00001 /********************************************************************** 00002 * File: ocrclass.h 00003 * Description: Class definitions and constants for the OCR API. 00004 * Author: Hewlett-Packard Co 00005 * 00006 * (C) Copyright 1996, Hewlett-Packard Co. 00007 ** Licensed under the Apache License, Version 2.0 (the "License"); 00008 ** you may not use this file except in compliance with the License. 00009 ** You may obtain a copy of the License at 00010 ** http://www.apache.org/licenses/LICENSE-2.0 00011 ** Unless required by applicable law or agreed to in writing, software 00012 ** distributed under the License is distributed on an "AS IS" BASIS, 00013 ** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 00014 ** See the License for the specific language governing permissions and 00015 ** limitations under the License. 00016 * 00017 **********************************************************************/ 00018 00019 /********************************************************************** 00020 * This file contains typedefs for all the structures used by 00021 * the HP OCR interface. 00022 * The code is designed to be used with either a C or C++ compiler. 00023 * The structures are designed to allow them to be used with any 00024 * structure alignment upto 8. 00025 **********************************************************************/ 00026 00027 #ifndef OCRCLASS_H 00028 #define OCRCLASS_H 00029 00030 #include <time.h> 00031 #ifdef __MSW32__ 00032 #include <windows.h> 00033 #endif 00034 #include "host.h" 00035 00036 /*Maximum lengths of various strings*/ 00037 #define MAX_FONT_NAME 34 /*name of font */ 00038 #define MAX_OCR_NAME 32 /*name of engine */ 00039 #define MAX_OCR_VERSION 17 /*version code of engine */ 00040 00041 /*Image parameters*/ 00042 #define MIN_IMAGE_SIZE 64 /*smallest image that will be passed */ 00043 #define IMAGE_ROUNDING 32 /*all sizes are multiple of this */ 00044 00045 #if defined(__SLOW_TIMES__) 00046 /*Maximum timeouts of various functions (in secs)*/ 00047 #define STARTUP_TIMEOUT 100 /*start of OCR engine */ 00048 #define SHUTDOWN_TIMEOUT 50 /*end of OCR engine */ 00049 #define SENDIM_TIMEOUT 50 /*send of image */ 00050 #define RELEASE_TIMEOUT 50 /*release of semaphore */ 00051 #define READIM_TIMEOUT 100 /*read of image */ 00052 #define READTEXT_TIMEOUT 50 /*read of text */ 00053 #define PROGRESS_TIMEOUT 30 /*progress every 3 seconds */ 00054 #define BADTIMES_TIMEOUT 7 /*max lack of progress */ 00055 #else 00056 /*Maximum timeouts of various functions (in secs)*/ 00057 #define STARTUP_TIMEOUT 10 /*start of OCR engine */ 00058 #define SHUTDOWN_TIMEOUT 6 /*end of OCR engine */ 00059 #define SENDIM_TIMEOUT 5 /*send of image */ 00060 #define RELEASE_TIMEOUT 5 /*release of semaphore */ 00061 #define READIM_TIMEOUT 10 /*read of image */ 00062 #define READTEXT_TIMEOUT 5 /*read of text */ 00063 #define PROGRESS_TIMEOUT 3 /*progress every 3 seconds */ 00064 #define BADTIMES_TIMEOUT 7 /*max lack of progress */ 00065 #endif 00066 00067 /*language definitions are identical to RTF*/ 00068 #define LANGE_NONE 0x0400 /*no language */ 00069 #define LANGE_ALBANIAN 0x041c /*Albanian */ 00070 #define LANGE_BRITISH 0x0809 /*International English */ 00071 #define LANGE_BULGARIAN 0x0402 /*Bulgarian */ 00072 #define LANGE_CROATIAN 0x041a /*Croatian(latin alphabet) */ 00073 #define LANGE_CZECH 0x0405 /*Czech */ 00074 #define LANGE_DANISH 0x0406 /*Danish */ 00075 #define LANGE_DUTCH 0x0413 /*Dutch */ 00076 #define LANGE_FINNISH 0x040b /*Finnish */ 00077 #define LANGE_FRENCH 0x040c /*French */ 00078 #define LANGE_GERMAN 0x0407 /*German */ 00079 #define LANGE_GREEK 0x0408 /*Greek */ 00080 #define LANGE_HUNGARIAN 0x040e /*Hungarian */ 00081 #define LANGE_ITALIAN 0x0410 /*Italian */ 00082 #define LANGE_JAPANESE 0x0411 /*Japanese */ 00083 #define LANGE_KOREAN 0x0412 /*Korean */ 00084 #define LANGE_NORWEGIAN 0x0414 /*Bokmal */ 00085 #define LANGE_POLISH 0x0415 /*Polish */ 00086 #define LANGE_PORTUGESE 0x0416 /*Brazilian