refactor: improve code quality, testing, and protocol robustness

LLMS.txt

@@ -0,0 +1,235 @@
+# python-icap
+
+> Pure Python ICAP client library with no external dependencies
+
+python-icap is a Python library for communicating with ICAP (Internet Content Adaptation Protocol) servers. It implements RFC 3507 for integrating with servers like c-icap and SquidClamav for antivirus scanning, content filtering, and data loss prevention.
+
+## Key Features
+
+- Pure Python implementation with zero runtime dependencies
+- Sync (`IcapClient`) and async (`AsyncIcapClient`) clients with full API parity
+- High-level file scanning: `scan_file()`, `scan_bytes()`, `scan_stream()`
+- SSL/TLS support with custom certificates and mutual TLS
+- Bundled pytest plugin for testing without a live server
+- Python 3.8+ support
+
+## Installation
+
+```bash
+pip install python-icap
+```
+
+## Quick Start
+
+```python
+from icap import IcapClient
+
+with IcapClient('localhost', port=1344) as client:
+    response = client.scan_file('/path/to/file.pdf')
+    if response.is_no_modification:
+        print("File is clean")
+    else:
+        print("Threat detected")
+```
+
+Async usage:
+
+```python
+import asyncio
+from icap import AsyncIcapClient
+
+async def scan():
+    async with AsyncIcapClient('localhost') as client:
+        response = await client.scan_bytes(b"content")
+        print(f"Clean: {response.is_no_modification}")
+
+asyncio.run(scan())
+```
+
+## Public API
+
+### Main Classes
+
+Import from `icap`:
+
+- `IcapClient(address, port=1344, timeout=10, ssl_context=None, max_response_size=104857600)` - Synchronous client
+- `AsyncIcapClient(address, port=1344, timeout=10.0, ssl_context=None, max_response_size=104857600)` - Async client (max_response_size limits response size to prevent DoS, default 100MB)
+- `IcapResponse` - Response object with `status_code`, `headers`, `body`, `is_success`, `is_no_modification`, `encapsulated`, `istag`
+- `EncapsulatedParts` - Dataclass for parsed Encapsulated header offsets (`req_hdr`, `req_body`, `res_hdr`, `res_body`, `null_body`, `opt_body`)
+- `CaseInsensitiveDict` - Case-insensitive dictionary for ICAP headers (per RFC 3507)
+
+### High-Level Methods (Recommended)
+
+- `scan_file(filepath, service='avscan', chunk_size=0)` - Scan file by path (chunk_size>0 streams large files)
+- `scan_bytes(data, service='avscan', filename=None)` - Scan bytes directly
+- `scan_stream(stream, service='avscan', filename=None, chunk_size=0)` - Scan file-like objects
+
+### Low-Level Methods
+
+- `options(service)` - Query server capabilities
+- `respmod(service, http_request, http_response, headers=None, preview=None)` - Response modification
+- `reqmod(service, http_request, http_body=None, headers=None)` - Request modification
+
+### IcapResponse Properties
+
+- `status_code` - ICAP status code (200, 204, etc.)
+- `headers` - Case-insensitive header dictionary
+- `body` - Response body bytes
+- `is_success` - True for 2xx status codes
+- `is_no_modification` - True for 204 (content is clean)
+- `encapsulated` - Parsed `EncapsulatedParts` from Encapsulated header (or None)
+- `istag` - ISTag header value for cache validation (RFC 3507 Section 4.7)
+
+### Connection Management
+
+- `connect()` / `disconnect()` - Manual connection control
+- `is_connected` - Property to check connection status
+- Context manager support: `with IcapClient(...) as client:`
+
+### Exceptions
+
+Import from `icap.exception`:
+
+- `IcapException` - Base exception class
+- `IcapConnectionError` - Connection failures
+- `IcapTimeoutError` - Operation timeouts
+- `IcapProtocolError` - Malformed responses
+- `IcapServerError` - Server 5xx errors
+
+## Pytest Plugin
+
+The bundled pytest plugin provides fixtures for testing ICAP integrations.
+
+### Mock Fixtures (no server required)
+
+- `mock_icap_client` - Returns clean (204) responses
+- `mock_async_icap_client` - Async version
+- `mock_icap_client_virus` - Returns virus detection responses
