Almathera Ten Pack 3: CDPD 3

home *** CD-ROM | disk | FTP | other *** search

/ Almathera Ten Pack 3: CDPD 3 / Almathera Ten on Ten - Disc 3: CDPD3.iso / jam / jamdisk6 / dspell.run / DSpell.DOC < prev next >

Wrap

Text File | 1991-09-18 | 39.5 KB | 1,224 lines

Documentation for DSPELL version 3.1 Copyright (C) Rod Young - September 1991 This programme is intended to be Shareware; that is: 1) You may copy it and distribute it in any way you feel providing that all the release files are included, BUT YOU MUST NOT SELL OR PROFIT FROM IT!! (A small fee for disks and postage may be levied by clubs, Buletin Boards etc. BUT THATS ALL!). 2) If you find yourself using it frequently you must register with me at the address given at the end of this document a) If you are waged, a registration fee of, say £15.00 would seem suitable b) If you are unwaged, a nice letter or a bit of your own work will suffice. c) If you are a business, an Institution or are just stinking rich, just think of how much you would have had to pay if this had been a commercial programme, and react accordingly! 3) No restrictions are made if you wish to have multiple copies, and there is no site licence rubbish to bother with. If you fail to comply with 1 and 2 above then your Agnes will melt with embarrasment! Compliance means free upgrades (if I make any) and reduced price copies of anything else I may come up with in the future. INTRODUCTION ~~~~~~~~~~~~ I have for many years now been fascinated by words, their use, misuse and construction, as a professional graphic artist responsible for the publication of seventeen quarterly and one monthly magazine, I have become increasingly alarmed by the great number of misspellings that are appearing in submitted articles, and in publications in general. The majority of these errors are of the type termed `Americanism'. I could not really understand this considering the power of modern word-processors and their associated spell-checking programmes. That is until I started using these spell-checkers on a regular basis and found to my utter dismay that even programmes that boast UK English dictionaries actually use cut-down American ones! You can normally catch these out with the following test: Skillfull Correct US English spelling. Skilful Correct UK English spelling. Skillful Cut down dictionary version illegal in both countries. This is a single and very simple example but demonstrates the American dictionary writers awareness of the single `l' double `l' difference between the two versions of our language, but their ignorance of the more subtle forms of UK English. Even WordStar and WordPerfect the so-called `industry standard' word-processors fall into this trap. I appeal to these companies to get their act together and realise that the only persons able to correctly cut an American dictionary into UK English are those born and brought up in the language. End of gripe and on to DSpell; the programme was written in response to MY need to check editorial from the UK, America and Europe for correct spelling. And as I was using an Amiga for recreational as well as professional work I decided to write the programme for the Amiga, and make it look like an Amiga programme despite, the fact that I was using IBMs and MACs exclusively at work (by the way, all the programmes I have used on these systems fail the above `skilful' test). I was only programming in Basic at the time and soon realised that this was far too slow to accomplish the task in hand, so, taking a deep breath I bought Lattice C and sat down to learn `C' programming from books. It has taken over two years to complete the programme with many a visit to the`Guru' on the way, nearly 60% of this time was spent checking words against main-stream dictionaries, a long task sadly overlooked in most of the PD spelling checkers available to date, programmers should remember that a programme is only as good as its data! Before going on to the programmes it would be wise to remember these words: NOTHING WILL CATCH A WHITE WORD IN A WRONG PLAICE!!! RUNNING THE PROGRAMMES ~~~~~~~~~~~~~~~~~~~~~~ Dspell is designed to be run from the CLI but you can run it from the Workbench using a program like Xicon. Upon booting dspell will sense if you have the interlace flag set in preferences and will open in interlace mode, but only if you have over 1.5 meg of memory. Dspell needs large amounts of memory for its data files so be aware of this at all times. If you are going to be doing a lot of spell checking in one go, I advise resetting the Amiga prior to the session to clear memory. If you are low on memory try reducing or removing any background tasks you might have running, screen blankers etc. The files in this release marked LARGE are the files that I am currently using with 3 meg, they WILL NOT run on a 1 meg machine, and you might as well give up now if you only have half a meg, this programme is SERIOUS!! The DSpell release should contain the following programmes: DSPELL The main spelling-checker. DSPELL.UDCT DSpell's user dictionary. DSPELL.DATA The compiled dictionary used by DSpell. DSPELL.PFX A dictionary of prefixes and combined forms that will be passed automatically. DSPELL.ABS A dictionary