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 "dictionary/structure/pt_common/bigram/bigram_list_read_write_utils.h"
18
19 #include "dictionary/utils/byte_array_utils.h"
20 #include "dictionary/utils/buffer_with_extendable_buffer.h"
21
22 namespace latinime {
23
24 const BigramListReadWriteUtils::BigramFlags BigramListReadWriteUtils::MASK_ATTRIBUTE_ADDRESS_TYPE =
25 0x30;
26 const BigramListReadWriteUtils::BigramFlags
27 BigramListReadWriteUtils::FLAG_ATTRIBUTE_ADDRESS_TYPE_ONEBYTE = 0x10;
28 const BigramListReadWriteUtils::BigramFlags
29 BigramListReadWriteUtils::FLAG_ATTRIBUTE_ADDRESS_TYPE_TWOBYTES = 0x20;
30 const BigramListReadWriteUtils::BigramFlags
31 BigramListReadWriteUtils::FLAG_ATTRIBUTE_ADDRESS_TYPE_THREEBYTES = 0x30;
32 const BigramListReadWriteUtils::BigramFlags
33 BigramListReadWriteUtils::FLAG_ATTRIBUTE_OFFSET_NEGATIVE = 0x40;
34 // Flag for presence of more attributes
35 const BigramListReadWriteUtils::BigramFlags BigramListReadWriteUtils::FLAG_ATTRIBUTE_HAS_NEXT =
36 0x80;
37 // Mask for attribute probability, stored on 4 bits inside the flags byte.
38 const BigramListReadWriteUtils::BigramFlags
39 BigramListReadWriteUtils::MASK_ATTRIBUTE_PROBABILITY = 0x0F;
40
getBigramEntryPropertiesAndAdvancePosition(const ReadOnlyByteArrayView buffer,BigramFlags * const outBigramFlags,int * const outTargetPtNodePos,int * const bigramEntryPos)41 /* static */ bool BigramListReadWriteUtils::getBigramEntryPropertiesAndAdvancePosition(
42 const ReadOnlyByteArrayView buffer, BigramFlags *const outBigramFlags,
43 int *const outTargetPtNodePos, int *const bigramEntryPos) {
44 if (static_cast<int>(buffer.size()) <= *bigramEntryPos) {
45 AKLOGE("Read invalid pos in getBigramEntryPropertiesAndAdvancePosition(). bufSize: %zd, "
46 "bigramEntryPos: %d.", buffer.size(), *bigramEntryPos);
47 return false;
48 }
49 const BigramFlags bigramFlags = ByteArrayUtils::readUint8AndAdvancePosition(buffer.data(),
50 bigramEntryPos);
51 if (outBigramFlags) {
52 *outBigramFlags = bigramFlags;
53 }
54 const int targetPos = getBigramAddressAndAdvancePosition(buffer, bigramFlags, bigramEntryPos);
55 if (outTargetPtNodePos) {
56 *outTargetPtNodePos = targetPos;
57 }
58 return true;
59 }
60
skipExistingBigrams(const ReadOnlyByteArrayView buffer,int * const bigramListPos)61 /* static */ bool BigramListReadWriteUtils::skipExistingBigrams(const ReadOnlyByteArrayView buffer,
62 int *const bigramListPos) {
63 BigramFlags flags;
64 do {
65 if (!getBigramEntryPropertiesAndAdvancePosition(buffer, &flags, 0 /* outTargetPtNodePos */,
66 bigramListPos)) {
67 return false;
68 }
69 } while(hasNext(flags));
70 return true;
71 }
72
getBigramAddressAndAdvancePosition(const ReadOnlyByteArrayView buffer,const BigramFlags flags,int * const pos)73 /* static */ int BigramListReadWriteUtils::getBigramAddressAndAdvancePosition(
74 const ReadOnlyByteArrayView buffer, const BigramFlags flags, int *const pos) {
75 int offset = 0;
76 const int origin = *pos;
77 switch (MASK_ATTRIBUTE_ADDRESS_TYPE & flags) {
78 case FLAG_ATTRIBUTE_ADDRESS_TYPE_ONEBYTE:
79 offset = ByteArrayUtils::readUint8AndAdvancePosition(buffer.data(), pos);
80 break;
81 case FLAG_ATTRIBUTE_ADDRESS_TYPE_TWOBYTES:
82 offset = ByteArrayUtils::readUint16AndAdvancePosition(buffer.data(), pos);
83 break;
84 case FLAG_ATTRIBUTE_ADDRESS_TYPE_THREEBYTES:
85 offset = ByteArrayUtils::readUint24AndAdvancePosition(buffer.data(), pos);
86 break;
87 }
88 if (isOffsetNegative(flags)) {
89 return origin - offset;
90 } else {
91 return origin + offset;
92 }
93 }
94
95 } // namespace latinime
96