fix: review comments

Author: bradjin8

api/export_api.py

@@ -22,8 +22,8 @@
 from services.workspace_db import (
     _build_composer_id_to_workspace_id,
     _collect_workspace_entries,
-    _load_bubble_map,
-    _load_code_block_diff_map,
+    load_bubble_map,
+    load_code_block_diff_map,
     _open_global_db,
 )
 from services.workspace_resolver import (
@@ -116,8 +116,8 @@ def export_chats():
             if global_db is None:
                 return jsonify({"error": "Cursor global storage not found"}), 404
 
-            bubble_map = _load_bubble_map(global_db)
-            code_block_diff_map = _load_code_block_diff_map(global_db)
+            bubble_map = load_bubble_map(global_db)
+            code_block_diff_map = load_code_block_diff_map(global_db)
 
             try:
                 composer_rows = global_db.execute(

api/workspaces.py

@@ -16,7 +16,7 @@
 from utils.workspace_path import resolve_workspace_path, get_cli_chats_path
 from utils.cli_chat_reader import list_cli_projects
 from utils.path_helpers import get_workspace_folder_paths, get_workspace_display_name
-from utils.workspace_descriptor import _read_json_file
+from utils.workspace_descriptor import read_json_file
 from services.workspace_resolver import (
     _infer_workspace_name_from_context,
     # Re-exported for back-compat with existing tests that import from api.workspaces
@@ -107,7 +107,7 @@ def get_workspace(workspace_id):
         folder = None
         workspace_name = workspace_id
         try:
-            wd = _read_json_file(wj_path)
+            wd = read_json_file(wj_path)
             folder_paths = get_workspace_folder_paths(wd)
             folder = folder_paths[0] if folder_paths else wd.get("folder")
             derived_name = get_workspace_display_name(wd)

scripts/export.py

@@ -58,9 +58,9 @@
     _build_composer_id_to_workspace_id,
     _collect_invalid_workspace_ids,
     _collect_workspace_entries,
-    _load_bubble_map,
-    _load_code_block_diff_map,
-    _load_project_layouts_map,
+    load_bubble_map,
+    load_code_block_diff_map,
+    load_project_layouts_map,
     _open_global_db,
 )
 from services.workspace_resolver import (  # noqa: E402
@@ -221,9 +221,9 @@ def main():
                 file=sys.stderr,
             )
         else:
-            project_layouts_map = _load_project_layouts_map(global_db)
-            bubble_map = _load_bubble_map(global_db)
-            code_block_diff_map = _load_code_block_diff_map(global_db)
+            project_layouts_map = load_project_layouts_map(global_db)
+            bubble_map = load_bubble_map(global_db)
+            code_block_diff_map = load_code_block_diff_map(global_db)
 
             try:
                 ide_composer_rows = global_db.execute(

services/workspace_db.py

@@ -1,13 +1,16 @@
 from __future__ import annotations
 
 import json
+import logging
 import os
 import sqlite3
 from contextlib import closing, contextmanager
 from pathlib import Path
 
+_logger = logging.getLogger(__name__)
+
 from utils.path_helpers import get_workspace_folder_paths
-from utils.workspace_descriptor import _read_json_file
+from utils.workspace_descriptor import read_json_file
 
 
 # ── Global-DB KV loaders ────────────────────────────────────────────────────
@@ -17,11 +20,11 @@
 # corrupt table cannot propagate to callers.
 
 
-def _load_bubble_map(global_db) -> dict[str, dict]:
+def load_bubble_map(global_db) -> dict[str, dict]:
     """Load all ``bubbleId:*`` KV entries into ``{bubble_id: bubble_dict}``.
 
