[Bug]: In-memory cache in data_loader.py is defined but never used

[anshul23102]

Describe the bug

utils/data_loader.py declares _projects_cache = None and a clear_cache() helper, but load_all_projects() never reads or writes that variable. As a result, every call to load_all_projects() opens and parses projects.json from disk, even when the data has not changed.

Every HTTP request that touches /, /api/recommend, or /project/<id> triggers at least one (often two or three) redundant file reads.

Current code (utils/data_loader.py)

_projects_cache = None          # defined but never populated

def load_all_projects():
    with open(DATA_FILE, "r", encoding="utf-8") as f:   # reads disk on every call
        return json.load(f)

def clear_cache():
    global _projects_cache
    _projects_cache = None      # clears a cache that never fills

Expected behaviour

projects.json is read once on the first call; every subsequent call returns the cached list. clear_cache() resets it (used in tests).

Proposed fix

Implement a thread-safe double-checked locking pattern using threading.Lock:

_projects_cache = None
_cache_lock = threading.Lock()

def load_all_projects():
    global _projects_cache
    if _projects_cache is not None:
        return _projects_cache
    with _cache_lock:
        if _projects_cache is None:
            with open(DATA_FILE, "r", encoding="utf-8") as f:
                _projects_cache = json.load(f)
    return _projects_cache

This ensures the file is read exactly once even under concurrent requests.

---

I am a GSSoC 2026 contributor and would like to work on this fix. Could you please assign this issue to me?

[github-actions]

Thank you for opening your first issue in DevPath.

Please make sure to:

• Read CONTRIBUTING.md
• Wait for assignment before starting work
• Follow repository guidelines carefully

Happy contributing.

[Nikkiraj4]

/assign gssoc

[anshul23102]

/assign gssoc

Note: I opened this issue and already have a draft PR (#272) ready with the fix. All 29 tests pass. Please assign this to me.

[Nisha-Sanap]

Hi maintainers
I’d like to work on this issue. I have reviewed the problem statement and understood the expected behavior. I’m interested in contributing a clean and well-tested fix for it.

Please assign this issue to me under GSSoC/OSS contribution.

[Prince2301p]

Hi,
I would like to work on resolving this issue. Please assign it to me. I’m a participant in GSSoC 2026 and would be happy to contribute.
Looking forward to collaborating with you.

[anshul23102]

Thank you for assigning this to me! PR #272 is ready for review, it implements the thread-safe double-checked locking pattern as described in the issue. All 29 existing tests pass. Please let me know if any changes are needed.

[anshul23102]

Just a heads up - PR #272 has been ready for review for 5 days and directly resolves this issue with the thread-safe double-checked locking pattern. All 29 existing tests pass. A separate enhancement PR (#513) was filed today for the same file, so flagging this to avoid duplicate work. Happy to address any review feedback on #272.