tesseract/ccutil/unicharset.h

986 lines
40 KiB
C++

///////////////////////////////////////////////////////////////////////
// File: unicharset.h
// Description: Unicode character/ligature set class.
// Author: Thomas Kielbus
// Created: Wed Jun 28 17:05:01 PDT 2006
//
// (C) Copyright 2006, Google Inc.
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
// http://www.apache.org/licenses/LICENSE-2.0
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
///////////////////////////////////////////////////////////////////////
#ifndef TESSERACT_CCUTIL_UNICHARSET_H__
#define TESSERACT_CCUTIL_UNICHARSET_H__
#include "errcode.h"
#include "genericvector.h"
#include "helpers.h"
#include "serialis.h"
#include "strngs.h"
#include "tesscallback.h"
#include "unichar.h"
#include "unicharmap.h"
// Enum holding special values of unichar_id. Every unicharset has these.
// Warning! Keep in sync with kSpecialUnicharCodes.
enum SpecialUnicharCodes {
UNICHAR_SPACE,
UNICHAR_JOINED,
UNICHAR_BROKEN,
SPECIAL_UNICHAR_CODES_COUNT
};
class CHAR_FRAGMENT {
public:
// Minimum number of characters used for fragment representation.
static const int kMinLen = 6;
// Maximum number of characters used for fragment representation.
static const int kMaxLen = 3 + UNICHAR_LEN + 2;
// Maximum number of fragments per character.
static const int kMaxChunks = 5;
// Setters and Getters.
inline void set_all(const char *unichar, int pos, int total, bool natural) {
set_unichar(unichar);
set_pos(pos);
set_total(total);
set_natural(natural);
}
inline void set_unichar(const char *uch) {
strncpy(this->unichar, uch, UNICHAR_LEN);
this->unichar[UNICHAR_LEN] = '\0';
}
inline void set_pos(int p) { this->pos = p; }
inline void set_total(int t) { this->total = t; }
inline const char* get_unichar() const { return this->unichar; }
inline int get_pos() const { return this->pos; }
inline int get_total() const { return this->total; }
// Returns the string that represents a fragment
// with the given unichar, pos and total.
static STRING to_string(const char *unichar, int pos, int total,
bool natural);
// Returns the string that represents this fragment.
STRING to_string() const {
return to_string(unichar, pos, total, natural);
}
// Checks whether a fragment has the same unichar,
// position and total as the given inputs.
inline bool equals(const char *other_unichar,
int other_pos, int other_total) const {
return (strcmp(this->unichar, other_unichar) == 0 &&
this->pos == other_pos && this->total == other_total);
}
inline bool equals(const CHAR_FRAGMENT *other) const {
return this->equals(other->get_unichar(),
other->get_pos(),
other->get_total());
}
// Checks whether a given fragment is a continuation of this fragment.
// Assumes that the given fragment pointer is not NULL.
inline bool is_continuation_of(const CHAR_FRAGMENT *fragment) const {
return (strcmp(this->unichar, fragment->get_unichar()) == 0 &&
this->total == fragment->get_total() &&
this->pos == fragment->get_pos() + 1);
}
// Returns true if this fragment is a beginning fragment.
inline bool is_beginning() const { return this->pos == 0; }
// Returns true if this fragment is an ending fragment.
inline bool is_ending() const { return this->pos == this->total-1; }
// Returns true if the fragment was a separate component to begin with,
// ie did not need chopping to be isolated, but may have been separated
// out from a multi-outline blob.
inline bool is_natural() const { return natural; }
void set_natural(bool value) { natural = value; }
// Parses the string to see whether it represents a character fragment
// (rather than a regular character). If so, allocates memory for a new
// CHAR_FRAGMENT instance and fills it in with the corresponding fragment
// information. Fragments are of the form:
// |m|1|2, meaning chunk 1 of 2 of character m, or
// |:|1n2, meaning chunk 1 of 2 of character :, and no chopping was needed
// to divide the parts, as they were already separate connected components.
//
// If parsing succeeded returns the pointer to the allocated CHAR_FRAGMENT
// instance, otherwise (if the string does not represent a fragment or it
// looks like it does, but parsing it as a fragment fails) returns NULL.