-    Skips rows whose JSON value is not a dict; JSON parse errors are silently
-    discarded so a single malformed row cannot block the rest.
+    Skips rows whose JSON value is not a dict; JSON parse errors are logged at
+    DEBUG level so a single malformed row cannot block the rest.
     """
     bubble_map: dict[str, dict] = {}
     try:
@@ -39,12 +42,12 @@ def _load_bubble_map(global_db) -> dict[str, dict]:
             b = json.loads(row["value"])
             if isinstance(b, dict):
                 bubble_map[bid] = b
-        except Exception:
-            pass
+        except (json.JSONDecodeError, ValueError, KeyError, TypeError) as e:
+            _logger.debug("Skipping malformed bubbleId row %s: %s", row["key"], e)
     return bubble_map
 
 
-def _load_project_layouts_map(global_db) -> dict[str, list]:
+def load_project_layouts_map(global_db) -> dict[str, list]:
     """Load ``projectLayouts`` from ``messageRequestContext:*`` KV entries.
 
     Returns ``{composer_id: [root_path_str, ...]}``.  String-encoded layout
@@ -73,14 +76,14 @@ def _load_project_layouts_map(global_db) -> dict[str, list]:
                     o = json.loads(layout) if isinstance(layout, str) else layout
                     if isinstance(o, dict) and o.get("rootPath"):
                         layouts_map[cid].append(o["rootPath"])
-                except Exception:
-                    pass
-        except Exception:
-            pass
+                except (json.JSONDecodeError, ValueError, KeyError, TypeError) as e:
+                    _logger.debug("Skipping malformed layout entry in %s: %s", row["key"], e)
+        except (json.JSONDecodeError, ValueError, KeyError, TypeError) as e:
+            _logger.debug("Skipping malformed messageRequestContext row %s: %s", row["key"], e)
     return layouts_map
 
 
-def _load_code_block_diff_map(global_db) -> dict[str, list]:
+def load_code_block_diff_map(global_db) -> dict[str, list]:
     """Load ``codeBlockDiff:*`` KV entries into ``{composer_id: [diff_dict]}``.
 
     Each diff dict contains all fields from the raw JSON value plus a
@@ -105,8 +108,8 @@ def _load_code_block_diff_map(global_db) -> dict[str, list]:
                     **d,
                     "diffId": parts[2] if len(parts) > 2 else None,
                 })
-        except Exception:
-            pass
+        except (json.JSONDecodeError, ValueError, KeyError, TypeError) as e:
+            _logger.debug("Skipping malformed codeBlockDiff row %s: %s", row["key"], e)
     return diff_map
 
 
@@ -133,7 +136,7 @@ def _collect_invalid_workspace_ids(workspace_entries: list[dict]) -> set[str]:
     invalid: set[str] = set()
     for entry in workspace_entries:
         try:
-            wd = _read_json_file(entry["workspaceJsonPath"])
+            wd = read_json_file(entry["workspaceJsonPath"])
             folders = get_workspace_folder_paths(wd)
             if not folders:
                 invalid.add(entry["name"])

services/workspace_listing.py

@@ -12,14 +12,14 @@
     normalize_file_path,
     to_epoch_ms,
 )
-from utils.workspace_descriptor import _read_json_file
+from utils.workspace_descriptor import read_json_file
 from utils.workspace_path import get_cli_chats_path
 from services.workspace_db import (
     _build_composer_id_to_workspace_id,
     _collect_invalid_workspace_ids,
     _collect_workspace_entries,
-    _load_bubble_map,
-    _load_project_layouts_map,
+    load_bubble_map,
+    load_project_layouts_map,
     _open_global_db,
 )
 from services.workspace_resolver import (
@@ -56,8 +56,8 @@ def _safe_fetchall(query: str, params: tuple = ()) -> list:
                     "SELECT key, value FROM cursorDiskKV WHERE key LIKE 'composerData:%' AND LENGTH(value) > 10"
                 )
 
-                project_layouts_map: dict[str, list] = _load_project_layouts_map(global_db)
-                bubble_map: dict[str, dict] = _load_bubble_map(global_db)
+                project_layouts_map: dict[str, list] = load_project_layouts_map(global_db)
+                bubble_map: dict[str, dict] = load_bubble_map(global_db)
 
                 invalid_workspace_aliases = _infer_invalid_workspace_aliases(
                     composer_rows=composer_rows,
@@ -109,7 +109,7 @@ def _safe_fetchall(query: str, params: tuple = ()) -> list:
     for entry in workspace_entries:
         norm_folder = ""
         try:
-            wd = _read_json_file(entry["workspaceJsonPath"])
+            wd = read_json_file(entry["workspaceJsonPath"])
             folders = get_workspace_folder_paths(wd)
             first_folder = folders[0] if folders else None
             if first_folder:

services/workspace_resolver.py

@@ -13,7 +13,7 @@
     get_workspace_folder_paths,
     normalize_file_path,
 )
-from utils.workspace_descriptor import _basename_from_pathish, _read_json_file
+from utils.workspace_descriptor import basename_from_pathish, read_json_file
 from services.workspace_db import _open_global_db
 from models import SchemaError, Workspace
 
