From e54938f5aed185d4b02ec3910ecba552f8435185 Mon Sep 17 00:00:00 2001
From: Kent Slaney <kent@slaney.org>
Date: Mon, 27 Nov 2023 20:37:29 -0800
Subject: [PATCH] moved cython server string parse to c

---
 include/Common.h    |  8 +++++++-
 libmc/_client.pyx   | 46 ++++++++++++++++-----------------------------
 src/Common.cpp      | 24 +++++++++++++++--------
 src/Connection.cpp  |  1 +
 tests/test_unix.cpp |  2 +-
 5 files changed, 41 insertions(+), 40 deletions(-)

diff --git a/include/Common.h b/include/Common.h
index 4e8538d9..da4142d8 100644
--- a/include/Common.h
+++ b/include/Common.h
@@ -225,9 +225,15 @@ typedef enum {
   QUIT_OP,
 } op_code_t;
 
+struct ServerSpec {
+  char* host;
+  char* port;
+  char* alias;
+};
+
 const char* errCodeToString(err_code_t err);
 bool isLocalSocket(const char* host);
-char** splitServerString(char* input);
+ServerSpec splitServerString(char* input);
 
 } // namespace mc
 } // namespace douban
diff --git a/libmc/_client.pyx b/libmc/_client.pyx
index e07c5d89..9ad7c2d6 100644
--- a/libmc/_client.pyx
+++ b/libmc/_client.pyx
@@ -3,6 +3,7 @@
 # cython: profile=False, c_string_type=unicode, c_string_encoding=utf8
 
 from libc.stdint cimport uint8_t, uint32_t, uint64_t, int64_t
+from libc.stdlib cimport atoi
 from libcpp cimport bool as bool_t
 from libcpp.string cimport string
 from libcpp.vector cimport vector
@@ -45,7 +46,12 @@ cdef extern from "Common.h" namespace "douban::mc":
         VERSION_OP
         QUIT_OP
 
-    #bool isLocalSocket(const char* host) nogil
+    cdef struct ServerSpec:
+        char* host
+        char* port
+        char* alias
+
+    ServerSpec splitServerString(char* input) nogil
 
 
 cdef extern from "Export.h":
@@ -380,39 +386,19 @@ cdef class PyClient:
 
         servers_ = []
         for srv in servers:
-            addr_alias = srv.rsplit(' ', 1)
-            addr = addr_alias[0]
-            if len(addr_alias) == 1:
-                alias = None
-            elif (len(addr) - len(addr.rstrip("\\"))) % 2 == 1:
-                addr = srv
-                alias = None
-            else:
-                alias = addr_alias[1]
-            
-            if addr.startswith("/"):
-                port = 0
-            else:
-                host_port = addr.split(':')
-                host = host_port[0]
-                if len(host_port) == 1:
-                    port = MC_DEFAULT_PORT
-                else:
-                    port = int(host_port[1])
             if PY_MAJOR_VERSION > 2:
-                host = PyUnicode_AsUTF8String(host)
-                alias = PyUnicode_AsUTF8String(alias) if alias else None
-            servers_.append((host, port, alias))
+                srv = PyUnicode_AsUTF8String(srv)
+            srv = PyString_AsString(srv)
+            servers_.append(srv)
 
         Py_INCREF(servers_)
         for i in range(n):
-            host, port, alias = servers_[i]
-            c_hosts[i] = PyString_AsString(host)
-            c_ports[i] = PyInt_AsLong(port)
-            if alias is None:
-                c_aliases[i] = NULL
-            else:
-                c_aliases[i] = PyString_AsString(alias)
+            c_split = splitServerString(servers_[i])
+            host, port, alias = c_split
+
+            c_hosts[i] = c_split.host
+            c_ports[i] = MC_DEFAULT_PORT if c_split.port == NULL else atoi(c_split.port)
+            c_aliases[i] = c_split.alias
 
         if init:
             rv = self._imp.init(c_hosts, c_ports, n, c_aliases)
diff --git a/src/Common.cpp b/src/Common.cpp
index 43ef3585..be1a98b6 100644
--- a/src/Common.cpp
+++ b/src/Common.cpp
@@ -50,28 +50,36 @@ const char* errCodeToString(err_code_t err) {
   }
 }
 
-bool isLocalPath(const char* host) {
+bool isLocalSocket(const char* host) {
   return host[0] == '/';
 }
 
-char** splitServerString(char* input) {
-  bool escaped = false;
-  char *res[3] = { input--, NULL, NULL };
+ServerSpec splitServerString(char* input) {
+  bool escaped = false, num = false;
+  ServerSpec res = { input--, NULL, NULL };
   for (;;) {
     switch (*(++input))
     {
       case ':': // invalid in a UNIX path
         *input = '\0';
-        res[1] = input + 1;
+        res.port = input + 1;
+        num = true;
       case '\0':
-        break;
+        return res;
       case ' ':
         if (!escaped) {
           *input = '\0';
-          res[2] = input + 1;
-          break;
+          res.alias = input + 1;
+          return res;
         }
       default:
+        if (num) {
+          *res.port = ':';
+          res.port = NULL;
+          num = false;
+        }
+      case '0': case '1': case '2': case '3': case '4':
+      case '5': case '6': case '7': case '8': case '9':
         escaped = false;
         continue;
       case '\\':
diff --git a/src/Connection.cpp b/src/Connection.cpp
index 2f6aacf4..9acc510a 100644
--- a/src/Connection.cpp
+++ b/src/Connection.cpp
@@ -11,6 +11,7 @@
 #include <fcntl.h>
 #include <queue>
 
+#include "Common.h"
 #include "Connection.h"
 #include "Keywords.h"
 
diff --git a/tests/test_unix.cpp b/tests/test_unix.cpp
index d2ed79d8..5341aae8 100644
--- a/tests/test_unix.cpp
+++ b/tests/test_unix.cpp
@@ -11,4 +11,4 @@ TEST(test_unix, establish_connection) {
   Client* client = newUnixClient();
   EXPECT_TRUE(client != NULL);
   delete client;
-}
\ No newline at end of file
+}