//
// Note: The caller is responsible for deallocating memory
// associated with the returned pointer.
static CHAR_FRAGMENT *parse_from_string(const char *str);
private:
char unichar[UNICHAR_LEN + 1];
// True if the fragment was a separate component to begin with,
// ie did not need chopping to be isolated, but may have been separated
// out from a multi-outline blob.
bool natural;
inT16 pos; // fragment position in the character
inT16 total; // total number of fragments in the character
};
// The UNICHARSET class is an utility class for Tesseract that holds the
// set of characters that are used by the engine. Each character is identified
// by a unique number, from 0 to (size - 1).
class UNICHARSET {
public:
// Custom list of characters and their ligature forms (UTF8)
// These map to unicode values in the private use area (PUC) and are supported
// by only few font families (eg. Wyld, Adobe Caslon Pro).
static const char* kCustomLigatures[][2];
// List of strings for the SpecialUnicharCodes. Keep in sync with the enum.
static const char* kSpecialUnicharCodes[SPECIAL_UNICHAR_CODES_COUNT];
// ICU 2.0 UCharDirection enum (from third_party/icu/include/unicode/uchar.h)
enum Direction {
U_LEFT_TO_RIGHT = 0,
U_RIGHT_TO_LEFT = 1,
U_EUROPEAN_NUMBER = 2,
U_EUROPEAN_NUMBER_SEPARATOR = 3,
U_EUROPEAN_NUMBER_TERMINATOR = 4,
U_ARABIC_NUMBER = 5,
U_COMMON_NUMBER_SEPARATOR = 6,
U_BLOCK_SEPARATOR = 7,
U_SEGMENT_SEPARATOR = 8,
U_WHITE_SPACE_NEUTRAL = 9,
U_OTHER_NEUTRAL = 10,
U_LEFT_TO_RIGHT_EMBEDDING = 11,
U_LEFT_TO_RIGHT_OVERRIDE = 12,
U_RIGHT_TO_LEFT_ARABIC = 13,
U_RIGHT_TO_LEFT_EMBEDDING = 14,
U_RIGHT_TO_LEFT_OVERRIDE = 15,
U_POP_DIRECTIONAL_FORMAT = 16,
U_DIR_NON_SPACING_MARK = 17,
U_BOUNDARY_NEUTRAL = 18,
U_CHAR_DIRECTION_COUNT
};
// Create an empty UNICHARSET
UNICHARSET();
~UNICHARSET();
// Return the UNICHAR_ID of a given unichar representation within the
// UNICHARSET.
const UNICHAR_ID unichar_to_id(const char* const unichar_repr) const;
// Return the UNICHAR_ID of a given unichar representation within the
// UNICHARSET. Only the first length characters from unichar_repr are used.
const UNICHAR_ID unichar_to_id(const char* const unichar_repr,
int length) const;
// Return the minimum number of bytes that matches a legal UNICHAR_ID,
// while leaving the rest of the string encodable. Returns 0 if the
// beginning of the string is not encodable.
// WARNING: this function now encodes the whole string for precision.
// Use encode_string in preference to repeatedly calling step.
int step(const char* str) const;
// Return whether the given UTF-8 string is encodable with this UNICHARSET.
// If not encodable, write the first byte offset which cannot be converted
// into the second (return) argument.
bool encodable_string(const char *str, int *first_bad_position) const;
// Encodes the given UTF-8 string with this UNICHARSET.
// Any part of the string that cannot be encoded (because the utf8 can't
// be broken up into pieces that are in the unicharset) then:
// if give_up_on_failure, stops and returns a partial encoding,
// else continues and inserts an INVALID_UNICHAR_ID in the returned encoding.
// Returns true if the encoding succeeds completely, false if there is at
// least one failure.
// If lengths is not NULL, then it is filled with the corresponding
// byte length of each encoded UNICHAR_ID.
// If encoded_length is not NULL then on return it contains the length of
// str that was encoded. (if give_up_on_failure the location of the first
// failure, otherwise strlen(str).)
bool encode_string(const char* str, bool give_up_on_failure,
GenericVector<UNICHAR_ID>* encoding,
GenericVector<char>* lengths,
int* encoded_length) const;
// Return the unichar representation corresponding to the given UNICHAR_ID
// within the UNICHARSET.
const char* const id_to_unichar(UNICHAR_ID id) const;
// Return the UTF8 representation corresponding to the given UNICHAR_ID after
// resolving any private encodings internal to Tesseract. This method is
// preferrable to id_to_unichar for outputting text that will be visible to
// external applications.
const char* const id_to_unichar_ext(UNICHAR_ID id) const;
// Return a STRING that reformats the utf8 str into the str followed
// by its hex unicodes.
static STRING debug_utf8_str(const char* str);
// Return a STRING containing debug information on the unichar, including
// the id_to_unichar, its hex unicodes and the properties.
