Regsitrar: bulk change - does not work with windows line change and without "Domain" in header row

[vohmar]

• bulk change reads the csv file incorrectly if the file has windows style LF+CR line endings by loosing domain names in transfer request for example. System needs to handle both windows and unix new line formats as there are still quite many windows users.
• first row as header is required by the import script, but for the bulk op to work the names of the first two columns must be exactly "Domain,Transfer code". This is undocumented and unnecessary requirement - first(header) row should be ignored.
• looks like domain names read from csv are case sensitive - that is wrong
• error messages need improving: 'domain not found', 'transfer code not found', 'transfer code is wrong'

[virtualfeline]

• To clarify, it looks like the file read itself does not fail as such but reading the domain fields fails, if the 1st column label in header is not "Domain", just like reading the transfer code fields fails when 2nd label is not "Transfer code".

• Also, the processing of domain names is case sensitive and the import expects them to be in lowercase. Uppercase input in domain names generates "not found" error for the name field. This is undesirable.

• Also, errors generated are worded too laconically and as such are no help to the user. For instance user cannot differentiate whether the 'not found' error refers to domain name or transfer code field. And in cases where the transfer contains domains the user already owns, the error should make mention of the fact instead of "transfer code is wrong" which it is, of course, but even if it were correct the op makes no sense and the script should check for it and tell you. It looks like it does not, if you upload a list from your own domains, it will just transfer it from yourself to yourself.

[vohmar]

case description updated accordingly