On Thu, 15 Oct 2020 14:30:54 GMT, Hannes Wallnöfer <hann...@openjdk.org> wrote:
>> Jonathan Gibbons has updated the pull request incrementally with one >> additional commit since the last revision: >> >> Remove redundant field >> (There is an equivalent field in `HtmlDocletWriter`) > > src/jdk.javadoc/share/classes/jdk/javadoc/internal/doclets/toolkit/util/IndexBuilder.java > line 169: > >> 167: // don't put summary-page items in the A-Z index: >> 168: // they are listed separately, at the top of the index page >> 169: >> itemsByFirstChar.computeIfAbsent(keyCharacter(item.getLabel()), > > You're preventing these summary pages from showing up in the alphabetic > index, but not from the by-category index > below. This means they are added to the javadoc search contents, which by > itself is a good thing. However, this means > we also need to think about the category under which the item is listed. For > example, you are using > `IndexItem.Category.MEMBER` for the Deprecated API list, but deprecation > applies to all element categories. I guess we > should use the `TAGS` category for summary pages unless there is a clear > element category they fit in? Initially, I put the summary pages in the alphabetic index as well, but it disrupted test files too much. It was sort-of silly to see `All Classes` and `All Packages` showing up under `A` in the alphabetic index when previously there was no `A` entry. Hence the change to the current behavior. See next comment about which category to use. > src/jdk.javadoc/share/classes/jdk/javadoc/internal/doclets/toolkit/util/Utils.java > line 2662: > >> 2660: * The entries may come from the AST and DocCommentParser, or may >> be autromatically >> 2661: * generated comments for mandated elements and JavaFX properties. >> 2662: ** @see CommentUtils#dcInfoMap > > I think you unintentionally removed a line break while fixing the reference. Oops ------------- PR: https://git.openjdk.java.net/jdk/pull/643