STRING debug_str(UNICHAR_ID id) const;
STRING debug_str(const char * unichar_repr) const {
return debug_str(unichar_to_id(unichar_repr));
}
// Add a unichar representation to the set.
void unichar_insert(const char* const unichar_repr);
// Return true if the given unichar id exists within the set.
// Relies on the fact that unichar ids are contiguous in the unicharset.
bool contains_unichar_id(UNICHAR_ID unichar_id) const {
return unichar_id != INVALID_UNICHAR_ID && unichar_id < size_used &&
unichar_id >= 0;
}
// Return true if the given unichar representation exists within the set.
bool contains_unichar(const char* const unichar_repr) const;
bool contains_unichar(const char* const unichar_repr, int length) const;
// Return true if the given unichar representation corresponds to the given
// UNICHAR_ID within the set.
bool eq(UNICHAR_ID unichar_id, const char* const unichar_repr) const;
// Delete CHAR_FRAGMENTs stored in properties of unichars array.
void delete_pointers_in_unichars() {
for (int i = 0; i < size_used; ++i) {
if (unichars[i].properties.fragment != NULL) {
delete unichars[i].properties.fragment;
unichars[i].properties.fragment = NULL;
}
}
}
// Clear the UNICHARSET (all the previous data is lost).
void clear() {
if (script_table != NULL) {
for (int i = 0; i < script_table_size_used; ++i)
delete[] script_table[i];
delete[] script_table;
script_table = NULL;
script_table_size_used = 0;
}
if (unichars != NULL) {
delete_pointers_in_unichars();
delete[] unichars;
unichars = NULL;
}
script_table_size_reserved = 0;
size_reserved = 0;
size_used = 0;
ids.clear();
top_bottom_set_ = false;
script_has_upper_lower_ = false;
script_has_xheight_ = false;
null_sid_ = 0;
common_sid_ = 0;
latin_sid_ = 0;
cyrillic_sid_ = 0;
greek_sid_ = 0;
han_sid_ = 0;
hiragana_sid_ = 0;
katakana_sid_ = 0;
}
// Return the size of the set (the number of different UNICHAR it holds).
int size() const {
return size_used;
}
// Reserve enough memory space for the given number of UNICHARS
void reserve(int unichars_number);
// Opens the file indicated by filename and saves unicharset to that file.
// Returns true if the operation is successful.
bool save_to_file(const char * const filename) const {
FILE* file = fopen(filename, "w+b");
if (file == NULL) return false;
bool result = save_to_file(file);
fclose(file);
return result;
}
// Saves the content of the UNICHARSET to the given file.
// Returns true if the operation is successful.
bool save_to_file(FILE *file) const {
STRING str;
if (!save_to_string(&str)) return false;
if (fwrite(&str[0], str.length(), 1, file) != 1) return false;
return true;
}
bool save_to_file(tesseract::TFile *file) const {
STRING str;
if (!save_to_string(&str)) return false;
if (file->FWrite(&str[0], str.length(), 1) != 1) return false;
return true;
}
// Saves the content of the UNICHARSET to the given STRING.
// Returns true if the operation is successful.
bool save_to_string(STRING *str) const;
// Load a unicharset from a unicharset file that has been loaded into
// the given memory buffer.
// Returns true if the operation is successful.
bool load_from_inmemory_file(const char* const memory, int mem_size,
bool skip_fragments);
// Returns true if the operation is successful.
bool load_from_inmemory_file(const char* const memory, int mem_size) {
return load_from_inmemory_file(memory, mem_size, false);
}
// Opens the file indicated by filename and loads the UNICHARSET
// from the given file. The previous data is lost.
// Returns true if the operation is successful.
bool load_from_file(const char* const filename, bool skip_fragments) {
FILE* file = fopen(filename, "rb");
if (file == NULL) return false;
bool result = load_from_file(file, skip_fragments);
fclose(file);
return result;
}
// returns true if the operation is successful.
bool load_from_file(const char* const filename) {
return load_from_file(filename, false);
}
// Loads the UNICHARSET from the given file. The previous data is lost.
// Returns true if the operation is successful.
bool load_from_file(FILE *file, bool skip_fragments);
bool load_from_file(FILE *file) { return load_from_file(file, false); }
bool load_from_file(tesseract::TFile *file, bool skip_fragments);
// Sets up internal data after loading the file, based on the char
// properties. Called from load_from_file, but also needs to be run
// during set_unicharset_properties.
void post_load_setup();
// Returns true if right_to_left scripts are significant in the unicharset,
// but without being so sensitive that "universal" unicharsets containing
// characters from many scripts, like orientation and script detection,
// look like they are right_to_left.
bool major_right_to_left() const;
// Set a whitelist and/or blacklist of characters to recognize.
