tesseract v5.3.3.20231005
wordstrboxrenderer.cpp
Go to the documentation of this file.
1/**********************************************************************
2 * File: wordstrboxrenderer.cpp
3 * Description: Renderer for creating box file with WordStr strings.
4 * based on the tsv renderer.
5 *
6 * (C) Copyright 2019, 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 <tesseract/baseapi.h> // for TessBaseAPI
20#include <tesseract/renderer.h>
21#include "tesseractclass.h" // for Tesseract
22
23namespace tesseract {
24
31char *TessBaseAPI::GetWordStrBoxText(int page_number = 0) {
32 if (tesseract_ == nullptr || (page_res_ == nullptr && Recognize(nullptr) < 0)) {
33 return nullptr;
34 }
35
36 std::string wordstr_box_str;
37 int left = 0, top = 0, right = 0, bottom = 0;
38
39 bool first_line = true;
40
42 while (!res_it->Empty(RIL_BLOCK)) {
43 if (res_it->Empty(RIL_WORD)) {
44 res_it->Next(RIL_WORD);
45 continue;
46 }
47
48 if (res_it->IsAtBeginningOf(RIL_TEXTLINE)) {
49 if (!first_line) {
50 wordstr_box_str += "\n\t " + std::to_string(right + 1);
51 wordstr_box_str += " " + std::to_string(image_height_ - bottom);
52 wordstr_box_str += " " + std::to_string(right + 5);
53 wordstr_box_str += " " + std::to_string(image_height_ - top);
54 wordstr_box_str += " " + std::to_string(page_number); // row for tab for EOL
55 wordstr_box_str += "\n";
56 } else {
57 first_line = false;
58 }
59 // Use bounding box for whole line for WordStr
60 res_it->BoundingBox(RIL_TEXTLINE, &left, &top, &right, &bottom);
61 wordstr_box_str += "WordStr " + std::to_string(left);
62 wordstr_box_str += " " + std::to_string(image_height_ - bottom);
63 wordstr_box_str += " " + std::to_string(right);
64 wordstr_box_str += " " + std::to_string(image_height_ - top);
65 wordstr_box_str += " " + std::to_string(page_number); // word
66 wordstr_box_str += " #";
67 }
68 do {
69 wordstr_box_str += std::unique_ptr<const char[]>(res_it->GetUTF8Text(RIL_WORD)).get();
70 wordstr_box_str += " ";
71 res_it->Next(RIL_WORD);
72 } while (!res_it->Empty(RIL_BLOCK) && !res_it->IsAtBeginningOf(RIL_WORD));
73 }
74
75 if (left != 0 && top != 0 && right != 0 && bottom != 0) {
76 wordstr_box_str += "\n\t " + std::to_string(right + 1);
77 wordstr_box_str += " " + std::to_string(image_height_ - bottom);
78 wordstr_box_str += " " + std::to_string(right + 5);
79 wordstr_box_str += " " + std::to_string(image_height_ - top);
80 wordstr_box_str += " " + std::to_string(page_number); // row for tab for EOL
81 wordstr_box_str += "\n";
82 }
83 char *ret = new char[wordstr_box_str.length() + 1];
84 strcpy(ret, wordstr_box_str.c_str());
85 delete res_it;
86 return ret;
87}
88
89/**********************************************************************
90 * WordStrBox Renderer interface implementation
91 **********************************************************************/
93 : TessResultRenderer(outputbase, "box") {}
94
96 const std::unique_ptr<const char[]> wordstrbox(api->GetWordStrBoxText(imagenum()));
97 if (wordstrbox == nullptr) {
98 return false;
99 }
100
101 AppendString(wordstrbox.get());
102
103 return true;
104}
105
106} // namespace tesseract.
int Recognize(ETEXT_DESC *monitor)
Definition: baseapi.cpp:834
PAGE_RES * page_res_
The page-level data.
Definition: baseapi.h:772
Tesseract * tesseract_
The underlying data object.
Definition: baseapi.h:765
char * GetWordStrBoxText(int page_number)
LTRResultIterator * GetLTRIterator()
Definition: baseapi.cpp:1320
char * GetUTF8Text(PageIteratorLevel level) const
virtual bool Next(PageIteratorLevel level)
virtual bool IsAtBeginningOf(PageIteratorLevel level) const
bool Empty(PageIteratorLevel level) const
bool BoundingBox(PageIteratorLevel level, int *left, int *top, int *right, int *bottom) const
void AppendString(const char *s)
Definition: renderer.cpp:111
bool AddImageHandler(TessBaseAPI *api) override
TessWordStrBoxRenderer(const char *outputbase)