[zanata/zanata-platform] 3421d6: fix(ZNTA-2247): Remove magic numbers for glossary ...

[GitHub]

  Branch: refs/heads/znta-2247-find-glossary-columns
  Home:   https://github.com/zanata/zanata-platform
  Commit: 3421d6bcc5c58619580565c3c882b9af652fe2ef
      https://github.com/zanata/zanata-platform/commit/3421d6bcc5c58619580565c3c882b9af652fe2ef
  Author: Damian Jansen <djansen at redhat.com>
  Date:   2017-10-22 (Sun, 22 Oct 2017)

  Changed paths:
    M common/zanata-adapter-glossary/src/main/java/org/zanata/adapter/glossary/GlossaryCSVReader.java

  Log Message:
  -----------
  fix(ZNTA-2247): Remove magic numbers for glossary csv columns

Still expects column 0 to be the source locale, but checks that the
PoS and Description columns are named. This prevents garbage in
said fields.
This change also prevents repeatedly setting these fields, only
doing so once per row.