// An empty or NULL whitelist enables everything (minus any blacklist).
// An empty or NULL blacklist disables nothing.
// An empty or NULL unblacklist has no effect.
// The blacklist overrides the whitelist.
// The unblacklist overrides the blacklist.
// Each list is a string of utf8 character strings. Boundaries between
// unicharset units are worked out automatically, and characters not in
// the unicharset are silently ignored.
void set_black_and_whitelist(const char* blacklist, const char* whitelist,
const char* unblacklist);
// Set the isalpha property of the given unichar to the given value.
void set_isalpha(UNICHAR_ID unichar_id, bool value) {
unichars[unichar_id].properties.isalpha = value;
}
// Set the islower property of the given unichar to the given value.
void set_islower(UNICHAR_ID unichar_id, bool value) {
unichars[unichar_id].properties.islower = value;
}
// Set the isupper property of the given unichar to the given value.
void set_isupper(UNICHAR_ID unichar_id, bool value) {
unichars[unichar_id].properties.isupper = value;
}
// Set the isdigit property of the given unichar to the given value.
void set_isdigit(UNICHAR_ID unichar_id, bool value) {
unichars[unichar_id].properties.isdigit = value;
}
// Set the ispunctuation property of the given unichar to the given value.
void set_ispunctuation(UNICHAR_ID unichar_id, bool value) {
unichars[unichar_id].properties.ispunctuation = value;
}
// Set the isngram property of the given unichar to the given value.
void set_isngram(UNICHAR_ID unichar_id, bool value) {
unichars[unichar_id].properties.isngram = value;
}
// Set the script name of the given unichar to the given value.
// Value is copied and thus can be a temporary;
void set_script(UNICHAR_ID unichar_id, const char* value) {
unichars[unichar_id].properties.script_id = add_script(value);
}
// Set other_case unichar id in the properties for the given unichar id.
void set_other_case(UNICHAR_ID unichar_id, UNICHAR_ID other_case) {
unichars[unichar_id].properties.other_case = other_case;
}
// Set the direction property of the given unichar to the given value.
void set_direction(UNICHAR_ID unichar_id, UNICHARSET::Direction value) {
unichars[unichar_id].properties.direction = value;
}
// Set mirror unichar id in the properties for the given unichar id.
void set_mirror(UNICHAR_ID unichar_id, UNICHAR_ID mirror) {
unichars[unichar_id].properties.mirror = mirror;
}
// Record normalized version of unichar with the given unichar_id.
void set_normed(UNICHAR_ID unichar_id, const char* normed) {
unichars[unichar_id].properties.normed = normed;
unichars[unichar_id].properties.normed_ids.truncate(0);
}
// Sets the normed_ids vector from the normed string. normed_ids is not
// stored in the file, and needs to be set when the UNICHARSET is loaded.
void set_normed_ids(UNICHAR_ID unichar_id);
// Return the isalpha property of the given unichar.
bool get_isalpha(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return false;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.isalpha;
}
// Return the islower property of the given unichar.
bool get_islower(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return false;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.islower;
}
// Return the isupper property of the given unichar.
bool get_isupper(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return false;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.isupper;
}
// Return the isdigit property of the given unichar.
bool get_isdigit(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return false;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.isdigit;
}
// Return the ispunctuation property of the given unichar.
bool get_ispunctuation(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return false;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.ispunctuation;
}
// Return the isngram property of the given unichar.
bool get_isngram(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return false;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.isngram;
}
// Returns whether the unichar id represents a unicode value in the private
// use area.
bool get_isprivate(UNICHAR_ID unichar_id) const;
// Returns true if the ids have useful min/max top/bottom values.
bool top_bottom_useful() const {
return top_bottom_set_;
}
// Sets all ranges to empty, so they can be expanded to set the values.
void set_ranges_empty();
// Sets all the properties for this unicharset given a src_unicharset with
// everything set. The unicharsets don't have to be the same, and graphemes
// are correctly accounted for.
void SetPropertiesFromOther(const UNICHARSET& src) {
PartialSetPropertiesFromOther(0, src);
}
// Sets properties from Other, starting only at the given index.
void PartialSetPropertiesFromOther(int start_index, const UNICHARSET& src);
// Expands the tops and bottoms and widths for this unicharset given a
// src_unicharset with ranges in it. The unicharsets don't have to be the
// same, and graphemes are correctly accounted for.
