x509: allow leading text in PEM files

[HonzaCholasta]

This fixes a regression introduced in commit
b8d6524.

https://fedorahosted.org/freeipa/ticket/4985

[tiran]

NACK

The ^ is correct because the regular expression must search for a line that starts with -----BEGIN CERTIFICATE-----. I cannot reproduce the issue locally. The regexp matches a cert with leading text:

>>> import re
>>> regexp = u"^-----BEGIN CERTIFICATE-----(.*?)-----END CERTIFICATE-----"
>>> pem = u"leading line\n-----BEGIN CERTIFICATE-----\nabcd\n-----END CERTIFICATE-----\ntrailing text"
>>> re.search(regexp, pem, re.MULTILINE | re.DOTALL)
<_sre.SRE_Match object at 0x7f667778d0a8>
>>> re.search(regexp, pem, re.MULTILINE | re.DOTALL).group(1)
u'\nabcd\n'

[HonzaCholasta]

Oops, didn't realize that ^ matches beginning of each line in multiline mode. I think we can keep the test, though.

[tiran]

Yes, please keep the test. It should pass with the current regular expression, too.

[HonzaCholasta]

Fixed upstream
master:
https://fedorahosted.org/freeipa/changeset/89dfbab3ca076812590f371c21abcb51b350170b