Change Log
Neighborhood Technologies
--------------------------

*Note: I thought I'd put a change log file here so developers can keep some
track of the changes they are making before something more concrete is done.
--------------------------
-
Changes Commited on 2007-02-04
-
Fixed bug [ 1645254 ] error in foldLine
Wasn't an error in the fold line per-say, but I found out that when parsing a VCard
it wasn't adding the note object to the VCard object. Also fixed a bug when adding
multiple extended values (it was adding a CRLF in the wrong StringBuilder). Added
support for the parser to read ENCODING=QUOTED-PRINTABLE or just QUOTED-PRINTABLE
they're both valid under VCard 2.1 standard, and by default cardme will print
them as ENCODING=QUOTED-PRINTABLE .. just seems proper.

-
Changes Commited on 2006-10-04
-
Fixed a bug with the ISO dates, where it would be 1 month behind.
Removed dependency on sun.* classes by using public domain Base64 class. It also seems to be faster than Sun's.
Improved some of the code in Base64Wrapper.


-
Changes Commited on 2006-10-04
-
Applied Chris Hubicks mime-dir-patch : [ 1568087 ] Implement Mime-Dir and vCard Java API
Cardme (CVS version) now depends on mime-dir-j-2.1.0-1 package. Developers can download
the source code from http://sourceforge.net/projects/mime-dir-j and compile the code and
add the jar file (as a dependency to cardme).


-
Changes Commited on 2006-10-03
-

1: Strings are initialized with "" and use compareTo() checks, rather htan just using null.
   - Fixed, now all classes use null for initialization and checks

2: Calling hasXXX(), getXXX(), setXXX(), or removeXXX() on list type methods has the side
   efffect of creating an empty list of that type.
   - NullPointerException will be thrown if any of those methods are accessed before the
   list is created. The list gets created automatically on "addXXX()" or if user sets a
   pre-made one using "setXXX()"

3: The setXXX() list methods do not allow one to set all elements to null.
   - It should now.

4: Double values use the 'double' primitive type, initialized with Double.MAX_VALUE,
   rather than using Double class initialized to null.
   - Using Double class now.

5: VCardWriter calls needsFolding() and then foldLine(), rather than just having foldLine()
   do a length check.
   - Fixed. foldLine() calls needsFolding()

6: All the classes implementing the Type interface don't have a common base class.
   - Now they do have one called AbstractType

7: The PreferredAddressImpl base class is not abstract
   -It is now abstract and renamed to AbstractPreferredAddress
   
8: All the classes implementing VCardImage subinterfaces(Logo/Photo) don't have a common base class.
   Likewise with VCardSound
   - They now share a common base class called AbstractVCardData, to which its subclasses can implement
     VCardData, Photo, Sound, or Logo interfaces.

9: VCard Parser could not parse inline data
   - VCardParser now parses inline data for Key, logo, photo, and sound tags.

10: Folding/Unfolding not consistent with MIME-DIR
    - This is fixed in this version. Cardme now folds and unfolds lines and labels with full MIME-DIR
      compliance. VCardReader unfolds the vcard. VCardWriter folds the vcard. The VCardParser assumes
      to get a fully unfolded VCard string.
-
Changes Commited on 2006-09-27
-
1: Added more bug fixes and moved the adding of the tokens from outside
   the VCardParser to inside it. By default the vcard parser will be
   initialized with v2.1 vcard tokens. Any extra extensions will have
   to be added manually.

2: Support for MIME-DIR line folding for lines that exceed 75 characters
   has been fully implemented. VCardReader should be used to read VCF
   files or strings and will unfold the entire VCard. Or one can alternatively
   use the VCardUtils to do folding and unfolding. Folding of lines is
   done individually before the vcard is written out as a string.

3: Added a "HOW_TO.TXT" to give a quick and dirty intro on how to use cardme.

TODO: The vcard parser still has problems reading in VCards that have
      BASE64 (binary) inline data. But writing out vcards with inline
      data is ok. Once I get it to recognize and parse inline data
      properly, I'll work on v3.0 RFC.

-
Changes Commited on 2006-09-19
-
The Cardme package implements the VCard RFC standard for v2.1
I tried to add as much support as I could possible. The VCard object
it self I think fully supports v2.1 vcards. It has support for adding
BASE64 encodes of images and sounds and any type of file ..etc.

I'll list the things that I didn't yet implement.

1: VCards are supposed to have a limit of 76 characters per line (or something)
   And if it goes over, the next line must be indented by 1 space.

2: I didn't add support for multiple character encodings. Though there is
   enough flexibility and scalability in the cardme library to do so. Right
   now we just assume everything is in English.

3: I wrote a parser to read and write VCards to .VCF format. It ouputs a
   String and the user writes that to a File. Writing the VCard works fine.
   Reading them in doesn't have support for reading BASE64 blocks of data.
   For now it works with giving the URL value. This is a feature I need to
   add in the VCard Reader/Parser.
   
4: There is a lot of Microsoft compatibility modes I added in the VCard Parser
   to read in and parse VCard files produced from MS Outlook 2003. Aparently
   someone doesn't follow the standards. Check in source comments for more
   details.

Aside from that the cardme library seems to work well with me. I enjoyed
working on it and hope that it evolves further.

Disclaimer:
-
I have worked on Cardme using Eclipse and my setup for it is a bit different
from the CVS. It looks like this structure is suited to using Maven. I have
not touched any of the Maven xml files or hibernate xml files. Everything
inside the "cardme" folder should compile all together cleanly with the
Java 1.5 compiler.