void ExpandRangesFromOther(const UNICHARSET& src);
// Makes this a copy of src. Clears this completely first, so the automattic
// ids will not be present in this if not in src.
void CopyFrom(const UNICHARSET& src);
// For each id in src, if it does not occur in this, add it, as in
// SetPropertiesFromOther, otherwise expand the ranges, as in
// ExpandRangesFromOther.
void AppendOtherUnicharset(const UNICHARSET& src);
// Returns true if the acceptable ranges of the tops of the characters do
// not overlap, making their x-height calculations distinct.
bool SizesDistinct(UNICHAR_ID id1, UNICHAR_ID id2) const;
// Returns the min and max bottom and top of the given unichar in
// baseline-normalized coordinates, ie, where the baseline is
// kBlnBaselineOffset and the meanline is kBlnBaselineOffset + kBlnXHeight
// (See normalis.h for the definitions).
void get_top_bottom(UNICHAR_ID unichar_id,
int* min_bottom, int* max_bottom,
int* min_top, int* max_top) const {
if (INVALID_UNICHAR_ID == unichar_id) {
*min_bottom = *min_top = 0;
*max_bottom = *max_top = 256; // kBlnCellHeight
return;
}
ASSERT_HOST(contains_unichar_id(unichar_id));
*min_bottom = unichars[unichar_id].properties.min_bottom;
*max_bottom = unichars[unichar_id].properties.max_bottom;
*min_top = unichars[unichar_id].properties.min_top;
*max_top = unichars[unichar_id].properties.max_top;
}
void set_top_bottom(UNICHAR_ID unichar_id,
int min_bottom, int max_bottom,
int min_top, int max_top) {
unichars[unichar_id].properties.min_bottom =
static_cast<uinT8>(ClipToRange(min_bottom, 0, MAX_UINT8));
unichars[unichar_id].properties.max_bottom =
static_cast<uinT8>(ClipToRange(max_bottom, 0, MAX_UINT8));
unichars[unichar_id].properties.min_top =
static_cast<uinT8>(ClipToRange(min_top, 0, MAX_UINT8));
unichars[unichar_id].properties.max_top =
static_cast<uinT8>(ClipToRange(max_top, 0, MAX_UINT8));
}
// Returns the width stats (as mean, sd) of the given unichar relative to the
// median advance of all characters in the character set.
void get_width_stats(UNICHAR_ID unichar_id,
float* width, float* width_sd) const {
if (INVALID_UNICHAR_ID == unichar_id) {
*width = 0.0f;
*width_sd = 0.0f;;
return;
}
ASSERT_HOST(contains_unichar_id(unichar_id));
*width = unichars[unichar_id].properties.width;
*width_sd = unichars[unichar_id].properties.width_sd;
}
void set_width_stats(UNICHAR_ID unichar_id, float width, float width_sd) {
unichars[unichar_id].properties.width = width;
unichars[unichar_id].properties.width_sd = width_sd;
}
// Returns the stats of the x-bearing (as mean, sd) of the given unichar
// relative to the median advance of all characters in the character set.
void get_bearing_stats(UNICHAR_ID unichar_id,
float* bearing, float* bearing_sd) const {
if (INVALID_UNICHAR_ID == unichar_id) {
*bearing = *bearing_sd = 0.0f;
return;
}
ASSERT_HOST(contains_unichar_id(unichar_id));
*bearing = unichars[unichar_id].properties.bearing;
*bearing_sd = unichars[unichar_id].properties.bearing_sd;
}
void set_bearing_stats(UNICHAR_ID unichar_id,
float bearing, float bearing_sd) {
unichars[unichar_id].properties.bearing = bearing;
unichars[unichar_id].properties.bearing_sd = bearing_sd;
}
// Returns the stats of the x-advance of the given unichar (as mean, sd)
// relative to the median advance of all characters in the character set.
void get_advance_stats(UNICHAR_ID unichar_id,
float* advance, float* advance_sd) const {
if (INVALID_UNICHAR_ID == unichar_id) {
*advance = *advance_sd = 0;
return;
}
ASSERT_HOST(contains_unichar_id(unichar_id));
*advance = unichars[unichar_id].properties.advance;
*advance_sd = unichars[unichar_id].properties.advance_sd;
}
void set_advance_stats(UNICHAR_ID unichar_id,
float advance, float advance_sd) {
unichars[unichar_id].properties.advance = advance;
unichars[unichar_id].properties.advance_sd = advance_sd;
}
// Returns true if the font metrics properties are empty.
bool PropertiesIncomplete(UNICHAR_ID unichar_id) const {
return unichars[unichar_id].properties.AnyRangeEmpty();
}
// Return the script name of the given unichar.
