Skip to content
This repository has been archived by the owner on Aug 8, 2023. It is now read-only.

Refactor explicit operator bool usage for some cases #8601

Merged
merged 4 commits into from
Apr 14, 2017

Conversation

brunoabinader
Copy link
Member

No description provided.

@brunoabinader brunoabinader added the Core The cross-platform C++ core, aka mbgl label Mar 31, 2017
@brunoabinader brunoabinader changed the title [all] Prefer 'valid()' over 'explicit operator bool()' [all] Prefer valid() over explicit operator bool() Mar 31, 2017
@brunoabinader brunoabinader changed the title [all] Prefer valid() over explicit operator bool() [all] Prefer valid() over explicit operator bool() Mar 31, 2017
@brunoabinader brunoabinader force-pushed the rip-explicit-operator-bool branch 2 times, most recently from 54ba98f to b411b8b Compare March 31, 2017 15:08
Copy link
Contributor

@jfirebaugh jfirebaugh left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

So, to add some additional nuance to #8562 (comment), not all our uses of operator bool() are implementing a validity check (though some of them definitely are). I would distinguish:

  • Use of member data sentinel values to create an ad hoc implementation of an "absent" state, e.g. in PositionedIcon, Shaping, Glyph, GlyphMetrics. These should typically be replaced with uses of optional<T> and elimination of the "invalid" state from the inner type.
  • Types which are acting similarly to a pointer or optional type, e.g. ExtensionFunction, CFHandle. These could arguably use optional too, but sometimes it's not convenient. Emulating a pointer by providing operator bool() (and usually operator* as well) is reasonable.
  • Result. This is in its own category because it's a fundamental type, on the level of optional. Providing operator bool() is necessary for error-handling ergonomics.
  • Types which have some other special state, but where that state is considered a valid member of the type's domain. These should typically use a named method specific to the state, e.g. TransitionOptions::present() or PropertyValue::defined(). (Note that the most natural name might name the complement of the special state -- all members of the type's domain other than the special state.)

@@ -34,7 +34,7 @@ class PropertyValue {
const T & asConstant() const { return value.template get< T >(); }
const CameraFunction<T>& asCameraFunction() const { return value.template get<CameraFunction<T>>(); }

explicit operator bool() const { return !isUndefined(); };
bool valid() const { return !isUndefined(); };
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

defined or present would be better; an undefined property value isn't invalid.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is actually not used, it seems isUndefined() suits the purpose already.

@@ -18,7 +18,7 @@ class TransitionOptions {
};
}

explicit operator bool() const {
bool valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same here.

@@ -48,7 +48,7 @@ class LatLng {
else if (longitude < 0 && end.longitude > 0) longitude += util::DEGREES_MAX;
}

explicit operator bool() const {
bool valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

#3802 tracks removing the invalid state from these classes entirely.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sure, I'm going to defer these to #3802.

@@ -17,7 +17,7 @@ class Size {
return width * height;
}

constexpr explicit operator bool() const {
constexpr bool valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Reverse the sense and call it empty or zero instead?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Deferring this to #8562.

@@ -17,7 +17,7 @@ class Result : private variant<T, Error> {
public:
using variant<T, Error>::variant;

explicit operator bool() const {
bool valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this is a reasonable use of operator bool and doesn't need to change. Of course, we should probably follow d7227e1 and eliminate this copy of Result entirely.

@@ -158,8 +158,8 @@ Statement &Statement::operator=(Statement &&other) {
Statement::~Statement() {
}

Statement::operator bool() const {
return impl.operator bool();
bool Statement::valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ditto.

@@ -18,7 +18,7 @@ class ExtensionFunction<R(Args...)> {
ExtensionFunction(const ProcAddress ptr_) : ptr(ptr_) {
}

explicit operator bool() const {
bool valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this is a valid use of operator bool and doesn't need to change. ExtensionFunction acts like a function pointer.

@@ -23,34 +23,37 @@ SymbolInstance::SymbolInstance(Anchor& anchor,
const std::size_t featureIndex_) :
point(anchor.point),
index(index_),
hasText(shapedTextOrientations.first || shapedTextOrientations.second),
hasIcon(shapedIcon),
hasText(shapedTextOrientations.first.valid() || shapedTextOrientations.second.valid()),
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

A better fix here is likely to use an optional.

@@ -16,7 +16,7 @@ namespace mbgl {
GlyphRange getGlyphRange(char16_t glyph);

struct GlyphMetrics {
explicit operator bool() const {
bool valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If "all 0" is being used as a proxy for "no metrics available", a better change is to use optional<GlyphMetrics>.

@@ -42,8 +42,8 @@ struct Glyph {
explicit Glyph(Rect<uint16_t> rect_, GlyphMetrics metrics_)
: rect(std::move(rect_)), metrics(std::move(metrics_)) {}

explicit operator bool() const {
return metrics || rect.hasArea();
bool valid() const {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same here -- optional<Glyph> instead.

@brunoabinader
Copy link
Member Author

@jfirebaugh let's wait for #3802 and #8562 to land, so I can rebase and address the review comments in one step.

@brunoabinader brunoabinader force-pushed the rip-explicit-operator-bool branch 2 times, most recently from a9c52fd to e677563 Compare April 12, 2017 16:15
@brunoabinader brunoabinader changed the title [all] Prefer valid() over explicit operator bool() Refactor explicit operator bool usage for some cases Apr 12, 2017
@brunoabinader brunoabinader added the ⚠️ DO NOT MERGE Work in progress, proof of concept, or on hold label Apr 12, 2017
@brunoabinader
Copy link
Member Author

@jfirebaugh I've applied the requested changes based on your comments - thank you for the detailed explaining on these!

The only exception is Shaping, which starts failing some render tests once I replace the explicit operator bool usage with optional. I'm going to skip it for now and get back to it once I figure out exactly why in some cases the symbols are not being inserted into the buckets when the text shaping is not valid.

@brunoabinader brunoabinader removed the ⚠️ DO NOT MERGE Work in progress, proof of concept, or on hold label Apr 13, 2017
@@ -21,7 +21,7 @@ class SDFGlyph {
AlphaImage bitmap;

// Glyph metrics
GlyphMetrics metrics;
optional<GlyphMetrics> metrics;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Once Glyph is optional, GlyphMetrics::operator bool() is unused, so we don't need to make this optional.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have an alternate fix for glyph-related operator bool() prepared as part of #8747, so I'll omit the last commit here and merge.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you @jfirebaugh 👍

@jfirebaugh jfirebaugh merged commit 98e2e59 into master Apr 14, 2017
@jfirebaugh jfirebaugh deleted the rip-explicit-operator-bool branch April 14, 2017 19:33
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
Core The cross-platform C++ core, aka mbgl
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants