Added case for literal (W)STRING to be handled by a presized buffer i…

…nstead of a single character

pyads/connection.py

@@ -50,7 +50,6 @@
     ads_type_to_ctype,
     PLCSimpleDataType,
     PLCDataType,
-    STRING_BUFFER,
 )
 from .filetimes import filetime_to_dt
 from .pyads_ex import (
@@ -449,7 +448,7 @@ def get_all_symbols(self) -> List[AdsSymbol]:
             symbol_size_msg = self.read(
                 ADSIGRP_SYM_UPLOADINFO2,
                 ADSIOFFS_DEVDATA_ADSSTATE,
-                PLCTYPE_STRING * STRING_BUFFER,
+                PLCTYPE_STRING,
                 return_ctypes=True,
             )
             sym_count = struct.unpack("I", symbol_size_msg[0:4])[0]

pyads/pyads_ex.py

@@ -840,7 +840,11 @@ def adsSyncReadReqEx2(
     index_offset_c = ctypes.c_ulong(index_offset)
 
     # Strings were handled specifically before, but their sizes are contained and we
-    # can proceed as normal:
+    # can proceed as normal
+    if data_type == PLCTYPE_STRING or data_type == PLCTYPE_WSTRING:
+        # This implies a string of size 1, which is so are we instead use a large fixed
+        # size buffer:
+        data_type = data_type * STRING_BUFFER
     data = data_type()
 
     data_pointer = ctypes.pointer(data)

tests/test_connection_class.py

@@ -231,9 +231,11 @@ def test_read_string(self):
         # Assert that the server received the correct command
         self.assert_command_id(requests[0], constants.ADSCOMMAND_READ)
 
-        # We are reading only a single character, which the test-server defaults at 0
-        expected_result = "\x00"
-        self.assertEqual(expected_result, result)
+        # The string buffer is 1024 bytes long, this will be filled with \x0F
+        # and null terminated with \x00 by our test server. The \x00 will get
+        # chopped off during parsing to python string type
+        expected_result = "\x0F" * 1023
+        self.assertEqual(result, expected_result)
 
     def test_write_uint(self):
         value = 100