Mercurial > hgrepos > Python2 > PyMuPDF

comparison mupdf-source/thirdparty/tesseract/src/ccstruct/fontinfo.cpp @ 2:b50eed0cc0ef upstream

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
ADD: MuPDF v1.26.7: the MuPDF source as downloaded by a default build of PyMuPDF 1.26.4. The directory name has changed: no version number in the expanded directory now.
author Franz Glasner <fzglas.hg@dom66.de>
date Mon, 15 Sep 2025 11:43:07 +0200
parents
children
comparison
equal deleted inserted replaced
1:1d09e1dec1d9 2:b50eed0cc0ef
1 ///////////////////////////////////////////////////////////////////////
2 // File: fontinfo.cpp
3 // Description: Font information classes abstracted from intproto.h/cpp.
4 // Author: rays@google.com (Ray Smith)
5 //
6 // (C) Copyright 2011, Google Inc.
7 // Licensed under the Apache License, Version 2.0 (the "License");
8 // you may not use this file except in compliance with the License.
9 // You may obtain a copy of the License at
10 // http://www.apache.org/licenses/LICENSE-2.0
11 // Unless required by applicable law or agreed to in writing, software
12 // distributed under the License is distributed on an "AS IS" BASIS,
13 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 // See the License for the specific language governing permissions and
15 // limitations under the License.
16 //
17 ///////////////////////////////////////////////////////////////////////
18
19 #include "fontinfo.h"
20 #include "bitvector.h"
21 #include "unicity_table.h"
22
23 namespace tesseract {
24
25 // Writes to the given file. Returns false in case of error.
26 bool FontInfo::Serialize(FILE *fp) const {
27 if (!write_info(fp, *this)) {
28 return false;
29 }
30 if (!write_spacing_info(fp, *this)) {
31 return false;
32 }
33 return true;
34 }
35 // Reads from the given file. Returns false in case of error.
36 // If swap is true, assumes a big/little-endian swap is needed.
37 bool FontInfo::DeSerialize(TFile *fp) {
38 if (!read_info(fp, this)) {
39 return false;
40 }
41 if (!read_spacing_info(fp, this)) {
42 return false;
43 }
44 return true;
45 }
46
47 FontInfoTable::FontInfoTable() {
48 using namespace std::placeholders; // for _1, _2
49 set_clear_callback(std::bind(FontInfoDeleteCallback, _1));
50 }
51
52 FontInfoTable::~FontInfoTable() = default;
53
54 // Writes to the given file. Returns false in case of error.
55 bool FontInfoTable::Serialize(FILE *fp) const {
56 return this->SerializeClasses(fp);
57 }
58 // Reads from the given file. Returns false in case of error.
59 // If swap is true, assumes a big/little-endian swap is needed.
60 bool FontInfoTable::DeSerialize(TFile *fp) {
61 truncate(0);
62 return this->DeSerializeClasses(fp);
63 }
64
65 // Returns true if the given set of fonts includes one with the same
66 // properties as font_id.
67 bool FontInfoTable::SetContainsFontProperties(int font_id,
68 const std::vector<ScoredFont> &font_set) const {
69 uint32_t properties = at(font_id).properties;
70 for (auto &&f : font_set) {
71 if (at(f.fontinfo_id).properties == properties) {
72 return true;
73 }
74 }
75 return false;
76 }
77
78 // Returns true if the given set of fonts includes multiple properties.
79 bool FontInfoTable::SetContainsMultipleFontProperties(
80 const std::vector<ScoredFont> &font_set) const {
81 if (font_set.empty()) {
82 return false;
83 }
84 int first_font = font_set[0].fontinfo_id;
85 uint32_t properties = at(first_font).properties;
86 for (unsigned f = 1; f < font_set.size(); ++f) {
87 if (at(font_set[f].fontinfo_id).properties != properties) {
88 return true;
89 }
90 }
91 return false;
92 }
93
94 // Moves any non-empty FontSpacingInfo entries from other to this.
95 void FontInfoTable::MoveSpacingInfoFrom(FontInfoTable *other) {
96 using namespace std::placeholders; // for _1, _2
97 set_clear_callback(std::bind(FontInfoDeleteCallback, _1));
98 for (unsigned i = 0; i < other->size(); ++i) {
99 std::vector<FontSpacingInfo *> *spacing_vec = other->at(i).spacing_vec;
100 if (spacing_vec != nullptr) {
101 int target_index = get_index(other->at(i));
102 if (target_index < 0) {
103 // Bit copy the FontInfo and steal all the pointers.
