Hide private enums from documentation

[quirkylemon]

This loops through all the enums and checks if it has a description (doc comment) or if it is public. it also renames the variable e to a more descriptive name
fixes #84119

[quirkylemon]

I think it is working properly

[daveTheOldCoder]

I tested this PR and it worked correctly.

[Mickeon]

Tested myself and works as intended.

[dalexeev]

As far as I understand, it is not guaranteed that cd.enums has key. Above:

if (cd.enums.has(key)) {
    if (cd.enums[key].description.strip_edges().is_empty() && E.key.begins_with("_")) {
        continue;
    }
}

[quirkylemon]

So should we still be checking if cd.enums.has(key) or should we skip any enum that doesn't have a key?
Edit: if I understand it correctly cd.enums will always have key

[AThousandShips]

Above the enums are filtered on script docs, but not here, is that an oversight?

Suggested change

	if (cd.enums[key].description.strip_edges().is_empty() && E.key.begins_with("_")) {
	if (cd.is_script_doc && cd.enums[key].description.strip_edges().is_empty() && E.key.begins_with("_")) {

[quirkylemon]

correct me if I'm wrong but we shouldn't have to check cd.is_script_doc every time because it doesn't change per enum.
the reason we are checking if the description is empty or if the name starts with "_" is it is possible for enums that don't need documentation generated to be in the for loop.

[AThousandShips]

Then there should be some other flag to control this as the filtering is on script docs only

[quirkylemon]

oh I didn't realize we only filter on script docs.

[AThousandShips]

Suggested change

	bool has_enums = false;
	if (enums.size()) {
	for (KeyValue<String, DocData::EnumDoc> &E : cd.enums) {
	if (cd.is_script_doc && cd.is_script_doc && E.key.begins_with("_") && E.value.description.strip_edges().is_empty()) {
	bool has_enums = enums.size() && !cd.is_script_doc;
	if (enums.size() && !has_enums) {
	for (KeyValue<String, DocData::EnumDoc> &E : cd.enums) {
	if (E.key.begins_with("_") && E.value.description.strip_edges().is_empty()) {

No need to check these every time, the loop accepts enums if they're non-script doc ones, so we can skip this part

Might be a more elegant way to do this just a suggestion, but the loop is unnecessary if it's not a script doc

[quirkylemon]

yeah you're right.

[AThousandShips]

LGTM, would need to verify the comment about the enum key though, I'd suggest keeping it as it was to avoid any weirdness

[quirkylemon]

Should I go ahead and combine the commits?
Also what does "LGTM" mean?

[AThousandShips]

Looks Good To Me :)

Yes please squash

[quirkylemon]

whenever I do git rebase it says there are 1382 commits to push. this has happened before and it didn't mess anything up but it did send a message in every pr that was in it. Is there any way to fix it?

[AThousandShips]

It's just because you're updating your branch, the others should be just the normal ones, just that the branch doesn't contain them upstream

[quirkylemon]

Squashed!!
and sorry about all the issues.

[AThousandShips]

No problem at all! It's a learning experience :)

[akien-mga]

Apologies for yet another nitpick, but could you amend the commit message to be a bit more explicit (and ideally start with a capital letter like other commits)? The title of this PR would be a good option, for a changelog reader I think the "from documentation" part is important to get the context.

I can do it myself too if needed.

[quirkylemon]

Done

[akien-mga]

Thanks!