mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2024-11-27 20:59:36 +08:00
Code Issues Actions 11 Packages Projects Releases Wiki Activity

Reduce scope of some local variables (reported by Codacy)

Browse Source 
Apply also some smaller optimizations and add TODO comments
for local variables which require further examination.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
This commit is contained in:
Stefan Weil 2024-06-11 08:58:25 +02:00
parent 560529332c
commit 96772c5761
9 changed files with 38 additions and 54 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
src/api/pdfrenderer.cpp
View File

@ -242,13 +242,13 @@ static void GetWordBaseline(int writing_direction, int ppi, int height, int word
double word_length;
double x, y;
{
int px = word_x1;
int py = word_y1;
double l2 = dist2(line_x1, line_y1, line_x2, line_y2);
if (l2 == 0) {
x = line_x1;
y = line_y1;
} else {
int px = word_x1;
int py = word_y1;
double t = ((px - line_x2) * (line_x2 - line_x1) + (py - line_y2) * (line_y2 - line_y1)) / l2;
x = line_x2 + t * (line_x2 - line_x1);
y = line_y2 + t * (line_y2 - line_y1);

3
src/ccmain/applybox.cpp
View File

@ -258,10 +258,10 @@ void Tesseract::MaximallyChopWord(const std::vector<TBOX> &boxes, BLOCK *block,
}
const double e = exp(1.0); // The base of natural logs.
unsigned blob_number;
int right_chop_index = 0;
if (!assume_fixed_pitch_char_segment) {
// We only chop if the language is not fixed pitch like CJK.
SEAM *seam = nullptr;
int right_chop_index = 0;
while ((seam = chop_one_blob(boxes, blob_choices, word_res, &blob_number)) != nullptr) {
word_res->InsertSeam(blob_number, seam);
BLOB_CHOICE *left_choice = blob_choices[blob_number];
@ -685,6 +685,7 @@ void Tesseract::SearchForText(const std::vector<BLOB_CHOICE_LIST *> *choices, in
void Tesseract::TidyUp(PAGE_RES *page_res) {
int ok_blob_count = 0;
int bad_blob_count = 0;
// TODO: check usage of ok_word_count.
int ok_word_count = 0;
int unlabelled_words = 0;
PAGE_RES_IT pr_it(page_res);

18
src/ccmain/control.cpp
View File

@ -949,6 +949,7 @@ bool Tesseract::ReassignDiacritics(int pass, PAGE_RES_IT *pr_it, bool *make_next
}
real_word->AddSelectedOutlines(wanted, wanted_blobs, wanted_outlines, nullptr);
AssignDiacriticsToNewBlobs(outlines, pass, real_word, pr_it, &word_wanted, &target_blobs);
// TODO: check code.
int non_overlapped = 0;
int non_overlapped_used = 0;
for (unsigned i = 0; i < word_wanted.size(); ++i) {
@ -1121,9 +1122,9 @@ bool Tesseract::SelectGoodDiacriticOutlines(int pass, float certainty_threshold,
C_BLOB *blob,
const std::vector<C_OUTLINE *> &outlines,
int num_outlines, std::vector<bool> *ok_outlines) {
std::string best_str;
float target_cert = certainty_threshold;
if (blob != nullptr) {
std::string best_str;
float target_c2;
target_cert = ClassifyBlobAsWord(pass, pr_it, blob, best_str, &target_c2);
if (debug_noise_removal) {
@ -1797,9 +1798,6 @@ not_a_word:
}
bool Tesseract::check_debug_pt(WERD_RES *word, int location) {
bool show_map_detail = false;
int16_t i;
if (!test_pt) {
return false;
}
@ -1811,6 +1809,7 @@ bool Tesseract::check_debug_pt(WERD_RES *word, int location) {
if (location < 0) {
return true; // For breakpoint use
}
bool show_map_detail = false;
tessedit_rejection_debug.set_value(true);
debug_x_ht_level.set_value(2);
tprintf("\n\nTESTWD::");
@ -1864,7 +1863,7 @@ bool Tesseract::check_debug_pt(WERD_RES *word, int location) {
tprintf("\n");
if (show_map_detail) {
tprintf("\"%s\"\n", word->best_choice->unichar_string().c_str());
for (i = 0; word->best_choice->unichar_string()[i] != '\0'; i++) {
for (unsigned i = 0; word->best_choice->unichar_string()[i] != '\0'; i++) {
tprintf("**** \"%c\" ****\n", word->best_choice->unichar_string()[i]);
word->reject_map[i].full_print(debug_fp);
}
@ -1891,13 +1890,12 @@ static void find_modal_font( // good chars in word
int16_t *font_out, // output font
int8_t *font_count // output count
) {
int16_t font; // font index
int32_t count; // pile count
if (fonts->get_total() > 0) {
font = static_cast<int16_t>(fonts->mode());
// font index
int16_t font = static_cast<int16_t>(fonts->mode());
*font_out = font;
count = fonts->pile_count(font);
// pile count
int32_t count = fonts->pile_count(font);
*font_count = count < INT8_MAX ? count : INT8_MAX;
fonts->add(font, -*font_count);
} else {

