Michael Miller
ef64fa466b
Avoid type cast issues
2022-03-30 20:44:43 -06:00
Michael Miller
39f5ea34eb
Another place that needs a cast
2022-03-30 20:33:19 -06:00
Michael Miller
613c969fbb
Match RSpec behavior
2022-03-30 20:28:36 -06:00
Michael Miller
a8e55e32d2
Works now due to lazy doubles
2022-03-30 19:13:14 -06:00
Michael Miller
1327d74040
Fix handling of stubs with the same name
2022-03-26 00:13:24 -06:00
Michael Miller
1199379833
Various fixes
2022-03-26 00:12:37 -06:00
Michael Miller
bbaa8816b8
Weird segfault when inspecting value type is back.
...
Working around with a crude rescue block.
2022-03-25 23:54:44 -06:00
Michael Miller
28ae03d56f
Remove expect_null_double
...
Workaround no longer needed.
2022-03-25 22:45:13 -06:00
Michael Miller
b28437ccc4
Match RSpec behavior
2022-03-25 22:32:27 -06:00
Michael Miller
28488d308e
Don't invoke fallback if there are stubs for the method
2022-03-25 22:30:39 -06:00
Michael Miller
fc51c1d730
More logging
2022-03-24 02:39:15 -06:00
Michael Miller
b97a367be7
Method return type restriction takes precedence over original return type
...
Raise TypeCastError on ambiguous stub values.
2022-03-24 02:01:25 -06:00
Michael Miller
5aceaf7d8c
Comment out problematic casts
...
Specs compile again.
2022-03-22 22:13:59 -06:00
Michael Miller
7dd2cffa8e
Inspecting the value's class seems to work now
2022-03-22 22:13:02 -06:00
Michael Miller
d6fcf128ba
Temporarily disable returning un-cast value
2022-03-22 22:11:34 -06:00
Michael Miller
78dea01abe
Manual method existence check
2022-03-20 21:49:21 -06:00
Michael Miller
0b6e44c227
Fix types of arguments given to common object methods
2022-03-20 15:39:19 -06:00
Michael Miller
0fb7be6722
Use stub instead of abstract_stub by default for stub_all
...
RSpec doesn't appear to raise unexpected message errors for standard object methods (like to_s).
There are also a lot of issues that arise when requiring a stub on common methods.
2022-03-20 15:38:26 -06:00
Michael Miller
c9f91299fe
Merge branch 'master' into mock-redesign
2022-03-19 23:01:52 -06:00
Michael Miller
419e7c1b8d
Use Box for Wrapper implementation
...
This won't raise on invalid casts.
It should reduce the amount of instantiated types and methods.
2022-03-19 23:00:32 -06:00
Michael Miller
e209fe3eb5
Use Box for Wrapper implementation
...
This won't raise on invalid casts.
It should reduce the amount of instantiated types and methods.
2022-03-19 22:53:50 -06:00
Michael Miller
18a03b02fe
Include reserved
2022-03-19 22:36:25 -06:00
Michael Miller
03a031ad38
Use absolute path to RESERVED_KEYWORDS
2022-03-19 22:00:12 -06:00
Michael Miller
b23a567ee3
Add require statements
2022-03-19 21:56:58 -06:00
Michael Miller
04cf211f24
Move reserved keywords to a separate file
...
Add finalize as a reserved keyword.
2022-03-19 21:56:18 -06:00
Michael Miller
04f3614ff7
Workaround issue with value not being found
...
For some reason the `value` method isn't found.
Event casting to TypedStub doesn't work.
2022-03-19 20:00:50 -06:00
Michael Miller
bed84b315d
Add lazy double
2022-03-19 19:32:41 -06:00
Michael Miller
162ad4df33
Report call and args like other error messages
2022-03-19 18:58:06 -06:00
Michael Miller
1657d3577d
Fix test of double name
2022-03-19 18:57:39 -06:00
Michael Miller
b3fa2e3bfe
Address Ameba issues
2022-03-19 17:22:13 -06:00
Michael Miller
b83c6b7b1e
Remove or skip mock, double, and stub tests
2022-03-19 17:20:15 -06:00
Michael Miller
0c7f72dc78
Test handling of blocks in doubles
2022-03-19 17:04:51 -06:00
Michael Miller
043ebd818f
Fix argument order
2022-03-19 16:42:14 -06:00
Michael Miller
2f1999b377
Fix call expansion in debug log
...
Macro expansion somehow generated an error when calling a missing method with a block.
```
syntax error in :1
Error: unknown token: 'n'
```
It seems like 'n' is from the '\n' that was previously generated:
```
Log.debug { "Got undefined method " + "baz do\n yield\nend" }
```
2022-03-19 16:32:25 -06:00
Michael Miller
17e97cb39a
Explode method expansion
...
Somewhere along the way the block argument gets dropped.
Even though `method` is a `Def` that accepts a block, the `&block` portion is dropped.
Possible Crystal compiler bug.
2022-03-19 16:13:11 -06:00
Michael Miller
15dd2ea6f1
Inject stubbable method that accepts a block
2022-03-19 16:09:35 -06:00
Michael Miller
99aa8afdce
Add as_null_object method
2022-03-19 14:41:45 -06:00
Michael Miller
5d21e4bb71
Remove double_dsl_spec
2022-03-19 14:27:00 -06:00
Michael Miller
70787ce2d4
Extend NullDouble from Double
2022-03-19 13:00:42 -06:00
Michael Miller
3fa6baea4d
Test mixing block and keyword stubs
2022-03-19 12:52:23 -06:00
Michael Miller
d40040c8ee
Remove extra block body leftover from debug
...
Update docs to show example of default method implementation.
2022-03-19 12:37:46 -06:00
Michael Miller
25f546245a
Handle injecting stubbed abstract method
2022-03-19 12:33:09 -06:00
Michael Miller
9091b8d027
Literally inject stubs from double block
2022-03-19 12:28:00 -06:00
Michael Miller
5a4e95e8a2
Split double macro into def_double and new_double
2022-03-19 11:41:53 -06:00
Michael Miller
f1daecd25f
Initial spec for double DSL
2022-03-19 11:41:29 -06:00
Michael Miller
4dff018330
Pass along block only if it exists
2022-03-19 11:28:36 -06:00
Michael Miller
a508aeb4a3
Use a shared macro to handle stub type cast
2022-03-17 02:45:50 -06:00
Michael Miller
751ea77172
Add docs explaining how stub macros work
2022-03-17 02:18:32 -06:00
Michael Miller
4f7a4a3453
Avoid issue with TypeCastError message
...
For some reason, attempting to get the class of the stubbed value causes a segfault.
This appears to get swallowed in some cases.
Trying to inspect the value was difficult without triggering some error.
In the failing case, the value is 42 (returned by a stubbed foo method).
The method is supposed to return a String, and this error checking should be triggered (line 119).
The value is a union, and inspecting the bytes contained in it looks like it's prefixed with the Crystal type ID for Int32.
However it looks like Crystal tries to treat it as a Bool? I'm not sure.
Very weird behavior and instead of wasting more time on it, I'm omitting the returned type from the error.
This type of casting error shouldn't be common.
This seems like a Crystal bug, but I'm having a hard time distilling the issue.
2022-03-16 22:40:24 -06:00
Michael Miller
b68c75dda5
Add tests around problematic issues in NullDouble to Double
2022-03-16 21:10:47 -06:00
