From de2e08257d704a3b7aa6aa82df1e4f044212013d Mon Sep 17 00:00:00 2001
From: shanjunjie <shanjunjie666@gmail.com>
Date: Mon, 6 Jul 2026 01:54:53 +0800
Subject: [PATCH] fix(dsml): preserve schema strings and robust e2e reads

---
 proxy/dsml_shim.py          |  2 ++
 test/test_dsml_shim.py      |  8 ++++++
 test/test_proxy_dsml_e2e.py | 50 ++++++++++++++++++++++++++++++++++++-
 3 files changed, 59 insertions(+), 1 deletion(-)

diff --git a/proxy/dsml_shim.py b/proxy/dsml_shim.py
index 802ee07..23099cb 100644
--- a/proxy/dsml_shim.py
+++ b/proxy/dsml_shim.py
@@ -61,6 +61,8 @@ def _coerce_param(pname, string_attr, raw, prop_schema):
     if string_attr == "true":
         return raw
     typ = (prop_schema or {}).get("type")
+    if typ == "string":
+        return raw
     try:
         if typ == "integer":
             return int(raw)
diff --git a/test/test_dsml_shim.py b/test/test_dsml_shim.py
index 9ec998e..164453c 100644
--- a/test/test_dsml_shim.py
+++ b/test/test_dsml_shim.py
@@ -110,6 +110,14 @@ def test_string_true_stays_string_even_if_numeric(self):
         blk = wrap_typed(P2, "calc", [("n", ' string="true"', "42")])
         self.assertEqual(ds.parse_dsml_tool_calls(blk, NUM)[0]["input"]["n"], "42")
 
+    def test_schema_string_stays_string_without_string_attr(self):
+        # DeepSeek can omit string="true" on string parameters. Schema type string
+        # must win over the JSON fallback, or numeric-looking text is coerced and
+        # the whole tool call gets discarded by validation.
+        blk = wrap_typed(P2, "web_search", [("query", "", "42")])
+        self.assertEqual(ds.parse_dsml_tool_calls(blk, WS),
+                         [{"name": "web_search", "input": {"query": "42"}}])
+
     def test_missing_string_attr_falls_back_json_then_str(self):
         blk = wrap_typed(P2, "calc", [("f", "", "3.5")])
         # f 是 number → 3.5；无 schema 属性的 json 兜底另测
diff --git a/test/test_proxy_dsml_e2e.py b/test/test_proxy_dsml_e2e.py
index 837eef2..045e946 100644
--- a/test/test_proxy_dsml_e2e.py
+++ b/test/test_proxy_dsml_e2e.py
@@ -77,7 +77,7 @@ def serve():
                 c, _ = srv.accept()
             except OSError:
                 return
-            req = c.recv(65536)
+            req = _read_http_request(c)
             is_stream = b'"stream": true' in req or b'"stream":true' in req
             if is_stream:
                 sse = _build_sse()
@@ -95,6 +95,54 @@ def serve():
     return f"http://127.0.0.1:{port}/up", srv
 
 
+def _read_http_request(sock):
+    data = b""
+    while b"\r\n\r\n" not in data:
+        chunk = sock.recv(65536)
+        if not chunk:
+            return data
+        data += chunk
+
+    head, sep, body = data.partition(b"\r\n\r\n")
+    content_length = 0
+    for line in head.split(b"\r\n"):
+        name, colon, value = line.partition(b":")
+        if colon and name.strip().lower() == b"content-length":
+            try:
+                content_length = int(value.strip())
+            except ValueError:
+                content_length = 0
+            break
+
+    while len(body) < content_length:
+        chunk = sock.recv(65536)
+        if not chunk:
+            break
+        body += chunk
+    return head + sep + body
+
+
+class _ChunkedSocket:
+    def __init__(self, chunks):
+        self._chunks = list(chunks)
+
+    def recv(self, _size):
+        if not self._chunks:
+            return b""
+        return self._chunks.pop(0)
+
+
+class HttpRequestReader(unittest.TestCase):
+    def test_reads_body_split_after_headers(self):
+        body = b'{"stream": true, "messages":["large enough to split"]}'
+        head = (b"POST /up HTTP/1.1\r\nHost: 127.0.0.1\r\n"
+                b"Content-Type: application/json\r\n"
+                + f"Content-Length: {len(body)}\r\n\r\n".encode("ascii"))
+        sock = _ChunkedSocket([head[:17], head[17:], body[:9], body[9:]])
+        req = _read_http_request(sock)
+        self.assertEqual(req, head + body)
+
+
 def raw_post(host, port, path, body):
     s = socket.create_connection((host, port), timeout=5)
     req = (f"POST {path} HTTP/1.1\r\nHost: {host}\r\n"