Tesseract
3.02
|
Go to the source code of this file.
Functions | |
FILE * | OpenBoxFile (const STRING &fname) |
bool | ReadNextBox (int *line_number, FILE *box_file, STRING *utf8_str, TBOX *bounding_box) |
bool | ReadNextBox (int target_page, int *line_number, FILE *box_file, STRING *utf8_str, TBOX *bounding_box) |
bool | ParseBoxFileStr (const char *boxfile_str, int *page_number, STRING *utf8_str, TBOX *bounding_box) |
void | MakeBoxFileStr (const char *unichar_str, const TBOX &box, int page_num, STRING *box_str) |
Variables | |
const int | kBoxReadBufSize = 1024 |
Definition at line 155 of file boxread.cpp.
{ *box_str = unichar_str; box_str->add_str_int(" ", box.left()); box_str->add_str_int(" ", box.bottom()); box_str->add_str_int(" ", box.right()); box_str->add_str_int(" ", box.top()); box_str->add_str_int(" ", page_num); }
FILE* OpenBoxFile | ( | const STRING & | fname | ) |
Definition at line 33 of file boxread.cpp.
{ STRING filename = fname; const char *lastdot = strrchr(filename.string(), '.'); if (lastdot != NULL) filename[lastdot - filename.string()] = '\0'; filename += ".box"; FILE* box_file = NULL; if (!(box_file = fopen(filename.string(), "rb"))) { CANTOPENFILE.error("read_next_box", TESSEXIT, "Cant open box file %s", filename.string()); } return box_file; }
bool ParseBoxFileStr | ( | const char * | boxfile_str, |
int * | page_number, | ||
STRING * | utf8_str, | ||
TBOX * | bounding_box | ||
) |
Definition at line 105 of file boxread.cpp.
{ *bounding_box = TBOX(); // Initialize it to empty. *utf8_str = ""; char uch[kBoxReadBufSize]; const char *buffptr = boxfile_str; // Read the unichar without messing up on Tibetan. // According to issue 253 the utf-8 surrogates 85 and A0 are treated // as whitespace by sscanf, so it is more reliable to just find // ascii space and tab. int uch_len = 0; while (*buffptr != '\0' && *buffptr != ' ' && *buffptr != '\t' && uch_len < kBoxReadBufSize - 1) { uch[uch_len++] = *buffptr++; } uch[uch_len] = '\0'; if (*buffptr != '\0') ++buffptr; int x_min, y_min, x_max, y_max; *page_number = 0; int count = sscanf(buffptr, "%d %d %d %d %d", &x_min, &y_min, &x_max, &y_max, page_number); if (count != 5 && count != 4) { tprintf("Bad box coordinates in boxfile string!\n"); return false; } // Test for long space-delimited string label. if (strcmp(uch, kMultiBlobLabelCode) == 0 && (buffptr = strchr(buffptr, '#')) != NULL) { strncpy(uch, buffptr + 1, kBoxReadBufSize); chomp_string(uch); uch_len = strlen(uch); } // Validate UTF8 by making unichars with it. int used = 0; while (used < uch_len) { UNICHAR ch(uch + used, uch_len - used); int new_used = ch.utf8_len(); if (new_used == 0) { tprintf("Bad UTF-8 str %s starts with 0x%02x at col %d\n", uch + used, uch[used], used + 1); return false; } used += new_used; } *utf8_str = uch; bounding_box->set_to_given_coords(x_min, y_min, x_max, y_max); return true; // Successfully read a box. }
Definition at line 59 of file boxread.cpp.
{ return ReadNextBox(-1, line_number, box_file, utf8_str, bounding_box); }
bool ReadNextBox | ( | int | target_page, |
int * | line_number, | ||
FILE * | box_file, | ||
STRING * | utf8_str, | ||
TBOX * | bounding_box | ||
) |
Definition at line 67 of file boxread.cpp.
{ int page = 0; char buff[kBoxReadBufSize]; // boxfile read buffer char *buffptr = buff; while (fgets(buff, sizeof(buff) - 1, box_file)) { (*line_number)++; buffptr = buff; const unsigned char *ubuf = reinterpret_cast<const unsigned char*>(buffptr); if (ubuf[0] == 0xef && ubuf[1] == 0xbb && ubuf[2] == 0xbf) buffptr += 3; // Skip unicode file designation. // Check for blank lines in box file while (*buffptr == ' ' || *buffptr == '\t') buffptr++; if (*buffptr != '\0') { if (!ParseBoxFileStr(buffptr, &page, utf8_str, bounding_box)) { tprintf("Box file format error on line %i; ignored\n", *line_number); continue; } if (target_page >= 0 && target_page != page) continue; // Not on the appropriate page. return true; // Successfully read a box. } } fclose(box_file); return false; // EOF }
const int kBoxReadBufSize = 1024 |