Mercurial > hgrepos > Python2 > PyMuPDF

comparison mupdf-source/thirdparty/tesseract/src/textord/colpartitionset.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: colpartitionset.h
3 // Description: Class to hold a list of ColPartitions of the page that
4 // correspond roughly to columns.
5 // Author: Ray Smith
6 //
7 // (C) Copyright 2008, Google Inc.
8 // Licensed under the Apache License, Version 2.0 (the "License");
9 // you may not use this file except in compliance with the License.
10 // You may obtain a copy of the License at
11 // http://www.apache.org/licenses/LICENSE-2.0
12 // Unless required by applicable law or agreed to in writing, software
13 // distributed under the License is distributed on an "AS IS" BASIS,
14 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
15 // See the License for the specific language governing permissions and
16 // limitations under the License.
17 //
18 ///////////////////////////////////////////////////////////////////////
19
20 #ifndef TESSERACT_TEXTORD_COLPARTITIONSET_H_
21 #define TESSERACT_TEXTORD_COLPARTITIONSET_H_
22
23 #include "colpartition.h" // For ColPartition_LIST.
24 #include "rect.h" // For TBOX.
25 #include "tabvector.h" // For BLOBNBOX_CLIST.
26
27 namespace tesseract {
28
29 class WorkingPartSet_LIST;
30 class ColSegment_LIST;
31 class ColPartitionSet;
32 using PartSetVector = std::vector<ColPartitionSet *>;
33
34 // ColPartitionSet is a class that holds a list of ColPartitions.
35 // Its main use is in holding a candidate partitioning of the width of the
36 // image into columns, where each member ColPartition is a single column.
37 // ColPartitionSets are used in building the column layout of a page.
38 class ColPartitionSet : public ELIST_LINK {
39 public:
40 ColPartitionSet() = default;
41 explicit ColPartitionSet(ColPartition_LIST *partitions);
42 explicit ColPartitionSet(ColPartition *partition);
43
44 ~ColPartitionSet() = default;
45
46 // Simple accessors.
47 const TBOX &bounding_box() const {
48 return bounding_box_;
49 }
50 bool Empty() const {
51 return parts_.empty();
52 }
53 int ColumnCount() const {
54 return parts_.length();
55 }
56
57 // Returns the number of columns of good width.
58 int GoodColumnCount() const;
59
60 // Return an element of the parts_ list from its index.
61 ColPartition *GetColumnByIndex(int index);
62
63 // Return the ColPartition that contains the given coords, if any, else
64 // nullptr.
65 ColPartition *ColumnContaining(int x, int y);
66
67 // Return the bounding boxes of columns at the given y-range
68 void GetColumnBoxes(int y_bottom, int y_top, ColSegment_LIST *segments);
69
70 // Extract all the parts from the list, relinquishing ownership.
71 void RelinquishParts();
72
73 // Attempt to improve this by adding partitions or expanding partitions.
74 void ImproveColumnCandidate(const WidthCallback &cb, PartSetVector *src_sets);
75
76 // If this set is good enough to represent a new partitioning into columns,
77 // add it to the vector of sets, otherwise delete it.
78 void AddToColumnSetsIfUnique(PartSetVector *column_sets,
79 const WidthCallback &cb);
80
81 // Return true if the partitions in other are all compatible with the columns
82 // in this.
83 bool CompatibleColumns(bool debug, ColPartitionSet *other,
84 const WidthCallback &cb);
85
86 // Returns the total width of all blobs in the part_set that do not lie
87 // within an approved column. Used as a cost measure for using this
88 // column set over another that might be compatible.
89 int UnmatchedWidth(ColPartitionSet *part_set);
90
91 // Return true if this ColPartitionSet makes a legal column candidate by
92 // having legal individual partitions and non-overlapping adjacent pairs.
93 bool LegalColumnCandidate();
94
95 // Return a copy of this. If good_only will only copy the Good ColPartitions.
96 ColPartitionSet *Copy(bool good_only);
97
98 // Display the edges of the columns at the given y coords.
99 void DisplayColumnEdges(int y_bottom, int y_top, ScrollView *win);
100
101 // Return the ColumnSpanningType that best explains the columns overlapped
102 // by the given coords(left,right,y), with the given margins.
103 // Also return the first and last column index touched by the coords and
104 // the leftmost spanned column.
105 // Column indices are 2n + 1 for real columns (0 based) and even values
106 // represent the gaps in between columns, with 0 being left of the leftmost.
107 // resolution refers to the ppi resolution of the image. It may be 0 if only
108 // the first_col and last_col are required.
109 ColumnSpanningType SpanningType(int resolution, int left, int right,
110 int height, int y, int left_margin,
111 int right_margin, int *first_col,
112 int *last_col, int *first_spanned_col);
113
114 // The column_set has changed. Close down all in-progress WorkingPartSets in
115 // columns that do not match and start new ones for the new columns in this.
116 // As ColPartitions are turned into BLOCKs, the used ones are put in
117 // used_parts, as they still need to be referenced in the grid.
118 void ChangeWorkColumns(const ICOORD &bleft, const ICOORD &tright,
119 int resolution, ColPartition_LIST *used_parts,
120 WorkingPartSet_LIST *working_set);
121
122 // Accumulate the widths and gaps into the given variables.
123 void AccumulateColumnWidthsAndGaps(int *total_width, int *width_samples,
124 int *total_gap, int *gap_samples);
125
126 // Provide debug output for this ColPartitionSet and all the ColPartitions.
127 void Print();
128
129 private:
130 // Add the given partition to the list in the appropriate place.
131 void AddPartition(ColPartition *new_part, ColPartition_IT *it);
132
133 // Compute the coverage and good column count. Coverage is the amount of the
134 // width of the page (in pixels) that is covered by ColPartitions, which are
135 // used to provide candidate column layouts.
136 // Coverage is split into good and bad. Good coverage is provided by
137 // ColPartitions of a frequent width (according to the callback function
138 // provided by TabFinder::WidthCB, which accesses stored statistics on the
139 // widths of ColPartitions) and bad coverage is provided by all other
140 // ColPartitions, even if they have tab vectors at both sides. Thus:
141 // |-----------------------------------------------------------------|
142 // | Double width heading |
143 // |-----------------------------------------------------------------|
144 // |-------------------------------| |-------------------------------|
145 // | Common width ColPartition | | Common width ColPartition |
146 // |-------------------------------| |-------------------------------|
147 // the layout with two common-width columns has better coverage than the
148 // double width heading, because the coverage is "good," even though less in
149 // total coverage than the heading, because the heading coverage is "bad."
150 void ComputeCoverage();
151
152 // Adds the coverage, column count and box for a single partition,
153 // without adding it to the list. (Helper factored from ComputeCoverage.)
154 void AddPartitionCoverageAndBox(const ColPartition &part);
155
156 // The partitions in this column candidate.
157 ColPartition_LIST parts_;
158 // The number of partitions that have a frequent column width.
159 int good_column_count_;
160 // Total width of all the good ColPartitions.
161 int good_coverage_;
162 // Total width of all the bad ColPartitions.
163 int bad_coverage_;
164 // Bounding box of all partitions in the set.
165 TBOX bounding_box_;
166 };
167
168 ELISTIZEH(ColPartitionSet)
169
170 } // namespace tesseract.
171
172 #endif // TESSERACT_TEXTORD_COLPARTITION_H_