RFR: JDK-8254627: Cleanup {Abstract, Single, Split}IndexWriter classes
Jonathan Gibbons
jjg at openjdk.java.net
Tue Oct 13 15:31:13 UTC 2020
On Tue, 13 Oct 2020 10:52:56 GMT, Hannes Wallnöfer <hannesw at openjdk.org> wrote:
>> Inspired by other recent work to cleanup index generation, this change is primarily about cleanup for
>> `AbstractIndexWriter`, `SingleIndexWriter`, `SplitIndexWriter` to reduce code duplication and to reduce the 3 class to
>> just 1 that can handle either single index files or split index files. As part of this cleanup, some anomalies were
>> uncovered in the naming of annotation types, enum constants and records, which have also been addressed.
>> The changes involved moving/merging code from `SingleIndexWriter` and `SplitIndexWriter` in to `AbstractIndexWriter`,
>> which is now no longer abstract and is thus renamed to just `IndexWriter`. Regrettably, the rename is not properly
>> tracked.
>
> src/jdk.javadoc/share/classes/jdk/javadoc/internal/doclets/formats/html/IndexWriter.java line 132:
>
>> 130: Content mainContent = new ContentBuilder();
>> 131: List<Character> allFirstCharacters = new ArrayList<>(mainIndex.getFirstCharacters());
>> 132: allFirstCharacters.sort(Comparator.naturalOrder());
>
> It seems a bit redundant to retrieve and sort the `allFirstCharacters` list in each invocation of `generateIndexFile`,
> maybe pass it as an argument from the `generate(HtmlConfiguration)` method?
Given your other comment (listed next in this page) that the sort is actually unnecessary, it may be not so important
to pass in the list of all first characters from the `generate` method, but it is easy enough to do so. I will check it
out.
-------------
PR: https://git.openjdk.java.net/jdk/pull/621
More information about the javadoc-dev
mailing list