mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2024-11-23 18:49:08 +08:00
Code Issues Actions 10 Packages Projects Releases Wiki Activity

backport dict.cpp and dict.h changes from commit da03e4e910

Browse Source 
to fix issue #1253
This commit is contained in:
Zdenko Podobný 2019-01-17 11:59:54 +01:00
parent 200886ee56
commit f907620a1a
2 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
dict/dict.cpp
View File

@ -241,7 +241,8 @@ void Dict::Load(const char *data_file_name, const STRING &lang) {
if (load_bigram_dawg) {
bigram_dawg_ = dawg_cache_->GetSquishedDawg(
lang, data_file_name, TESSDATA_BIGRAM_DAWG, dawg_debug_level);
if (bigram_dawg_) dawgs_ += bigram_dawg_;
// The bigram_dawg_ is NOT used like the other dawgs! DO NOT add to the
// dawgs_!!
}
if (load_freq_dawg) {
freq_dawg_ = dawg_cache_->GetSquishedDawg(
@ -332,6 +333,7 @@ void Dict::End() {
delete dawgs_[i];
}
}
dawg_cache_->FreeDawg(bigram_dawg_);
if (dawg_cache_is_ours_) {
delete dawg_cache_;
dawg_cache_ = NULL;
@ -350,7 +352,7 @@ void Dict::End() {
int Dict::def_letter_is_okay(void* void_dawg_args,
UNICHAR_ID unichar_id,
bool word_end) const {
DawgArgs *dawg_args = reinterpret_cast<DawgArgs*>(void_dawg_args);
DawgArgs *dawg_args = reinterpret_cast<DawgArgs *>(void_dawg_args);
if (dawg_debug_level >= 3) {
tprintf("def_letter_is_okay: current unichar=%s word_end=%d"

4
dict/dict.h
View File

@ -528,14 +528,14 @@ class Dict {
DawgVector dawgs_;
SuccessorListsVector successors_;
Trie *pending_words_;
/// The following pointers are only cached for convenience.
/// The dawgs will be deleted when dawgs_ vector is destroyed.
// bigram_dawg_ points to a dawg of two-word bigrams which always supercede if
// any of them are present on the best choices list for a word pair.
// the bigrams are stored as space-separated words where:
// (1) leading and trailing punctuation has been removed from each word and
// (2) any digits have been replaced with '?' marks.
Dawg *bigram_dawg_;
/// The following pointers are only cached for convenience.
/// The dawgs will be deleted when dawgs_ vector is destroyed.
// TODO(daria): need to support multiple languages in the future,
// so maybe will need to maintain a list of dawgs of each kind.
Dawg *freq_dawg_;