Merge pull request #188 from gentlementlegen/fix/link-credit

fix: unique link credits
ubiquity-os-marketplace · Nov 26, 2024 · 74efe55 · 74efe55
2 parents 9831cdf + 90a648b
commit 74efe55
Show file tree

Hide file tree

Showing 13 changed files with 1,449 additions and 1,178 deletions.
diff --git a/dist/index.js b/dist/index.js
diff --git a/src/configuration/formatting-evaluator-config.ts b/src/configuration/formatting-evaluator-config.ts
@@ -15,6 +15,7 @@ export const commentType = stringLiteralUnion(
 );
 
 export const wordRegex = /\b\w+\b/;
+export const urlRegex = /https?:\/\/\S+/g;
 
 const htmlEntity = Type.Object({
   score: Type.Number(),

diff --git a/src/parser/content-evaluator-module.ts b/src/parser/content-evaluator-module.ts
@@ -109,7 +109,7 @@ export class ContentEvaluatorModule extends BaseModule {
     );
 
     if (Object.keys(relevancesByAi).length !== commentsToEvaluate.length + prCommentsToEvaluate.length) {
-      throw this.context.logger.fatal("Relevance / Comment length mismatch!", {
+      throw this.context.logger.error("Relevance / Comment length mismatch!", {
         relevancesByAi,
         commentsToEvaluate,
         prCommentsToEvaluate,

diff --git a/src/parser/data-purge-module.ts b/src/parser/data-purge-module.ts
@@ -33,6 +33,8 @@ export class DataPurgeModule extends BaseModule {
           .replace(/^\/.+/g, "")
           // Remove HTML comments
           .replace(/<!--[\s\S]*?-->/g, "")
+          // Remove the footnotes
+          .replace(/^###### .*?\[\^\d+\^][\s\S]*$/gm, "")
           // Keep only one new line needed by markdown-it package to convert to html
           .replace(/\n\s*\n/g, "\n")
           .trim();

diff --git a/src/parser/formatting-evaluator-module.ts b/src/parser/formatting-evaluator-module.ts
@@ -6,6 +6,7 @@ import { commentEnum, CommentType } from "../configuration/comment-types";
 import {
   FormattingEvaluatorConfiguration,
   formattingEvaluatorConfigurationType,
+  urlRegex,
   wordRegex,
 } from "../configuration/formatting-evaluator-config";
 import { IssueActivity } from "../issue-activity";
@@ -151,6 +152,7 @@ export class FormattingEvaluatorModule extends BaseModule {
   _classifyTagsWithWordCount(htmlElement: HTMLElement, commentType: GithubCommentScore["type"]) {
     const formatting: Record<string, { score: number; elementCount: number }> = {};
     const elements = htmlElement.getElementsByTagName("*");
+    const urlSet = new Set<string>();
 
     for (const element of elements) {
       const tagName = element.tagName.toLowerCase();
@@ -168,9 +170,26 @@ export class FormattingEvaluatorModule extends BaseModule {
         element.remove();
         continue;
       }
-      this._updateTagCount(formatting, tagName, score);
+      if (tagName === "a") {
+        const url = element.getAttribute("href");
+        if (url) {
+          urlSet.add(url.split(/[#?]/)[0]);
+        }
+      } else {
+        const bodyContent = element.textContent;
+        const matches = bodyContent?.match(urlRegex);
+        matches?.map((url) => url.split(/[#?]/)[0]).forEach((url) => urlSet.add(url));
+        this._updateTagCount(formatting, tagName, score);
+      }
     }
-    const words = this._countWordsFromRegex(htmlElement.textContent ?? "", this._multipliers[commentType]?.wordValue);
+    urlSet.forEach(() => {
+      this._updateTagCount(formatting, "a", this._multipliers[commentType].html["a"].score ?? 0);
+    });
+    const words = this._countWordsFromRegex(
+      htmlElement.textContent?.replace(urlRegex, "") ?? "",
+      this._multipliers[commentType]?.wordValue
+    );
+
     return { formatting, words };
   }