From 0dfa9814ec998487c4f76e3418e0a232dbcceb7d Mon Sep 17 00:00:00 2001
From: Joe Mou <dev@mou.fo>
Date: Thu, 22 Jan 2026 18:35:27 -0500
Subject: Add blob view for displaying file contents

---
 src/scraper.ts | 77 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 77 insertions(+)

(limited to 'src/scraper.ts')

diff --git a/src/scraper.ts b/src/scraper.ts
index add03b2..fa8d655 100644
--- a/src/scraper.ts
+++ b/src/scraper.ts
@@ -15,6 +15,20 @@ export interface GitHubTree {
   };
 }
 
+export interface GitHubBlob {
+  path: string;
+  branch: string;
+  content: string;
+  rawContent: string;
+  language: string | null;
+  size: number;
+  repo: {
+    name: string;
+    owner: string;
+    isPublic: boolean;
+  };
+}
+
 export async function getGitHubTree(
   owner: string,
   repo: string,
@@ -89,3 +103,66 @@ export async function getGitHubTree(
     },
   };
 }
+
+export async function getGitHubBlob(
+  owner: string,
+  repo: string,
+  branch: string,
+  path: string,
+): Promise<GitHubBlob> {
+  const url = `https://github.com/${owner}/${repo}/blob/${branch}/${path}`;
+
+  const response = await fetch(url, {
+    headers: {
+      "User-Agent":
+        "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36",
+      Accept:
+        "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8",
+      "Accept-Language": "en-US,en;q=0.5",
+    },
+  });
+  const html = await response.text();
+
+  // GitHub embeds file data in a JSON script tag
+  let payload;
+
+  const match = html.match(
+    /<script type="application\/json" data-target="react-app\.embeddedData">([^<]+)<\/script>/,
+  );
+  if (match) {
+    try {
+      const data = JSON.parse(match[1]);
+      if (data.payload?.blob) {
+        payload = data.payload;
+      }
+    } catch (e) {
+      // Skip invalid JSON
+    }
+  }
+
+  if (!payload) {
+    throw new Error("Could not find blob data in embedded JSON");
+  }
+
+  const blob = payload.blob;
+  const rawLines = blob.rawLines || [];
+  const rawContent = rawLines.join("\n");
+  const highlightedLines = blob.highlightedLines || [];
+  const content = highlightedLines
+    .map((line: any) => line.text || "")
+    .join("\n");
+
+  return {
+    path: payload.path,
+    branch: payload.refInfo.name,
+    content: content || rawContent,
+    rawContent,
+    language: blob.language || null,
+    size: blob.size,
+    repo: {
+      name: payload.repo.name,
+      owner: payload.repo.ownerLogin,
+      isPublic: payload.repo.public,
+    },
+  };
+}
-- 
cgit v1.3.1