CardDavMATE really shorts all vCard attributes because then I can find/remove duplicate

lines using "uniq()", but the END:VCARD is not sorted if your vCard is correct ...

valid vCard looks like:

BEGIN:VCARD\r\n

...

...

END:VCARD\r\n

are you sure that your vCard not looks like this invalid vCard:

BEGIN:VCARD\r\n

...

...

END:VCARD\r\n

?

JM

p.s.: the next version will not sort vCard attributes, there is another algorithm to

remove duplicate attribute/value pairs

On 02 Dec 2013, at 15:18, Johan Vromans <jvromans@squirrel.nl> wrote:

Johan Vromans <jvromans@squirrel.nl> writes:

[...] I keep getting a 'vCard is
not conformant to RFC 2426' error on every entry in my addresses
database. New entries initially appear correctly, but after a short
while (30 secs?) the entry on the screen is replaced with the same
non-conformant error message.

Some debugging shows that the vCard entry that is processed internally
(addressbook.js, line 1206) looks like:

 <![CDATA[BEGIN:VCARD
          ADR;TYPE=HOME:**********
          CATEGORIES:Unfiled
          EMAIL;TYPE=INTERNET:***************
          END:VCARD
          FN:**********
          N:**************
          REV:2001-07-07T13:00:00+01:00
          TEL;TYPE=HOME:***********
          UID:palm-addressbook-00c57015-00003880-Johan Vromans@phoenix.squirrel.nl
          VERSION:3.0
          X-ALSO:***]]>

In other words, the attributes are sorted alphabetically, causing
parsing to fail (I assume) due to a misplaced END:VCARD.

When I retrieve the entry directly from DAViCal, using wget:

 BEGIN:VCARD
 VERSION:3.0
 PRODID:-//Johan Vromans//NONSGML pdb2vcard 0.01//EN
 UID:palm-addressbook-00c57015-00003880-Johan Vromans@phoenix.squirrel.nl

 CATEGORIES:Unfiled
 FN:*******
 N:*******
 ADR;TYPE=HOME:**********
 TEL;TYPE=HOME:**********
 EMAIL:********
 X-ALSO:******
 REV:2001-07-07T13:00:00+01:00
 END:VCARD

(The empty line contains a single space, apparently a continuation line
from the UID.)

Any idea what is going on here?

-- Johan