feat: support multi-CSV profiling, ratio analysis, and code guidance

[rad1092]

Motivation

• Add a workflow to analyze multiple CSVs together so users can compute column-level ratios (missing/unique/top values), numeric sign distributions, and discover shared/union columns across files.
• Provide actionable next steps by auto-generating example pandas code and a human-readable markdown report so users can continue with visualization or feature engineering.
• Improve the CLI and desktop experience to expose multi-file analysis and local environment diagnostics for smoother end-to-end usage.

Description

• Added bitnet_tools/multi_csv.py which implements analyze_multiple_csv, per-file column profiling (missing_ratio, unique_ratio, top_values), numeric sign distribution, build_code_guidance, build_multi_csv_markdown, and result_to_json.
• Extended the streaming summarizer with summarize_reader in bitnet_tools/analysis.py and added build_markdown_report to produce single-file markdown outputs.
• Extended CLI (bitnet_tools/cli.py) with a new multi-analyze subcommand and support for report, desktop, and doctor flows, and added bitnet-desktop entry point to pyproject.toml.
• Implemented a simple Windows desktop UI (bitnet_tools/desktop.py and bitnet_desktop.pyw) and an environment diagnostic helper (bitnet_tools/doctor.py) and added BitNet_Desktop_Start.bat to simplify Windows launch.
• Updated README.md to document multi-CSV capabilities and workflow, and added tests in tests/test_analysis.py and tests/test_cli.py covering the new multi-CSV report builder and CLI behavior.

Testing

• Ran pytest -q and all tests passed (11 passed).
• Verified CLI help shows multi-analyze with python -m bitnet_tools.cli --help.
• Executed a sample multi-file run python -m bitnet_tools.cli multi-analyze /tmp/a.csv /tmp/b.csv --question "다중 csv 분석" --out-json /tmp/multi.json --out-report /tmp/multi.md which produced both JSON and markdown outputs successfully.

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4f343c900b

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Escape newline sequences in generated pandas snippet

build_code_guidance currently uses \n inside quoted literals in the template string, so the generated pandas_example contains actual line breaks inside '...' and fails to compile (SyntaxError: unterminated string literal) when users run the suggested code from JSON/markdown output. This breaks the new “code guidance” workflow for any multi-analyze result that is copied into Python.

Useful? React with 👍 / 👎.