@mrpatronz/nexusflow 0.2.17 → 0.2.19

package/src/analyzers/readme-summarizer.ts

@@ -1,102 +1,102 @@
-/**
- * @module analyzers/readme-summarizer
- * Extracts a summary from a repository's README.md file.
- */
-
-import * as fs from 'node:fs/promises';
-import * as path from 'node:path';
-
-/**
- * Reads a repository's README.md and extracts the first meaningful
- * prose paragraph as a summary. Looks for README.md (case-insensitive).
- *
- * It automatically strips badges, TOC lists, headers, HTML tags, and redacts ClientIDs/Secrets.
- *
- * @param repoPath - Absolute path to the repository root.
- * @returns The extracted summary text, or null if no README is found.
- */
-export async function extractReadmeSummary(
-  repoPath: string,
-): Promise<string | null> {
-  // Try common README filenames
-  const candidates = ['README.md', 'readme.md', 'Readme.md', 'README.MD', 'README'];
-
-  let content: string | null = null;
-
-  for (const filename of candidates) {
-    try {
-      content = await fs.readFile(path.join(repoPath, filename), 'utf-8');
-      break;
-    } catch {
-      continue;
-    }
-  }
-
-  if (!content) return null;
-
-  // 1. Strip code blocks to avoid false matching on secrets/prose
-  content = content.replace(/```[\s\S]*?```/g, '');
-
-  // 2. Scrub ClientIDs/Secrets/Credentials
-  // Match standard secret/ID patterns: GUIDs, client_id/secret variables, hex keys
-  const clientSecretsRegex = /(client_id|client_secret|appid|secret|password|key|token|credential)\s*[:=]\s*["']?[a-zA-Z0-9-_\/\+\.]{16,}["']?/gi;
-  const guidRegex = /[a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{12}/g;
-  
-  content = content
-    .replace(clientSecretsRegex, (match, p1) => `${p1}: [REDACTED]`)
-    .replace(guidRegex, '[REDACTED_ID]');
-
-  // 3. Process line-by-line to find the first prose paragraph
-  const lines = content.split('\n');
-  const proseLines: string[] = [];
-
-  for (const line of lines) {
-    const trimmed = line.trim();
-
-    // Skip empty lines
-    if (!trimmed) {
-      if (proseLines.length > 0) {
-        // We found a paragraph and hit an empty line. Let's finish!
-        break;
-      }
-      continue;
-    }
-
-    // Skip HTML tags
-    if (trimmed.startsWith('<') && trimmed.endsWith('>')) continue;
-
-    // Skip badges
-    if (trimmed.startsWith('[![') || (trimmed.startsWith('![') && trimmed.includes('http'))) continue;
-
-    // Skip horizontal rules
-    if (/^(-{3,}|={3,}|\*{3,})$/.test(trimmed)) continue;
-
-    // Skip headers (Markdown # )
-    if (trimmed.startsWith('#')) continue;
-
-    // Detect Table of Contents (TOC) lists
-    // Skip lines that look like: - [About](#about) or 1. [About](#about) or * 1. [Section](#section)
-    if (/^(?:[-*+]\s*|\d+\.\s*)(\d+\.)?\s*\[[^\]]+\]\(#[^)]+\)/.test(trimmed)) {
-      continue;
-    }
-
-    // Skip link-only lines (e.g., "[API Docs](https://...)")
-    if (/^\[[^\]]+\]\([^)]+\)$/.test(trimmed)) {
-      continue;
-    }
-
-    // Skip standard bullet and numbered lists if we are searching for prose (e.g. at the top of README before prose)
-    if (proseLines.length === 0 && (/^[-*+]\s+/.test(trimmed) || /^\d+\.\s+/.test(trimmed))) {
-      continue;
-    }
-
-    // Accumulate prose lines
-    proseLines.push(line);
-  }
-
-  const summary = proseLines.join(' ').replace(/\s+/g, ' ').trim();
-  if (!summary) return null;
-  
-  // Limit the summary to a concise paragraph (first 250 characters)
-  return summary.length > 250 ? summary.slice(0, 250) + '...' : summary;
-}
+/**
+ * @module analyzers/readme-summarizer
+ * Extracts a summary from a repository's README.md file.
