:00
56402976cd
more fixes for fasta analyzer integration
...
* fix scanner adapter setSourceStart
* fix fasta token copy() methods
* replace "is TokenWithComment" with "?.preceedingComments != null"
* update scanner test
R=brianwilkerson@google.com
Review-Url: https://codereview.chromium.org/2928773003 .
2017-06-08 18:02:09 -04:00
029771d746
Add type inference for complex assignments whose LHS is an index expression.
...
In later CLs I will expand on this code to handle other kinds of
complex assignments.
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2926763003 .
2017-06-08 09:25:24 -07:00
001ac97389
Implement initializer asserts.
...
Closes https://github.com/dart-lang/sdk/issues/29762
R=kasperl@google.com
Review-Url: https://codereview.chromium.org/2930673002 .
2017-06-08 13:46:26 +02:00
e551bef8ba
add check of preconditions in run step.
...
BUG=
R=ahe@google.com
Review-Url: https://codereview.chromium.org/2914773003 .
2017-06-07 15:45:55 -07:00
6e8155ae6f
Follow up improvement for checking dart:_builtin
...
R=ahe@google.com
Review-Url: https://codereview.chromium.org/2915153002 .
2017-06-07 15:38:18 -07:00
cb5f3842c4
Skip directly to the last library to set the entry point and check importUri
...
instead of fileUri.
On some scenarios fileUri is relative and entryUri was absolute, but importUri
should match for the entry library.
BUG=
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2926933002 .
2017-06-07 14:50:01 -07:00
aecfab68aa
Add an incremental reloader example and a utility tool to trigger a reload by
...
hand.
The example includes:
- an interactive UI that lets you trigger reloads without typing commands
- a wrapper of the incremental kernel generator that invalidates files based on
modification time stamps
BUG=
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2928483005 .
2017-06-07 13:44:56 -07:00
ebdb0af64f
Add integration test: hot reload + incremental compiler
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2925953002 .
2017-06-07 12:47:27 -07:00
c510737b17
Make a copy of files to invalidate to avoid concurrent modifications.
...
In contrast to Analysis Driver, in Kernel Driver FileState.refresh()
is asynchronous, so we might (and actually do) get concurrent watch
events that update the set of invalidate files.
R=ahe@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2924943005 .
2017-06-07 11:10:50 -07:00
400c7a0cb5
Set main procedure for Program.
...
Why is it called main _method_ when it is actually Procedure?
R=ahe@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2931603003 .
2017-06-07 10:59:32 -07:00
710af5af38
Measure time spent building graph of files.
...
It's about 1000 ms for analyzer_cli, which I use for benchmarking.
R=ahe@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2932463002 .
2017-06-07 10:26:09 -07:00
5c3dd4b11b
Change MethodInvocation.interfaceTarget from a Procedure to a Member.
...
It's possible for the interface target to be a field, e.g.:
typedef void F();
class C {
F f;
}
void g(C c) {
c.f(); // Interface target is C::f
}
Also, this CL fixes MethodInvocation.getStaticType() in the case where
the interface target is a getter.
R=dmitryas@google.com , kmillikin@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2923653003 .
2017-06-07 09:20:53 -07:00
c569a4a8b8
Compute all canonical names only for libraries being serialized.
...
This makes compilation with Kernel Driver about 20% faster than Fasta,
about 5.8 seconds using Kernel Driver vs. 7.2 seconds using Fasta. Note
that this is initial compilation, i.e. nothing read from the cache yet.
R=ahe@google.com , kmillikin@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2927463003 .
2017-06-07 08:50:04 -07:00
b0374f214c
update analyzer parser tests for fasta scanner
...
R=brianwilkerson@google.com
Review-Url: https://codereview.chromium.org/2930453002 .
2017-06-07 11:05:44 -04:00
c4a8b146e0
Semantic checks on switch cases and fall-through errors.
...
R=paulberry@google.com
Review-Url: https://codereview.chromium.org/2926533004 .
2017-06-07 11:26:34 +02:00
d6a766bb14
Improve NSM handling.
...
R=paulberry@google.com
Review-Url: https://codereview.chromium.org/2927613002 .
2017-06-07 09:28:13 +02:00
b18c325f68
Add int.toString() to analyzer mock SDK.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2922383003 .
2017-06-06 13:35:36 -07:00
49110ba0d6
Fix implementation of fold() in analyzer's mock SDK.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2924853003 .
2017-06-06 13:23:45 -07:00
7ab22d7ad2
Properly flatten FutureOr types when inferring the type of an async method.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2923733004 .
2017-06-06 13:13:08 -07:00
f7f9d1766f
Fix corner cases of type inference with implicit references to .call.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2928613003 .