+- `mock_icap_client_timeout` - Raises `IcapTimeoutError`
+- `mock_icap_client_connection_error` - Raises `IcapConnectionError`
+
+### Response Builder
+
+```python
+from icap.pytest_plugin import IcapResponseBuilder
+
+response = IcapResponseBuilder().clean().build()  # 204 No Modification
+response = IcapResponseBuilder().virus("Trojan.Gen").build()  # Virus detected
+response = IcapResponseBuilder().error(503, "Unavailable").build()  # Error
+```
+
+### Pre-built Response Fixtures
+
+- `icap_response_clean` - 204 No Modification response
+- `icap_response_virus` - Virus detection response (X-Virus-ID: EICAR-Test)
+- `icap_response_options` - OPTIONS response with server capabilities
+- `icap_response_error` - 500 Server Error response
+- `icap_response_builder` - Factory fixture returning `IcapResponseBuilder`
+
+### MockIcapClient
+
+```python
+from icap.pytest_plugin import MockIcapClient, IcapResponseBuilder
+
+client = MockIcapClient()
+client.on_respmod(IcapResponseBuilder().virus("Trojan").build())
+response = client.scan_bytes(b"content")
+client.assert_called("scan_bytes", times=1)
+```
+
+### Markers
+
+```python
+@pytest.mark.icap_mock(response="clean")
+def test_clean(icap_mock):
+    assert icap_mock.scan_bytes(b"data").is_no_modification
+
+@pytest.mark.icap_mock(response="virus", virus_name="EICAR")
+def test_virus(icap_mock):
+    assert not icap_mock.scan_bytes(b"data").is_no_modification
+```
+
+### Advanced Mock Features
+
+```python
+from icap.pytest_plugin import MockIcapClient, IcapResponseBuilder
+
+# Conditional responses based on scanned content
+client = MockIcapClient()
+client.when(data_contains=b"EICAR").respond(
+    IcapResponseBuilder().virus("EICAR-Test").build()
+)
+client.on_any(IcapResponseBuilder().clean().build())  # fallback for everything else
+
+# Call inspection
+response = client.scan_bytes(b"test")
+call = client.last_call
+assert call.method == "scan_bytes"
+assert call.data == b"test"  # or call.kwargs["data"]
+
+# Strict mode (fails if no matching response)
+@pytest.mark.icap_mock(strict=True)
+def test_strict(icap_mock):
+    icap_mock.on_respmod(IcapResponseBuilder().clean().build())
+    # MockResponseExhaustedError if called more times than configured
+```
+
+## Project Structure
+
+```
+src/icap/
+├── __init__.py          # Public API exports
+├── icap.py              # Synchronous IcapClient
+├── async_icap.py        # AsyncIcapClient
+├── response.py          # IcapResponse, EncapsulatedParts, CaseInsensitiveDict
+├── exception.py         # Custom exceptions
+├── _protocol.py         # Shared protocol utilities
+└── pytest_plugin/       # Bundled pytest plugin
+    ├── __init__.py      # Plugin entry point: fixtures and markers
+    ├── builder.py       # IcapResponseBuilder
+    ├── mock_client.py   # MockIcapClient (sync)
+    ├── mock_async.py    # MockAsyncIcapClient
+    ├── matchers.py      # ResponseMatcher, MatcherBuilder
+    ├── call_record.py   # MockCall for call inspection
+    ├── protocols.py     # ResponseCallback protocols
+    └── mock.py          # Re-exports for backward compatibility
+```
+
+## Common Service Names
+
+- `"avscan"` or `"srv_clamav"` - ClamAV virus scanning
+- `"squidclamav"` - SquidClamav service
+- `"echo"` - Echo service (testing)
+
+## Testing
+
+```bash
+# Unit tests
+pytest -m "not integration"
+
+# Integration tests (requires Docker)
+docker compose -f docker/docker-compose.yml up -d
+pytest -m integration
+```
+
+## EICAR Test String
+
+Standard test string for triggering antivirus detection:
+
+```python
+EICAR = b'X5O!P%@AP[4\\PZX54(P^)7CC)7}$EICAR-STANDARD-ANTIVIRUS-TEST-FILE!$H+H*'
+```
+
+## Protocol Reference
+
+- RFC 3507: Internet Content Adaptation Protocol
+- Default port: 1344
+- Methods: OPTIONS, REQMOD, RESPMOD
+
+## Links
+
+- Repository: https://github.com/CaptainDriftwood/python-icap
+- PyPI: https://pypi.org/project/python-icap/
+- c-icap: https://c-icap.sourceforge.net/
+- SquidClamav: https://squidclamav.darold.net/
+- ClamAV: https://www.clamav.net/

