Do not try to extract text for empty files

[mmoossen]

If you have an empty file, it will be sent to tika, and tika will tell:

Failed tasks while import & analysis (ETL):
X-TIKA:EXCEPTION:runtime
enhance_extract_text_tika_server
X-TIKA_EXCEPTION_runtime:
org.apache.tika.exception.ZeroByteFileException: InputStream must have > 0 bytes at

one can argue that it does not make any sense to index empty files, but in our use case we want the metadata, for instance file name and path, of the file to be indexed.

i think the best is to check the file size before sending it to tika.