email

src/servers/queue/handlers/gmail-sync.ts

@@ -1,6 +1,7 @@
-import { mkdirSync, readdirSync, writeFileSync } from 'node:fs';
+import { mkdirSync, readFileSync, readdirSync, writeFileSync } from 'node:fs';
 import { join } from 'node:path';
 import { homedir } from 'node:os';
+import type { EmailSummary } from 'types';
 import type { JobHandler } from '../types';
 import { registerHandler } from '../handler-registry';
 import { DATA_PATH } from '../../data-path';
@@ -117,6 +118,129 @@ function buildEmlFilename(id: string, internalDate: string | undefined, rawEmail
   return `${dateStr}_${slugify(subject)}_${id}.eml`;
 }
 
+function decodeMimeWords(text: string): string {
+  return text.replace(/=\?([^?]+)\?(B|Q)\?([^?]+)\?=/gi, (_, _charset, encoding, encoded) => {
+    try {
+      if (encoding.toUpperCase() === 'B') {
+        return Buffer.from(encoded, 'base64').toString('utf-8');
+      }
+      const bytes: number[] = [];
+      for (let i = 0; i < encoded.length; i++) {
+        if (encoded[i] === '_') {
+          bytes.push(0x20);
+        } else if (encoded[i] === '=' && i + 2 < encoded.length) {
+          bytes.push(parseInt(encoded.slice(i + 1, i + 3), 16));
+          i += 2;
+        } else {
+          bytes.push(encoded.charCodeAt(i));
+        }
+      }
+      return Buffer.from(bytes).toString('utf-8');
+    } catch {
+      return encoded;
+    }
+  });
+}
+
+function extractHeader(raw: string, name: string): string {
+  const match = raw.match(new RegExp(`^${name}:\\s*(.+)$`, 'mi'));
+  return match?.[1]?.trim() ? decodeMimeWords(match[1].trim()) : '';
+}
+
+function findHeaderEnd(text: string): number {
+  const crlf = text.indexOf('\r\n\r\n');
+  const lf = text.indexOf('\n\n');
+  if (crlf !== -1) return crlf + 4;
+  if (lf !== -1) return lf + 2;
+  return -1;
+}
+
+function extractSnippet(raw: string): string {
+  const idx = findHeaderEnd(raw);
+  if (idx === -1) return '';
+
+  let body = raw.slice(idx);
+
+  // If multipart, skip boundary line + part headers to reach actual content
+  if (body.trimStart().startsWith('--')) {
+    const afterBoundary = body.slice(body.indexOf('\n') + 1);
+    const partBodyStart = findHeaderEnd(afterBoundary);
+    if (partBodyStart !== -1) body = afterBoundary.slice(partBodyStart);
+  }
+
+  // Stop at next MIME boundary
+  const nextBoundary = body.indexOf('\n--');
+  if (nextBoundary !== -1) body = body.slice(0, nextBoundary);
+
+  return body.replace(/\s+/g, ' ').trim().slice(0, 120);
+}
+
+function countAttachments(raw: string): number {
+  const matches = raw.match(/^Content-Disposition:\s*attachment/gim);
+  return matches?.length ?? 0;
+}
+
+function parseEmlToSummary(raw: string, id: string): EmailSummary {
+  const dateStr = extractHeader(raw, 'Date');
+  const date = dateStr ? new Date(dateStr).toISOString() : new Date(0).toISOString();
+  const attachmentCount = countAttachments(raw);
+  return {
+    id,
+    from: extractHeader(raw, 'From'),
+    to: extractHeader(raw, 'To'),
+    subject: extractHeader(raw, 'Subject') || '(no subject)',
+    date,
+    snippet: extractSnippet(raw),
+    ...(attachmentCount > 0 ? { attachmentCount } : {}),
+  };
+}
+
+type EmailIndex = { v: number; entries: EmailSummary[] };
+const INDEX_VERSION = 3;
+
+export function rebuildIndex(emailsDir: string): EmailSummary[] {
+  let filenames: string[];
+  try {
+    filenames = readdirSync(emailsDir).filter((f) => f.endsWith('.eml'));
+  } catch {
+    return [];
+  }
+
+  const indexPath = join(emailsDir, 'index.json');
+  const existing = new Map<string, EmailSummary>();
+  try {
+    const raw = JSON.parse(readFileSync(indexPath, 'utf-8'));
+    const index = (Array.isArray(raw) ? null : raw) as EmailIndex | null;
+    if (index?.v === INDEX_VERSION) {
+      for (const entry of index.entries) {
+        existing.set(entry.id, entry);
+      }
+    }
+  } catch {
+    /* no existing index or corrupted */
+  }
+
+  const entries: EmailSummary[] = [];
+  for (const filename of filenames) {
+    const id = filename.replace(/\.eml$/, '');
+    const cached = existing.get(id);
+    if (cached) {
+      entries.push(cached);
+    } else {
+      try {
+        const raw = readFileSync(join(emailsDir, filename), 'utf-8');
+        entries.push(parseEmlToSummary(raw, id));
+      } catch {
+        /* skip unreadable files */
+      }
+    }
+  }
+
+  entries.sort((a, b) => new Date(b.date).getTime() - new Date(a.date).getTime());
+  writeFileSync(indexPath, JSON.stringify({ v: INDEX_VERSION, entries }));
+  return entries;
+}
+
 type SyncProgress = { saved: number; skipped: number; errors: number; page: number };
 type OnProgress = (progress: SyncProgress) => void;
 
@@ -263,6 +387,9 @@ const gmailSyncHandler: JobHandler = {
         }
 
         console.log(`[gmail-sync] Saved ${totalSaved}, skipped ${totalSkipped}, errors ${totalErrors}`);
+
+        rebuildIndex(outputDir);
+        console.log('[gmail-sync] Index rebuilt');
       },
     },
   ],