1 /*
2 * Copyright (C) 2013 The Android Open Source Project
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 *
8 * http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
15 */
16
17 #include "char_utils.h"
18 #include "binary_format.h"
19 #include "defines.h"
20 #include "digraph_utils.h"
21
22 namespace latinime {
23
24 const DigraphUtils::digraph_t DigraphUtils::GERMAN_UMLAUT_DIGRAPHS[] =
25 { { 'a', 'e', 0x00E4 }, // U+00E4 : LATIN SMALL LETTER A WITH DIAERESIS
26 { 'o', 'e', 0x00F6 }, // U+00F6 : LATIN SMALL LETTER O WITH DIAERESIS
27 { 'u', 'e', 0x00FC } }; // U+00FC : LATIN SMALL LETTER U WITH DIAERESIS
28 const DigraphUtils::digraph_t DigraphUtils::FRENCH_LIGATURES_DIGRAPHS[] =
29 { { 'a', 'e', 0x00E6 }, // U+00E6 : LATIN SMALL LETTER AE
30 { 'o', 'e', 0x0153 } }; // U+0153 : LATIN SMALL LIGATURE OE
31 const DigraphUtils::DigraphType DigraphUtils::USED_DIGRAPH_TYPES[] =
32 { DIGRAPH_TYPE_GERMAN_UMLAUT, DIGRAPH_TYPE_FRENCH_LIGATURES };
33
hasDigraphForCodePoint(const int dictFlags,const int compositeGlyphCodePoint)34 /* static */ bool DigraphUtils::hasDigraphForCodePoint(
35 const int dictFlags, const int compositeGlyphCodePoint) {
36 const DigraphUtils::DigraphType digraphType = getDigraphTypeForDictionary(dictFlags);
37 if (DigraphUtils::getDigraphForDigraphTypeAndCodePoint(digraphType, compositeGlyphCodePoint)) {
38 return true;
39 }
40 return false;
41 }
42
43 // Returns the digraph type associated with the given dictionary.
getDigraphTypeForDictionary(const int dictFlags)44 /* static */ DigraphUtils::DigraphType DigraphUtils::getDigraphTypeForDictionary(
45 const int dictFlags) {
46 if (BinaryFormat::REQUIRES_GERMAN_UMLAUT_PROCESSING & dictFlags) {
47 return DIGRAPH_TYPE_GERMAN_UMLAUT;
48 }
49 if (BinaryFormat::REQUIRES_FRENCH_LIGATURES_PROCESSING & dictFlags) {
50 return DIGRAPH_TYPE_FRENCH_LIGATURES;
51 }
52 return DIGRAPH_TYPE_NONE;
53 }
54
55 // Retrieves the set of all digraphs associated with the given dictionary flags.
56 // Returns the size of the digraph array, or 0 if none exist.
getAllDigraphsForDictionaryAndReturnSize(const int dictFlags,const DigraphUtils::digraph_t ** const digraphs)57 /* static */ int DigraphUtils::getAllDigraphsForDictionaryAndReturnSize(
58 const int dictFlags, const DigraphUtils::digraph_t **const digraphs) {
59 const DigraphUtils::DigraphType digraphType = getDigraphTypeForDictionary(dictFlags);
60 return getAllDigraphsForDigraphTypeAndReturnSize(digraphType, digraphs);
61 }
62
63 // Returns the digraph codepoint for the given composite glyph codepoint and digraph codepoint index
64 // (which specifies the first or second codepoint in the digraph).
getDigraphCodePointForIndex(const int compositeGlyphCodePoint,const DigraphCodePointIndex digraphCodePointIndex)65 /* static */ int DigraphUtils::getDigraphCodePointForIndex(const int compositeGlyphCodePoint,
66 const DigraphCodePointIndex digraphCodePointIndex) {
67 if (digraphCodePointIndex == NOT_A_DIGRAPH_INDEX) {
68 return NOT_A_CODE_POINT;
69 }
70 const DigraphUtils::digraph_t *const digraph =
71 DigraphUtils::getDigraphForCodePoint(compositeGlyphCodePoint);
72 if (!digraph) {
73 return NOT_A_CODE_POINT;
74 }
75 if (digraphCodePointIndex == FIRST_DIGRAPH_CODEPOINT) {
76 return digraph->first;
77 } else if (digraphCodePointIndex == SECOND_DIGRAPH_CODEPOINT) {
78 return digraph->second;
79 }
80 ASSERT(false);
81 return NOT_A_CODE_POINT;
82 }
83
84 // Retrieves the set of all digraphs associated with the given digraph type.
85 // Returns the size of the digraph array, or 0 if none exist.
getAllDigraphsForDigraphTypeAndReturnSize(const DigraphUtils::DigraphType digraphType,const DigraphUtils::digraph_t ** const digraphs)86 /* static */ int DigraphUtils::getAllDigraphsForDigraphTypeAndReturnSize(
87 const DigraphUtils::DigraphType digraphType,
88 const DigraphUtils::digraph_t **const digraphs) {
89 if (digraphType == DigraphUtils::DIGRAPH_TYPE_GERMAN_UMLAUT) {
90 *digraphs = GERMAN_UMLAUT_DIGRAPHS;
91 return NELEMS(GERMAN_UMLAUT_DIGRAPHS);
92 }
93 if (digraphType == DIGRAPH_TYPE_FRENCH_LIGATURES) {
94 *digraphs = FRENCH_LIGATURES_DIGRAPHS;
95 return NELEMS(FRENCH_LIGATURES_DIGRAPHS);
96 }
97 return 0;
98 }
99
100 /**
101 * Returns the digraph for the input composite glyph codepoint, or 0 if none exists.
102 * compositeGlyphCodePoint: the method returns the digraph corresponding to this codepoint.
103 */
getDigraphForCodePoint(const int compositeGlyphCodePoint)104 /* static */ const DigraphUtils::digraph_t *DigraphUtils::getDigraphForCodePoint(
105 const int compositeGlyphCodePoint) {
106 for (size_t i = 0; i < NELEMS(USED_DIGRAPH_TYPES); i++) {
107 const DigraphUtils::digraph_t *const digraph = getDigraphForDigraphTypeAndCodePoint(
108 USED_DIGRAPH_TYPES[i], compositeGlyphCodePoint);
109 if (digraph) {
110 return digraph;
111 }
112 }
113 return 0;
114 }
115
116 /**
117 * Returns the digraph for the input composite glyph codepoint, or 0 if none exists.
118 * digraphType: the type of digraphs supported.
119 * compositeGlyphCodePoint: the method returns the digraph corresponding to this codepoint.
120 */
getDigraphForDigraphTypeAndCodePoint(const DigraphUtils::DigraphType digraphType,const int compositeGlyphCodePoint)121 /* static */ const DigraphUtils::digraph_t *DigraphUtils::getDigraphForDigraphTypeAndCodePoint(
122 const DigraphUtils::DigraphType digraphType, const int compositeGlyphCodePoint) {
123 const DigraphUtils::digraph_t *digraphs = 0;
124 const int compositeGlyphLowerCodePoint = toLowerCase(compositeGlyphCodePoint);
125 const int digraphsSize =
126 DigraphUtils::getAllDigraphsForDictionaryAndReturnSize(digraphType, &digraphs);
127 for (int i = 0; i < digraphsSize; i++) {
128 if (digraphs[i].compositeGlyph == compositeGlyphLowerCodePoint) {
129 return &digraphs[i];
130 }
131 }
132 return 0;
133 }
134
135 } // namespace latinime
136