| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|
|
|
|
|
| |
Class loaders have problems with referring to static inner classes (e.g.
state classes) when they cannot load the declaring class. To deal with
this the components have been moved to their own packages and state/rpc
classes will be defined in their own Java files.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
| |
conversion from Property classes
|
|
|
|
|
| |
string was interpreted as a valid date
Also added test for the Resolution enum static methods
|
| |
|
| |
|
| |
|
| |
|
| |
|
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
src/com/vaadin/Application.java
src/com/vaadin/data/Property.java
src/com/vaadin/data/util/PropertyFormatter.java
src/com/vaadin/terminal/ParameterHandler.java
src/com/vaadin/terminal/URIHandler.java
src/com/vaadin/terminal/gwt/client/HistoryImplIEVaadin.java
src/com/vaadin/terminal/gwt/client/ui/layout/ChildComponentContainer.java
src/com/vaadin/terminal/gwt/server/ApplicationPortlet.java
src/com/vaadin/terminal/gwt/server/PortletApplicationContext.java
src/com/vaadin/ui/Field.java
src/com/vaadin/ui/LoginForm.java
src/com/vaadin/ui/UriFragmentUtility.java
|
| |
| |
| |
| | |
svn changeset:22317/svn branch:6.7
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Eliminate the redundant method Validator.isValid(). Validator.validate()
does the same but throws an exception including the validation error
message.
Migration:
- replace implementations of isValid() with internalIsValid() in custom
subclasses of AbstractValidator
- use Validator.validate() and catch exceptions rather than call
Validator.isValid()
- note that AbstractField still provides an isValid() method that is
internally based on validate()
|
| |
| |
| |
| | |
Eliminate some more warnings for use of raw types (Field, Property,
Class), mostly by changing API method return types.
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The interfaces Property and Field and most of their implementations are
now parameterized with the type of their value. The method getValue()
returns the Property type but setValue() takes Object as its parameter.
No implicit conversions between value type and strings are performed in
most locations (fields).
Among others, AbstractTextField, DateField and RichTextArea not have
specific value types and do not accept arbitrary values.
Most locations requiring migration will be visible as compilation
errors, with the exception of some cases where a non-parameterized
Property or Field (or one parametrized with Object) is used.
Not yet done:
- Label
- converters
- setValue() parameterization (causes much more migration effort)
|
|
|
|
|
|
| |
because it has an invalid date string
svn changeset:20851/svn branch:6.7
|
|
|
|
| |
svn changeset:20536/svn branch:6.7
|
|
|
|
|
|
| |
other fields and AbstractField declaration of default for invalidAllowed
svn changeset:18480/svn branch:6.6
|
|
|
|
| |
svn changeset:16851/svn branch:6.5
|
|
|
|
| |
svn changeset:16830/svn branch:6.5
|
|
|
|
| |
svn changeset:16816/svn branch:6.5
|
|
|
|
|
|
| |
error. Kind of a built in validator that is run even though the value of the field is null.
svn changeset:16808/svn branch:6.5
|
|
|
|
| |
svn changeset:16806/svn branch:6.5
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Changes summarized:
- removed invalid implementation of isEmpty(). Can't work with validation/buffering. Added in #4311, fixes #5927
- removed invalid overridden implementation of valueChange. Added in #5810 and #5277.
- moved a lot of flag updates and other internal field updates to setInternalValue. Also tasks removed from valueChange(), now possible since #6002 is fixed.
- n non-reported bugs fixed (at least some related to read buffering)
- m regressions caused just a moment before large milestone?
Trivial minor enhancements to improve maintainability:
- removed obsolete toString() method (same as in AbstractField)
- added <?> to getType() method as in super implementation, fixes a compilation warning
- decreased visibility of setValue(Object, boolean), should be exposed to extending classes only - probably opened accidentally
- removed obsolete implementation of setValue(Object), did exactly the same thing as the parent
- javadoc typo
Sorry for a large patch. Don't like these either. I just don't see how this could have been fixed in smaller patches (except for the trivial fixes).
svn changeset:16797/svn branch:6.5
|
|
|
|
| |
svn changeset:16667/svn branch:6.5
|
|
|
|
| |
svn changeset:16482/svn branch:6.5
|
|
|
|
|
|
| |
button and select for one-to-one mapping
svn changeset:15867/svn branch:6.5
|
|
|
|
| |
svn changeset:15371/svn branch:6.5
|
|
|
|
| |
svn changeset:14128/svn branch:6.4
|
|
|
|
| |
svn changeset:14022/svn branch:6.4
|
|
|
|
| |
svn changeset:13724/svn branch:6.4
|
|
|
|
| |
svn changeset:13359/svn branch:6.4
|
|
|
|
| |
svn changeset:12055/svn branch:6.3
|
|
|
|
| |
svn changeset:12051/svn branch:6.3
|
|
|
|
| |
svn changeset:12004/svn branch:6.3
|
|
|
|
|
|
|
|
| |
* #4263 DateField - returning valid Date in handleUnparsableDateString method works only at the first attempt
and
* #4304 Resetting the value in DateField.handleUnparsableDate does not update the client side text
svn changeset:11820/svn branch:6.3
|
|
|
|
| |
svn changeset:10131/svn branch:6.2
|
|
|
|
| |
svn changeset:9624/svn branch:6.2
|
|
|
|
| |
svn changeset:9622/svn branch:6.2
|
|
|
|
| |
svn changeset:9095/svn branch:2009-09-widget-packaging_3332
|
|
|
|
|
|
| |
refining.
svn changeset:8930/svn branch:2009-09-widget-packaging_3332
|
|
|
|
| |
svn changeset:8200/svn branch:6.0
|
|
|
|
| |
svn changeset:7984/svn branch:6.0
|
|
|
|
| |
svn changeset:7854/svn branch:6.0
|