Hi, Petr.
Not an expert here, just a few notes.
- I guess textTypeToNative must be initialized only once and can be
replaced by Collections.unmodifiableMap(q);
- Please check all related tests on all supported platforms, since we
have a lots of regressions here.
On 4/21/14 5:39 PM, Petr Pchelko wrote:
Hello, Anthony.
I've updated the fix: http://cr.openjdk.java.net/~pchelko/9/8027148/webrev.01/
2. I'm not sure if the new formatting for htmlDocumntTypes at new lines 887 -
889 looks better than the old one.
Reverted.
1. src/share/classes/java/awt/datatransfer/SystemFlavorMap.java
118 private Map<String, LinkedHashSet<DataFlavor>> getNativeToFlavor() {
Usually we use a generic interface, such as Set, instead of a concrete
implementation class (LinkedHashSet) in generic types declarations to avoid
dependencies on concrete implementations that may change in the future. Would
that be possible to do the same for method return type declarations in this
file?
I've done this on purpose here. It is VERY important for the collection used
here to have a predictable iteration order, but it must not be sorted. We do
not have any generic interface for this type of Set, so I've decided to place
an exact class here so that nobody could by accident use a wrong Set and break
everything.
If a generic Set would be used someone could easily change it to HashSet
somewhere in the overrides and not notice the bug.
3. In setNativesForFlavor() and addFlavorForUnencodedNative() we used to
remove(null) from the getNativesForFlavorCache and getFlavorsForNativeCache
caches. Now we don't. What is the story behind the nulls? How did they get
there previously, and how do we avoid them now?
I've made a wrapper class for this cache (see the bottom of the file). With the
wrapper we can create the cache lazily and the cache logic is now in one place.
With best regards. Petr.
On 21.04.2014, at 17:20, Anthony Petrov <[email protected]> wrote:
Hi Petr,
1. src/share/classes/java/awt/datatransfer/SystemFlavorMap.java
118 private Map<String, LinkedHashSet<DataFlavor>> getNativeToFlavor() {
Usually we use a generic interface, such as Set, instead of a concrete
implementation class (LinkedHashSet) in generic types declarations to avoid
dependencies on concrete implementations that may change in the future. Would
that be possible to do the same for method return type declarations in this
file?
2. I'm not sure if the new formatting for htmlDocumntTypes at new lines 887 -
889 looks better than the old one.
3. In setNativesForFlavor() and addFlavorForUnencodedNative() we used to
remove(null) from the getNativesForFlavorCache and getFlavorsForNativeCache
caches. Now we don't. What is the story behind the nulls? How did they get
there previously, and how do we avoid them now?
Otherwise the fix looks okay. Note that I'm not an expert in this code, so I
may have missed some issues in the logic changes.
--
best regards,
Anthony
On 4/18/2014 5:49 PM, Petr Pchelko wrote:
Hello, AWT Team.
Please review the fix for the issue:
https://bugs.openjdk.java.net/browse/JDK-8027148
The fix is available at:
http://cr.openjdk.java.net/~pchelko/9/8027148/webrev/
Sorry for the long text, but it’s quite tangled)
The problem:
The flavor map contains some predefined mappings which are stored in the
flavormap.properties file. But we can extend these mappings using
addUnencodedNativeForFlavor method. Javadoc states, that these new mappings would have
lower priority that standard mappings. But in the current implementation this was not the
case, because getNativesForFlavor method relied on the fact, that standard text mappings
were stored as FlavorBaseType<->Native and newly added mappings were stored as
DataFlavor<->Native, but after some fix in Java 8 this is not the case any more.
Everything is stored as a DataFlavor as a key. This is important only for text flavors,
because we support different text charsets and can reencode the text on the fly. So each
native text format could be represented in many different DataFlavors with different
encodings and representation classes. When we generate the set of DataFlavor’s that a text
native can be translated to we no longer know how to distinguish the standard mappings and
additional mappings and the get shuffled when we generate missing mappings for text formats.
The solution:
I’ve added an additional map for standard text mappings. With this map we can now take
natives for mime types directly and not "find all flavors for a mime-type and than
all natives for each flavor". This is not only faster, but we can distinguish
standard text mappings from custom and return the list in the correct order. The new hash
map contains only a few elements.
Also I’ve replaced the ArrayList as a collection of natives for a particular
Flavor with a LinkedHashSet, because this list could not contain duplicated
which we were enforcing ourselves. Now it works out of the box and some code
can be removed.
I’ve measured the performance of a couple of most hot methods and on average
the new implementation is 1.7 times faster.
The test is being open sources.
I’ve tested this with JCK and our regression tests, everything looks good. Also
I’ve tested with a couple of hand-made toys.
Thank you.
With best regards. Petr.
--
Best regards, Sergey.