// The returned pointer will always be the same for the same script, it's
// managed by unicharset and thus MUST NOT be deleted
int get_script(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return null_sid_;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.script_id;
}
// Return the character properties, eg. alpha/upper/lower/digit/punct,
// as a bit field of unsigned int.
unsigned int get_properties(UNICHAR_ID unichar_id) const;
// Return the character property as a single char. If a character has
// multiple attributes, the main property is defined by the following order:
// upper_case : 'A'
// lower_case : 'a'
// alpha : 'x'
// digit : '0'
// punctuation: 'p'
char get_chartype(UNICHAR_ID unichar_id) const;
// Get other_case unichar id in the properties for the given unichar id.
UNICHAR_ID get_other_case(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return INVALID_UNICHAR_ID;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.other_case;
}
// Returns the direction property of the given unichar.
Direction get_direction(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return UNICHARSET::U_OTHER_NEUTRAL;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.direction;
}
// Get mirror unichar id in the properties for the given unichar id.
UNICHAR_ID get_mirror(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return INVALID_UNICHAR_ID;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.mirror;
}
// Returns UNICHAR_ID of the corresponding lower-case unichar.
UNICHAR_ID to_lower(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return INVALID_UNICHAR_ID;
ASSERT_HOST(contains_unichar_id(unichar_id));
if (unichars[unichar_id].properties.islower) return unichar_id;
return unichars[unichar_id].properties.other_case;
}
// Returns UNICHAR_ID of the corresponding upper-case unichar.
UNICHAR_ID to_upper(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return INVALID_UNICHAR_ID;
ASSERT_HOST(contains_unichar_id(unichar_id));
if (unichars[unichar_id].properties.isupper) return unichar_id;
return unichars[unichar_id].properties.other_case;
}
// Returns true if this UNICHARSET has the special codes in
// SpecialUnicharCodes available. If false then there are normal unichars
// at these codes and they should not be used.
bool has_special_codes() const {
return get_fragment(UNICHAR_BROKEN) != NULL &&
strcmp(id_to_unichar(UNICHAR_BROKEN),
kSpecialUnicharCodes[UNICHAR_BROKEN]) == 0;
}
// Returns true if there are any repeated unicodes in the normalized
// text of any unichar-id in the unicharset.
bool AnyRepeatedUnicodes() const;
// Return a pointer to the CHAR_FRAGMENT class if the given
// unichar id represents a character fragment.
const CHAR_FRAGMENT *get_fragment(UNICHAR_ID unichar_id) const {
if (INVALID_UNICHAR_ID == unichar_id) return NULL;
ASSERT_HOST(contains_unichar_id(unichar_id));
return unichars[unichar_id].properties.fragment;
}
// Return the isalpha property of the given unichar representation.
bool get_isalpha(const char* const unichar_repr) const {
return get_isalpha(unichar_to_id(unichar_repr));
}
// Return the islower property of the given unichar representation.
bool get_islower(const char* const unichar_repr) const {
return get_islower(unichar_to_id(unichar_repr));
}
// Return the isupper property of the given unichar representation.
bool get_isupper(const char* const unichar_repr) const {
return get_isupper(unichar_to_id(unichar_repr));
}
// Return the isdigit property of the given unichar representation.
bool get_isdigit(const char* const unichar_repr) const {
return get_isdigit(unichar_to_id(unichar_repr));
}
// Return the ispunctuation property of the given unichar representation.
bool get_ispunctuation(const char* const unichar_repr) const {
return get_ispunctuation(unichar_to_id(unichar_repr));
}
// Return the character properties, eg. alpha/upper/lower/digit/punct,
// of the given unichar representation
unsigned int get_properties(const char* const unichar_repr) const {
return get_properties(unichar_to_id(unichar_repr));
}
char get_chartype(const char* const unichar_repr) const {
return get_chartype(unichar_to_id(unichar_repr));
}
// Return the script name of the given unichar representation.
// The returned pointer will always be the same for the same script, it's
// managed by unicharset and thus MUST NOT be deleted
int get_script(const char* const unichar_repr) const {
return get_script(unichar_to_id(unichar_repr));
}
// Return a pointer to the CHAR_FRAGMENT class struct if the given
// unichar representation represents a character fragment.
const CHAR_FRAGMENT *get_fragment(const char* const unichar_repr) const {
if (unichar_repr == NULL || unichar_repr[0] == '\0' ||
!ids.contains(unichar_repr)) {
return NULL;
}
return get_fragment(unichar_to_id(unichar_repr));
}
// Return the isalpha property of the given unichar representation.