of common abbreviations. DSPELL.BAD A cross-checking dictionary to catch illegal words otherwise passed in prefix checking. DED The ASCII dictionary editor. MAIN044.DCT The main ASCII dictionary file. DASH The full dictionary compiler. DASH3 The `streamlined' dictionary compiler. _ DCount | DChop | DExpand | FlagExp |- Dictionary maintenance tools DoubCut | RootStrip _| _ Report | Strip |- File Tools CURSON A small C-programme to restore your cursor should you abort Dash, or any tool programme by using Ctrl-C. Dspell will also generate two files called DSPELL.LOG and LAST.LOG INSTALLATION ~~~~~~~~~~~~ Floppy Disk Users ~~~~~~~~~~~~~~~~~ DSpell was developed to run on a hard disk for best results, but will run from a floppy disk, however the load times for the dictionaries and document will be considerably longer. DSpell looks for its dictionary files on the logical device DSpell:, if you are running from a single floppy disk it must be named DSpell and these programmes must be on it as well as DSpell itself. If you have two drives I suggest that you have a separate disk called DSpell for the dictionary files, and another (called anything you like except DSpell) for Dspell and the files to be checked. In both cases copy all the file/dictionary support tools and the ASCII dictionary itself to a separate disk for editing. It is important to keep as few files as possible on the DSpell: disk as room is required for the user dictionary and the log file to be saved at the end of a session, these files can become quite large so be careful. Hard Disk Users ~~~~~~~~~~~~~~~ DSpell can be installed anywhere on the hard disk but I suggest you put it in the SYS:C directory where it can be called at any time. The files DSPELL.UDCT and DSPELL.DATA must be placed in a directory and the following assignment made in either your startup-sequence or from the CLI: Assign DSpell: <path>:<directory-name> for example: Assign DSpell: DH3:Text.Progs/Spell_Checkers where the directory Spell_Checkers is a sub-directory of Text.Progs on partition DH3:. *********************************************************** THE PROGRAMMES *********************************************************** DSPELL ~~~~~~ FEATURES ~~~~~~~~ DSpell has the following features: 1) Fully intuitionised,supporting mouse and keyboard input. No menus are used - everything you require is available with a single mouse click or key press. 2) A fifteen line display of your document (45 lines in interlace mode), with a moving cursor to allow you to see the progress of the checking, and inspect errors in context with the rest of the document. 3) Instant editing of suspect words with re-checking of replaced words. 4) Ability to replace or ignore suspect words globally or singularly. 5) Full Capitalisation checking from both the dictionaries and on sentence boundaries. 6) Checking for illegal usage on certain words. 7) Hyphenation Checking for prefixes. 8) Fully user-editable dictionary with its own specialised Editing programme. 9) Size of dictionaries and documents limited only by disk space and memory PREPARATION ~~~~~~~~~~~ DSpell will only successfully process ASCII files with no imbedded formatting code (Bold, Underline etc), it is tolerant of tabs although a highlighted word may appear out of position in the display. This will not affect the final document when your corrections are saved to disk. Amiga codes can also cause unusual display upsets but normally these are restricted to colour changes. If your document does contain formatting code the results are unpredictable, anything from simple misplaced highlighting to the complete display suddenly vanishing. Two tools for checking files for formatting codes and removing them - Report and Strip - are included on the disk. Remember the Abort gadget is always active even if invisible and will return you to the file requester. SYNTAX ~~~~~~ DSpell [?] [x] [a] [s] [nc] [s line no.] DSpell by itself runs the programme with all defaults in operation, the flags modify these as follows ? Brings up a syntax reminder for the rest of the flags. x Turns off global replacement completely (global ignore is always available). This option is in effect for the whole of the current spelling session. a This one for dictionary building buffs. It turns off all gadgets except Abort and Save & Exit and runs through a file adding every word, abbreviation or acronym not in the main dictionary to the user dictionary. Floppy users beware - this can create enormous files very quickly. This option is in effect for the whole of the current session. nc Turns off all capitalisation checking for the whole session. s Start checking from the line number set in [line no.]