Test the preview server's request handling

Only split_cgi_output was ever run. The handler, which builds the CGI
environment, guards the asset directory, caps the body, times cgit out,
turns a failed run into a 500 and answers HEAD, ran in no test at all.
The server listens on a port the system picks, so parallel runs cannot
collide.

2 files changed, 227 insertions, 4 deletions

diff --git a/tests/serve-handler-check.py b/tests/serve-handler-check.pynew file mode 100644index 0000000..cc45b45--- /dev/null+++ b/tests/serve-handler-check.py@@ -0,0 +1,179 @@+"""Check how tools/serve.py hands a request to the CGI and the answer back.++The server is started on a free loopback port with a shell script in place of+cgit, which writes back the CGI environment it was given, or fails or stalls+when the query asks it to. Each case sends one request and checks one part of+the way through. Run with the repository root, a scratch directory and the+name of one case.+"""++import http.client+import importlib.util+import socket+import sys+import threading+from pathlib import Path++# Loading serve.py would otherwise leave a __pycache__ directory behind in+# tools/, compiled bytecode the repository has no use for.+sys.dont_write_bytecode = True++STUB = """#!/bin/sh+case "$QUERY_STRING" in+fail) exit 3 ;;+stall) exec sleep 5 ;;+esac+printf 'Status: 200 OK\\nContent-Type: text/plain\\nX-Stub: yes\\n\\n'+printf 'method=%s\\n' "$REQUEST_METHOD"+printf 'script=[%s]\\n' "$SCRIPT_NAME"+printf 'path=%s\\n' "$PATH_INFO"+printf 'query=%s\\n' "$QUERY_STRING"+printf 'cookie=%s\\n' "$HTTP_COOKIE"+printf 'length=%s\\n' "$CONTENT_LENGTH"+printf 'config=%s\\n' "$CGIT_CONFIG"+printf 'body='+cat+"""+++def load_serve(root):+    path = root + "/tools/serve.py"+    spec = importlib.util.spec_from_file_location("serve", path)+    module = importlib.util.module_from_spec(spec)+    spec.loader.exec_module(module)+    return module+++def environment(request, scratch, port):+    status, headers, body = request(+        "GET", "/foo%2bbar/log/?h=main", headers={"Cookie": "a=b"}+    )+    lines = body.decode().splitlines()+    return (+        status == 200+        and headers.get("X-Stub") == "yes"+        and "method=GET" in lines+        and "script=[]" in lines+        and "path=/foo+bar/log/" in lines+        and "query=h=main" in lines+        and "cookie=a=b" in lines+        and "config=%s" % (scratch / "cgitrc") in lines+    )+++def post_body(request, scratch, port):+    status, _, body = request("POST", "/p/?login", body=b"user=me")+    lines = body.decode().splitlines()+    return status == 200 and "length=7" in lines and "body=user=me" in lines+++def asset(request, scratch, port):+    status, headers, body = request("GET", "/cgit.css")+    return (+        status == 200+        and headers.get("Content-Type") == "text/css"+        and body == b"div#cgit {}\n"+    )+++def repository_file(request, scratch, port):+    status, headers, body = request("GET", "/foo/tree/cgit.css")+    return status == 200 and b"path=/foo/tree/cgit.css" in body+++def asset_link_outside(request, scratch, port):+    status, headers, body = request("GET", "/outside.css")+    return status == 200 and b"path=/outside.css" in body+++def failing_cgi(request, scratch, port):+    status, _, _ = request("GET", "/?fail")+    return status == 500+++def stalled_cgi(request, scratch, port):+    status, _, _ = request("GET", "/?stall")+    return status == 504+++def oversized_body(request, scratch, port):+    status, _, _ = request("POST", "/", body=b"x" * 17)+    return status == 413+++# A client drops whatever follows the headers of an answer to HEAD, so this+# one reads the raw answer, which the server ends by closing the connection.+def head(request, scratch, port):+    with socket.create_connection(("127.0.0.1", port), timeout=30) as raw:+        raw.sendall(b"HEAD /foo/ HTTP/1.0\r\n\r\n")+        answer = b""+        while chunk := raw.recv(4096):+            answer += chunk+    headers, _, body = answer.partition(b"\r\n\r\n")+    return (+        headers.startswith(b"HTTP/1.0 200")+        and b"X-Stub: yes" in headers+        and b"Content-Length: 0" not in headers+        and body == b""+    )+++CASES = {+    "environment": environment,+    "post-body": post_body,+    "asset": asset,+    "repository-file": repository_file,+    "asset-link-outside": asset_link_outside,+    "failing-cgi": failing_cgi,+    "stalled-cgi": stalled_cgi,+    "oversized-body": oversized_body,+    "head": head,+}+++def main():+    root, scratch, name = sys.argv[1:]+    check = CASES[name]+    serve = load_serve(root)+    # Small bounds, so the timeout and the body cap are reached quickly, and+    # the type main() pins for the stylesheet.+    serve.CGI_TIMEOUT = 1+    serve.MAX_BODY_BYTES = 16+    serve.mimetypes.add_type("text/css", ".css")++    scratch = Path(scratch).resolve()+    data = scratch / "assets"+    data.mkdir()+    (data / "cgit.css").write_text("div#cgit {}\n")+    (scratch / "outside.css").write_text("not an asset\n")+    (data / "outside.css").symlink_to(scratch / "outside.css")+    (scratch / "cgitrc").write_text("")+    stub = scratch / "cgit"+    stub.write_text(STUB)+    stub.chmod(0o755)++    server = serve.CgitServer(("127.0.0.1", 0), scratch / "cgitrc", stub, data)+    thread = threading.Thread(target=server.serve_forever)+    thread.start()++    def request(method, path, body=None, headers={}):+        connection = http.client.HTTPConnection(+            "127.0.0.1", server.server_port, timeout=30+        )+        try:+            connection.request(method, path, body=body, headers=headers)+            response = connection.getresponse()+            return response.status, dict(response.getheaders()), response.read()+        finally:+            connection.close()++    try:+        ok = check(request, scratch, server.server_port)+    finally:+        server.shutdown()+        server.server_close()+        thread.join()+    return 0 if ok else 1+++sys.exit(main())
diff --git a/tests/serve-handler-check.py b/tests/serve-handler-check.py
new file mode 100644
index 0000000..cc45b45
--- /dev/null
+++ b/tests/serve-handler-check.py
@@ -0,0 +1,179 @@
+"""Check how tools/serve.py hands a request to the CGI and the answer back.
+
+The server is started on a free loopback port with a shell script in place of
+cgit, which writes back the CGI environment it was given, or fails or stalls
+when the query asks it to. Each case sends one request and checks one part of
+the way through. Run with the repository root, a scratch directory and the
+name of one case.
+"""
+
+import http.client
+import importlib.util
+import socket
+import sys
+import threading
+from pathlib import Path
+
+# Loading serve.py would otherwise leave a __pycache__ directory behind in
+# tools/, compiled bytecode the repository has no use for.
+sys.dont_write_bytecode = True
+
+STUB = """#!/bin/sh
+case "$QUERY_STRING" in
+fail) exit 3 ;;
+stall) exec sleep 5 ;;
+esac
+printf 'Status: 200 OK\\nContent-Type: text/plain\\nX-Stub: yes\\n\\n'
+printf 'method=%s\\n' "$REQUEST_METHOD"
+printf 'script=[%s]\\n' "$SCRIPT_NAME"
+printf 'path=%s\\n' "$PATH_INFO"
+printf 'query=%s\\n' "$QUERY_STRING"
+printf 'cookie=%s\\n' "$HTTP_COOKIE"
+printf 'length=%s\\n' "$CONTENT_LENGTH"
+printf 'config=%s\\n' "$CGIT_CONFIG"
+printf 'body='
+cat
+"""
+
+
+def load_serve(root):
+    path = root + "/tools/serve.py"
+    spec = importlib.util.spec_from_file_location("serve", path)
+    module = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(module)
+    return module
+
+
+def environment(request, scratch, port):
+    status, headers, body = request(
+        "GET", "/foo%2bbar/log/?h=main", headers={"Cookie": "a=b"}
+    )
+    lines = body.decode().splitlines()
+    return (
+        status == 200
+        and headers.get("X-Stub") == "yes"
+        and "method=GET" in lines
+        and "script=[]" in lines
+        and "path=/foo+bar/log/" in lines
+        and "query=h=main" in lines
+        and "cookie=a=b" in lines