28
src/ccmain/docqual.cpp
View File

@ -60,10 +60,10 @@ int16_t Tesseract::word_blob_quality(WERD_RES *word) {
}
int16_t Tesseract::word_outline_errs(WERD_RES *word) {
int16_t i = 0;
int16_t err_count = 0;
if (word->rebuild_word != nullptr) {
int16_t i = 0;
for (unsigned b = 0; b < word->rebuild_word->NumBlobs(); ++b) {
TBLOB *blob = word->rebuild_word->blobs[b];
err_count += count_outline_errs(word->best_choice->unichar_string()[i], blob->NumOutlines());
@ -209,13 +209,8 @@ void Tesseract::unrej_good_quality_words( // unreject potential
void Tesseract::doc_and_block_rejection( // reject big chunks
PAGE_RES_IT &page_res_it, bool good_quality_doc) {
int16_t block_no = 0;
int16_t row_no = 0;
BLOCK_RES *current_block;
ROW_RES *current_row;
bool rej_word;
bool prev_word_rejected;
int16_t char_quality = 0;
int16_t accepted_char_quality;
@ -238,7 +233,7 @@ void Tesseract::doc_and_block_rejection( // reject big chunks
WERD_RES *word;
while ((word = page_res_it.word()) != nullptr) {
current_block = page_res_it.block();
block_no = current_block->block->pdblk.index();
int16_t block_no = current_block->block->pdblk.index();
if (current_block->char_count > 0 &&
(current_block->rej_count * 100.0 / current_block->char_count) >
tessedit_reject_block_percent) {
@ -246,8 +241,9 @@ void Tesseract::doc_and_block_rejection( // reject big chunks
tprintf("REJECTING BLOCK %d #chars: %d; #Rejects: %d\n", block_no,
current_block->char_count, current_block->rej_count);
}
prev_word_rejected = false;
bool prev_word_rejected = false;
while ((word = page_res_it.word()) != nullptr && (page_res_it.block() == current_block)) {
bool rej_word;
if (tessedit_preserve_blk_rej_perfect_wds) {
rej_word = word->reject_map.reject_count() > 0 ||
word->reject_map.length() < tessedit_preserve_min_wd_len;
@ -284,9 +280,9 @@ void Tesseract::doc_and_block_rejection( // reject big chunks
}
/* Walk rows in block testing for row rejection */
row_no = 0;
int16_t row_no = 0;
while (page_res_it.word() != nullptr && page_res_it.block() == current_block) {
current_row = page_res_it.row();
ROW_RES *current_row = page_res_it.row();
row_no++;
/* Reject whole row if:
fraction of chars on row which are rejected exceed a limit AND
@ -302,9 +298,10 @@ void Tesseract::doc_and_block_rejection( // reject big chunks
tprintf("REJECTING ROW %d #chars: %d; #Rejects: %d\n", row_no,
current_row->char_count, current_row->rej_count);
}
prev_word_rejected = false;
bool prev_word_rejected = false;
while ((word = page_res_it.word()) != nullptr && page_res_it.row() == current_row) {
/* Preserve words on good docs unless they are mostly rejected*/
bool rej_word;
if (!tessedit_row_rej_good_docs && good_quality_doc) {
rej_word = word->reject_map.reject_count() /
static_cast<float>(word->reject_map.length()) >
@ -448,8 +445,6 @@ void Tesseract::tilde_crunch(PAGE_RES_IT &page_res_it) {
}
bool Tesseract::terrible_word_crunch(WERD_RES *word, GARBAGE_LEVEL garbage_level) {
float rating_per_ch;
int adjusted_len;
int crunch_mode = 0;
if (word->best_choice->unichar_string().empty() ||
@ -457,11 +452,11 @@ bool Tesseract::terrible_word_crunch(WERD_RES *word, GARBAGE_LEVEL garbage_level
word->best_choice->unichar_string().size())) {
crunch_mode = 1;
} else {
adjusted_len = word->reject_map.length();
int adjusted_len = word->reject_map.length();
if (adjusted_len > crunch_rating_max) {
adjusted_len = crunch_rating_max;
}
rating_per_ch = word->best_choice->rating() / adjusted_len;
float rating_per_ch = word->best_choice->rating() / adjusted_len;
if (rating_per_ch > crunch_terrible_rating) {
crunch_mode = 2;
@ -528,7 +523,6 @@ bool Tesseract::potential_word_crunch(WERD_RES *word, GARBAGE_LEVEL garbage_leve
}
void Tesseract::tilde_delete(PAGE_RES_IT &page_res_it) {
WERD_RES *word;
PAGE_RES_IT copy_it;
bool deleting_from_bol = false;
bool marked_delete_point = false;
@ -539,7 +533,7 @@ void Tesseract::tilde_delete(PAGE_RES_IT &page_res_it) {
page_res_it.restart_page();
while (page_res_it.word() != nullptr) {
word = page_res_it.word();
WERD_RES *word = page_res_it.word();
delete_mode = word_deletable(word, debug_delete_mode);
if (delete_mode != CR_NONE) {

