<char code='0a'/><char code='0a'/> in the 1.0.2 import files is causing the parser to strip out text. Put in a replacement string to code around the problem
Moved check for define later to earlier in the call - it was setting "content in use" before checking define later and this would end up with content in use and define later = true