+        and "config=%s" % (scratch / "cgitrc") in lines
+    )
+
+
+def post_body(request, scratch, port):
+    status, _, body = request("POST", "/p/?login", body=b"user=me")
+    lines = body.decode().splitlines()
+    return status == 200 and "length=7" in lines and "body=user=me" in lines
+
+
+def asset(request, scratch, port):
+    status, headers, body = request("GET", "/cgit.css")
+    return (
+        status == 200
+        and headers.get("Content-Type") == "text/css"
+        and body == b"div#cgit {}\n"
+    )
+
+
+def repository_file(request, scratch, port):
+    status, headers, body = request("GET", "/foo/tree/cgit.css")
+    return status == 200 and b"path=/foo/tree/cgit.css" in body
+
+
+def asset_link_outside(request, scratch, port):
+    status, headers, body = request("GET", "/outside.css")
+    return status == 200 and b"path=/outside.css" in body
+
+
+def failing_cgi(request, scratch, port):
+    status, _, _ = request("GET", "/?fail")
+    return status == 500
+
+
+def stalled_cgi(request, scratch, port):
+    status, _, _ = request("GET", "/?stall")
+    return status == 504
+
+
+def oversized_body(request, scratch, port):
+    status, _, _ = request("POST", "/", body=b"x" * 17)
+    return status == 413
+
+
+# A client drops whatever follows the headers of an answer to HEAD, so this
+# one reads the raw answer, which the server ends by closing the connection.
+def head(request, scratch, port):
+    with socket.create_connection(("127.0.0.1", port), timeout=30) as raw:
+        raw.sendall(b"HEAD /foo/ HTTP/1.0\r\n\r\n")
+        answer = b""
+        while chunk := raw.recv(4096):
+            answer += chunk
+    headers, _, body = answer.partition(b"\r\n\r\n")
+    return (
+        headers.startswith(b"HTTP/1.0 200")
+        and b"X-Stub: yes" in headers
+        and b"Content-Length: 0" not in headers
+        and body == b""
+    )
+
+
+CASES = {
+    "environment": environment,
+    "post-body": post_body,
+    "asset": asset,
+    "repository-file": repository_file,
+    "asset-link-outside": asset_link_outside,
+    "failing-cgi": failing_cgi,
+    "stalled-cgi": stalled_cgi,
+    "oversized-body": oversized_body,
+    "head": head,
+}
+
+
+def main():
+    root, scratch, name = sys.argv[1:]
+    check = CASES[name]
+    serve = load_serve(root)
+    # Small bounds, so the timeout and the body cap are reached quickly, and
+    # the type main() pins for the stylesheet.
+    serve.CGI_TIMEOUT = 1
+    serve.MAX_BODY_BYTES = 16
+    serve.mimetypes.add_type("text/css", ".css")
+
+    scratch = Path(scratch).resolve()
+    data = scratch / "assets"
+    data.mkdir()
+    (data / "cgit.css").write_text("div#cgit {}\n")
+    (scratch / "outside.css").write_text("not an asset\n")
+    (data / "outside.css").symlink_to(scratch / "outside.css")
+    (scratch / "cgitrc").write_text("")
+    stub = scratch / "cgit"
+    stub.write_text(STUB)
+    stub.chmod(0o755)
+
+    server = serve.CgitServer(("127.0.0.1", 0), scratch / "cgitrc", stub, data)
+    thread = threading.Thread(target=server.serve_forever)
+    thread.start()
+
+    def request(method, path, body=None, headers={}):
+        connection = http.client.HTTPConnection(
+            "127.0.0.1", server.server_port, timeout=30
+        )
+        try:
+            connection.request(method, path, body=body, headers=headers)
+            response = connection.getresponse()
+            return response.status, dict(response.getheaders()), response.read()
+        finally:
+            connection.close()
+
+    try:
+        ok = check(request, scratch, server.server_port)
+    finally:
+        server.shutdown()
+        server.server_close()
+        thread.join()
+    return 0 if ok else 1
+
+
+sys.exit(main())
