ALL: Browse Folders sorting uppercase before lowercase

[GoogleCodeExporter]

Your model / PRS+ version?
PRS 650 / Nightly R028

What steps will reproduce the problem?
1. Look at the attached screenshot

2.1. Sorting makes Upper-case go before lower-case

2.2. Deleting doesn't work. I click "Term" to no result.

2.3. More over, both this window title and the previous option menu item use 
"Remove collections" string (STR_NODE_TITLE_DELETE_COLLECTIONS or 
STR_UI_MENU_DELETE_COLLECTIONS). But here we do not have folders but 
collections.

Original issue reported on code.google.com by alr...@gmail.com on 3 Mar 2012 at 8:01

Attachments:

• 2012-3-2_0.jpg

[GoogleCodeExporter]

Folders are made to mimic collection nodes to enable search. That makes the 
regular collection options, such as 'delete collections' show up as well, but 
since folders are not actually collections, they don't (and shouldn't) work - 
just ignore them.

The sorting does seem to be off though, even after removing my search 
modifications. Mikheil: perhaps you know more about this?

Original comment by quisvir on 3 Mar 2012 at 3:55

• Changed title: Browse Folders sorting uppercase before lowercase

[GoogleCodeExporter]

Folders use "unicode" sort order. (why have mixed case folders anyway?)
I don't quite get what collections have to do with it...

Original comment by msukhias...@gmail.com on 3 Mar 2012 at 8:06

[GoogleCodeExporter]

Why have mixed case folders anyway?
>> Because default Sony has. Look at Internal_Memory.jpg with database, Digital 
Editions and ScreenShots

OTOH, I think you should do some kind of "smart" sorting process. I mean, look 
at Sorting_Test_Page1 and Sorting_Test_Page2. For a plain user, who does not 
know about "unicode", they are just WRONG.

You're current system does not correctly sort neither ç nor ñ. I'm not sure 
about ç, as it's not used in Spanish (although I think it's mixed with c, I 
mean, for sorting order, c and ç are equivalent). French or Portuguese help 
needed.

But I can tell you that ñ is a letter with all the same rights any other one 
has and its order is between n and o. Spanish alphabet goes ...-l-m-n-ñ-o-p-...

Original comment by alr...@gmail.com on 4 Mar 2012 at 6:09

Attachments:

• Internal_Memory.jpg
• Sorting_Test_Page1.jpg
• Sorting_Test_Page2.jpg

[GoogleCodeExporter]

(Double post but for a good reason).

I'll post it also in FR thread, but there I cannot attach screenshots so I use 
bug-tracker.

Sony sorting is also wrong. Look at the screenshot.

+ Sony correctly mixes upper and lower case.
+ But it doesn't detect Ç and Ñ as normal latin letters.
+ And I've just noticed that accented letters are also wrong. Look at the last 
books in "*" section. (Forget about the starting "El", it's a Spanish article 
and it's correctly filtered out since the fixes made for 
http://code.google.com/p/prs-plus/issues/detail?id=287). "ú" should be treated 
like "u", but it appears as an "out-of-alphabet" letter. (And ü, á, é, í, 
ó...)

Original comment by alr...@gmail.com on 4 Mar 2012 at 6:38

Attachments:

• Wrong_Sony_Sorting.jpg

[GoogleCodeExporter]

There is a new google-doc spreedsheet PRS+Alphabet&Sort I've set up.
Reason is to collect Sortorder and possible popup-characters for the keyboard
Don't know if a custon sort could be implemented without tweaking the binaries 
but lets start to collect the needed Infos
https://docs.google.com/spreadsheet/ccc?key=0AvEpfPLpz-UKdDB0MUtSaUJjaFRDVzFkcjB
FX0xmRkE

Original comment by Mark.N...@gentlemansclub.de on 4 Mar 2012 at 9:09

[GoogleCodeExporter]

Ok I'll download and modify and update es version ASAP.

Nevertheless one suggestion about sorting which your spreadsheet doesn't seem 
to cover:

+ There are some characters which are equivalent for sorting. Example lower and 
upper case and accented not accented letters. 

I mean: sorting "ultimo" (1st singular person in present of "ultimar" - to give 
the final touches to something) before "última" (last in feminine) is wrong. 
It doesn't matter if "u" uses accent or not. You can chose if you want "ultimo" 
before or after "último", (last in masculine), but "última" and "ultima" (3d 
singular person in present of "ultimar") always should go before "ultimo" and 
"último". 

+ OTOH, "ñ" does always go after "n" and before "o". (And as I said before, I 
don't know if this also happens with "ç" or if "ç" is like an accented "c").

So I think you need more info than what you are covering in your spreadsheet.

Original comment by alr...@gmail.com on 4 Mar 2012 at 9:41

[GoogleCodeExporter]

>You're current system does not correctly sort neither ç nor ñ
They come _after_ lower/upper case Latin letters in Unicode tables and are 
sorted accordingly.

Please note that sorting ("collation") is a complex subject, there is no 
universal sorting that fits all, each language has its quirks even if you stick 
to the same alphabet.

Original comment by msukhias...@gmail.com on 4 Mar 2012 at 9:47

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Sony's code (in general) mixes uppercase and lowercase simply by adding 
toLowerCase(), which we could also do. For complex characters, Mikheil is 
absolutely right in saying it's incredibly complicated (I was not aware of 
this, tbh). See for example:
http://www.unicode.org/reports/tr10/

We could consider implementing something like this:
http://www.indiscripts.com/post/2010/10/alphabetical-sort-in-javascript-and-inde
sign
but I reckon these things very quickly become overkill for our purposes (due to 
memory/cpu limits).

Original comment by quisvir on 4 Mar 2012 at 3:12

[GoogleCodeExporter]

The problem here is, in case of "books" view, it's not JS, but binary function. 
We currently use porkupan's binary, without having source for it.

Putting it on hold for a while, we might get back to it later.

Original comment by msukhias...@gmail.com on 9 Mar 2012 at 12:26

• Changed title: ALL: Browse Folders sorting uppercase before lowercase
• Changed state: WontFix