mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2024-11-25 03:29:05 +08:00
Code Issues Actions 11 Packages Projects Releases Wiki Activity

INCOMPATIBLE fix to hOCR line height information - fixes #225.

Browse Source 
This fixes the duplicate line IDs caused by inserting height information
into the middle of the ID and it moves the line height info into
the title attribute like everything else, rather than using non-standard
HTML attributes (which won't validate).

This change may break consumers of the HTML output, but 3.04 has only
been in the wild for 6 months and the current HTML is invalid, so I 
believe the benefit outweighs the cost for the fix.
This commit is contained in:
Tom Morris 2016-02-15 18:02:46 -05:00
parent 3422059e09
commit 431786276c
1 changed files with 42 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

67
api/baseapi.cpp
View File

@ -1378,18 +1378,42 @@ static void AddBaselineCoordsTohOCR(const PageIterator *it,
hocr_str->add_str_double(" ", round(p0 * 1000.0) / 1000.0);
}
static void AddBoxTohOCR(const PageIterator *it,
static void AddIdTohOCR(STRING* hocr_str, const std::string base, int num1, int num2) {
unsigned long bufsize = base.length() + 2 * kMaxIntSize;
char id_buffer[bufsize];
if (num2 >= 0) {
snprintf(id_buffer, bufsize - 1, "%s_%d_%d", base.c_str(), num1, num2);
} else {
snprintf(id_buffer, bufsize - 1, "%s_%d", base.c_str(), num1);
}
id_buffer[bufsize - 1] = '\0';
*hocr_str += " id='";
*hocr_str += id_buffer;
*hocr_str += "'";
}
static void AddBoxTohOCR(const ResultIterator *it,
PageIteratorLevel level,
STRING* hocr_str) {
int left, top, right, bottom;
it->BoundingBox(level, &left, &top, &right, &bottom);
hocr_str->add_str_int("' title=\"bbox ", left);
// This is the only place we use double quotes instead of single quotes,
// but it may too late to change for consistency
hocr_str->add_str_int(" title=\"bbox ", left);
hocr_str->add_str_int(" ", top);
hocr_str->add_str_int(" ", right);
hocr_str->add_str_int(" ", bottom);
// Add baseline coordinates for textlines only.
if (level == RIL_TEXTLINE)
// Add baseline coordinates & heights for textlines only.
if (level == RIL_TEXTLINE) {
AddBaselineCoordsTohOCR(it, level, hocr_str);
// add custom height measures
float row_height, descenders, ascenders; // row attributes
it->RowAttributes(&row_height, &descenders, &ascenders);
// TODO: Do we want to limit these to a single decimal place?
hocr_str->add_str_double("; x_size ", row_height);
hocr_str->add_str_double("; x_descenders ", descenders * -1);
hocr_str->add_str_double("; x_ascenders ", ascenders);
}
*hocr_str += "\">";
}
@ -1420,7 +1444,6 @@ char* TessBaseAPI::GetHOCRText(struct ETEXT_DESC* monitor, int page_number) {
int lcnt = 1, bcnt = 1, pcnt = 1, wcnt = 1;
int page_id = page_number + 1; // hOCR uses 1-based page numbers.
float row_height, descenders, ascenders; // row attributes
bool font_info = false;
GetBoolVariable("hocr_font_info", &font_info);
@ -1446,8 +1469,9 @@ char* TessBaseAPI::GetHOCRText(struct ETEXT_DESC* monitor, int page_number) {
delete[] utf8_str;
#endif
hocr_str.add_str_int(" <div class='ocr_page' id='page_", page_id);
hocr_str += "' title='image \"";
hocr_str += " <div class='ocr_page'";
AddIdTohOCR(&hocr_str, "page", page_id, -1);
hocr_str += " title='image \"";
if (input_file_) {
hocr_str += HOcrEscape(input_file_->string());
} else {
@ -1469,36 +1493,29 @@ char* TessBaseAPI::GetHOCRText(struct ETEXT_DESC* monitor, int page_number) {
// Open any new block/paragraph/textline.
if (res_it->IsAtBeginningOf(RIL_BLOCK)) {
hocr_str.add_str_int(" <div class='ocr_carea' id='block_", page_id);
hocr_str.add_str_int("_", bcnt);
hocr_str += " <div class='ocr_carea'";
AddIdTohOCR(&hocr_str, "block", page_id, bcnt);
AddBoxTohOCR(res_it, RIL_BLOCK, &hocr_str);
}
if (res_it->IsAtBeginningOf(RIL_PARA)) {
hocr_str += "\n <p class='ocr_par'";
if (res_it->ParagraphIsLtr()) {
hocr_str.add_str_int("\n <p class='ocr_par' dir='ltr' id='par_",
page_id);
hocr_str.add_str_int("_", pcnt);
hocr_str += " dir='ltr'";
} else {
hocr_str.add_str_int("\n <p class='ocr_par' dir='rtl' id='par_",
page_id);
hocr_str.add_str_int("_", pcnt);
hocr_str += " dir='rtl'";
}
AddIdTohOCR(&hocr_str, "par", page_id, pcnt);
AddBoxTohOCR(res_it, RIL_PARA, &hocr_str);
}
if (res_it->IsAtBeginningOf(RIL_TEXTLINE)) {
int fontsize;
hocr_str.add_str_int("\n <span class='ocr_line' id='line_", page_id);
res_it->RowAttributes(&row_height, &descenders, &ascenders);
hocr_str.add_str_int("' size='", row_height);
hocr_str.add_str_int("' descenders='", descenders * -1);
hocr_str.add_str_int("' ascenders='", ascenders);
hocr_str.add_str_int("_", lcnt);
hocr_str += "\n <span class='ocr_line'";
AddIdTohOCR(&hocr_str, "line", page_id, lcnt);
AddBoxTohOCR(res_it, RIL_TEXTLINE, &hocr_str);
}
// Now, process the word...
hocr_str.add_str_int("<span class='ocrx_word' id='word_", page_id);
hocr_str.add_str_int("_", wcnt);
hocr_str += "<span class='ocrx_word'";
AddIdTohOCR(&hocr_str, "word", page_id, wcnt);
int left, top, right, bottom;
bool bold, italic, underlined, monospace, serif, smallcaps;
int pointsize, font_id;
@ -1507,7 +1524,7 @@ char* TessBaseAPI::GetHOCRText(struct ETEXT_DESC* monitor, int page_number) {
font_name = res_it->WordFontAttributes(&bold, &italic, &underlined,
&monospace, &serif, &smallcaps,
&pointsize, &font_id);
hocr_str.add_str_int("' title='bbox ", left);
hocr_str.add_str_int(" title='bbox ", left);
hocr_str.add_str_int(" ", top);
hocr_str.add_str_int(" ", right);
hocr_str.add_str_int(" ", bottom);