Add conditional HTTP requests and cache validators

README.md

@@ -13,6 +13,7 @@
 
 ## Table of Contents
 
+- [What's new (2026-06-22) — Conditional HTTP Requests & Cache Validators](#whats-new-2026-06-22--conditional-http-requests--cache-validators)
 - [What's new (2026-06-22) — Cookie Jar (HTTP Session Carry)](#whats-new-2026-06-22--cookie-jar-http-session-carry)
 - [What's new (2026-06-22) — HTTP Content Negotiation & Decompression](#whats-new-2026-06-22--http-content-negotiation--decompression)
 - [What's new (2026-06-22) — multipart/form-data Build & Parse](#whats-new-2026-06-22--multipartform-data-build--parse)
@@ -144,6 +145,12 @@
 
 ---
 
+## What's new (2026-06-22) — Conditional HTTP Requests & Cache Validators
+
+Skip re-downloading unchanged resources (ETag / 304). Full reference: [`docs/source/Eng/doc/new_features/v92_features_doc.rst`](docs/source/Eng/doc/new_features/v92_features_doc.rst).
+
+- **`store_validators` / `conditioned_call` / `is_fresh` / `parse_cache_control` / `is_not_modified`** (`AC_parse_cache_control`, `AC_store_validators`): `http_request` never sent `If-None-Match`/`If-Modified-Since` nor read `Cache-Control`, so every poll re-downloaded. This extracts validators, parses `Cache-Control` (max-age/no-store/…), decides freshness by an explicit age, conditions the next request, and detects `304 Not Modified`. Pure-stdlib, deterministic.
+
 ## What's new (2026-06-22) — Cookie Jar (HTTP Session Carry)
 
 Carry a session across HTTP calls. Full reference: [`docs/source/Eng/doc/new_features/v91_features_doc.rst`](docs/source/Eng/doc/new_features/v91_features_doc.rst).

README/README_zh-CN.md

@@ -12,6 +12,7 @@
 
 ## 目录
 
+- [本次更新 (2026-06-22) — 条件式 HTTP 请求与缓存验证子](#本次更新-2026-06-22--条件式-http-请求与缓存验证子)
 - [本次更新 (2026-06-22) — Cookie Jar(HTTP 会话携带)](#本次更新-2026-06-22--cookie-jarhttp-会话携带)
 - [本次更新 (2026-06-22) — HTTP 内容协商与解压](#本次更新-2026-06-22--http-内容协商与解压)
 - [本次更新 (2026-06-22) — multipart/form-data 构建与解析](#本次更新-2026-06-22--multipartform-data-构建与解析)
@@ -143,6 +144,12 @@
 
 ---
 
+## 本次更新 (2026-06-22) — 条件式 HTTP 请求与缓存验证子
+
+跳过重新下载未变更的资源(ETag / 304)。完整参考:[`docs/source/Zh/doc/new_features/v92_features_doc.rst`](../docs/source/Zh/doc/new_features/v92_features_doc.rst)。
+
+- **`store_validators` / `conditioned_call` / `is_fresh` / `parse_cache_control` / `is_not_modified`**(`AC_parse_cache_control`、`AC_store_validators`):`http_request` 从不发 `If-None-Match`/`If-Modified-Since` 也不读 `Cache-Control`,因此每次轮询都重新下载。本功能提取验证子、解析 `Cache-Control`(max-age/no-store/…)、以明确 age 判定新鲜度、为下一个请求加上条件标头,并检测 `304 Not Modified`。纯标准库、确定。
+
 ## 本次更新 (2026-06-22) — Cookie Jar(HTTP 会话携带)
 
 跨 HTTP 调用携带会话。完整参考:[`docs/source/Zh/doc/new_features/v91_features_doc.rst`](../docs/source/Zh/doc/new_features/v91_features_doc.rst)。

