feat: improve _build_problem_tags_cache

Author: wislertt

.amazonq/rules/problem-creation.md

@@ -13,10 +13,11 @@ When user requests a problem by **number** or **name/slug**, the assistant will:
     - Images provide crucial visual context, especially for tree and graph problems
     - Always verify images are included in `readme_examples` and accessible
 4. **Create** JSON file in `leetcode_py/cli/resources/leetcode/json/problems/{problem_name}.json`
-5. **Update** Makefile with `PROBLEM ?= {problem_name}`
-6. **Generate** problem structure using `make p-gen`
-7. **Verify** with `make p-lint` - fix template issues in JSON if possible, or manually fix generated files if template limitations
-8. **Iterate** if JSON fixes: re-run `make p-gen PROBLEM={problem_name} FORCE=1` and `make p-lint` until passes to ensure reproducibility
+5. **Update tags.json5** - If user specifies tags, manually add problem name to corresponding tag arrays in `leetcode_py/cli/resources/leetcode/json/tags.json5`
+6. **Update** Makefile with `PROBLEM ?= {problem_name}`
+7. **Generate** problem structure using `make p-gen`
+8. **Verify** with `make p-lint` - fix template issues in JSON if possible, or manually fix generated files if template limitations
+9. **Iterate** if JSON fixes: re-run `make p-gen PROBLEM={problem_name} FORCE=1` and `make p-lint` until passes to ensure reproducibility
 
 ## Scraping Commands
 

leetcode_py/cli/utils/problem_finder.py

@@ -59,6 +59,23 @@ def _add_problem_to_tag_map(
     problem_tags_map[problem_name].append(tag_name)
 
 
+def _process_tag_reference(
+    tags_data: dict, item: dict, tag_name: str, problem_tags_map: dict[str, list[str]]
+) -> None:
+    for problem_name in tags_data.get(item["tag"], []):
+        if isinstance(problem_name, str):
+            _add_problem_to_tag_map(problem_tags_map, problem_name, tag_name)
+
+
+def _process_tag_item(
+    tags_data: dict, item: str | dict, tag_name: str, problem_tags_map: dict[str, list[str]]
+) -> None:
+    if isinstance(item, dict) and "tag" in item:
+        _process_tag_reference(tags_data, item, tag_name, problem_tags_map)
+    elif isinstance(item, str):
+        _add_problem_to_tag_map(problem_tags_map, item, tag_name)
+
+
 @lru_cache(maxsize=1)
 def _build_problem_tags_cache() -> dict[str, list[str]]:
     try:
@@ -68,17 +85,9 @@ def _build_problem_tags_cache() -> dict[str, list[str]]:
         problem_tags_map: dict[str, list[str]] = {}
 
         for tag_name, problems in tags_data.items():
-            if not isinstance(problems, list):
-                continue
-
-            for item in problems:
-                if isinstance(item, dict) and "tag" in item:
-                    # Include all problems from referenced tag
-                    for problem_name in tags_data.get(item["tag"], []):
-                        if isinstance(problem_name, str):
-                            _add_problem_to_tag_map(problem_tags_map, problem_name, tag_name)
-                elif isinstance(item, str):
-                    _add_problem_to_tag_map(problem_tags_map, item, tag_name)
+            if isinstance(problems, list):
+                for item in problems:
+                    _process_tag_item(tags_data, item, tag_name, problem_tags_map)
 
         return problem_tags_map
     except (ValueError, OSError, KeyError):