104 push_back(other->at(i));
105 other->at(i).name = nullptr;
106 } else {
107 delete at(target_index).spacing_vec;
108 at(target_index).spacing_vec = other->at(i).spacing_vec;
109 }
110 other->at(i).spacing_vec = nullptr;
111 }
112 }
113 }
114
115 // Moves this to the target unicity table.
116 void FontInfoTable::MoveTo(UnicityTable<FontInfo> *target) {
117 target->clear();
118 using namespace std::placeholders; // for _1, _2
119 target->set_clear_callback(std::bind(FontInfoDeleteCallback, _1));
120 for (unsigned i = 0; i < size(); ++i) {
121 // Bit copy the FontInfo and steal all the pointers.
122 target->push_back(at(i));
123 at(i).name = nullptr;
124 at(i).spacing_vec = nullptr;
125 }
126 }
127
128 // Callbacks for GenericVector.
129 void FontInfoDeleteCallback(FontInfo f) {
130 if (f.spacing_vec != nullptr) {
131 for (auto data : *f.spacing_vec) {
132 delete data;
133 }
134 delete f.spacing_vec;
135 f.spacing_vec = nullptr;
136 }
137 delete[] f.name;
138 f.name = nullptr;
139 }
140
141 /*---------------------------------------------------------------------------*/
142 // Callbacks used by UnicityTable to read/write FontInfo/FontSet structures.
143 bool read_info(TFile *f, FontInfo *fi) {
144 uint32_t size;
145 if (!f->DeSerialize(&size)) {
146 return false;
147 }
148 char *font_name = new char[size + 1];
149 fi->name = font_name;
150 if (!f->DeSerialize(font_name, size)) {
151 return false;
152 }
153 font_name[size] = '\0';
154 return f->DeSerialize(&fi->properties);
155 }
156
157 bool write_info(FILE *f, const FontInfo &fi) {
158 int32_t size = strlen(fi.name);
159 return tesseract::Serialize(f, &size) && tesseract::Serialize(f, &fi.name[0], size) &&
160 tesseract::Serialize(f, &fi.properties);
161 }
162
163 bool read_spacing_info(TFile *f, FontInfo *fi) {
164 int32_t vec_size, kern_size;
165 if (!f->DeSerialize(&vec_size)) {
166 return false;
167 }
168 ASSERT_HOST(vec_size >= 0);
169 if (vec_size == 0) {
170 return true;
171 }
172 fi->init_spacing(vec_size);
173 for (int i = 0; i < vec_size; ++i) {
174 auto *fs = new FontSpacingInfo();
175 if (!f->DeSerialize(&fs->x_gap_before) || !f->DeSerialize(&fs->x_gap_after) ||
176 !f->DeSerialize(&kern_size)) {
177 delete fs;
178 return false;
179 }
180 if (kern_size < 0) { // indication of a nullptr entry in fi->spacing_vec
181 delete fs;
182 continue;
183 }
184 if (kern_size > 0 &&
185 (!f->DeSerialize(fs->kerned_unichar_ids) || !f->DeSerialize(fs->kerned_x_gaps))) {
186 delete fs;
187 return false;
188 }
189 fi->add_spacing(i, fs);
190 }
191 return true;
192 }
193
194 bool write_spacing_info(FILE *f, const FontInfo &fi) {
195 int32_t vec_size = (fi.spacing_vec == nullptr) ? 0 : fi.spacing_vec->size();
196 if (!tesseract::Serialize(f, &vec_size)) {
197 return false;
198 }
199 int16_t x_gap_invalid = -1;
200 for (int i = 0; i < vec_size; ++i) {
201 FontSpacingInfo *fs = fi.spacing_vec->at(i);
202 int32_t kern_size = (fs == nullptr) ? -1 : fs->kerned_x_gaps.size();
203 if (fs == nullptr) {
204 // Writing two invalid x-gaps.
205 if (!tesseract::Serialize(f, &x_gap_invalid, 2) || !tesseract::Serialize(f, &kern_size)) {
206 return false;
207 }
208 } else {
209 if (!tesseract::Serialize(f, &fs->x_gap_before) ||
210 !tesseract::Serialize(f, &fs->x_gap_after) || !tesseract::Serialize(f, &kern_size)) {
211 return false;
212 }
213 }
214 if (kern_size > 0 &&
215 (!Serialize(f, fs->kerned_unichar_ids) || !Serialize(f, fs->kerned_x_gaps))) {
216 return false;
217 }
218 }
219 return true;
220 }
221
222 bool write_set(FILE *f, const FontSet &fs) {
223 int size = fs.size();
224 return tesseract::Serialize(f, &size) &&
225 (size > 0 ? tesseract::Serialize(f, &fs[0], size) : true);
226 }
227
228 } // namespace tesseract.