README/README_zh-TW.md

@@ -12,6 +12,7 @@
 
 ## 目錄
 
+- [本次更新 (2026-06-22) — 條件式 HTTP 請求與快取驗證子](#本次更新-2026-06-22--條件式-http-請求與快取驗證子)
 - [本次更新 (2026-06-22) — Cookie Jar(HTTP 工作階段攜帶)](#本次更新-2026-06-22--cookie-jarhttp-工作階段攜帶)
 - [本次更新 (2026-06-22) — HTTP 內容協商與解壓縮](#本次更新-2026-06-22--http-內容協商與解壓縮)
 - [本次更新 (2026-06-22) — multipart/form-data 建立與解析](#本次更新-2026-06-22--multipartform-data-建立與解析)
@@ -143,6 +144,12 @@
 
 ---
 
+## 本次更新 (2026-06-22) — 條件式 HTTP 請求與快取驗證子
+
+略過重新下載未變更的資源(ETag / 304)。完整參考:[`docs/source/Zh/doc/new_features/v92_features_doc.rst`](../docs/source/Zh/doc/new_features/v92_features_doc.rst)。
+
+- **`store_validators` / `conditioned_call` / `is_fresh` / `parse_cache_control` / `is_not_modified`**(`AC_parse_cache_control`、`AC_store_validators`):`http_request` 從不送 `If-None-Match`/`If-Modified-Since` 也不讀 `Cache-Control`,因此每次輪詢都重新下載。本功能擷取驗證子、解析 `Cache-Control`(max-age/no-store/…)、以明確 age 判定新鮮度、為下一個請求加上條件標頭,並偵測 `304 Not Modified`。純標準函式庫、具決定性。
+
 ## 本次更新 (2026-06-22) — Cookie Jar(HTTP 工作階段攜帶)
 
 跨 HTTP 呼叫攜帶工作階段。完整參考:[`docs/source/Zh/doc/new_features/v91_features_doc.rst`](../docs/source/Zh/doc/new_features/v91_features_doc.rst)。

docs/source/Eng/doc/new_features/v92_features_doc.rst

@@ -0,0 +1,47 @@
+Conditional HTTP Requests & Cache Validators
+============================================
+
+``http_request`` never sends ``If-None-Match`` / ``If-Modified-Since`` nor reads
+``Cache-Control``, so every poll re-downloads an unchanged resource. This
+extracts caching validators from a response, parses ``Cache-Control``, decides
+freshness, and conditions the next request so the server can answer ``304 Not
+Modified``.
+
+Pure standard library; imports no ``PySide6``. Freshness takes an explicit age
+(no wall clock), so the logic is fully deterministic in CI.
+
+Headless API
+------------
+
+.. code-block:: python
+
+    from je_auto_control import (
+        store_validators, conditioned_call, is_fresh, is_not_modified,
+        parse_cache_control, build_call,
+    )
+
+    response = http_request(url)
+    validators = store_validators(response)
+    if is_fresh(validators, age_seconds=now - stored_at):
+        use_cached()
+    else:
+        revalidation = conditioned_call(build_call(url), validators)
+        fresh = perform(revalidation)
+        if is_not_modified(fresh):
+            use_cached()           # 304 → the stored body is still valid
+
+``store_validators`` pulls ``etag`` / ``last_modified`` / ``date`` and the parsed
+``cache_control`` from a response. ``parse_cache_control`` turns the header into
+a directive dict (``max-age`` as an int, flags as ``True``). ``conditioned_call``
+adds ``If-None-Match`` / ``If-Modified-Since`` to a ``build_call`` dict.
+``is_fresh`` reports whether a cached entry is still fresh for a given age
+(``no-store`` / ``no-cache`` are never fresh). ``is_not_modified`` detects a
+``304`` response.
+
+Executor commands
+-----------------
+
+``AC_parse_cache_control`` returns ``{directives}`` for ``headers``;
+``AC_store_validators`` returns ``{validators}`` for a ``response``. Both are
+exposed as MCP tools (``ac_parse_cache_control`` / ``ac_store_validators``) and
+as Script Builder commands under **Data**.

