tp: open files with utf-8 encoding in gen_amalgamted_sql.py (google#3343)

Fixes google#3342

----

In Korean locales it is not unusual that the platform default encoding
on Windows is the EUC-KR codepage, so on these and other systems that
don't default to UTF-8, it would help to have any build script that
loads UTF-8 files make that explicit in their code.

Signed-off-by: Christian W. Damus <cdamus.ext@eclipsesource.com>

Author: cdamus

tools/gen_amalgamated_sql.py

@@ -87,7 +87,7 @@ def main():
 
   sql_files = []
   if args.input_list_file:
-    with open(args.input_list_file, 'r') as input_list_file:
+    with open(args.input_list_file, 'r', encoding='utf-8') as input_list_file:
       for line in input_list_file.read().splitlines():
         sql_files.append(line)
   else:
@@ -102,7 +102,7 @@ def main():
   # Extract the SQL output from each file.
   sql_outputs = {}
   for file_name in sql_files:
-    with open(file_name, 'r') as f:
+    with open(file_name, 'r', encoding='utf-8') as f:
       relpath = os.path.relpath(file_name, root_dir)
 
       # We've had bugs (e.g. b/264711057) when Soong's common path logic breaks
@@ -111,7 +111,7 @@ def main():
       assert '../' not in relpath
       sql_outputs[relpath] = f.read()
 
-  with open(args.cpp_out, 'w+') as output:
+  with open(args.cpp_out, 'w+', encoding='utf-8') as output:
     output.write(REPLACEMENT_HEADER)
     output.write(NAMESPACE_BEGIN.format(args.namespace))