. This option is in effect for the first file only. DSpell is the main spell-checker. Upon booting DSpell takes some seconds to initialise and load the main dictionaries. A load 'meter' lets you know how thing are progressing. Once the dictionaries are loaded you are presented with a standard file requester. Those of you familiar with the use of file requesters can skip the next section, just remember the Delete gadget works!!! The Requester ~~~~~~~~~~~~~ The file requester was written by Anders Bjerin and is public domain. In my `umble opinion it is one of the best requesters I've ever used, better than many you get with many commercial programmes. Choosing a Path: ~~~~~~~~~~~~~~~~ Simply click on one of the drive gadgets at the top of the requester and the contents of that drive are displayed in the main area. Directories are at the top and are coloured yellow, with files below in white. To access a directory just click on its name. If the drive you require is not listed, simply type in its name in the Drawer gadget. Next to the Drawer gadget is a small box with a green `P' (small joke) in it, this is the `Parent' gadget that returns you to the previous level in the directory structure. Choosing a file: ~~~~~~~~~~~~~~~~ At the top of the requester is a small area labelled Ext:, entering a file extension here (eg .doc) will limit the requester to displaying only files with that extension. Clicking on a file name pastes the file into the File: gadget, you then click on LOAD to start checking that file. You can also scroll through the list of files with the scroll gadget to the right. When saving you can edit the name of the file in the File area to save it under a different name, or simply click SAVE to overwrite the existing file. DELETE is a file management option and does just that, the programme prompts on all dangerous operations to verify your choice. QUIT will not prompt if you haven't saved. After selecting a file and clicking on LOAD Dspell immediately starts checking the document. The main display area is split into two parts as follows: The Bottom Display ~~~~~~~~~~~~~~~~~~ The bottom section displays the document being checked so you can see possible errors in context. This display is 15/45 lines by 75 characters. The cursor moves through the text so you can watch progress. When DSpell comes across a word that is not in the dictionary the word is highlighted and the programme waits for you to decide how to deal with the word. The display may not accurately reflect the layout of your document; the programme has been designed to suit documents prepared for publication (DTP), where the only carriage returns are at the end of paragraphs; if your document contains `terminated' lines that are longer than the display width, the end words will wrap round on to a line of their own. The Top Display ~~~~~~~~~~~~~~~ The top section is where all the action occurs. The very top row contains the Abort gadget, which is active at all times; a reminder of the file-name being checked, and a line counter. The line counter is based on the display width of 72 characters and tells you how far you have got in the document, make a note of the line number should you wish to abort spelling in the middle of a document, and restart at the same place later (see `s' flag). The file-name reminder area is also used to display alert and information messages. EDITING ~~~~~~~ DSpell performs multiple checks on any single word as follows: 1) The word is checked against the main dictionary. 2) If (1) fails the word is checked to see if it is a valid plural. Plurals with the forms `s', `es' and `ies' are supported. The word is reconstructed into its `root' form (i.e. verifies becomes verify and re-checked. 3) If this fails the word is checked to see if it ends in one of the `LY' forms, a similar reconstruction to item 2 takes place (i.e. formally becomes formal) and the word is re-checked. 4) If this fails the word is then checked to see if it begins with a valid prefix (re, de, pro, under etc) if so the prefix is removed and the `root' word is re-checked. 5) If all checks fail entirely the word is passed to the user. Plural and `ly' checking is very reliable so no record is kept however prefix checking may pass the odd unusual word or words where hyphens are needed. These are reported to the file DSPELL.LOG which should be read after every session. DSpell creates this file if it does not exist and backs it up to a file called LAST.LOG if it does. Any unusual words passed by the programme should be entered separately in the file DSPELL.BAD and any future occurrences will be caught by the programme. When DSpell detects an unknown word it is highlighted in the text and a copy is automatically pasted into the edit area (the area that appears sunken on the display). The programme halts while it waits for your decision on what to do and the programme may also prompt for a certain course of action to be taken, all such prompts appear in the file reminder area at the top of the screen. The suspect word is immediately available for editing using the keyboard, or for other operations using the mouse. The forward and back arrow keys, the delete and backspace keys all produce the standard editing effects, (note the `Right Amiga X' combination is not implemented to clear the line). Partial line editing is available using the backslash, if you have the following line in your text - Roger was unable. to respell his own name The error here is the unwanted full stop and also you might want `respell' to just read `spell'. Dspell will stop on the word `to' with an alert message: Capital Letter Expected. `to' will be automatically capitalised and placed in the edit area; edit the word to look like this \\ to \\\ the first two backslashes remove the full stop and the space before the word, note the space after the slashes, this replaces the removed one. The back three slashes remove the space after the word and the first two letters of the following word, again note the space in the replacement string. After any edits have been made to a word the following options are available (Note that all options refer to the word in the edit area and not the highlighted word in the text): Save & Exit: ~~~~~~~~~~~~ Saves any edits made to the file and returns you to the file-requester. Ignore Once: ~~~~~~~~~~~~ Skips the occurrence of the suspect word once only. It will be picked up again the next time it is encountered. Ignore for Rest of File: ~~~~~~~~~~~~~~~~~~~~~~~~ Skips all occurrences of the unknown word throughout the text, essential when checking such things as programme documents where unusual names (e.g. DSpell) occur frequently in the text. Add to User Dictionary: ~~~~~~~~~~~~~~~~~~~~~~~~ If the unknown word is a proper word you can add it to the user dictionary (DSPELL.UDCT), whether the word is proper or not is entirely at your discretion. The user dictionary is an ASCII file that can be loaded directly into the dictionary editor (DED) for processing prior to inclusion into the master dictionary. If you have edited the word it will be replaced in the text and all further occurrences of that word will be automatically replaced Please be aware that the number of words in the user dictionary bears a direct relationship to the speed of programme execution; when the user dictionary contains more than 150 words a small yellow square appears in the top right-hand corner of the display to alert you that its about time to process the dictionary. Here I should add that DSpell should always be used in conjunction with a good UK dictionary for accurate double checking. Replace Once: ~~~~~~~~~~~~ Replaces the highlighted word in the text area with the current contents of the edit area. Replace In Rest of File: ~~~~~~~~~~~~~~~~~~~~~~~~~ Replaces the current contents of the edit area into the text and when DSpell encounters another occurrence of the misspelling it will automatically replace the word without prompting, this operation is case sensitive, so DSpell will treat `Skillful', `skillful' and `SKILLFUL' as separate misspellings of skilful. Note that this option can be disabled entirely by using the `x' flag option. In both this option and `Replace Once' the replaced word is re-checked before you can continue. This allows for many goes at getting the word right before resorting to the Guess Spelling option. NOTE: DSpell is not a word-processor, there is a 38 character limit to the size of your replacement copy to accommodate words such as floccinaucinihilipilification and antidisestablishmentarianism, so you could technically replace a word with a short phrase, but be aware that all occurrences of that word will be replaced by the phrase. Such major changes should be done with your normal word-processor. DSpell will flash the screen if you attempt to overwrite the limit. Replaced lines that overwrite the display width will appear truncated with a red arrow to indicate the overflow. Keyboard Shortcuts ~~~~~~~~~~~~~~~~~~ Hitting the Return key will emulate the `Ignore Once' option if no edits have been made to the word and `Replace Once' if the word has been edited. Guess Spelling ~~~~~~~~~~~~~~ This is activated by the HELP key (situated between the main and numeric keypads) and opens a small window. The programme then searches the entire dictionary for possible words (this may take several seconds, in the case of prefixed words the dictionary is actually searched twice). You will be alerted when a prefix search is about to take place and can abort it by clicking anywhere in the Guess Spelling window while the alert is visible. If the word to be guessed is a plural or end in one of the `ly' forms (or both) this alert will appear: CAUTION! Constructed Words Dspell will have cut the plural from the original and now attempts to pluralise the words it is searching for. Occasionally it can make some very odd plurals up! hence the warning, make sure of your word before replacing. The list of possible words is then displayed in the window, with capitalisation to reflect the word in the text, words generated with prefixes are displayed last and in a different colour. Click on the word you require; the word is pasted into the edit area; close the Guess Spelling window using it's close gadget. The word is then available for further editing or direct replacement into the text. Note: no other tools are available while the Guess spelling window is open. OTHER TOPICS ~~~~~~~~~~~~ Funny Words ~~~~~~~~~~~ During Guess spelling DSpell always looks for the complete word first, after that it looks for a common prefix (under,pre, re, over, for example) and cuts it off. It then looks for the remaining part of the word. For example you may have mistyped the word underground as: Undergroind Dspell fails to find this so cuts off under and searches for `groind', when it finds a possible match, such as `ground' it replaces the prefix and displays the word `underground'. However because the