docs/source/Eng/eng_index.rst

@@ -114,6 +114,7 @@ Comprehensive guides for all AutoControl features.
    doc/new_features/v89_features_doc
    doc/new_features/v90_features_doc
    doc/new_features/v91_features_doc
+   doc/new_features/v92_features_doc
    doc/ocr_backends/ocr_backends_doc
    doc/observability/observability_doc
    doc/operations_layer/operations_layer_doc

docs/source/Zh/doc/new_features/v92_features_doc.rst

@@ -0,0 +1,41 @@
+條件式 HTTP 請求與快取驗證子
+========================
+
+``http_request`` 從不送出 ``If-None-Match`` / ``If-Modified-Since`` 也不讀取 ``Cache-Control``,因此每次
+輪詢都會重新下載未變更的資源。本功能從回應擷取快取驗證子、解析 ``Cache-Control``、判定新鮮度,並為下一個
+請求加上條件標頭,讓伺服器可回應 ``304 Not Modified``。
+
+純標準函式庫;不匯入 ``PySide6``。新鮮度判定接受明確的 age(不使用 wall clock),因此邏輯在 CI 中完全
+具決定性。
+
+無頭 API
+--------
+
+.. code-block:: python
+
+    from je_auto_control import (
+        store_validators, conditioned_call, is_fresh, is_not_modified,
+        parse_cache_control, build_call,
+    )
+
+    response = http_request(url)
+    validators = store_validators(response)
+    if is_fresh(validators, age_seconds=now - stored_at):
+        use_cached()
+    else:
+        revalidation = conditioned_call(build_call(url), validators)
+        fresh = perform(revalidation)
+        if is_not_modified(fresh):
+            use_cached()           # 304 → 已儲存的內文仍有效
+
+``store_validators`` 從回應取出 ``etag`` / ``last_modified`` / ``date`` 與解析後的 ``cache_control``。
+``parse_cache_control`` 把標頭轉成 directive dict(``max-age`` 為 int,旗標為 ``True``)。``conditioned_call``
+把 ``If-None-Match`` / ``If-Modified-Since`` 加到 ``build_call`` dict。``is_fresh`` 在給定 age 下回報快取項
+是否仍新鮮(``no-store`` / ``no-cache`` 永不新鮮)。``is_not_modified`` 偵測 ``304`` 回應。
+
+執行器命令
+----------
+
+``AC_parse_cache_control`` 對 ``headers`` 回傳 ``{directives}``;``AC_store_validators`` 對 ``response``
+回傳 ``{validators}``。兩者皆以 MCP 工具(``ac_parse_cache_control`` / ``ac_store_validators``)以及
+Script Builder 中 **Data** 分類下的命令提供。

docs/source/Zh/zh_index.rst

@@ -114,6 +114,7 @@ AutoControl 所有功能的完整使用指南。
    doc/new_features/v89_features_doc
    doc/new_features/v90_features_doc
    doc/new_features/v91_features_doc
+   doc/new_features/v92_features_doc
    doc/ocr_backends/ocr_backends_doc
    doc/observability/observability_doc
    doc/operations_layer/operations_layer_doc

je_auto_control/__init__.py

@@ -431,6 +431,11 @@
 )
 # RFC 6265 cookie jar (carry a session across HTTP calls)
 from je_auto_control.utils.cookie_jar import CookieJar, parse_set_cookie
+# Conditional HTTP requests + RFC 9111 cache validators (ETag / 304)
+from je_auto_control.utils.http_conditional import (
+    conditioned_call, is_fresh, is_not_modified, parse_cache_control,
+    store_validators,
+)
 # W3C Trace Context propagation (traceparent / tracestate)
 from je_auto_control.utils.trace_context import (
     SpanContext, TraceContextError, child_context, extract_context,
@@ -958,6 +963,8 @@ def start_autocontrol_gui(*args, **kwargs):
     "build_accept", "build_accept_encoding", "decode_body", "negotiated_call",
     "parse_quality_values",
     "CookieJar", "parse_set_cookie",
+    "conditioned_call", "is_fresh", "is_not_modified", "parse_cache_control",
+    "store_validators",
     "SpanContext", "TraceContextError", "child_context", "extract_context",
     "format_traceparent", "format_tracestate", "inject_context",
     "new_root_context", "new_span_id", "new_trace_id", "parse_traceparent",