README.md

@@ -12,6 +12,7 @@
 [![Python 3.8 | 3.9 | 3.10 | 3.11 | 3.12 | 3.13 | 3.14](https://img.shields.io/badge/python-3.8%20%7C%203.9%20%7C%203.10%20%7C%203.11%20%7C%203.12%20%7C%203.13%20%7C%203.14-blue.svg)](https://www.python.org/downloads/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
 [![No Dependencies](https://img.shields.io/badge/dependencies-none-brightgreen.svg)](pyproject.toml)
+[![LLMS.txt](https://img.shields.io/badge/LLMS-txt-blue)](https://github.com/CaptainDriftwood/python-icap/blob/master/LLMS.txt)
 
 [![uv](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/uv/main/assets/badge/v0.json)](https://github.com/astral-sh/uv)
 [![Ruff](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/ruff/main/assets/badge/v2.json)](https://github.com/astral-sh/ruff)
@@ -21,6 +22,7 @@ A pure Python ICAP (Internet Content Adaptation Protocol) client with no externa
 
 ## Table of Contents
 
+- [Quick Reference](#quick-reference)
 - [Overview](#overview)
 - [What is ICAP?](#what-is-icap)
   - [Key Differences from HTTP](#key-differences-from-http)
@@ -39,6 +41,7 @@ A pure Python ICAP (Internet Content Adaptation Protocol) client with no externa
   - [Scanning Content with RESPMOD](#scanning-content-with-respmod)
   - [Scanning Files](#scanning-files)
   - [Manual File Scanning (lower-level API)](#manual-file-scanning-lower-level-api)
+  - [Preview Mode](#preview-mode)
 - [Async Usage](#async-usage)
   - [Basic Async Example](#basic-async-example)
   - [Concurrent Scanning](#concurrent-scanning)
@@ -75,6 +78,39 @@ python-icap provides a clean, Pythonic API for integrating ICAP into your applic
 - **Pytest plugin** - Mock clients and fixtures for testing without a live server
 - **Zero dependencies** - Pure Python stdlib implementation
 
+## Quick Reference
+
+```python
+# Common imports
+from icap import IcapClient, AsyncIcapClient, IcapResponse
+from icap.exception import IcapException, IcapConnectionError, IcapTimeoutError
+
+# Scan bytes (simplest)
+with IcapClient("localhost") as client:
+    response = client.scan_bytes(b"content")
+    is_clean = response.is_no_modification
+
+# Scan file
+with IcapClient("localhost") as client:
+    response = client.scan_file("/path/to/file.pdf")
+
+# Async scan
+async with AsyncIcapClient("localhost") as client:
+    response = await client.scan_bytes(b"content")
+
+# Check server capabilities
+with IcapClient("localhost") as client:
+    response = client.options("avscan")
+    preview_size = response.headers.get("Preview")  # bytes for preview mode
+
+# Response properties
+response.status_code       # 200, 204, etc.
+response.is_success        # True for 2xx
+response.is_no_modification  # True for 204 (clean)
+response.headers           # Dict of ICAP headers
+response.body              # Response body bytes
+```
+
 ## What is ICAP?
 
 **ICAP (Internet Content Adaptation Protocol)** is a simple protocol that lets network devices (like proxies) send HTTP content to a separate server for inspection or modification before passing it along.
@@ -188,10 +224,11 @@ This allows the ICAP server to efficiently parse the message without scanning th
 
 ## Installation
 
-> **Note:** This package is not yet published to PyPI due to a name collision. Install directly from source.
-
 ```bash
-# Standard installation
+# Install from PyPI
+pip install python-icap
+
+# Or install from source
 pip install .
 
 # Development installation (editable)
@@ -322,6 +359,51 @@ else:
     print("File contains threats")
 ```
 
+### Preview Mode
+
+ICAP servers can advertise a preview size via OPTIONS, allowing clients to send just the beginning of a file for initial scanning. If the server can determine the content is clean from the preview alone, it returns 204; otherwise it requests the full content with 100 Continue.
+
+```python
+from icap import IcapClient
+
+with IcapClient('localhost', port=1344) as client:
+    # Step 1: Query server capabilities to get preview size
+    options_response = client.options('avscan')
+    preview_size = options_response.headers.get('Preview')
+
+    if preview_size:
+        print(f"Server supports preview mode: {preview_size} bytes")
+    else:
+        print("Server does not advertise preview support")
+
+    # Step 2: Send RESPMOD with preview enabled
+    # The client handles the 100 Continue flow automatically
+    content = b"Large file content here..." * 1000
+
+    response = client.scan_bytes(
+        content,
+        service='avscan',
+        preview=int(preview_size) if preview_size else None
+    )
+
+    if response.is_no_modification:
+        print("Content is clean")
+    else:
+        print("Threat detected")
+```
+
+**How Preview Mode Works:**
+
+1. Client sends OPTIONS to discover the server's preview size
+2. Client sends RESPMOD with only the first N bytes (preview)
+3. Server analyzes the preview:
+   - Returns **204 No Modification** if the preview is enough to determine content is clean
+   - Returns **100 Continue** to request the remaining data
+4. If 100 Continue received, client sends the rest of the content
+5. Server returns final verdict (200 or 204)
+
+Preview mode reduces bandwidth and latency when scanning large files that are obviously clean (or obviously malicious) from their headers.
+
 ## Async Usage
 
 python-icap includes an async client (`AsyncIcapClient`) for use with `asyncio`. The async client provides the same API as the sync client but with `async`/`await` syntax.
@@ -596,8 +678,8 @@ python-icap/
 │   ├── async_icap.py     # Asynchronous ICAP client
 │   ├── _protocol.py      # Shared protocol constants
 │   ├── response.py       # Response handling
-│   └── exception.py      # Custom exceptions
-├── pytest_src/icap/        # Pytest plugin for ICAP testing
+│   ├── exception.py      # Custom exceptions
+│   └── pytest_plugin/    # Bundled pytest plugin for testing
 ├── tests/                # Unit tests
 ├── examples/             # Usage examples
 ├── docker/               # Docker setup for integration testing