+ */
+
+import * as fs from 'node:fs/promises';
+import * as path from 'node:path';
+
+/**
+ * Reads a repository's README.md and extracts the first meaningful
+ * prose paragraph as a summary. Looks for README.md (case-insensitive).
+ *
+ * It automatically strips badges, TOC lists, headers, HTML tags, and redacts ClientIDs/Secrets.
+ *
+ * @param repoPath - Absolute path to the repository root.
+ * @returns The extracted summary text, or null if no README is found.
+ */
+export async function extractReadmeSummary(
+  repoPath: string,
+): Promise<string | null> {
+  // Try common README filenames
+  const candidates = ['README.md', 'readme.md', 'Readme.md', 'README.MD', 'README'];
+
+  let content: string | null = null;
+
+  for (const filename of candidates) {
+    try {
+      content = await fs.readFile(path.join(repoPath, filename), 'utf-8');
+      break;
+    } catch {
+      continue;
+    }
+  }
+
+  if (!content) return null;
+
+  // 1. Strip code blocks to avoid false matching on secrets/prose
+  content = content.replace(/```[\s\S]*?```/g, '');
+
+  // 2. Scrub ClientIDs/Secrets/Credentials
+  // Match standard secret/ID patterns: GUIDs, client_id/secret variables, hex keys
+  const clientSecretsRegex = /(client_id|client_secret|appid|secret|password|key|token|credential)\s*[:=]\s*["']?[a-zA-Z0-9-_\/\+\.]{16,}["']?/gi;
+  const guidRegex = /[a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{12}/g;
+  
+  content = content
+    .replace(clientSecretsRegex, (match, p1) => `${p1}: [REDACTED]`)
+    .replace(guidRegex, '[REDACTED_ID]');
+
+  // 3. Process line-by-line to find the first prose paragraph
+  const lines = content.split('\n');
+  const proseLines: string[] = [];
+
+  for (const line of lines) {
+    const trimmed = line.trim();
+
+    // Skip empty lines
+    if (!trimmed) {
+      if (proseLines.length > 0) {
+        // We found a paragraph and hit an empty line. Let's finish!
+        break;
+      }
+      continue;
+    }
+
+    // Skip HTML tags
+    if (trimmed.startsWith('<') && trimmed.endsWith('>')) continue;
+
+    // Skip badges
+    if (trimmed.startsWith('[![') || (trimmed.startsWith('![') && trimmed.includes('http'))) continue;
+
+    // Skip horizontal rules
+    if (/^(-{3,}|={3,}|\*{3,})$/.test(trimmed)) continue;
+
+    // Skip headers (Markdown # )
+    if (trimmed.startsWith('#')) continue;
+
+    // Detect Table of Contents (TOC) lists
+    // Skip lines that look like: - [About](#about) or 1. [About](#about) or * 1. [Section](#section)
+    if (/^(?:[-*+]\s*|\d+\.\s*)(\d+\.)?\s*\[[^\]]+\]\(#[^)]+\)/.test(trimmed)) {
+      continue;
+    }
+
+    // Skip link-only lines (e.g., "[API Docs](https://...)")
+    if (/^\[[^\]]+\]\([^)]+\)$/.test(trimmed)) {
+      continue;
+    }
+
+    // Skip standard bullet and numbered lists if we are searching for prose (e.g. at the top of README before prose)
+    if (proseLines.length === 0 && (/^[-*+]\s+/.test(trimmed) || /^\d+\.\s+/.test(trimmed))) {
+      continue;
+    }
+
+    // Accumulate prose lines
+    proseLines.push(line);
+  }
+
+  const summary = proseLines.join(' ').replace(/\s+/g, ' ').trim();
+  if (!summary) return null;
+  
+  // Limit the summary to a concise paragraph (first 250 characters)
+  return summary.length > 250 ? summary.slice(0, 250) + '...' : summary;
+}