Load concatenated split CSV caches when primary file is missing and add tests

[hemanth-asirvatham]

Motivation

• Allow the API to recover cached outputs when a single final CSV is not present but split parts like name_1.csv, name_2.csv, ... exist.
• Ensure deterministic ordering when reassembling split files so downstream consumers get a consistent DataFrame.
• Provide unit tests that cover both the split-file fallback and the preference for a primary file when it exists.

Description

• Added Path import and a helper _find_split_parts that discovers and numerically sorts split files matching the pattern stem_<index><suffix> in the same directory as the target path.
• Extended _load_cached_dataframe to detect and load split CSV parts, concat them with pd.concat(..., ignore_index=True), and print a diagnostic message when using split results.
• Preserved the existing behavior of preferring the primary final_path file when present and raising FileNotFoundError only if neither the primary nor split parts are available.
• Added tests/test_api_cached_loading.py with two tests that validate the split-file fallback ordering and the primary-file preference.

Testing

• Ran the new unit tests in tests/test_api_cached_loading.py using pytest, and both tests passed.
• Existing code paths that read a single CSV remain covered by the new tests which confirm the primary file is preferred when present.

---

Codex Task

[github-actions]

Thank you for your submission, we really appreciate it. Like many open-source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution. You can sign the CLA by just posting a Pull Request Comment same as the below format.

---

I have read the CLA Document and I hereby sign the CLA

---

_{You can retrigger this bot by commenting recheck in this Pull Request. Posted by the CLA Assistant Lite bot.}