Fix JSON parsing for chunked responses in subprocess transport

[couloirextreme]

Problem

Fixes #6 - The Claude Code SDK fails with JSONDecodeError when Claude CLI returns large responses that get split across multiple lines in the subprocess stdout stream.

This manifests as:

json.decoder.JSONDecodeError: Unterminated string starting at: line 1 column 131 (char 130)
claude_code_sdk._errors.CLIJSONDecodeError: Failed to decode JSON:
{"type":"user","message":{"role":"user","content":[{"tool_use_id":"toolu_01...

Root Cause

The current implementation in SubprocessCLITransport.receive_messages() attempts to parse each line as complete JSON with json.loads(line_str), but chunked responses contain incomplete JSON fragments.

Solution

Implements a stateful JSON parser that:

• Maintains a buffer to accumulate partial JSON across lines
• Tracks brace/bracket nesting to identify complete JSON objects
• Handles string escaping and quotes properly
• Yields complete JSON objects as they're reconstructed
• Backward compatible - no API changes

Testing

✅ Tested with real-world Linear issue analysis tool that previously caused TaskGroup errors
✅ Successfully processes large responses (4000+ chars) that were failing
✅ Maintains compatibility with simple single-line JSON responses

Before/After

Before: unhandled errors in a TaskGroup (1 sub-exception)
After: ✅ Cost: $0.4161, Duration: 97010ms, Response: [full analysis]

[Copilot]

Pull Request Overview

This PR refactors the JSON parsing logic in the CLI transport to robustly handle JSON fragments spread across multiple lines.

• Introduces a stateful JSON parser that accumulates buffer data and parses complete JSON objects
• Replaces the per-line JSON parsing with a buffer-based approach in receive_messages

[Copilot]

Modifying the outer 'json_buffer' inside the nested function may lead to an UnboundLocalError; consider declaring 'nonlocal json_buffer' at the start of the nested function 'read_stderr'.

[ltawfik]

thanks @couloirextreme . This issue was already fixed in PR #5. The current implementation correctly handles multiple JSON objects.