@@ -24,7 +24,7 @@ def _get_workspace_display_name(workspace_path: str, workspace_id: str) -> str:
         return "Other chats"
     wj_path = os.path.join(workspace_path, workspace_id, "workspace.json")
     try:
-        workspace = Workspace.from_dict(_read_json_file(wj_path), workspace_id=workspace_id)
+        workspace = Workspace.from_dict(read_json_file(wj_path), workspace_id=workspace_id)
         name = get_workspace_display_name(workspace.raw)
         if name:
             return name
@@ -103,7 +103,7 @@ def _infer_workspace_name_from_context(workspace_path: str, workspace_id: str) -
                         obj = layout
                     if not isinstance(obj, dict):
                         continue
-                    hint = _basename_from_pathish(obj.get("rootPath"))
+                    hint = basename_from_pathish(obj.get("rootPath"))
                     if hint:
                         counts[hint] = counts.get(hint, 0) + 1
 
@@ -121,7 +121,7 @@ def _get_project_from_file_path(
     best_len = 0
     for entry in workspace_entries:
         try:
-            wd = _read_json_file(entry["workspaceJsonPath"])
+            wd = read_json_file(entry["workspaceJsonPath"])
             for folder in get_workspace_folder_paths(wd):
                 wp = normalize_file_path(folder)
                 try:
@@ -140,7 +140,7 @@ def _create_project_name_to_workspace_id_map(workspace_entries):
     mapping = {}
     for entry in workspace_entries:
         try:
-            wd = _read_json_file(entry["workspaceJsonPath"])
+            wd = read_json_file(entry["workspaceJsonPath"])
             for folder in get_workspace_folder_paths(wd):
                 wp = re.sub(r"^file://", "", folder)
                 parts = wp.replace("\\", "/").split("/")
@@ -156,7 +156,7 @@ def _create_workspace_path_to_id_map(workspace_entries):
     out = {}
     for entry in workspace_entries:
         try:
-            wd = _read_json_file(entry["workspaceJsonPath"])
+            wd = read_json_file(entry["workspaceJsonPath"])
             for folder in get_workspace_folder_paths(wd):
                 normalized = normalize_file_path(folder)
                 out[normalized] = entry["name"]
@@ -269,7 +269,7 @@ def _determine_project_for_conversation(
     folder_name_to_ws = []
     for entry in workspace_entries:
         try:
-            wd = _read_json_file(entry["workspaceJsonPath"])
+            wd = read_json_file(entry["workspaceJsonPath"])
             for folder in get_workspace_folder_paths(wd):
                 name = re.sub(r"^file://", "", folder).replace("\\", "/").split("/")[-1]
                 if name:

services/workspace_tabs.py

@@ -14,13 +14,13 @@
 from utils.exclusion_rules import build_searchable_text, is_excluded_by_rules
 from utils.text_extract import extract_text_from_bubble
 from utils.tool_parser import parse_tool_call as _parse_tool_call
-from utils.workspace_descriptor import _read_json_file
+from utils.workspace_descriptor import read_json_file
 from models import Bubble, Composer, SchemaError
 from services.workspace_db import (
     _build_composer_id_to_workspace_id,
     _collect_invalid_workspace_ids,
     _collect_workspace_entries,
-    _load_code_block_diff_map,
+    load_code_block_diff_map,
     _open_global_db,
 )
 from services.workspace_resolver import (
@@ -64,7 +64,7 @@ def assemble_workspace_tabs(
         target_folder = ""
         wj_path = os.path.join(workspace_path, workspace_id, "workspace.json")
         try:
-            wd = _read_json_file(wj_path)
+            wd = read_json_file(wj_path)
             folders = get_workspace_folder_paths(wd)
             first_folder = folders[0] if folders else None
             if first_folder:
@@ -74,7 +74,7 @@ def assemble_workspace_tabs(
         if target_folder:
             for entry in workspace_entries:
                 try:
-                    wd2 = _read_json_file(entry["workspaceJsonPath"])
+                    wd2 = read_json_file(entry["workspaceJsonPath"])
                     folders2 = get_workspace_folder_paths(wd2)
                     f2 = folders2[0] if folders2 else None
                     if f2 and normalize_file_path(f2) == target_folder:
@@ -116,7 +116,7 @@ def _safe_fetchall(query: str, params: tuple = ()) -> list:
                     print(f"Schema drift in bubble {bid}: {e}")
 
         # Load codeBlockDiffs
-        code_block_diff_map = _load_code_block_diff_map(global_db)
+        code_block_diff_map = load_code_block_diff_map(global_db)
 
         # Load messageRequestContext rows once; build both
         # message_request_context_map and project_layouts_map from the same pass.

utils/cursor_md_exporter.py

@@ -208,7 +208,7 @@ def cursor_ide_chat_to_markdown(
         into ``code_block_diff_map``.
     bubble_map:
         Global ``{bubble_id: bubble_dict}`` map loaded from
-        ``cursorDiskKV`` (see ``services.workspace_db._load_bubble_map``).
+        ``cursorDiskKV`` (see ``services.workspace_db.load_bubble_map``).
     code_block_diff_map:
         Optional ``{composer_id: [diff_dict]}`` map.  When ``None`` no code
         edit bubbles are appended.
@@ -364,7 +364,7 @@ def cursor_ide_chat_to_markdown(
 
     # ── Frontmatter ───────────────────────────────────────────────────────────
     fm_lines = ["---"]
-    fm_lines.append(f"log_id: {composer_id}")
+    fm_lines.append(f"log_id: {json.dumps(composer_id, ensure_ascii=False)}")
     fm_lines.append("log_type: chat")
     fm_lines.append(f"title: {json.dumps(title, ensure_ascii=False)}")
     fm_lines.append(f"created_at: {datetime.fromtimestamp(created_ms / 1000).isoformat()}")
@@ -374,14 +374,14 @@ def cursor_ide_chat_to_markdown(
     fm_lines.append(f"workspace: {ws_slug}")
     fm_lines.append(f"workspace_name: {json.dumps(ws_display_name, ensure_ascii=False)}")
     if model_name and model_name != "default":
-        fm_lines.append(f"model: {model_name}")
+        fm_lines.append(f"model: {json.dumps(model_name, ensure_ascii=False)}")
     fm_lines.append(f"message_count: {len(bubbles)}")
     if total_tool_calls:
         fm_lines.append(f"total_tool_calls: {total_tool_calls}")
     if tool_breakdown:
         fm_lines.append("tool_call_breakdown:")
         for tn, cnt in sorted(tool_breakdown.items(), key=lambda x: -x[1]):
-            fm_lines.append(f"  {tn}: {cnt}")
+            fm_lines.append(f"  {json.dumps(tn, ensure_ascii=False)}: {cnt}")
     total_think = sum(1 for bub in bubbles if bub.get("thinking"))
     if total_think:
         fm_lines.append(f"thinking_count: {total_think}")

utils/workspace_descriptor.py

@@ -7,7 +7,7 @@
 from urllib.parse import unquote, urlparse
 
 
-def _read_json_file(path: str):
+def read_json_file(path: str):
     """Read a workspace.json with Cursor indirection applied."""
     return _resolve_workspace_descriptor(path)
 
@@ -70,7 +70,7 @@ def _resolve_workspace_descriptor(path: str, depth: int = 0):
     return out
 
 
-def _basename_from_pathish(path_value: str | None) -> str | None:
+def basename_from_pathish(path_value: str | None) -> str | None:
     """Extract a readable leaf folder name from file URI or filesystem path."""
     if not path_value:
         return None