G L O S O R -= DICTIONARY-PRACTICE =- (C) Copyright 1996-97 Joakim Ek ver 3.20 Bulgarian/English/Swedish The program shows the word either in Bulgarian, and wants the answer in English or Swedish, or ii asks in Swedish and/or English, and wants the answer in Bulgarian, either spelled, pronounced or latinized. The questions are made at random from all the words in the current database. There is a prompt saying "BUL->S/E" for Bulgarian to Swedish/Englis or "S/E->BUL" for the other way. You may also make the color change depending on the language of the question. See options futher down. Your answer must match exactly the answer in the database, either just the normal+hilited, or the complete word(s) including the dim. It will consider it wrong if it is misspelled, words are swapped, or only a part of the dimmed is typed in. It is not case-sensitive though, but normally you shall use only upper-case letters for everything. When using codepage jek850 or jek852 it will turn on CAPSLOCK when the answer is to be typed in latin, and turn it off when it wants cyrillic. This is becaus the keyboard-driver KBJEK has a mode, 4, that puts uppercase latin on uppercase-keys and upper-case cyrillic on lower-case keys. Currently only theese two special codepages are supported for cyrillic. You will have to use the drivers KBJEK and JEKFONT to support them. If you do not have them you can still use the program with just the latinized words. The latinization of words are made as phonetic in correspondense to swedish pronounciation, wich differs slightly from the normal ones used. Especially because the letters and is used, that are not present in English alphabet. The following table describes the cyrillic and corresponding latin letter. The key that is used to type when using jek850 is also displayed: CYR 850:a b c d e f g h i j k l m n o p q r s t u v w x y z CYR 852: SOUND: A B ZH D E F G CH I J K L M N O P SHT R S T JO V SH TS TJ Z JA KEY: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z ENGLISH KEY: (A-Z SAME AS SWEDISH) [ ; ' ] On the line between the questions there is a ( ) with numbers inside. First (or only) is the number of remaining words in the database. The second number, if present, are the number of words that have not been correctly answered. On the first row beneath the line is the number of asked questions, the number of right answers and the right answers in %. The Bulgarian is always at the left half of the screen, and the Swedish/English at the Right, the answer-field therefore is moved between right and left. When a question is answered correctly a * will mark that question. You may also set an option that will turn the color of right answered questions green, and wrong answers red. If you think an answer is right, even though the computer thinks you are wrong (like omitting one of two options) you might override it by beginning the next answer with + or , The previous question will be marked with + (instead of *) and, if enabled, the color made green. The score will also be added by 1. Note that if you use + after a question that already was answered right, the score will be wrong. The Bulgarian word is normally displayed with one letter hilighted, this is the letter/syllable that should be stressed. There are also some dim letters/words. This is either used to display information like SING for singlyular, V for verb and others, and also to make some words optional in the answer if there are several words corresponding to one word in the other language. It does not mean that the dim word is lesser correct, even though the most used word tends to be put first. Words separated with , are of the same meaning while ; separates words of diffrent meaning. Sometimes there are extra words added, like "LIKE {TO}", where TO is dim, and the same additional word is dim,ed in Bulgarian. Hopefully it will be clear what is ment with what words, dims, separators and paranteses. (This is also slightly diffrent in the diffrent data-files). XXX is used for where another word(s) should be fitted. When all words in the database is done, it will restart. Depending on the option it will ask qll words again, or just the ones you answered wrong. You may also specify wheather the ones corrected using + should be asked again. The score-counter is not reset. You will have to exit the program if you want to begin at 0 again. If all words are answered correctly it will start over with all words again. On the line between