Portugese */ 00087 #define LANGE_ROMANIAN 0x0418 /*Romanian */ 00088 #define LANGE_RUSSIAN 0x0419 /*Russian */ 00089 #define LANGE_SCHINESE 0x0804 /*Simplified Chinese */ 00090 #define LANGE_SLOVAK 0x041b /*Slovak */ 00091 #define LANGE_SPANISH 0x040a /*Castilian */ 00092 #define LANGE_SWEDISH 0x041d /*Swedish */ 00093 #define LANGE_TCHINESE 0x0404 /*Traditional Chinese */ 00094 #define LANGE_TURKISH 0x041f /*Turkish */ 00095 #define LANGE_USENGLISH 0x0409 /*American */ 00096 00097 /*font family definitions are identical to RTF*/ 00098 #define FFAM_NONE 0 /*unknown */ 00099 #define FFAM_ROMAN 1 /*serifed prop */ 00100 #define FFAM_SWISS 2 /*sans-serif prop */ 00101 #define FFAM_MODERN 3 /*fixed pitch */ 00102 00103 /*character set definitions are identical to RTF*/ 00104 #define CHSET_ANSI 0 /*Ansi efigs */ 00105 #define CHSET_SHIFT_JIS 128 /*JIS X 0208-1990 */ 00106 #define CHSET_KOREAN 129 /*KS C 5601-1992 */ 00107 #define CHSET_SCHINESE 134 /*GB 2312-80 */ 00108 #define CHSET_BIG5 136 /*Big Five */ 00109 #define CHSET_CYRILLIC 204 /*Cyrillic */ 00110 #define CHSET_EEUROPE 238 /*Eastern Europe */ 00111 00112 /*pitch set definitions are identical to RTF*/ 00113 #define PITCH_DEF 0 /*default */ 00114 #define PITCH_FIXED 1 /*fixed pitch */ 00115 #define PITCH_VAR 2 /*variable pitch */ 00116 00117 /*Bitmasks for character enhancements. 00118 OR these together for enhancement in ocr_append_char*/ 00119 #define EUC_BOLD 1 /*bold character */ 00120 #define EUC_ITALIC 2 /*italic char */ 00121 #define EUC_UNDERLINE 4 /*underlined char */ 00122 #define EUC_SUBSCRIPT 8 /*subscript char */ 00123 #define EUC_SUPERSCRIPT 16 /*superscript char */ 00124 00125 /*enum for character rendering direction*/ 00126 enum OCR_CHAR_DIRECTION 00127 { 00128 OCR_CDIR_RIGHT_LEFT, /*right to left horizontal */ 00129 OCR_CDIR_LEFT_RIGHT, /*left to right horizontal */ 00130 OCR_CDIR_TOP_BOTTOM, /*top to bottom vertical */ 00131 OCR_CDIR_BOTTOM_TOP /*bottom to top vertical */ 00132 }; 00133 00134 /*enum for line rendering direction*/ 00135 enum OCR_LINE_DIRECTION 00136 { 00137 OCR_LDIR_DOWN_RIGHT, /*horizontal lines go down */ 00138 /*vertical lines go right */ 00139 OCR_LDIR_UP_LEFT /*horizontal lines go up */ 00140 }; 00141 00142 /*enum for newline type*/ 00143 enum OCR_NEWLINE_TYPE 00144 { 00145 OCR_NL_NONE, /*not a newline */ 00146 OCR_NL_NEWLINE, /*this is a newline but not new para */ 00147 OCR_NL_NEWPARA /*this is a newline and a new para */ 00148 }; 00149 00150 /*error codes that can be returned from the API functions other than OKAY and HPERR*/ 00151 #define OCR_API_NO_MEM (-2) /*filled output buffer */ 00152 #define OCR_API_BAD_CHAR (-3) /*whitespace sent to ocr_append_char */ 00153 #define OCR_API_BAD_STATE (-4) /*invalid call sequence */ 00154 00155 /*error codes used for passing errors back to the HP side*/ 00156 enum OCR_ERR_CODE 00157 { 00158 OCR_ERR_NONE, /*no error */ 00159 OCR_ERR_CLEAN_EXIT, /*no error */ 00160 OCR_ERR_NO_MEM, /*out of memory */ 00161 OCR_ERR_FILE_READ, /*failed to read data file */ 00162 OCR_ERR_TMP_WRITE, /*failed to write temp file */ 00163 OCR_ERR_TMP_READ, /*failed to read temp file */ 00164 OCR_ERR_BAD_DLL, /*missing or invalid dll subcomponent */ 00165 OCR_ERR_BAD_EXE, /*missing or invalid exe subcomponent */ 00166 OCR_ERR_BAD_LOAD, /*failed to load subcomponent */ 00167 OCR_ERR_BAD_LANG, /*unable to recognize requested language */ 00168 OCR_ERR_BAD_STATE, /*engine did call out of sequence */ 00169 OCR_ERR_INTERNAL1, /*internal error type 1 */ 00170 OCR_ERR_INTERNAL2, /*internal error type 1 */ 00171 OCR_ERR_INTERNAL3, /*internal error type 1 */ 00172 OCR_ERR_INTERNAL4, /*internal error type 1 */ 00173 OCR_ERR_INTERNAL5, /*internal error type 1 */ 00174 OCR_ERR_INTERNAL6, /*internal error type 1 */ 00175 OCR_ERR_INTERNAL7, /*internal error type 1 */ 00176 OCR_ERR_INTERNAL8, /*internal error type 1 */ 00177 OCR_ERR_TIMEOUT /*timed out in comms */ 00178 }; /*for calls to ocr_error */ 00179 00180 /********************************************************************** 00181 * EFONT_DESC 00182 * Description of one font. 00183 * The information required is basically that used by RTF. 00184 * The name may be either a valid font on the system or the empty string. 