New top language algorithm implementation (#1732)

* Reduced vercel maxDuration

* Implemented new algorithm for Top Langs

* Revert "Reduced vercel maxDuration"

This reverts commit b0bc626efe12c738cf5005e7f11c7d2a07b6387a.

* Added documentation

* Fixed broken implementation

* Update fetchTopLanguages.test.js

Changed tests

* Now uses the general formula

The parameters p and q can be set by the user.

* Updated tests and added new test

* Added new test

New test for order by repo count.

* Updated documentation

Added explanation and examples for new options.

* Updated documentation

This was overwritten in the merge commit.

* docs: improve docs and fix tests

* Renamed parameters

Renamed `p` and `q` to `size_weight` and `count_weight`, respectively.

* Updated the documentation

Changes introduced in f2516d60a442dfdbb9e24ddda8743664bcb8064d

---------

Co-authored-by: rickstaa <rick.staa@outlook.com>

api/top-langs.js

@@ -25,6 +25,8 @@ export default async (req, res) => {
     layout,
     langs_count,
     exclude_repo,
+    size_weight,
+    count_weight,
     custom_title,
     locale,
     border_radius,
@@ -46,6 +48,8 @@ export default async (req, res) => {
     const topLangs = await fetchTopLanguages(
       username,
       parseArray(exclude_repo),
+      size_weight,
+      count_weight,
     );
 
     const cacheSeconds = clampValue(

readme.md

@@ -310,6 +310,8 @@ You can provide multiple comma-separated values in the bg_color option to render
 -   `custom_title` - Sets a custom title for the card _(string)_. Default `Most Used Languages`.
 -   `disable_animations` - Disables all animations in the card _(boolean)_. Default: `false`.
 -   `hide_progress` - It uses the compact layout option, hides percentages, and removes the bars. Default: `false`.
+-   `size_weight` - Configures language stats algorithm _(number)_ (see [Language stats algorithm](#Language-stats-algorithm)), defaults to 1.
+-   `count_weight` - Configures language stats algorithm _(number)_ (see [Language stats algorithm](#Language-stats-algorithm)), defaults to 0.
 
 > **Warning**
 > Language names should be URI-escaped, as specified in [Percent Encoding](https://en.wikipedia.org/wiki/Percent-encoding)
@@ -359,7 +361,25 @@ Use [show_owner](#customization) variable to include the repo's owner username
 The top languages card shows a GitHub user's most frequently used top language.
 
 > **Note**
-> Top Languages does not indicate my skill level or anything like that; it's a GitHub metric to determine which languages have the most code on GitHub. It is a new feature of github-readme-stats.
+> Top Languages does not indicate the user's skill level or anything like that; it's a GitHub metric to determine which languages have the most code on GitHub. It is a new feature of github-readme-stats.
+
+### Language stats algorithm
+
+We use the following algorithm to calculate the languages percentages on the language card:
+
+```js
+ranking_index = (byte_count ^ size_weight) * (repo_count ^ count_weight)
+```
+
+By default, only the byte count is used for determining the languages percentages shown on the language card (i.e. `size_weight=1` and `count_weight=0`). You can, however, use the `&size_weight=` and `&count_weight=` options to weight the language usage calculation. The values must be positive real numbers. [More details about the algorithm can be found here](https://github.com/anuraghazra/github-readme-stats/issues/1600#issuecomment-1046056305).
+
+-   `&size_weight=1&count_weight=0` - _(default)_ Orders by byte count.
+-   `&size_weight=0.5&count_weight=0.5` - _(recommended)_ Uses both byte and repo count for ranking
+-   `&size_weight=0&count_weight=1` - Orders by repo count
+
+```md
+[![Top Langs](https://github-readme-stats.vercel.app/api/top-langs/?username=anuraghazra&size_weight=0.5&count_weight=0.5)](https://github.com/anuraghazra/github-readme-stats)
+```
 
 ### Usage
 
@@ -419,7 +439,7 @@ You can use the `&hide_progress=true` option to hide the percentages and the pro
 
 [![Top Langs](https://github-readme-stats.vercel.app/api/top-langs/?username=anuraghazra&layout=compact)](https://github.com/anuraghazra/github-readme-stats)
 
-- Hidden progress bars
+-   Hidden progress bars
 
 [![Top Langs](https://github-readme-stats.vercel.app/api/top-langs/?username=anuraghazra&hide_progress=true)](https://github.com/anuraghazra/github-readme-stats)
 
@@ -564,14 +584,14 @@ Since the GitHub API only allows 5k requests per hour, my `https://github-readme
 <details>
 <summary><b>:hammer_and_wrench: Step-by-step guide for deploying on other platforms</b></summary>
 