je_auto_control/gui/script_builder/command_schema.py

@@ -1800,6 +1800,22 @@ def _add_resilience_specs(specs: List[CommandSpec]) -> None:
         ),
         description="Parse one Set-Cookie header into name/value/attributes.",
     ))
+    specs.append(CommandSpec(
+        "AC_parse_cache_control", "Data", "HTTP Conditional: Cache-Control",
+        fields=(
+            FieldSpec("headers", FieldType.STRING,
+                      placeholder='{"Cache-Control": "max-age=60, public"}'),
+        ),
+        description="Parse a Cache-Control header into directives.",
+    ))
+    specs.append(CommandSpec(
+        "AC_store_validators", "Data", "HTTP Conditional: Store Validators",
+        fields=(
+            FieldSpec("response", FieldType.STRING,
+                      placeholder='{"headers": {"ETag": "\\"abc\\""}}'),
+        ),
+        description="Extract ETag / Last-Modified / Cache-Control validators.",
+    ))
     specs.append(CommandSpec(
         "AC_resolve_config", "Data", "Layered Config: Resolve",
         fields=(

je_auto_control/utils/executor/action_executor.py

@@ -3030,6 +3030,24 @@ def _redact_secret_text(text: str, mask: str = "***") -> Dict[str, Any]:
     return {"text": redact_secret_text(text, mask=mask)}
 
 
+def _parse_cache_control(headers: Any) -> Dict[str, Any]:
+    """Adapter: parse a Cache-Control header into {directives}."""
+    import json
+    from je_auto_control.utils.http_conditional import parse_cache_control
+    if isinstance(headers, str):
+        headers = json.loads(headers)
+    return {"directives": parse_cache_control(headers)}
+
+
+def _store_validators(response: Any) -> Dict[str, Any]:
+    """Adapter: extract cache validators from an HTTP response."""
+    import json
+    from je_auto_control.utils.http_conditional import store_validators
+    if isinstance(response, str):
+        response = json.loads(response)
+    return {"validators": store_validators(response)}
+
+
 def _cookie_header(set_cookies: Any) -> Dict[str, Any]:
     """Adapter: build a Cookie header from one/many Set-Cookie strings."""
     import json
@@ -4372,6 +4390,8 @@ def __init__(self):
             "AC_parse_quality_values": _parse_quality_values,
             "AC_cookie_header": _cookie_header,
             "AC_parse_set_cookie": _parse_set_cookie,
+            "AC_parse_cache_control": _parse_cache_control,
+            "AC_store_validators": _store_validators,
             "AC_profile_rows": _profile_rows,
             "AC_infer_schema": _infer_schema,
             "AC_parse_problem": _parse_problem,

je_auto_control/utils/http_conditional/__init__.py

@@ -0,0 +1,10 @@
+"""Conditional HTTP requests and cache validators for AutoControl."""
+from je_auto_control.utils.http_conditional.http_conditional import (
+    conditioned_call, is_fresh, is_not_modified, parse_cache_control,
+    store_validators,
+)
+
+__all__ = [
+    "conditioned_call", "is_fresh", "is_not_modified", "parse_cache_control",
+    "store_validators",
+]

