Mercurial > hgrepos > Python2 > PyMuPDF

comparison mupdf-source/thirdparty/tesseract/src/training/pango/boxchar.h @ 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: boxchar.h
3 * Description: Simple class to associate a Tesseract classification unit with
4 * its bounding box so that the boxes can be rotated as the image
5 * is rotated for degradation. Also includes routines to output
6 * the character-tagged boxes to a boxfile.
7 * Author: Ray Smith
8 *
9 * (C) Copyright 2013, Google Inc.
10 * Licensed under the Apache License, Version 2.0 (the "License");
11 * you may not use this file except in compliance with the License.
12 * You may obtain a copy of the License at
13 * http://www.apache.org/licenses/LICENSE-2.0
14 * Unless required by applicable law or agreed to in writing, software
15 * distributed under the License is distributed on an "AS IS" BASIS,
16 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
17 * See the License for the specific language governing permissions and
18 * limitations under the License.
19 *
20 **********************************************************************/
21
22 #ifndef TESSERACT_TRAINING_BOXCHAR_H_
23 #define TESSERACT_TRAINING_BOXCHAR_H_
24
25 #include <string>
26 #include <vector>
27
28 #include <allheaders.h> // for Leptonica API
29 #if (LIBLEPT_MAJOR_VERSION == 1 && LIBLEPT_MINOR_VERSION >= 83) || LIBLEPT_MAJOR_VERSION > 1
30 #include <pix_internal.h> // for fast access to Box geometry
31 #endif
32 #include <tesseract/export.h>
33
34 namespace tesseract {
35
36 class BoxChar {
37 public:
38 BoxChar(const char *utf8_str, int len);
39
40 ~BoxChar();
41
42 // Accessors.
43 const std::string &ch() const {
44 return ch_;
45 }
46 const Box *box() const {
47 return box_;
48 }
49 const int &page() const {
50 return page_;
51 }
52 void set_rtl_index(int index) {
53 rtl_index_ = index;
54 }
55 const int &rtl_index() const {
56 return rtl_index_;
57 }
58
59 // Set the box_ member.
60 void AddBox(int x, int y, int width, int height);
61
62 void set_page(int page) {
63 page_ = page;
64 }
65
66 std::string *mutable_ch() {
67 return &ch_;
68 }
69 Box *mutable_box() {
70 return box_;
71 }
72
73 // Sort function for sorting by left edge of box. Note that this will not
74 // work properly until after InsertNewlines and InsertSpaces.
75 bool operator<(const BoxChar &other) const {
76 if (box_ == nullptr) {
77 return true;
78 }
79 if (other.box_ == nullptr) {
80 return false;
81 }
82 return box_->x < other.box_->x;
83 }
84 // Increments *num_rtl and *num_ltr according to the directionality of
85 // characters in the box.
86 void GetDirection(int *num_rtl, int *num_ltr) const;
87 // Reverses the order of unicodes within the box. If Pango generates a
88 // ligature, these will get reversed on output, so reverse now.
89 void ReverseUnicodesInBox();
90
91 static void TranslateBoxes(int xshift, int yshift, std::vector<BoxChar *> *boxes);
92
93 // Prepares for writing the boxes to a file by inserting newlines, spaces,
94 // and re-ordering so the boxes are strictly left-to-right.
95 static void PrepareToWrite(std::vector<BoxChar *> *boxes);
96 // Inserts newline (tab) characters into the vector at newline positions.
97 static void InsertNewlines(bool rtl_rules, bool vertical_rules, std::vector<BoxChar *> *boxes);
98 // Converts nullptr boxes to space characters, with appropriate bounding
99 // boxes.
100 static void InsertSpaces(bool rtl_rules, bool vertical_rules, std::vector<BoxChar *> *boxes);
101 // Reorders text in a right-to-left script in left-to-right order.
102 static void ReorderRTLText(std::vector<BoxChar *> *boxes);
103 // Returns true if the vector contains mostly RTL characters.
104 static bool ContainsMostlyRTL(const std::vector<BoxChar *> &boxes);
105 // Returns true if the text is mostly laid out vertically.
106 static bool MostlyVertical(const std::vector<BoxChar *> &boxes);
107
108 // Returns the total length of all the strings in the boxes.
109 static int TotalByteLength(const std::vector<BoxChar *> &boxes);
110
111 // Rotate the vector of boxes between start and end by the given rotation.
112 // The rotation is in radians clockwise about the given center.
113 static void RotateBoxes(float rotation, int xcenter, int ycenter, int start_box, int end_box,
114 std::vector<BoxChar *> *boxes);
115
116 // Create a tesseract box file from the vector of boxes. The image height
117 // is needed to convert to tesseract coordinates.
118 static void WriteTesseractBoxFile(const std::string &name, int height,
119 const std::vector<BoxChar *> &boxes);
120 // Gets the tesseract box file as a string from the vector of boxes.
121 // The image height is needed to convert to tesseract coordinates.
122 static std::string GetTesseractBoxStr(int height, const std::vector<BoxChar *> &boxes);
123
124 private:
125 std::string ch_;
126 Box *box_;
127 int page_;
128 // If the box is an RTL character, contains the original position in the
129 // array of boxes (before reversal), otherwise -1.
130 int rtl_index_;
131 };
132
133 // Sort predicate to sort a vector of BoxChar*.
134 struct BoxCharPtrSort {
135 bool operator()(const BoxChar *box1, const BoxChar *box2) const {
136 if (box1->rtl_index() >= 0 && box2->rtl_index() >= 0) {
137 return box2->rtl_index() < box1->rtl_index();
138 }
139 return *box1 < *box2;
140 }
141 };
142
143 } // namespace tesseract
144
145 #endif // TESSERACT_TRAINING_BOXCHAR_H_