diff --git a/tests/t0401-serve.sh b/tests/t0401-serve.shindex ca15e1c..23b8697 100755--- a/tests/t0401-serve.sh+++ b/tests/t0401-serve.sh@@ -1,9 +1,11 @@ #!/bin/sh -# Checks how tools/serve.py, the preview server that runs the built binary and-# hands its output back over HTTP, tells the CGI headers from the body. The-# cases live in serve-split-check.py beside this file, and nothing here touches-# a repository, so the fixtures stay unbuilt.+# Checks tools/serve.py, the preview server that runs the built binary and+# hands its output back over HTTP. The cases live beside this file, in+# serve-split-check.py for how the CGI headers are told from the body and in+# serve-handler-check.py for a request served on a free loopback port with a+# stub in place of cgit. Nothing here touches a repository, so the fixtures+# stay unbuilt.  test_description='Check the preview server splits CGI output correctly' CGIT_TEST_NO_CREATE_REPOS=YesPlease@@ -51,4 +53,46 @@ test_expect_success 'output with no blank line at all is all headers' ' 	split_check no-blank-line ' +handler_check() {+	mkdir "$1" &&+	"$PYTHON" -I "$TEST_OUTPUT_DIRECTORY/serve-handler-check.py" \+		"$TEST_OUTPUT_DIRECTORY/.." "$PWD/$1" "$1"+}++test_expect_success 'a request reaches the CGI with the environment it needs' '+	handler_check environment+'++test_expect_success 'a request body reaches the CGI with its length' '+	handler_check post-body+'++test_expect_success 'a root-level asset is served from the data directory' '+	handler_check asset+'++test_expect_success 'a file of that name inside a repository goes to the CGI' '+	handler_check repository-file+'++test_expect_success 'a link out of the data directory is no asset' '+	handler_check asset-link-outside+'++test_expect_success 'a CGI that fails is answered with 500' '+	handler_check failing-cgi+'++test_expect_success 'a CGI that stalls is answered with 504' '+	handler_check stalled-cgi+'++test_expect_success 'a body over the cap is refused with 413' '+	handler_check oversized-body+'++test_expect_success 'a HEAD request gets the headers alone' '+	handler_check head+'+ test_done
diff --git a/tests/t0401-serve.sh b/tests/t0401-serve.sh
index ca15e1c..23b8697 100755
--- a/tests/t0401-serve.sh
+++ b/tests/t0401-serve.sh
@@ -1,9 +1,11 @@
 #!/bin/sh
 
-# Checks how tools/serve.py, the preview server that runs the built binary and
-# hands its output back over HTTP, tells the CGI headers from the body. The
-# cases live in serve-split-check.py beside this file, and nothing here touches
-# a repository, so the fixtures stay unbuilt.
+# Checks tools/serve.py, the preview server that runs the built binary and
+# hands its output back over HTTP. The cases live beside this file, in
+# serve-split-check.py for how the CGI headers are told from the body and in
+# serve-handler-check.py for a request served on a free loopback port with a
+# stub in place of cgit. Nothing here touches a repository, so the fixtures
+# stay unbuilt.
 
 test_description='Check the preview server splits CGI output correctly'
 CGIT_TEST_NO_CREATE_REPOS=YesPlease
@@ -51,4 +53,46 @@ test_expect_success 'output with no blank line at all is all headers' '
 	split_check no-blank-line
 '
 
+handler_check() {
+	mkdir "$1" &&
+	"$PYTHON" -I "$TEST_OUTPUT_DIRECTORY/serve-handler-check.py" \
+		"$TEST_OUTPUT_DIRECTORY/.." "$PWD/$1" "$1"
+}
+
+test_expect_success 'a request reaches the CGI with the environment it needs' '
+	handler_check environment
+'
+
+test_expect_success 'a request body reaches the CGI with its length' '
+	handler_check post-body
+'
+
+test_expect_success 'a root-level asset is served from the data directory' '
+	handler_check asset
+'
+
+test_expect_success 'a file of that name inside a repository goes to the CGI' '
+	handler_check repository-file
+'
+
+test_expect_success 'a link out of the data directory is no asset' '
+	handler_check asset-link-outside
+'
+
+test_expect_success 'a CGI that fails is answered with 500' '
+	handler_check failing-cgi
+'
+
+test_expect_success 'a CGI that stalls is answered with 504' '
+	handler_check stalled-cgi
+'
+
+test_expect_success 'a body over the cap is refused with 413' '
+	handler_check oversized-body
+'
+
+test_expect_success 'a HEAD request gets the headers alone' '
+	handler_check head
+'
+
 test_done