feat: replace Codex estimated usage with precise JSONL session log parsing

Replace the 70/30 input/output token estimation from state_5.sqlite
with precise parsing of Codex CLI JSONL session logs (~/.codex/sessions/).

- Parse event_msg (token_count), turn_context, and session_meta events
- Compute exact input/output/cached token deltas from cumulative totals
- Reuse session_log_sync table for incremental file scanning
- Pre-filter lines with string contains() before JSON deserialization
- Add codex_session data source to DataSourceBar with i18n (zh/en/ja)

src/i18n/locales/en.json

@@ -1023,7 +1023,8 @@
     "dataSource": {
       "proxy": "Proxy",
       "session_log": "Session Log",
-      "codex_db": "Codex DB"
+      "codex_db": "Codex DB",
+      "codex_session": "Codex Session"
     },
     "sessionSync": {
       "trigger": "Sync session logs",

src/i18n/locales/ja.json

@@ -1023,7 +1023,8 @@
     "dataSource": {
       "proxy": "プロキシ",
       "session_log": "セッションログ",
-      "codex_db": "Codex DB"
+      "codex_db": "Codex DB",
+      "codex_session": "Codex セッション"
     },
     "sessionSync": {
       "trigger": "セッションログを同期",

src/i18n/locales/zh.json

@@ -1023,7 +1023,8 @@
     "dataSource": {
       "proxy": "代理",
       "session_log": "会话日志",
-      "codex_db": "Codex 数据库"
+      "codex_db": "Codex 数据库",
+      "codex_session": "Codex 会话日志"
     },
     "sessionSync": {
       "trigger": "同步会话日志",