-1. Fork or clone this repo as per your needs
-2. Add `express` to the dependencies section of `package.json`
-https://github.com/anuraghazra/github-readme-stats/blob/ba7c2f8b55eac8452e479c8bd38b044d204d0424/package.json#L54-L61
-3. Run `npm i` if needed (initial setup)
-4. Run `node express.js` to start the server, or set the entry point to `express.js` in `package.json` if you're deploying on a managed service
-https://github.com/anuraghazra/github-readme-stats/blob/ba7c2f8b55eac8452e479c8bd38b044d204d0424/package.json#L11
-5. You're done 🎉
-</details>
+1.  Fork or clone this repo as per your needs
+2.  Add `express` to the dependencies section of `package.json`
+    <https://github.com/anuraghazra/github-readme-stats/blob/ba7c2f8b55eac8452e479c8bd38b044d204d0424/package.json#L54-L61>
+3.  Run `npm i` if needed (initial setup)
+4.  Run `node express.js` to start the server, or set the entry point to `express.js` in `package.json` if you're deploying on a managed service
+    <https://github.com/anuraghazra/github-readme-stats/blob/ba7c2f8b55eac8452e479c8bd38b044d204d0424/package.json#L11>
+5.  You're done 🎉
+    </details>
 
 ### Keep your fork up to date
 

src/fetchers/top-languages-fetcher.js

@@ -54,7 +54,12 @@ const fetcher = (variables, token) => {
  * @param {string[]} exclude_repo List of repositories to exclude.
  * @returns {Promise<import("./types").TopLangData>} Top languages data.
  */
-const fetchTopLanguages = async (username, exclude_repo = []) => {
+const fetchTopLanguages = async (
+  username,
+  exclude_repo = [],
+  size_weight = 1,
+  count_weight = 0,
+) => {
   if (!username) throw new MissingParamError(["username"]);
 
   const res = await retryer(fetcher, { login: username });
@@ -101,6 +106,8 @@ const fetchTopLanguages = async (username, exclude_repo = []) => {
     .sort((a, b) => b.size - a.size)
     .filter((name) => !repoToHide[name.name]);
 
+  let repoCount = 0;
+
   repoNodes = repoNodes
     .filter((node) => node.languages.edges.length > 0)
     // flatten the list of language nodes
@@ -111,9 +118,14 @@ const fetchTopLanguages = async (username, exclude_repo = []) => {
 
       // if we already have the language in the accumulator
       // & the current language name is same as previous name
-      // add the size to the language size.
+      // add the size to the language size and increase repoCount.
       if (acc[prev.node.name] && prev.node.name === acc[prev.node.name].name) {
         langSize = prev.size + acc[prev.node.name].size;
+        repoCount += 1;
+      } else {
+        // reset repoCount to 1
+        // language must exist in at least one repo to be detected
+        repoCount = 1;
       }
       return {
         ...acc,
@@ -121,10 +133,18 @@ const fetchTopLanguages = async (username, exclude_repo = []) => {
           name: prev.node.name,
           color: prev.node.color,
           size: langSize,
+          count: repoCount,
         },
       };
     }, {});
 
+  Object.keys(repoNodes).forEach((name) => {
+    // comparison index calculation
+    repoNodes[name].size =
+      Math.pow(repoNodes[name].size, size_weight) *
+      Math.pow(repoNodes[name].count, count_weight);
+  });
+
   const topLangs = Object.keys(repoNodes)
     .sort((a, b) => repoNodes[b].size - repoNodes[a].size)
     .reduce((result, key) => {

tests/fetchTopLanguages.test.js

@@ -60,20 +60,22 @@ const error = {
 };
 
 describe("FetchTopLanguages", () => {
-  it("should fetch correct language data", async () => {
+  it("should fetch correct language data while using the new calculation", async () => {
     mock.onPost("https://api.github.com/graphql").reply(200, data_langs);
 
-    let repo = await fetchTopLanguages("anuraghazra");
+    let repo = await fetchTopLanguages("anuraghazra", [], 0.5, 0.5);
     expect(repo).toStrictEqual({
       HTML: {
         color: "#0f0",
+        count: 2,
         name: "HTML",
-        size: 200,
+        size: 20.000000000000004,
       },
       javascript: {
         color: "#0ff",
+        count: 2,
         name: "javascript",
-        size: 200,
+        size: 20.000000000000004,
       },
     });
   });
@@ -85,17 +87,59 @@ describe("FetchTopLanguages", () => {
     expect(repo).toStrictEqual({
       HTML: {
         color: "#0f0",
+        count: 1,
         name: "HTML",
         size: 100,
       },
       javascript: {
         color: "#0ff",
+        count: 2,
         name: "javascript",
         size: 200,
       },
     });
   });
 
+  it("should fetch correct language data while using the old calculation", async () => {
+    mock.onPost("https://api.github.com/graphql").reply(200, data_langs);
+
+    let repo = await fetchTopLanguages("anuraghazra", [], 1, 0);
+    expect(repo).toStrictEqual({
+      HTML: {
+        color: "#0f0",
+        count: 2,
+        name: "HTML",
+        size: 200,
+      },
+      javascript: {
+        color: "#0ff",
+        count: 2,
+        name: "javascript",
+        size: 200,
+      },
+    });
+  });
+
+  it("should rank languages by the number of repositories they appear in", async () => {
+    mock.onPost("https://api.github.com/graphql").reply(200, data_langs);
+
+    let repo = await fetchTopLanguages("anuraghazra", [], 0, 1);
+    expect(repo).toStrictEqual({
+      HTML: {
+        color: "#0f0",
+        count: 2,
+        name: "HTML",
+        size: 2,
+      },
+      javascript: {
+        color: "#0ff",
+        count: 2,
+        name: "javascript",
+        size: 2,
+      },
+    });
+  });
+
   it("should throw error", async () => {
     mock.onPost("https://api.github.com/graphql").reply(200, error);