// Only the first length characters from unichar_repr are used.
bool get_isalpha(const char* const unichar_repr,
int length) const {
return get_isalpha(unichar_to_id(unichar_repr, length));
}
// Return the islower property of the given unichar representation.
// Only the first length characters from unichar_repr are used.
bool get_islower(const char* const unichar_repr,
int length) const {
return get_islower(unichar_to_id(unichar_repr, length));
}
// Return the isupper property of the given unichar representation.
// Only the first length characters from unichar_repr are used.
bool get_isupper(const char* const unichar_repr,
int length) const {
return get_isupper(unichar_to_id(unichar_repr, length));
}
// Return the isdigit property of the given unichar representation.
// Only the first length characters from unichar_repr are used.
bool get_isdigit(const char* const unichar_repr,
int length) const {
return get_isdigit(unichar_to_id(unichar_repr, length));
}
// Return the ispunctuation property of the given unichar representation.
// Only the first length characters from unichar_repr are used.
bool get_ispunctuation(const char* const unichar_repr,
int length) const {
return get_ispunctuation(unichar_to_id(unichar_repr, length));
}
// Returns normalized version of unichar with the given unichar_id.
const char *get_normed_unichar(UNICHAR_ID unichar_id) const {
if (unichar_id == UNICHAR_SPACE && has_special_codes()) return " ";
return unichars[unichar_id].properties.normed.string();
}
// Returns a vector of UNICHAR_IDs that represent the ids of the normalized
// version of the given id. There may be more than one UNICHAR_ID in the
// vector if unichar_id represents a ligature.
const GenericVector<UNICHAR_ID>& normed_ids(UNICHAR_ID unichar_id) const {
return unichars[unichar_id].properties.normed_ids;
}
// Return the script name of the given unichar representation.
// Only the first length characters from unichar_repr are used.
// The returned pointer will always be the same for the same script, it's
// managed by unicharset and thus MUST NOT be deleted
int get_script(const char* const unichar_repr,
int length) const {
return get_script(unichar_to_id(unichar_repr, length));
}
// Return the (current) number of scripts in the script table
int get_script_table_size() const {
return script_table_size_used;
}
// Return the script string from its id
const char* get_script_from_script_id(int id) const {
if (id >= script_table_size_used || id < 0)
return null_script;
return script_table[id];
}
// Returns the id from the name of the script, or 0 if script is not found.
// Note that this is an expensive operation since it involves iteratively
// comparing strings in the script table. To avoid dependency on STL, we
// won't use a hash. Instead, the calling function can use this to lookup
// and save the ID for relevant scripts for fast comparisons later.
int get_script_id_from_name(const char* script_name) const;
// Return true if the given script is the null script
bool is_null_script(const char* script) const {
return script == null_script;
}
// Uniquify the given script. For two scripts a and b, if strcmp(a, b) == 0,
// then the returned pointer will be the same.
// The script parameter is copied and thus can be a temporary.
int add_script(const char* script);
// Return the enabled property of the given unichar.
bool get_enabled(UNICHAR_ID unichar_id) const {
return unichars[unichar_id].properties.enabled;
}
int null_sid() const { return null_sid_; }
int common_sid() const { return common_sid_; }
int latin_sid() const { return latin_sid_; }
int cyrillic_sid() const { return cyrillic_sid_; }
int greek_sid() const { return greek_sid_; }
int han_sid() const { return han_sid_; }
int hiragana_sid() const { return hiragana_sid_; }
int katakana_sid() const { return katakana_sid_; }
int default_sid() const { return default_sid_; }
// Returns true if the unicharset has the concept of upper/lower case.
bool script_has_upper_lower() const {
return script_has_upper_lower_;
}
// Returns true if the unicharset has the concept of x-height.
