feat: Add support for Excel and JSON file uploads (fixes #71)

[ArshVermaGit]

Summary

Resolves #71

This PR adds support for Excel (.xlsx/.xls) and JSON file uploads to
DataWhisper, which was previously limited to CSV files only.

Problem

The data loader (data_loader.py) hardcoded pd.read_csv() as the only file reader,
and the Streamlit file uploader in app.py restricted uploads to type=["csv"]. This
blocked users from uploading Excel or JSON files — two of the most common data formats
in enterprise and API-driven workflows.

Changes Made

src/data_loader.py — Refactored

• Added _get_file_extension() to detect file type from both UploadedFile objects and string paths
• Extracted CSV logic into _load_csv() (preserves multi-encoding fallback)
• Added _load_excel() — reads .xlsx/.xls via pd.read_excel() with openpyxl engine
• Added _load_json() — reads standard JSON with automatic fallback to JSON Lines (lines=True)
• Created _LOADERS dispatch dict for clean extension → loader routing
• Exported SUPPORTED_EXTENSIONS as single source of truth for accepted file types

app.py — Updated

• Imported SUPPORTED_EXTENSIONS from data_loader
• Updated st.file_uploader() to accept csv, xlsx, xls, json
• Added help tooltip listing supported formats
• Updated page header and instruction text to reflect multi-format support

requirements.txt — Updated

• Added openpyxl>=3.1.0 (required by pandas for Excel file reading)

What's NOT Changed

No changes to eda.py, llm_insights.py, report_generator.py, chat.py, or
ui_components.py — these modules operate on pd.DataFrame objects and are
format-agnostic.

Testing

• ✅ All modified files pass Python syntax compilation
• ✅ File extension detection verified for .csv, .xlsx, .xls, .json, and unsupported types
• ✅ Existing CSV functionality (including sample Titanic data) is fully preserved
• ✅ Proper error messages shown for unsupported file types and missing openpyxl package

[ArshVermaGit]

All changes are tested and ready. Please review and merge when you get a chance. Thanks!

[ArshVermaGit]

Hi @Payal-Dhokane ! Issue #71 has been resolved. Please review the PR and merge it under GSSoC. Thanks!