00185 **********************************************************************/ 00186 00187 typedef struct /*font description */ 00188 { 00189 uinT16 language; /*default language */ 00190 uinT8 font_family; /*serif/not, fixed/not */ 00191 uinT8 char_set; /*character set standard */ 00192 uinT8 pitch; /*fixed or prop */ 00193 inT8 name[MAX_FONT_NAME + 1]; /*plain ascii name */ 00194 } EFONT_DESC; /*font description */ 00195 00196 /********************************************************************** 00197 * EOCR_DESC 00198 * Description of the OCR engine provided at startup. 00199 * The name and version may be reported to the user at some point. 00200 * The fonts array should indicate the fonts that the OCR system 00201 * can recognize. 00202 **********************************************************************/ 00203 00204 typedef struct /*startup info */ 00205 { 00206 inT32 protocol; /*interface version */ 00207 uinT32 font_count; /*number of fonts */ 00208 uinT16 language; /*default language */ 00209 uinT16 name[MAX_OCR_NAME + 1]; /*name of engine */ 00210 /*version of engine */ 00211 uinT16 version[MAX_OCR_VERSION + 1]; 00212 EFONT_DESC fonts[1]; /*array of fonts */ 00213 } EOCR_DESC; /*startup info */ 00214 00215 /********************************************************************** 00216 * ESTRIP_DESC 00217 * Description of the image strip as it is passed to the engine. 00218 * The image is always 1 bit, with 1=black. 00219 * The width is always a multiple of 32, so padding is always OK. 00220 * The height of the full image is always a multiple of 32. 00221 * The top y coordinate is 0, and increases down. 00222 * The top leftmost pixel is in the most significant bit of the first byte. 00223 **********************************************************************/ 00224 00225 typedef struct /*bitmap strip */ 00226 { 00227 inT16 x_size; /*width in pixels */ 00228 inT16 y_size; /*of full image */ 00229 inT16 strip_size; /*of this strip */ 00230 inT16 resolution; /*pixels per inch */ 00231 uinT8 data[8]; /*image data */ 00232 } ESTRIP_DESC; /*bitmap strip */ 00233 00234 /********************************************************************** 00235 * EANYCODE_CHAR 00236 * Description of a single character. The character code is defined by 00237 * the character set of the current font. 00238 * Output text is sent as an array of these structures. 00239 * Spaces and line endings in the output are represented in the 00240 * structures of the surrounding characters. They are not directly 00241 * represented as characters. 00242 * The first character in a word has a positive value of blanks. 00243 * Missing information should be set to the defaults in the comments. 00244 * If word bounds are known, but not character bounds, then the top and 00245 * bottom of each character should be those of the word. The left of the 00246 * first and right of the last char in each word should be set. All other 00247 * lefts and rights should be set to -1. 00248 * If set, the values of right and bottom are left+width and top+height. 00249 * Most of the members come directly from the parameters to ocr_append_char. 00250 * The formatting member uses the enhancement parameter and combines the 00251 * line direction stuff into the top 3 bits. 00252 * The coding is 0=RL char, 1=LR char, 2=DR NL, 3=UL NL, 4=DR Para, 00253 * 5=UL Para, 6=TB char, 7=BT char. API users do not need to know what 00254 * the coding is, only that it is backwards compatible with the previous 00255 * version. 