Examples of MorphData

Examples of com.code972.hebmorph.MorphData

103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
    }

    public static WordType isRecognizedWord(final String word, final boolean tolerate) {
        byte prefLen = 0;
        Integer prefixMask;
        MorphData md;

        if (customWords != null) {
            try {
                if (customWords.lookup(word) != null) return WordType.CUSTOM;
            } catch (IllegalArgumentException e) {
            }

            while (true) {
                // Make sure there are at least 2 letters left after the prefix (the words של, שלא for example)
                if (word.length() - prefLen < 2)
                    break;

                try {
                    prefixMask = prefixesTree.lookup(word.substring(0, ++prefLen));
                } catch (IllegalArgumentException e) {
                    break;
                }

                try {
                    md = customWords.lookup(word.substring(prefLen));
                } catch (IllegalArgumentException e) {
                    md = null;
                }
                if ((md != null) && ((md.getPrefixes() & prefixMask) > 0)) {
                    for (int result = 0; result < md.getLemmas().length; result++) {
                        if ((LingInfo.DMask2ps(md.getDescFlags()[result]) & prefixMask) > 0) {
                            return WordType.CUSTOM_WITH_PREFIX;
                        }
                    }
                }
            }
        }

        if (!isHebrewWord(word))
            return WordType.NON_HEBREW;

        try {
            if (dictRadix.lookup(word) != null) return WordType.HEBREW;
        } catch (IllegalArgumentException e) {
        }

        if (word.endsWith("'")) { // Try ommitting closing Geresh
            try {
                if (dictRadix.lookup(word.substring(0, word.length() - 1)) != null) return WordType.HEBREW;
            } catch (IllegalArgumentException e) {
            }
        }

        prefLen = 0;
        while (true) {
            // Make sure there are at least 2 letters left after the prefix (the words של, שלא for example)
            if (word.length() - prefLen < 2)
                break;

            try {
                prefixMask = prefixesTree.lookup(word.substring(0, ++prefLen));
            } catch (IllegalArgumentException e) {
                break;
            }

            try {
                md = dictRadix.lookup(word.substring(prefLen));
            } catch (IllegalArgumentException e) {
                md = null;
            }
            if ((md != null) && ((md.getPrefixes() & prefixMask) > 0)) {
                for (int result = 0; result < md.getLemmas().length; result++) {
                    if ((LingInfo.DMask2ps(md.getDescFlags()[result]) & prefixMask) > 0) {
                        return WordType.HEBREW_WITH_PREFIX;
                    }
                }
            }
        }
View Full Code Here

Examples of com.code972.hebmorph.MorphData

110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
    }

    public static WordType isRecognizedWord(final String word, final boolean tolerate) {
        byte prefLen = 0;
        Integer prefixMask;
        MorphData md;

        if (customWords != null) {
            try {
                if (customWords.lookup(word) != null) return WordType.CUSTOM;
            } catch (IllegalArgumentException ignored_ex) {
            }

            while (true) {
                // Make sure there are at least 2 letters left after the prefix (the words של, שלא for example)
                if (word.length() - prefLen < 2)
                    break;

                if ((prefixMask = prefixesTree.get(word.substring(0, ++prefLen))) == null)
                    break;

                try {
                    md = customWords.lookup(word.substring(prefLen));
                } catch (IllegalArgumentException ignored_ex) {
                    md = null;
                }
                if ((md != null) && ((md.getPrefixes() & prefixMask) > 0)) {
                    for (int result = 0; result < md.getLemmas().length; result++) {
                        if ((LingInfo.DMask2ps(md.getDescFlags()[result]) & prefixMask) > 0) {
                            return WordType.CUSTOM_WITH_PREFIX;
                        }
                    }
                }
            }
        }

        if (!isHebrewWord(word))
            return WordType.NON_HEBREW;

        try {
            if (dictRadix.lookup(word) != null) return WordType.HEBREW;
        } catch (IllegalArgumentException ignored_ex) {
        }

        if (word.endsWith("'")) { // Try ommitting closing Geresh
            try {
                if (dictRadix.lookup(word.substring(0, word.length() - 1)) != null) return WordType.HEBREW;
            } catch (IllegalArgumentException ignored_ex) {
            }
        }

        prefLen = 0;
        while (true) {
            // Make sure there are at least 2 letters left after the prefix (the words של, שלא for example)
            if (word.length() - prefLen < 2)
                break;

            if ((prefixMask = prefixesTree.get(word.substring(0, ++prefLen))) == null)
                break;

            try {
                md = dictRadix.lookup(word.substring(prefLen));
            } catch (IllegalArgumentException e) {
                md = null;
            }
            if ((md != null) && ((md.getPrefixes() & prefixMask) > 0)) {
                for (int result = 0; result < md.getLemmas().length; result++) {
                    if ((LingInfo.DMask2ps(md.getDescFlags()[result]) & prefixMask) > 0) {
                        return WordType.HEBREW_WITH_PREFIX;
                    }
                }
            }
        }
View Full Code Here

Examples of com.code972.hebmorph.MorphData

102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
    }

    public static WordType isRecognizedWord(final String word, final boolean tolerate) {
        byte prefLen = 0;
        Integer prefixMask;
        MorphData md;

        try {
            if (customWords.lookup(word) != null) return WordType.CUSTOM;
        } catch (IllegalArgumentException e) {
        }

        while (true) {
            // Make sure there are at least 2 letters left after the prefix (the words של, שלא for example)
            if (word.length() - prefLen < 2)
                break;

            try {
                prefixMask = prefixesTree.lookup(word.substring(0, ++prefLen));
            } catch (IllegalArgumentException e) {
                break;
            }

            try {
                md = customWords.lookup(word.substring(prefLen));
            } catch (IllegalArgumentException e) {
                md = null;
            }
            if ((md != null) && ((md.getPrefixes() & prefixMask) > 0)) {
                for (int result = 0; result < md.getLemmas().length; result++) {
                    if ((LingInfo.DMask2ps(md.getDescFlags()[result]) & prefixMask) > 0) {
                        return WordType.CUSTOM_WITH_PREFIX;
                    }
                }
            }
        }

        try {
            if (dictRadix.lookup(word) != null) return WordType.HEBREW;
        } catch (IllegalArgumentException e) {
        }

        if (word.endsWith("'")) { // Try ommitting closing Geresh
            try {
                if (dictRadix.lookup(word.substring(0, word.length() - 1)) != null) return WordType.HEBREW;
            } catch (IllegalArgumentException e) {
            }
        }

        prefLen = 0;
        while (true) {
            // Make sure there are at least 2 letters left after the prefix (the words של, שלא for example)
            if (word.length() - prefLen < 2)
                break;

            try {
                prefixMask = prefixesTree.lookup(word.substring(0, ++prefLen));
            } catch (IllegalArgumentException e) {
                break;
            }

            try {
                md = dictRadix.lookup(word.substring(prefLen));
            } catch (IllegalArgumentException e) {
                md = null;
            }
            if ((md != null) && ((md.getPrefixes() & prefixMask) > 0)) {
                for (int result = 0; result < md.getLemmas().length; result++) {
                    if ((LingInfo.DMask2ps(md.getDescFlags()[result]) & prefixMask) > 0) {
                        return WordType.HEBREW_WITH_PREFIX;
                    }
                }
            }
        }
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.