2017-06-06 13:00:26 -07:00
a0fb4d9aef
Replace ClosedWorldClassHierarchy with IncrementalClassHierarchy in SourceLoader.
...
This does not cause any measurable performance changes for Fasta.
This speeds up initial compilation using Kernel Driver from 16 seconds
to about 7.6 seconds. Just using Fasta takes 7 seconds, so Kernel
Driver is just about 10% slower now.
R=ahe@google.com , paulberry@google.com , sigmund@google.com
BUG=
Review-Url: https://codereview.chromium.org/2927523002 .
2017-06-06 12:38:23 -07:00
32e78868bd
Record method invocation targets that are not procedures.
...
We don't yet store the targets in the kernel representation, since
kernel currently requires method invocation targets to be procedures
(see https://codereview.chromium.org/2923653003/ ). But we record the
target that we *would* store in the kernel to instrumentation so that
we can pass tests.
R=ahe@google.com
Review-Url: https://codereview.chromium.org/2920223008 .
2017-06-06 12:27:34 -07:00
470127ae46
Setup correct scope for initializers and complain about fields initialized more than once.
...
R=johnniwinther@google.com
Review-Url: https://codereview.chromium.org/2916333002 .
2017-06-06 19:35:35 +02:00
47d50e51a3
Add type inference for this expressions.
...
R=ahe@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2925603002 .
2017-06-06 09:54:19 -07:00
7df768af0d
Restore invariant that stack is empty in parseFunctionBody.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2914373002 .
2017-06-06 16:58:31 +02:00
09e755196a
Pass a Target instance to DillTarget instead of its name
...
R=ahe@google.com
Review-Url: https://codereview.chromium.org/2924833002 .
2017-06-06 16:03:49 +02:00
c36936f37f
translate remaining fasta scanner error codes
...
This updates the fasta scanner to translate the remaining
previously untranslated error codes so that they will be seen
by the analyzer when the fasta scanner is used by analyzer.
R=ahe@google.com
Review-Url: https://codereview.chromium.org/2923113002 .
2017-06-06 09:08:13 -04:00
b5883633f7
Do not annotate type parameters of instantiated function types.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2925693002 .
2017-06-05 13:26:12 -07:00
2513c2b93f
Adjust expectations for parameterized closures to match front_end behavior.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2925583003 .
2017-06-05 12:51:37 -07:00
2162b2da9b
Improve annotation of type variables in type inference tests.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2918373003 .
2017-06-05 11:47:07 -07:00
3285dcf039
Remove type parameters from type promotion classes.
...
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2923613002 .
2017-06-05 11:42:10 -07:00
32677b631f
option to enable fasta scanner
...
* add cmdline option so that analyzer will use the fasta scanner
* refactor handling of unterminated strings
Use -DuseFastaScanner=true to use the fasta scanner
rather than the analyzer scanner
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2919163002 .
2017-06-05 14:13:08 -04:00
416a860de5
Mark some more type inference tests as passing.
...
These tests were already working--they just needed expectations files
(and in one case, a `main` method).
R=scheglov@google.com
Review-Url: https://codereview.chromium.org/2926433002 .
2017-06-05 10:07:27 -07:00
c4449efa33
improve handling of missing interpolation identifier
...
When the fasta scanner finds a STRING_INTERPOLATION_IDENTIFIER token '$'
without an identifier, it now inserts a synthetic identifier
R=ahe@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2922563002 .
2017-06-02 15:16:44 -04:00
5b6d7653fe
fix unterminated interpolation expression handling
...
R=ahe@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2914213003 .
2017-06-02 13:47:08 -04:00
506389a744
Remove an unnecessary TODO from PropertyGet type inference.
...
R=ahe@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2921913002 .
2017-06-02 08:57:48 -07:00
b61f9f2343
improve fasta unterminated string recovery
...
This updates the fasta scanner
to append synthetic closing quotes to unterminated strings.
Rework of https://codereview.chromium.org/2912693002/
R=ahe@google.com , scheglov@google.com
Review-Url: https://codereview.chromium.org/2915093002 .
2017-06-02 11:51:11 -04:00
1e2f06f8da
Reapply "Use backend targets to run Kernel transformations in Fasta"
...
This reverts commit 0b424ca5d2R=ahe@google.com
Review-Url: https://codereview.chromium.org/2919003003 .
2017-06-02 16:10:28 +02:00
a79a59d19c
Improve position information.
...
R=paulberry@google.com
Review-Url: https://codereview.chromium.org/2915123002 .
2017-06-02 15:14:53 +02:00
Peter von der Ahé