the questions "OMSTART" indicates restart, and "OMSTART, ALLA ORD" a restart with all words. To exit the program answer with just Q or 0. It uses DOS to input keystrokes, this is because no internal functions in QB allowed ascii 224 to be entered. (used for cyrillic A in my jek850). This means that some keys don't work normally. Like F-keys send strange letters, you'll just have to not hitting them! Also it made me add the option to hit the TAB-key to close the DOS-key-read, to be able to call up RELEX (an excellent TSR dictionary, not by me, not for free) by hitting ALT+1. You cannot type any letters before activationg the normal key-routine again by once more hitting TAB. A little tune indicates that TAB is hit on or off. While in off-mode you might move the cursor though. Use arrows to move one step or PgUP/PgDown to move one question up_/down or Home/End to move to the first column of Swedish/English or Bulgarian. This is used to take advantage of that RELEX can search for the word under the cursor when hitting ALT+1. If you have RELEX you can replace the RELEX.CHR with my RELEXCHR.850 to use my jek850 codepage in relex. I also supply RELEXCHR.BDS which is the original BDS codepage file for relex. I cannot supply RELEX since it is not my product. There is a BAT.file with example how to start GLOSOR.EXE. To get the help displayed type GLOSOR /? This help is displayed in swedish. You might make your own data-base of words. The program uses a special format internally, but it has a conversion included that converts to its own format from a text-file of some formats. GLOSOR /KONV? will display some help about this, in swedish. Here are a description in English: If no ##-line (described later) is present the conversion defaults to a file with the words in the order Bulgarian, English , Swedish. The bulgarian should be in jek852 codepage (cyrillic on lowercase latin, as described in the pronounciation-table above, the KEYS-line) and the fields should be separated with comma , Each word/question is on one line, that means CR/LF separates each word&question. Several words with the same meaning should be separated with . (wich is then converted to , that is used in the program). The words can be entered in 'TON'-format, that is [] around characters that should be hilighted (stressed syllables) and {} around letters/words that should be dimmed. There are some special things to have in mind when using [] and {}: If using a [] inside a {} you should type it like: NORMAL{,D[I]}{MMED} That is a closing } should always come directly after a closing ], then you can put another opening { if there is more dim to come. If this is not made the last half of the word will be displayed in normal, not dimmed. Also note that when dimming a word you should also dim the apropriate spaces to make it possible to answer without the dim. Like: ONE{ TWO} THREE This will make one space optional, and "ONE THREE" will be ok. If you type ONE {TWO} THREE you will have to answer "ONE TWO THREE" or "ONE THREE", with two spaces, to get it right. (This is currently not made correctly in the 1600-word data-file). Example of a source-file: v[]da,WATER,VATTEN nq {F},NIGHT Convert it using GLOSOR /KONV (If just this is specified, without filename, it will look for GLOSOR.TON in the above described format) You may specify a filename like GLOSOR /KONV:MINFIL.EXT and it will convert this file to GLOSOR.DAT (the internally used file). Note that it will empty the remaining records of a previous GLOSOR.DAT but not treuncate the file. Therefore it is normally best to delete GLOSOR.DAT before making a conversion. If you start the first line in the source-file with ## this indicates that a parameter-line is present. This line tells the conversion how to handle the file. If any lines in the file begins with # that line will be ignored. (exept of course the first one) This can only be done if a ## line is present first. If a line in either file-format does not contain a Bulgarian (spelled) word or neither of Swedish nor English that line will be skipped. Description of the ## parameter line: ##852TON.;,BULENGSVE ^ ^ ^ ^ ^ | | | | | | | | | 2,3 or 4 fields in the order they appear in the source-file. | | | | BUL=bulgarian(spell), ENG=english, SVE=swedish, UTL=Bulgarian | | | | pronounciation. (If UTL is not present pronounciation is calculated | | | | using grammar rules that normally gives the right result) | | | Separators. First the one used between words of the same meaning (, | | | in the program) then the one used between words of diffrent meaning | | | (;) and third the one used between the language-fields. | | | If # is entered as the language-field-separator the file is considered | | | to be fixed-spaceed instead of character-deliminated. Se later. | | | (If you dont use word-separators enter the default , and ; ) | | Formatting-type. 