Small cleanups

Change-Id: Ib66507b8934bc8019a762d24d5311411e044ec84
main
Ken Wakasa 2012-08-17 13:06:28 +09:00
parent e2ac3932e3
commit de8a9a8227
4 changed files with 17 additions and 16 deletions

View File

@ -29,8 +29,6 @@ class BigramDictionary {
BigramDictionary(const unsigned char *dict, int maxWordLength, int maxPredictions);
int getBigrams(const int32_t *word, int length, int *inputCodes, int codesSize,
unsigned short *outWords, int *frequencies, int *outputTypes) const;
int getBigramListPositionForWord(const int32_t *prevWord, const int prevWordLength,
const bool forceLowerCaseSearch) const;
void fillBigramAddressToFrequencyMapAndFilter(const int32_t *prevWord, const int prevWordLength,
std::map<int, int> *map, uint8_t *filter) const;
bool isValidBigram(const int32_t *word1, int length1, const int32_t *word2, int length2) const;
@ -45,6 +43,8 @@ class BigramDictionary {
bool getFirstBitOfByte(int *pos) { return (DICT[*pos] & 0x80) > 0; }
bool getSecondBitOfByte(int *pos) { return (DICT[*pos] & 0x40) > 0; }
bool checkFirstCharacter(unsigned short *word, int *inputCodes) const;
int getBigramListPositionForWord(const int32_t *prevWord, const int prevWordLength,
const bool forceLowerCaseSearch) const;
const unsigned char *DICT;
const int MAX_WORD_LENGTH;

View File

@ -61,13 +61,6 @@ class BinaryFormat {
static const int FLAG_ATTRIBUTE_ADDRESS_TYPE_TWOBYTES = 0x20;
static const int FLAG_ATTRIBUTE_ADDRESS_TYPE_THREEBYTES = 0x30;
private:
DISALLOW_IMPLICIT_CONSTRUCTORS(BinaryFormat);
const static int32_t MINIMAL_ONE_BYTE_CHARACTER_VALUE = 0x20;
const static int32_t CHARACTER_ARRAY_TERMINATOR = 0x1F;
const static int MULTIPLE_BYTE_CHARACTER_ADDITIONAL_SIZE = 2;
public:
const static int UNKNOWN_FORMAT = -1;
// Originally, format version 1 had a 16-bit magic number, then the version number `01'
// then options that must be 0. Hence the first 32-bits of the format are always as follow
@ -94,7 +87,6 @@ class BinaryFormat {
static int skipFrequency(const uint8_t flags, const int pos);
static int skipShortcuts(const uint8_t *const dict, const uint8_t flags, const int pos);
static int skipBigrams(const uint8_t *const dict, const uint8_t flags, const int pos);
static int skipAllAttributes(const uint8_t *const dict, const uint8_t flags, const int pos);
static int skipChildrenPosAndAttributes(const uint8_t *const dict, const uint8_t flags,
const int pos);
static int readChildrenPosition(const uint8_t *const dict, const uint8_t flags, const int pos);
@ -118,6 +110,13 @@ class BinaryFormat {
REQUIRES_FRENCH_LIGATURES_PROCESSING = 0x4
};
const static unsigned int NO_FLAGS = 0;
private:
DISALLOW_IMPLICIT_CONSTRUCTORS(BinaryFormat);
const static int32_t MINIMAL_ONE_BYTE_CHARACTER_VALUE = 0x20;
const static int32_t CHARACTER_ARRAY_TERMINATOR = 0x1F;
const static int MULTIPLE_BYTE_CHARACTER_ADDITIONAL_SIZE = 2;
static int skipAllAttributes(const uint8_t *const dict, const uint8_t flags, const int pos);
};
inline int BinaryFormat::detectFormat(const uint8_t *const dict) {

View File

@ -889,7 +889,7 @@ static int compare_pair_capital(const void *a, const void *b) {
- static_cast<int>((static_cast<const struct LatinCapitalSmallPair *>(b))->capital);
}
unsigned short latin_tolower(unsigned short c) {
unsigned short latin_tolower(const unsigned short c) {
struct LatinCapitalSmallPair *p =
static_cast<struct LatinCapitalSmallPair *>(bsearch(&c, SORTED_CHAR_MAP,
sizeof(SORTED_CHAR_MAP) / sizeof(SORTED_CHAR_MAP[0]),

View File

@ -17,21 +17,23 @@
#ifndef LATINIME_CHAR_UTILS_H
#define LATINIME_CHAR_UTILS_H
#include <cctype>
namespace latinime {
inline static int isAsciiUpper(unsigned short c) {
return c >= 'A' && c <= 'Z';
inline static bool isAsciiUpper(unsigned short c) {
return isupper(static_cast<int>(c)) != 0;
}
inline static unsigned short toAsciiLower(unsigned short c) {
return c - 'A' + 'a';
}
inline static int isAscii(unsigned short c) {
return c <= 127;
inline static bool isAscii(unsigned short c) {
return isascii(static_cast<int>(c)) != 0;
}
unsigned short latin_tolower(unsigned short c);
unsigned short latin_tolower(const unsigned short c);
/**
* Table mapping most combined Latin, Greek, and Cyrillic characters