Mercurial > hgrepos > Python2 > PyMuPDF

comparison mupdf-source/thirdparty/tesseract/src/ccstruct/ocrrow.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: ocrrow.cpp (Formerly row.c)
3 * Description: Code for the ROW class.
4 * Author: Ray Smith
5 *
6 * (C) Copyright 1991, Hewlett-Packard Ltd.
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 automatically generated configuration file if running autoconf.
20 #ifdef HAVE_CONFIG_H
21 # include "config_auto.h"
22 #endif
23
24 #include "blobbox.h"
25 #include "ocrrow.h"
26
27 namespace tesseract {
28
29 /**********************************************************************
30 * ROW::ROW
31 *
32 * Constructor to build a ROW. Only the stats stuff are given here.
33 * The words are added directly.
34 **********************************************************************/
35 ROW::ROW( // constructor
36 int32_t spline_size, // no of segments
37 int32_t *xstarts, // segment boundaries
38 double *coeffs, // coefficients
39 float x_height, // line height
40 float ascenders, // ascender size
41 float descenders, // descender drop
42 int16_t kern, // char gap
43 int16_t space // word gap
44 )
45 : baseline(spline_size, xstarts, coeffs), para_(nullptr) {
46 kerning = kern; // just store stuff
47 spacing = space;
48 xheight = x_height;
49 ascrise = ascenders;
50 bodysize = 0.0f;
51 descdrop = descenders;
52 has_drop_cap_ = false;
53 lmargin_ = 0;
54 rmargin_ = 0;
55 }
56
57 /**********************************************************************
58 * ROW::ROW
59 *
60 * Constructor to build a ROW. Only the stats stuff are given here.
61 * The words are added directly.
62 **********************************************************************/
63
64 ROW::ROW( // constructor
65 TO_ROW *to_row, // source row
66 int16_t kern, // char gap
67 int16_t space // word gap
68 )
69 : para_(nullptr) {
70 kerning = kern; // just store stuff
71 spacing = space;
72 xheight = to_row->xheight;
73 bodysize = to_row->body_size;
74 ascrise = to_row->ascrise;
75 descdrop = to_row->descdrop;
76 baseline = to_row->baseline;
77 has_drop_cap_ = false;
78 lmargin_ = 0;
79 rmargin_ = 0;
80 }
81
82 // Returns the bounding box including the desired combination of upper and
83 // lower noise/diacritic elements.
84 TBOX ROW::restricted_bounding_box(bool upper_dots, bool lower_dots) const {
85 TBOX box;
86 // This is a read-only iteration of the words in the row.
87 WERD_IT it(const_cast<WERD_LIST *>(&words));
88 for (it.mark_cycle_pt(); !it.cycled_list(); it.forward()) {
89 box += it.data()->restricted_bounding_box(upper_dots, lower_dots);
90 }
91 return box;
92 }
93
94 /**********************************************************************
95 * ROW::recalc_bounding_box
96 *
97 * Set the bounding box correctly
98 **********************************************************************/
99
100 void ROW::recalc_bounding_box() { // recalculate BB
101 WERD *word; // current word
102 WERD_IT it = &words; // words of ROW
103 int16_t left; // of word
104 int16_t prev_left; // old left
105
106 if (!it.empty()) {
107 word = it.data();
108 prev_left = word->bounding_box().left();
109 it.forward();
110 while (!it.at_first()) {
111 word = it.data();
112 left = word->bounding_box().left();
113 if (left < prev_left) {
114 it.move_to_first();
115 // words in BB order
116 it.sort(word_comparator);
117 break;
118 }
119 prev_left = left;
120 it.forward();
121 }
122 }
123 for (it.mark_cycle_pt(); !it.cycled_list(); it.forward()) {
124 word = it.data();
125 if (it.at_first()) {
126 word->set_flag(W_BOL, true);
127 } else {
128 // not start of line
129 word->set_flag(W_BOL, false);
130 }
131 if (it.at_last()) {
132 word->set_flag(W_EOL, true);
133 } else {
134 // not end of line
135 word->set_flag(W_EOL, false);
136 }
137 // extend BB as reqd
138 bound_box += word->bounding_box();
139 }
140 }
141
142 /**********************************************************************
143 * ROW::move
144 *
145 * Reposition row by vector
146 **********************************************************************/
147
148 void ROW::move( // reposition row
149 const ICOORD vec // by vector
150 ) {
151 WERD_IT it(&words); // word iterator
152
153 for (it.mark_cycle_pt(); !it.cycled_list(); it.forward()) {
154 it.data()->move(vec);
155 }
156
157 bound_box.move(vec);
158 baseline.move(vec);
159 }
160
161 /**********************************************************************
162 * ROW::print
163 *
164 * Display members
165 **********************************************************************/
166
167 void ROW::print( // print
168 FILE *fp // file to print on
169 ) const {
170 tprintf("Kerning= %d\n", kerning);
171 tprintf("Spacing= %d\n", spacing);
172 bound_box.print();
173 tprintf("Xheight= %f\n", xheight);
174 tprintf("Ascrise= %f\n", ascrise);
175 tprintf("Descdrop= %f\n", descdrop);
176 tprintf("has_drop_cap= %d\n", has_drop_cap_);
177 tprintf("lmargin= %d, rmargin= %d\n", lmargin_, rmargin_);
178 }
179
180 /**********************************************************************
181 * ROW::plot
182 *
183 * Draw the ROW in the given colour.
184 **********************************************************************/
185
186 #ifndef GRAPHICS_DISABLED
187 void ROW::plot( // draw it
188 ScrollView *window, // window to draw in
189 ScrollView::Color colour // colour to draw in
190 ) {
191 WERD *word; // current word
192 WERD_IT it = &words; // words of ROW
193
194 for (it.mark_cycle_pt(); !it.cycled_list(); it.forward()) {
195 word = it.data();
196 word->plot(window, colour); // all in one colour
197 }
198 }
199
200 /**********************************************************************
201 * ROW::plot
202 *
203 * Draw the ROW in rainbow colours.
204 **********************************************************************/
205
206 void ROW::plot( // draw it
207 ScrollView *window // window to draw in
208 ) {
209 WERD *word; // current word
210 WERD_IT it = &words; // words of ROW
211
212 for (it.mark_cycle_pt(); !it.cycled_list(); it.forward()) {
213 word = it.data();
214 word->plot(window); // in rainbow colours
215 }
216 }
217 #endif // !GRAPHICS_DISABLED
218
219 /**********************************************************************
220 * ROW::operator=
221 *
222 * Assign rows by duplicating the row structure but NOT the WERDLIST
223 **********************************************************************/
224
225 ROW &ROW::operator=(const ROW &source) {
226 this->ELIST_LINK::operator=(source);
227 kerning = source.kerning;
228 spacing = source.spacing;
229 xheight = source.xheight;
230 bodysize = source.bodysize;
231 ascrise = source.ascrise;
232 descdrop = source.descdrop;
233 if (!words.empty()) {
234 words.clear();
235 }
236 baseline = source.baseline; // QSPLINES must do =
237 bound_box = source.bound_box;
238 has_drop_cap_ = source.has_drop_cap_;
239 lmargin_ = source.lmargin_;
240 rmargin_ = source.rmargin_;
241 para_ = source.para_;
242 return *this;
243 }
244
245 } // namespace tesseract