Github user kaknikhil commented on a diff in the pull request:
https://github.com/apache/madlib/pull/271#discussion_r191587596
--- Diff: src/madpack/madpack.py ---
@@ -131,10 +141,73 @@ def _get_relative_maddir(maddir, port):
return maddir
#
------------------------------------------------------------------------------
+def _cleanup_comments_in_sqlfile(output_filename, upgrade):
+ """
+ @brief Remove comments in the sql script
+ """
+ if not upgrade:
+ with open(output_filename, 'r+') as output_filehandle:
+ full_sql = output_filehandle.read()
+ pattern =
re.compile(r"""(/\*(.|[\r\n])*?\*/)|(--(.*|[\r\n]))""")
+ res = ''
+ lines = re.split(r'[\r\n]+', full_sql)
+ for line in lines:
+ tmp = line
+ if not tmp.strip().startswith("E'"):
+ line = re.sub(pattern, '', line)
+ res += line + '\n'
+ full_sql = res.strip()
+ full_sql = re.sub(pattern, '', full_sql).strip()
+ # Re-write the cleaned-up sql to a new file. Python does not let us
--- End diff --
can we move the new file creation and the renaming logic to a different
function? This way the function will have a single responsibility of just
cleaning the input.
---