// script_has_xheight can be true even if script_has_upper_lower is not,
// when the script has a sufficiently predominant top line with ascenders,
// such as Devanagari and Thai.
bool script_has_xheight() const {
return script_has_xheight_;
}
private:
struct UNICHAR_PROPERTIES {
UNICHAR_PROPERTIES();
// Initializes all properties to sensible default values.
void Init();
// Sets all ranges wide open. Initialization default in case there are
// no useful values available.
void SetRangesOpen();
// Sets all ranges to empty. Used before expanding with font-based data.
void SetRangesEmpty();
// Returns true if any of the top/bottom/width/bearing/advance ranges/stats
// is emtpy.
bool AnyRangeEmpty() const;
// Expands the ranges with the ranges from the src properties.
void ExpandRangesFrom(const UNICHAR_PROPERTIES& src);
// Copies the properties from src into this.
void CopyFrom(const UNICHAR_PROPERTIES& src);
bool isalpha;
bool islower;
bool isupper;
bool isdigit;
bool ispunctuation;
bool isngram;
bool enabled;
// Possible limits of the top and bottom of the bounding box in
// baseline-normalized coordinates, ie, where the baseline is
// kBlnBaselineOffset and the meanline is kBlnBaselineOffset + kBlnXHeight
// (See normalis.h for the definitions).
uinT8 min_bottom;
uinT8 max_bottom;
uinT8 min_top;
uinT8 max_top;
// Statstics of the widths of bounding box, relative to the median advance.
float width;
float width_sd;
// Stats of the x-bearing and advance, also relative to the median advance.
float bearing;
float bearing_sd;
float advance;
float advance_sd;
int script_id;
UNICHAR_ID other_case; // id of the corresponding upper/lower case unichar
Direction direction; // direction of this unichar
// Mirror property is useful for reverse DAWG lookup for words in
// right-to-left languages (e.g. "(word)" would be in
// '[open paren]' 'w' 'o' 'r' 'd' '[close paren]' in a UTF8 string.
// However, what we want in our DAWG is
// '[open paren]', 'd', 'r', 'o', 'w', '[close paren]' not
// '[close paren]', 'd', 'r', 'o', 'w', '[open paren]'.
UNICHAR_ID mirror;
// A string of unichar_ids that represent the corresponding normed string.
// For awkward characters like em-dash, this gives hyphen.
// For ligatures, this gives the string of normal unichars.
GenericVector<UNICHAR_ID> normed_ids;
STRING normed; // normalized version of this unichar
// Contains meta information about the fragment if a unichar represents
// a fragment of a character, otherwise should be set to NULL.
// It is assumed that character fragments are added to the unicharset
// after the corresponding 'base' characters.
CHAR_FRAGMENT *fragment;
};
struct UNICHAR_SLOT {
char representation[UNICHAR_LEN + 1];
UNICHAR_PROPERTIES properties;
};
// Internal recursive version of encode_string above.
// str is the start of the whole string.
// str_index is the current position in str.
// str_length is the length of str.
// encoding is a working encoding of str.
// lengths is a working set of lengths of each element of encoding.
// best_total_length is the longest length of str that has been successfully
// encoded so far.
// On return:
// best_encoding contains the encoding that used the longest part of str.
// best_lengths (may be null) contains the lengths of best_encoding.
void encode_string(const char* str, int str_index, int str_length,
GenericVector<UNICHAR_ID>* encoding,
GenericVector<char>* lengths,
int* best_total_length,
GenericVector<UNICHAR_ID>* best_encoding,
GenericVector<char>* best_lengths) const;
// Gets the properties for a grapheme string, combining properties for
// multiple characters in a meaningful way where possible.
// Returns false if no valid match was found in the unicharset.
// NOTE that script_id, mirror, and other_case refer to this unicharset on
// return and will need redirecting if the target unicharset is different.
bool GetStrProperties(const char* utf8_str,
UNICHAR_PROPERTIES* props) const;
// Load ourselves from a "file" where our only interface to the file is
// an implementation of fgets(). This is the parsing primitive accessed by
// the public routines load_from_file() and load_from_inmemory_file().
bool load_via_fgets(TessResultCallback2<char *, char *, int> *fgets_cb,
bool skip_fragments);
UNICHAR_SLOT* unichars;
UNICHARMAP ids;
int size_used;
int size_reserved;
char** script_table;
int script_table_size_used;
int script_table_size_reserved;
const char* null_script;
// True if the unichars have their tops/bottoms set.
bool top_bottom_set_;
// True if the unicharset has significant upper/lower case chars.
bool script_has_upper_lower_;
// True if the unicharset has a significant mean-line with significant
// ascenders above that.
bool script_has_xheight_;
// A few convenient script name-to-id mapping without using hash.
// These are initialized when unicharset file is loaded. Anything
// missing from this list can be looked up using get_script_id_from_name.
int null_sid_;
int common_sid_;
int latin_sid_;
int cyrillic_sid_;
int greek_sid_;
int han_sid_;
int hiragana_sid_;
int katakana_sid_;
// The most frequently occurring script in the charset.
int default_sid_;
};
#endif // TESSERACT_CCUTIL_UNICHARSET_H__