2017-03-20 14:20:22 by Makoto Fujiwara | Files touched by this commit (1) |
Log message:
Add following lines for make test
BUILD_DEPENDS+= p5-Sub-Info-[0-9]*:../../devel/p5-Sub-Info
BUILD_DEPENDS+= p5-Term-Table-[0-9]*:../../devel/p5-Term-Table
|
2017-01-25 13:17:08 by Thomas Klausner | Files touched by this commit (2) |
Log message:
Updated p5-Params-ValidationCompiler to 0.23.
0.23 2017-01-23
- Trying to create a validator for positional parameters where a required
param came after one with a default did not throw an exception.
- A positional params validator with a slurpy type which had coercions did not
return the coerced values. It returned the original values instead.
|
2017-01-01 15:48:24 by Thomas Klausner | Files touched by this commit (2) |
Log message:
Updated p5-Params-ValidationCompiler to 0.22.
0.22 2016-12-31
- Explicitly load the B module. Previously the code relied on this already
being loaded by something else. Fixed by Tomasz Konojacki. PR #11.
- Removed the alpha warning from the docs.
|
2016-12-12 10:45:19 by Thomas Klausner | Files touched by this commit (2) |
Log message:
Updated p5-Params-ValidationCompiler to 0.21.
0.21 2016-12-06
- Switched to using GitHub issues.
0.20 2016-12-05
- The keys for parameter specifications are now validated. If an unknown key
is seen then an exception will be thrown. This will help you catch typos in
your parameter specification. Implemented by Greg Oschwald. PR #8.
|
2016-11-28 14:52:39 by Thomas Klausner | Files touched by this commit (2) |
Log message:
Updated p5-Params-ValidationCompiler to 0.19.
0.19 2016-11-21
- Non-inlinable Specio types caused a syntax error when used with positional
params.
- Positional params with coercions and defaults did not work properly. The
coerced value and the default would simply not be returned in any case.
|
2016-11-16 11:23:55 by Thomas Klausner | Files touched by this commit (2) |
Log message:
Updated p5-Params-ValidationCompiler to 0.18.
0.18 2016-11-13
- Using coercions with positional parameters could cause a "Modification of a
read-only value attempted" exception when the generated code tried to assign
to elements of @_. This is now fixed by making a copy if any of the types
have a coercion.
- Using Moose types with coercions in a positional params check would cause
invalid code to be generated. This could also happen with Type::Tiny if
either the type or a coercion could not be inlined.
|
2016-11-07 11:45:25 by Thomas Klausner | Files touched by this commit (2) |
Log message:
Updated p5-Params-ValidationCompiler to 0.17.
Add missing test dependencies.
0.17 2016-11-04
- When using positional parameters, parameters with a default are now
optional. For named parameters, this was already the case.
0.16 2016-11-03
- Moose and Specio types (and coercions) which provide variables to close over
when being inlined did not always compile properly. Most notable, this was
not being handled at all for Moose types, and not completely handled for
Specio coercions.
0.15 2016-11-03
- Previously, using a default with a positional parameter would result in an
error when compiling the validator subroutine. Defaults now work with
positional parameters. Implemented by Greg Oschwald. Based on PR #5.
0.14 2016-11-02
- Added a "named_to_list" option to support returning only the parameter
values from a named parameter validator rather than the key-value
pairs. Implemented by Greg Oschwald. Based on PR #4.
- Errors from calls to validation_for() now use croak so as to show up at the
call site, rather than in the internals
|
2016-10-01 14:54:08 by Thomas Klausner | Files touched by this commit (1) |
Log message:
Add some missing DEPENDS.
Add some missing test dependencies, mostly commented out since not
yet packaged.
|
2016-09-18 08:32:17 by Makoto Fujiwara | Files touched by this commit (3) |
Log message:
Import p5-Params-ValidationCompiler-0.13 as devel/p5-Params-ValidationCompiler.
Create a customized, optimized, non-lobotomized, uncompromised, and
thoroughly specialized parameter checking subroutine.
(make test is NOT confirmed yet)
|