'TON' is [] around hilighted letters, 'U22' is ASCII 22 | | around hilighted. Both use {} around dim letters/words. 'TXT' means no | | formatting (hilight or dim) is present in the source. | CodePage (EK.CPI) Only '850' or '852' is supported. (jek850 and jek852). | This is for the BUL and UTL fields. ENG and SVE should be uppercase-ascii. ## Indicates that this is a parameter-line. The above is for character-deliminated files, for fixed-space files enter # as the field-separator character, and put the language-fields with # in front on the startposition of each field, like this: ##852TON,;#BUL #ENG #SVE voda WATER VATTEN The first field, #BUL in the example, begins at position 1 but should be entered directly after the options on the parameterline as showed here. You have to have a Bulgarian, BUL, field. Even if you do not intend to use cyrillic. Though you might just enter anything, like a . ,and then enter a pronounciation in the UTL field if using only latinized when practising. If you have my xxx.G files and the G.BAT you can see some examples of conversion. (This files is either supplied with the program or in a special archive probably called G.ZIP from the same internet-site) COMMANDLINE-OPTIONS ------------------- There are several options that can be used to modify the way GLOSOR works. First there are the conversion-options, that should be used alone, and that do not start the actual practis-program. Above we explain: /KONV[:fil] Convert text-file (Default GLOSOR.TON) to GLOSOR.DAT (Database) There are one more conversion available: /SPAB Export words to modified SPAB Windows program This will make files for a special modified version of a windows-program that was made for german-words. Since this is not my program, and therefore not available from me, i don't bother to explain more about this option. Then there are all options and toggles that modifies the behaviour of the practice-program. They can be entered in any order: /CP[:NNN] CodePage NNN for cyrillic is loaded and should be used If /CP isa not specified only latinized bulgarian will be used, If /CP is used without number it will default to 852. You may specify 850 (jek850), 852 (jek852), 0 or 437 (only latin). You will have to load the cyrillic driver before starting the program. (See BAT-file for examples using KBJEK/JEKFONT) /CGA Use CGA graphic-mode (Monochrome). For some cyrillic-drivers. /SVE Always ask questions in Swedish/English, answer in Bulgarian. /BUL Always ask questions in Bulgarian, answer in English or Swedish /EJLAT Do not show latinized translation of the cyrillic words (Only valid if /CP is used, Latin is always showed with CP 0/437) /BULGKEY (Or /BK) Translate sedish keyboard to standard bulgarian layout (if no bulgarian keyboard-driver is present. Might be used if only DISPLAY.SYS/NLSFUNC/JEK.CPI) /FRG Show diffrent colors depending of the questions language /FELFRG Change the color depending on if the answer was correct or not /MONO Only uses black and white colors, for monochrome displays. /KANINTE Only reuse words that was incorrectly answerd on the next round /+KANINTE As above but do not reuse words that are + corrected either /PIP Make a diffrent sound indicating if an answer was right or wrong. This program is not primarilly done to be released, i made it for my own use. (thats one reason i haven't bothered to make support for any other codepages but my own, since I normally use only them) but if you like you can make suggestions. I won't say I'll do anything you ask for, but maby. My e-mail address is probalby: jek@swipnet.se or jocke@ecedata.se This program, and KBJEK and JEKFONT, can most likley be downloaded from: http://www.ecedata.se/jocke/files There will be links to this progrm, GLOSOR, and some other files. THIS PROGRAM IS DISTRIBUTED AS FREEWARE 'AS-IS' WITHOUT ANY WARRANTY WHATSOEVER REGARDING FUNCTIONALITY OR ANYTHING ELSE. I DO NOT TAKE RESPONSIBILITY FOR ANYTHING THAT CAN HAPPEN WHILE USING THIS PROGRAM, NOR CAN I GUARANTEE THAT THE CONTENTS OF THE DATA-FIELS IS CORRECT. Regards jek 97-09-24