Article on lambdas #32

Eczbek · 2025-04-12T00:23:14Z

No description provided.

jeremy-rifkin

Thank you so much for taking the time to put this together! This looks really good! I am about halfway through reading, dropping some comments below.

jeremy-rifkin · 2025-04-12T02:29:13Z

src/.vitepress/config.mts

                        { text: "Standards", link: "/resources/general/standards" },
+                        {
+                            text: "C++ Resources",
+                            items: [{ text: "Lambdas", link: "/resources/general/cpp/lambdas" }],


Suggested change

items: [{ text: "Lambdas", link: "/resources/general/cpp/lambdas" }],

items: [{ text: "Lambdas", link: "/resources/cpp/lambdas" }],

jeremy-rifkin · 2025-04-12T02:30:08Z

src/resources/cpp/lambdas.md

@@ -0,0 +1,365 @@
+## Abstract


Every page should start with an h1, that's how pages get page titles. E.g.

vs

I would also say let's not make an "abstract" section, this isn't a paper :)

jeremy-rifkin · 2025-04-12T02:34:24Z

src/resources/cpp/lambdas.md

+captures by reference), making lambdas ideal for short, context-dependent operations like custom comparisons, filters,
+or event handlers.
+
+For example:


This example comes after a couple paragraphs on captures, but doesn't use any captures. Maybe it'd be good to show this code example early on, then talk about captures, then show an example using captures?

jeremy-rifkin · 2025-04-12T02:35:41Z

src/resources/cpp/lambdas.md

+For example:
+
+```cpp
+// Define a function in namespace scope


I'm inclined to say this comment and others can be dropped, however, if anyone strongly feels this would be helpful for beginners reading the article then it's ok to keep. I think this article is a bit more of a deep-dive than beginner content, though.

The comments are meant to emphasize and explain the difference between the code snippets.

I think it's pretty clear from just reading the code. Thoughts?

jeremy-rifkin · 2025-04-12T02:36:40Z

src/resources/cpp/lambdas.md

+}
+```
+
+&nbsp;


Please don't try to manually insert space like this :) This will introduce inconsistency with how the site presents itself.

I use this as a break between paragraphs. Should I use something else or avoid breaks entirely?

I’d say avoid them entirely. The site styling already handled spacing between paragraphs and headers.

jeremy-rifkin · 2025-04-12T02:37:33Z

src/resources/cpp/lambdas.md

+The basic syntax of a lambda expression looks like this:
+
+```
+[captures](params) -> ReturnType {


Suggested change

[captures](params) -> ReturnType {

[captures](params) -> return_type {

jeremy-rifkin · 2025-04-12T02:41:14Z

src/resources/cpp/lambdas.md

+
+&nbsp;
+
+Each lambda expression has its own unique, unnameable type:


I'd recommend moving this part to a later section as it's an implementation detail that doesn't pertain to most uses of lambdas

Where should I move this to? Also, I think showing that a lambda is really similar to a struct with an operator() is useful early on, so should I write something else here instead?

Maybe a later section on implementation details

jeremy-rifkin · 2025-04-12T02:43:36Z

src/resources/cpp/lambdas.md

+- `[x...]` - Capture a pack `x` by value.
+- `[&x...]` - Capture a pack `x` by reference.


This is a cool bit of trivia and I had to go check the grammar about it. While this is technically its own special syntax in the standard, it's not different from a user perspective. Maybe these could be moved to a note along the lines of "Captures even work with packs!"

jeremy-rifkin · 2025-04-12T02:46:05Z

src/resources/cpp/lambdas.md

+	}();
+}();
+```
+


I think another cool thing to mention in this section about captures would be that lambdas under the hood only capture what they actually use, so [&] and [=] can be used without fear of bloat or overhead

jeremy-rifkin · 2025-04-12T02:47:26Z

src/resources/cpp/lambdas.md

+Lambda parameters work the same way as normal function parameters, and `auto` parameters make a lambda's `operator()`
+implicitly templated.
+
+If a lambda takes no parameters, the parameter list may be omitted entirely:


I'd probably move this to the end of the section, just so things flow more naturally (discussion about parameters followed by an example of parameters instead of a discussion of parameters followed by an example of no parameters)

jeremy-rifkin · 2025-04-12T16:13:52Z

src/resources/cpp/lambdas.md

+- `constexpr` - Explicitly specifies that a lambda's `operator()` is a
+  [constexpr function](https://en.cppreference.com/w/cpp/language/constexpr#constexpr_function).
+  - Mutually exclusive with `consteval`.
+  - Lambdas are implicitly marked `constexpr`, if possible.