je_auto_control/utils/http_conditional/http_conditional.py

@@ -0,0 +1,77 @@
+"""Conditional HTTP requests and RFC 9111 cache validators.
+
+``http_request`` never sends ``If-None-Match`` / ``If-Modified-Since`` nor reads
+``Cache-Control``, so every poll re-downloads an unchanged resource. This
+extracts caching validators from a response, parses ``Cache-Control``, decides
+freshness, and conditions the next request so the server can answer ``304 Not
+Modified``.
+
+Pure standard library; imports no ``PySide6``. Freshness takes an explicit age
+(no wall clock), so the logic is fully deterministic in CI.
+"""
+from typing import Any, Dict, Mapping, Optional
+
+
+def _header(headers: Optional[Mapping[str, Any]], name: str) -> str:
+    for key, value in (headers or {}).items():
+        if str(key).lower() == name:
+            return str(value)
+    return ""
+
+
+def parse_cache_control(headers: Optional[Mapping[str, Any]]) -> Dict[str, Any]:
+    """Parse a ``Cache-Control`` header into a directive dict."""
+    directives: Dict[str, Any] = {}
+    for part in _header(headers, "cache-control").split(","):
+        cleaned = part.strip()
+        if not cleaned:
+            continue
+        key, sep, value = cleaned.partition("=")
+        name = key.strip().lower()
+        if not sep:
+            directives[name] = True
+            continue
+        token = value.strip().strip('"')
+        try:
+            directives[name] = int(token)
+        except ValueError:
+            directives[name] = token
+    return directives
+
+
+def store_validators(response: Mapping[str, Any]) -> Dict[str, Any]:
+    """Extract cache validators from an ``http_request`` response."""
+    headers = response.get("headers") or {}
+    return {"etag": _header(headers, "etag") or None,
+            "last_modified": _header(headers, "last-modified") or None,
+            "date": _header(headers, "date") or None,
+            "cache_control": parse_cache_control(headers)}
+
+
+def conditioned_call(call: Mapping[str, Any],
+                     validators: Mapping[str, Any]) -> Dict[str, Any]:
+    """Return a copy of a ``build_call`` dict with conditional headers added."""
+    headers = dict(call.get("headers") or {})
+    if validators.get("etag"):
+        headers["If-None-Match"] = validators["etag"]
+    if validators.get("last_modified"):
+        headers["If-Modified-Since"] = validators["last_modified"]
+    out = dict(call)
+    out["headers"] = headers
+    return out
+
+
+def is_fresh(validators: Mapping[str, Any], age_seconds: float) -> bool:
+    """Whether a cached entry is still fresh ``age_seconds`` after storing."""
+    cache_control = validators.get("cache_control") or {}
+    if cache_control.get("no-store") or cache_control.get("no-cache"):
+        return False
+    max_age = cache_control.get("max-age")
+    if isinstance(max_age, int):
+        return age_seconds < max_age
+    return False
+
+
+def is_not_modified(response: Mapping[str, Any]) -> bool:
+    """Whether ``response`` is a ``304 Not Modified``."""
+    return int(response.get("status", 0)) == 304

je_auto_control/utils/mcp_server/tools/_factories.py

@@ -3343,6 +3343,28 @@ def rate_limit_tools() -> List[MCPTool]:
     ]
 
 
+def http_conditional_tools() -> List[MCPTool]:
+    return [
+        MCPTool(
+            name="ac_parse_cache_control",
+            description=("Parse the 'headers' Cache-Control into {directives} "
+                         "(max-age as int, flags as true)."),
+            input_schema=schema({"headers": {"type": "object"}}, ["headers"]),
+            handler=h.parse_cache_control,
+            annotations=READ_ONLY,
+        ),
+        MCPTool(
+            name="ac_store_validators",
+            description=("Extract cache validators (etag, last_modified, date, "
+                         "cache_control) from an HTTP 'response'. Returns "
+                         "{validators}."),
+            input_schema=schema({"response": {"type": "object"}}, ["response"]),
+            handler=h.store_validators,
+            annotations=READ_ONLY,
+        ),
+    ]
+
+
 def cookie_jar_tools() -> List[MCPTool]:
     return [
         MCPTool(
@@ -5296,7 +5318,7 @@ def media_assert_tools() -> List[MCPTool]:
     data_profile_tools, http_problem_tools, dotenv_tools,
     sse_client_tools, layered_config_tools, data_drift_tools,
     dataset_diff_tools, referential_tools, link_header_tools, multipart_tools,
-    http_content_tools, cookie_jar_tools,
+    http_content_tools, cookie_jar_tools, http_conditional_tools,
     saga_tools, decision_table_tools, locator_repair_tools,
     pii_text_tools, sarif_tools,
     screen_record_tools,

je_auto_control/utils/mcp_server/tools/_handlers.py

@@ -1787,6 +1787,17 @@ def parse_set_cookie(header):
     return _parse_set_cookie(header)
 
 
+def parse_cache_control(headers):
+    from je_auto_control.utils.executor.action_executor import (
+        _parse_cache_control)
+    return _parse_cache_control(headers)
+
+
+def store_validators(response):
+    from je_auto_control.utils.executor.action_executor import _store_validators
+    return _store_validators(response)
+
+
 def redact_config(obj, mask="***"):
     from je_auto_control.utils.executor.action_executor import _redact_config
     return _redact_config(obj, mask)