20
src/ccmain/fixspace.cpp
View File

@ -171,7 +171,6 @@ void Tesseract::fix_fuzzy_spaces(ETEXT_DESC *monitor, int32_t word_count, PAGE_R
void Tesseract::fix_fuzzy_space_list(WERD_RES_LIST &best_perm, ROW *row, BLOCK *block) {
int16_t best_score;
WERD_RES_LIST current_perm;
int16_t current_score;
bool improved = false;
best_score = eval_word_spacing(best_perm); // default score
@ -183,7 +182,7 @@ void Tesseract::fix_fuzzy_space_list(WERD_RES_LIST &best_perm, ROW *row, BLOCK *
while ((best_score != PERFECT_WERDS) && !current_perm.empty()) {
match_current_words(current_perm, row, block);
current_score = eval_word_spacing(current_perm);
int16_t current_score = eval_word_spacing(current_perm);
dump_words(current_perm, current_score, 2, improved);
if (current_score > best_score) {
best_perm.clear();
@ -201,11 +200,10 @@ void Tesseract::fix_fuzzy_space_list(WERD_RES_LIST &best_perm, ROW *row, BLOCK *
void initialise_search(WERD_RES_LIST &src_list, WERD_RES_LIST &new_list) {
WERD_RES_IT src_it(&src_list);
WERD_RES_IT new_it(&new_list);
WERD_RES *src_wd;
WERD_RES *new_wd;
for (src_it.mark_cycle_pt(); !src_it.cycled_list(); src_it.forward()) {
src_wd = src_it.data();
WERD_RES *src_wd = src_it.data();
if (!src_wd->combination) {
new_wd = WERD_RES::deep_copy(src_wd);
new_wd->combination = false;
@ -393,8 +391,6 @@ void transform_to_next_perm(WERD_RES_LIST &words) {
WERD_RES_IT prev_word_it(&words);
WERD_RES *word;
WERD_RES *prev_word;
WERD_RES *combo;
WERD *copy_word;
int16_t prev_right = -INT16_MAX;
TBOX box;
int16_t gap;
@ -425,12 +421,13 @@ void transform_to_next_perm(WERD_RES_LIST &words) {
gap = box.left() - prev_right;
if (gap <= min_gap) {
prev_word = prev_word_it.data();
WERD_RES *combo;
if (prev_word->combination) {
combo = prev_word;
} else {
/* Make a new combination and insert before
* the first word being joined. */
copy_word = new WERD;
auto *copy_word = new WERD;
*copy_word = *(prev_word->word);
// deep copy
combo = new WERD_RES(copy_word);
@ -546,7 +543,6 @@ void Tesseract::fix_sp_fp_word(WERD_RES_IT &word_res_it, ROW *row, BLOCK *block)
WERD_RES *word_res;
WERD_RES_LIST sub_word_list;
WERD_RES_IT sub_word_list_it(&sub_word_list);
int16_t blob_index;
int16_t new_length;
float junk;
@ -556,7 +552,7 @@ void Tesseract::fix_sp_fp_word(WERD_RES_IT &word_res_it, ROW *row, BLOCK *block)
return;
}
blob_index = worst_noise_blob(word_res, &junk);
auto blob_index = worst_noise_blob(word_res, &junk);
if (blob_index < 0) {
return;
}
@ -623,7 +619,6 @@ void Tesseract::break_noisiest_blob_word(WERD_RES_LIST &words) {
WERD_RES_IT worst_word_it;
float worst_noise_score = 9999;
int worst_blob_index = -1; // Noisiest blob of noisiest wd
int blob_index; // of wds noisiest blob
float noise_score; // of wds noisiest blob
WERD_RES *word_res;
C_BLOB_IT blob_it;
@ -636,7 +631,7 @@ void Tesseract::break_noisiest_blob_word(WERD_RES_LIST &words) {
int16_t i;
for (word_it.mark_cycle_pt(); !word_it.cycled_list(); word_it.forward()) {
blob_index = worst_noise_blob(word_it.data(), &noise_score);
auto blob_index = worst_noise_blob(word_it.data(), &noise_score);
if (blob_index > -1 && worst_noise_score > noise_score) {
worst_noise_score = noise_score;
worst_blob_index = blob_index;
@ -806,7 +801,6 @@ float Tesseract::blob_noise_score(TBLOB *blob) {
void fixspace_dbg(WERD_RES *word) {
TBOX box = word->word->bounding_box();
const bool show_map_detail = false;
int16_t i;
box.print();
tprintf(" \"%s\" ", word->best_choice->unichar_string().c_str());
@ -816,7 +810,7 @@ void fixspace_dbg(WERD_RES *word) {
tprintf("\n");
if (show_map_detail) {
tprintf("\"%s\"\n", word->best_choice->unichar_string().c_str());
for (i = 0; word->best_choice->unichar_string()[i] != '\0'; i++) {
for (unsigned i = 0; word->best_choice->unichar_string()[i] != '\0'; i++) {
tprintf("**** \"%c\" ****\n", word->best_choice->unichar_string()[i]);
word->reject_map[i].full_print(debug_fp);
}