00256 **********************************************************************/ 00257 00258 typedef struct /*single character */ 00259 { 00260 // It should be noted that the format for char_code for version 2.0 and beyond is UTF8 00261 // which means that ASCII characters will come out as one structure but other characters 00262 // will be returned in two or more instances of this structure with a single byte of the 00263 // UTF8 code in each, but each will have the same bounding box. 00264 // Programs which want to handle languagues with different characters sets will need to 00265 // handle extended characters appropriately, but *all* code needs to be prepared to 00266 // receive UTF8 coded characters for characters such as bullet and fancy quotes. 00267 uinT16 char_code; /*character itself */ 00268 inT16 left; /*of char (-1) */ 00269 inT16 right; /*of char (-1) */ 00270 inT16 top; /*of char (-1) */ 00271 inT16 bottom; /*of char (-1) */ 00272 inT16 font_index; /*what font (0) */ 00273 uinT8 confidence; /*0=perfect, 100=reject (0/100) */ 00274 uinT8 point_size; /*of char, 72=i inch, (10) */ 00275 inT8 blanks; /*no of spaces before this char (1) */ 00276 uinT8 formatting; /*char formatting (0) */ 00277 } EANYCODE_CHAR; /*single character */ 00278 00279 /********************************************************************** 00280 * ETEXT_DESC 00281 * Description of the output of the OCR engine. 00282 * This structure is used as both a progress monitor and the final 00283 * output header, since it needs to be a valid progress monitor while 00284 * the OCR engine is storing its output to shared memory. 00285 * During progress, all the buffer info is -1. 00286 * Progress starts at 0 and increases to 100 during OCR. No other constraint. 00287 * Every progress callback, the OCR engine must set ocr_alive to 1. 00288 * The HP side will set ocr_alive to 0. Repeated failure to reset 00289 * to 1 indicates that the OCR engine is dead. 00290 * If the cancel function is not null then it is called with the number of 00291 * user words found. If it returns true then operation is cancelled. 00292 **********************************************************************/ 00293 typedef bool (*CANCEL_FUNC)(void* cancel_this, int words); 00294 00295 typedef struct ETEXT_STRUCT /*output header */ 00296 { 00297 inT16 count; /*chars in this buffer(0) */ 00298 inT16 progress; /*percent complete increasing (0-100) */ 00299 inT8 more_to_come; /*true if not last */ 00300 inT8 ocr_alive; /*ocr sets to 1, HP 0 */ 00301 inT8 err_code; /*for errcode use */ 00302 CANCEL_FUNC cancel; /*returns true to cancel */ 00303 void* cancel_this; /*this or other data for cancel*/ 00304 clock_t end_time; /*time to stop if not 0*/ 00305 EANYCODE_CHAR text[1]; /*character data */ 00306 } ETEXT_DESC; /*output header */ 00307 00308 #ifdef __MSW32__ 00309 /********************************************************************** 00310 * ESHM_INFO 00311 * This data structure is used internally to the API to hold the handles 00312 * to the operating system tools used for interprocess communications. 00313 * API users do not access this structure directly. 00314 **********************************************************************/ 00315 typedef struct /*shared mem info */ 00316 { 00317 HANDLE shm_hand; /*handle to shm */ 00318 HANDLE mutex; /*alive check */ 00319 HANDLE ocr_sem; /*ocr semaphore */ 00320 HANDLE hp_sem; /*hp semaphore */ 00321 void *shm_mem; /*shared memory */ 00322 inT32 shm_size; /*size of shm */ 00323 } ESHM_INFO; /*shared mem info */ 00324 #elif defined (__MAC__) 00325 typedef struct /*shared mem info */ 00326 { 00327 Boolean mutex; /*alive check */ 00328 Boolean ocr_sem; /*ocr semaphore */ 00329 Boolean hp_sem; /*hp semaphore */ 00330 void *shm_mem; /*shared memory */ 00331 inT32 shm_size; /*size of shm */ 00332 inT16 language; 00333 00334 // Process management information follows: 00335 ProcessSerialNumber IPEProcess; 00336 ProcessSerialNumber OCRProcess; 00337 } ESHM_INFO; 00338 #elif defined (__UNIX__) 00339 typedef struct /*shared mem info */ 00340 { 00341 void *shm_mem; /*shared memory */ 00342 inT32 shm_size; /*size of shm */ 00343 } ESHM_INFO; 00344 #endif 00345 #endif