I think it'd be helpful to link to the criteria that determines if this is possible, either directly or as a footnote

jeremy-rifkin

I gave the rest a read, just a couple comments

jeremy-rifkin · 2025-04-12T16:18:27Z

src/resources/cpp/lambdas.md

+- `static` - Makes a lambda's `operator()` a
+  [static member function](https://en.cppreference.com/w/cpp/language/static#Static_member_functions).
+  - Mutually exclusive with `mutable`.
+  - Cannot be used if the captures list is not empty, or an explicit `this` parameter is present.


I think it would be helpful to describe why this might be useful. Also it might be worth noting the lambda -> function pointer trick in some other section, and how this works even without static:

auto* ptr = +[]{ return 4; };

jeremy-rifkin · 2025-04-12T16:20:55Z

src/resources/cpp/lambdas.md

+Lambdas may be given attributes that apply to their `operator()`s since C++23:
+
+```cpp
+auto very_important_number = [][[nodiscard]] { return 4; };


Very minor style nit but I'd suggest putting a space between the attribute, and the capture block, otherwise it's just a big block of square brackets [][[

Suggested change

auto very_important_number = [][[nodiscard]] { return 4; };

auto very_important_number = [] [[nodiscard]] { return 4; };

jeremy-rifkin · 2025-04-12T16:23:56Z

src/resources/cpp/lambdas.md

+
+&nbsp;
+
+### Capturing function parameters


I think this section is way too niche :)

jeremy-rifkin · 2025-04-12T16:29:12Z

src/resources/cpp/lambdas.md

+Making a type alias with an in-line lambda in a header file or module interface violates the
+[One-Definition Rule](https://en.cppreference.com/w/cpp/language/definition) because aliases are not a "definable item",
+so lambdas in them are not allowed to match with other lambda declarations in other
+[translation units](https://en.cppreference.com/w/cpp/language/translation_phases#Translation_process):


While I can imagine problems here I'm not seeing why this is an ODR violation, because this isn't a definable https://eel.is/c++draft/basic.def.odr#15 doesn't apply. This is an ODR violation though: https://eel.is/c++draft/basic.def.odr#18.

jeremy-rifkin · 2025-04-12T16:37:09Z

src/resources/cpp/lambdas.md

+
+&nbsp;
+
+### In-line partial specialization


I'm afraid I don't see why this is useful. It's also not really a specialization, which can be most clearly seen in the index sequence example. I would recommend keeping that example since that is a common useful pattern.

jeremy-rifkin · 2025-07-12T20:47:55Z

Converting to draft until ready

Create article about lambdas

81a7618

Eczbek force-pushed the main branch from a627f20 to 81a7618 Compare April 12, 2025 00:41

jeremy-rifkin reviewed Apr 12, 2025

View reviewed changes

jeremy-rifkin marked this pull request as draft July 12, 2025 20:47

	items: [{ text: "Lambdas", link: "/resources/general/cpp/lambdas" }],
	items: [{ text: "Lambdas", link: "/resources/cpp/lambdas" }],

+              }
+              ```
+              &nbsp;

	[captures](params) -> ReturnType {
	[captures](params) -> return_type {




		Each lambda expression has its own unique, unnameable type:

		- `[x...]` - Capture a pack `x` by value.
		- `[&x...]` - Capture a pack `x` by reference.

	auto very_important_number = [][[nodiscard]] { return 4; };
	auto very_important_number = [] [[nodiscard]] { return 4; };

Article on lambdas #32

Are you sure you want to change the base?

Article on lambdas #32

Uh oh!

Conversation

Eczbek commented Apr 12, 2025

Uh oh!

jeremy-rifkin left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jeremy-rifkin left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jeremy-rifkin commented Jul 12, 2025

Uh oh!

Uh oh!