2
src/ccmain/output.cpp
View File

@ -101,11 +101,11 @@ void Tesseract::write_results(PAGE_RES_IT &page_res_it,
bool force_eol) { // override tilde crunch?
WERD_RES *word = page_res_it.word();
const UNICHARSET &uchset = *word->uch_set;
bool need_reject = false;
UNICHAR_ID space = uchset.unichar_to_id(" ");
if ((word->unlv_crunch_mode != CR_NONE || word->best_choice->empty()) &&
!tessedit_zero_kelvin_rejection && !tessedit_word_for_word) {
bool need_reject = false;
if ((word->unlv_crunch_mode != CR_DELETE) &&
(!stats_.tilde_crunch_written ||
((word->unlv_crunch_mode == CR_KEEP_SPACE) && (word->word->space() > 0) &&

2
src/ccmain/paragraphs.cpp
View File

@ -2407,8 +2407,8 @@ static void InitializeTextAndBoxesPreRecognition(const MutableIterator &it, RowI
// Set up text, lword_text, and rword_text (mostly for debug printing).
std::string fake_text;
PageIterator pit(static_cast<const PageIterator &>(it));
bool first_word = true;
if (!pit.Empty(RIL_WORD)) {
bool first_word = true;
do {
fake_text += "x";
if (first_word) {

9
src/ccmain/pgedit.cpp
View File

@ -703,9 +703,7 @@ bool Tesseract::word_display(PAGE_RES_IT *pr_it) {
WERD_RES *word_res = pr_it->word();
WERD *word = word_res->word;
TBOX word_bb; // word bounding box
int word_height; // ht of word BB
bool displayed_something = false;
float shift; // from bot left
if (color_mode != CM_RAINBOW && word_res->box_word != nullptr) {
# ifndef DISABLED_LEGACY_ENGINE
@ -842,13 +840,14 @@ bool Tesseract::word_display(PAGE_RES_IT *pr_it) {
if (text.length() > 0) {
word_bb = word->bounding_box();
image_win->Pen(ScrollView::RED);
word_height = word_bb.height();
int text_height = 0.50 * word_height;
auto word_height = word_bb.height();
int text_height = word_height / 2;
if (text_height > 20) {
text_height = 20;
}
image_win->TextAttributes("Arial", text_height, false, false, false);
shift = (word_height < word_bb.width()) ? 0.25 * word_height : 0.0f;
// from bot left
float shift = (word_height < word_bb.width()) ? 0.25f * word_height : 0.0f;
image_win->Text(word_bb.left() + shift, word_bb.bottom() + 0.25 * word_height, text.c_str());
if (blame.length() > 0) {
image_win->Text(word_bb.left() + shift, word_bb.bottom() + 0.25 * word_height - text_height,

6
src/ccmain/reject.cpp
View File

@ -293,8 +293,6 @@ bool Tesseract::one_ell_conflict(WERD_RES *word_res, bool update_map) {
int16_t i;
int16_t offset;
bool non_conflict_set_char; // non conf set a/n?
bool conflict = false;
bool allow_1s;
ACCEPTABLE_WERD_TYPE word_type;
bool dict_perm_type;
bool dict_word_ok;
@ -411,11 +409,11 @@ bool Tesseract::one_ell_conflict(WERD_RES *word_res, bool update_map) {
Else reject all conflict chs
*/
if (word_contains_non_1_digit(word, lengths)) {
allow_1s =
bool allow_1s =
(alpha_count(word, lengths) == 0) || (word_res->best_choice->permuter() == NUMBER_PERM);
int16_t offset;
conflict = false;
bool conflict = false;
for (i = 0, offset = 0; word[offset] != '\0';
offset += word_res->best_choice->unichar_lengths()[i++]) {
if ((!allow_1s || (word[offset] != '1')) &&