author | Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com> |
Thu, 07 Jan 2010 13:31:38 +0200 | |
changeset 37 | cb62a4f66ebe |
parent 36 | 0ed94ceaa377 |
child 38 | 6297cdf66332 |
permissions | -rw-r--r-- |
36
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
1 |
/* |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
2 |
* Copyright (c) 2009 Nokia Corporation and/or its subsidiary(-ies). |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
3 |
* All rights reserved. |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
4 |
* This component and the accompanying materials are made available |
37
cb62a4f66ebe
Revision: 200951
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
36
diff
changeset
|
5 |
* under the terms of the License "Eclipse Public License v1.0" |
36
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
6 |
* which accompanies this distribution, and is available |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
7 |
* at the URL "http://www.eclipse.org/legal/epl-v10.html". |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
8 |
* |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
9 |
* Initial Contributors: |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
10 |
* Nokia Corporation - initial contribution. |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
11 |
* |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
12 |
* Contributors: |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
13 |
* |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
14 |
* Description: |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
15 |
* |
0ed94ceaa377
Revision: 200948
Dremov Kirill (Nokia-D-MSW/Tampere) <kirill.dremov@nokia.com>
parents:
0
diff
changeset
|
16 |
*/ |
0
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
17 |
/************************************************* |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
18 |
* libucp - Unicode Property Table handler * |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
19 |
*************************************************/ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
20 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
21 |
/* Internal header file defining the layout of compact nodes in the tree. */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
22 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
23 |
typedef struct cnode { |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
24 |
unsigned short int f0; |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
25 |
unsigned short int f1; |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
26 |
unsigned short int f2; |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
27 |
} cnode; |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
28 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
29 |
/* Things for the f0 field */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
30 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
31 |
#define f0_leftexists 0x8000 /* Left child exists */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
32 |
#define f0_typemask 0x3f00 /* Type bits */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
33 |
#define f0_typeshift 8 /* Type shift */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
34 |
#define f0_chhmask 0x00ff /* Character high bits */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
35 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
36 |
/* Things for the f2 field */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
37 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
38 |
#define f2_rightmask 0xf000 /* Mask for right offset bits */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
39 |
#define f2_rightshift 12 /* Shift for right offset */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
40 |
#define f2_casemask 0x0fff /* Mask for case offset */ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
41 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
42 |
/* The tree consists of a vector of structures of type cnode, with the root |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
43 |
node as the first element. The three short ints (16-bits) are used as follows: |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
44 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
45 |
(f0) (1) The 0x8000 bit of f0 is set if a left child exists. The child's node |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
46 |
is the next node in the vector. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
47 |
(2) The 0x4000 bits of f0 is spare. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
48 |
(3) The 0x3f00 bits of f0 contain the character type; this is a number |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
49 |
defined by the enumeration in ucp.h (e.g. ucp_Lu). |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
50 |
(4) The bottom 8 bits of f0 contain the most significant byte of the |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
51 |
character's 24-bit codepoint. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
52 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
53 |
(f1) (1) The f1 field contains the two least significant bytes of the |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
54 |
codepoint. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
55 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
56 |
(f2) (1) The 0xf000 bits of f2 contain zero if there is no right child of this |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
57 |
node. Otherwise, they contain one plus the exponent of the power of |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
58 |
two of the offset to the right node (e.g. a value of 3 means 8). The |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
59 |
units of the offset are node items. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
60 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
61 |
(2) The 0x0fff bits of f2 contain the signed offset from this character to |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
62 |
its alternate cased value. They are zero if there is no such |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
63 |
character. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
64 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
65 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
66 |
----------------------------------------------------------------------------- |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
67 |
||.|.| type (6) | ms char (8) || ls char (16) ||....| case offset (12) || |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
68 |
----------------------------------------------------------------------------- |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
69 |
| | | |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
70 |
| |-> spare | |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
71 |
| exponent of right |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
72 |
|-> left child exists child offset |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
73 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
74 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
75 |
The upper/lower casing information is set only for characters that come in |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
76 |
pairs. There are (at present) four non-one-to-one mappings in the Unicode data. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
77 |
These are ignored. They are: |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
78 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
79 |
1FBE Greek Prosgegrammeni (lower, with upper -> capital iota) |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
80 |
2126 Ohm |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
81 |
212A Kelvin |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
82 |
212B Angstrom |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
83 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
84 |
Certainly for the last three, having an alternate case would seem to be a |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
85 |
mistake. I don't know any Greek, so cannot comment on the first one. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
86 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
87 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
88 |
When searching the tree, proceed as follows: |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
89 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
90 |
(1) Start at the first node. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
91 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
92 |
(2) Extract the character value from f1 and the bottom 8 bits of f0; |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
93 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
94 |
(3) Compare with the character being sought. If equal, we are done. |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
95 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
96 |
(4) If the test character is smaller, inspect the f0_leftexists flag. If it is |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
97 |
not set, the character is not in the tree. If it is set, move to the next |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
98 |
node, and go to (2). |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
99 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
100 |
(5) If the test character is bigger, extract the f2_rightmask bits from f2, and |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
101 |
shift them right by f2_rightshift. If the result is zero, the character is |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
102 |
not in the tree. Otherwise, calculate the number of nodes to skip by |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
103 |
shifting the value 1 left by this number minus one. Go to (2). |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
104 |
*/ |
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
105 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
106 |
|
dd21522fd290
Revision: 200911
Kiiskinen Klaus (Nokia-D-MSW/Tampere) <klaus.kiiskinen@nokia.com>
parents:
diff
changeset
|
107 |
/* End of internal.h */ |