pattern matching is `blind' it will also provide you with such juicy words as: Undergrind Underbroiled Underbroiler These are always placed last in the list and are in a different colour, `real' words are always displayed in white, so if all the words in the window are coloured they have all been `manufactured'. If the word being searched is a plural DSpell will pluralise all the words in the Guess Spelling window. At the bottom of the window you will see either the word MORE or the word AGAIN, explained thus: MORE: clicking on it brings up another window full of possibilities. AGAIN: you have gone through the whole available list and can go back to the beginning. Double Wording ~~~~~~~~~~~~~~ Dspell automatically checks for occurrences of double wording providing no characters appear between the two words, if you completely remove the second word DSpell automatically attempts to adjust the word spacing to suit, this is effective in most cases. Capitalisation ~~~~~~~~~~~~~~ DSpell checks for instances of the first person singular (I, I've, I'll etc.) and ensures that they are properly capitalised by stopping with the alert "Capital Letter Expected", however, say you are checking a programming file that uses a variable `i' in about 300 places with white space either side, you will be in for a tough time as this check overrides the `Ignore For Rest' option, try the `nc' flag option in the command line. DSpell also checks that the first letter of a new sentence is a capital. The programme defines the end of a sentence (EOFS) as being a period(.), an exclamation mark(!) or a question mark(?). Abbreviations ending in periods i.e. (that's one) e.g., Dept., Ltd., Approx., etc. (that's another) are treated as not being EOFS's. You can place common abbreviations in the file DSPELL.ABS. In the case of trailing dots... ...these can be a `lead-in` or an EOFS, the programme treats them as being EOFS's to be safe. If they are a lead-in in your document simply click ignore. DSpell supports full capitalisation checking dependent on how the words are capitalised in your dictionary. Be careful here, you will capitalise words such as January and February of course, but NOT may (as in: may not) and march (as in: the march of time). Two types of capitalisation are supported. If the word begins with a capital in the dictionary but not in the text, you will be prompted for the word to be replaced with the alert: Capital Letter Expected! If the second letter is a capital DSpell assumes the whole word is capitalised, and you will be prompted with the alert: All Capitals Expected! In both cases the correctly capitalised word is placed in the edit area for quick replacement. `Ignore For rest' will ignore all further occurrences of the word in lowercase. Prefixes, Hyphenation and the Log File ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ During the course of a spell-checking session the programme checks words by cutting possible prefixes as mentioned earlier. Many prefixes can just be cut and the root word passed, however there are quite a few which will require hyphenation for clarity. With this type the programme will stop with the alert message: Hyphen Suggested! if the last character of the prefix matches the first character of the root word. For all other cases DSpell creates a file called DSPELL.LOG into which it writes the word in question along with its position in the file by line and word count. NOTE the lines refer to DSpell's 72 character line set-up, you must set your word processor to this width to accurately trace the word by this method. There is another group of prefixes which nearly always require hyphens, the programme will always stop at these with the alert: Hyphen Usually Required! In both cases when the programme stops, a ready hyphenated example is placed in the edit area for quick replacement. You can place hyphenated words n the main dictionary in which case the programme will stop with the message: Word Requires Hyphenation! QUITTING DSPELL ~~~~~~~~~~~~~~~ During the loading of the dictionaries you can leave the programme by clicking anywhere in the top area otherwise the QUIT gadget in the file-requester is the only way of leaving the programme (short of re-booting that is). If during loading of the file to be checked you click in the upper area you will be returned to the file-requester. When DSpell comes to the end of a document your are presented with the file-requester again and you can choose to save or quit. At any other time the Save & Exit or the Abort gadgets can be used to return to the file-requester. DICTIONARIES USED ~~~~~~~~~~~~~~~~~ Every care has been taken to ensure the words in the dictionary are correct UK English. When in doubt they have been checked against the following dictionaries: Chambers Concise Dictionary ~~~~~~~~~~~~~~~~~~~~~~~~~~~ © W & R Chambers Ltd and Cambridge University Press 1988 ISBN 1-85296-011-6. This is a good dictionary to have as it contains many examples of correct suffix usage. The Oxford Spelling Dictionary ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ © Oxford University Press 1986 ISBN 0-19-282670-0 A good quick checker - spellings only, no definitions The Concise Oxford Dictionary ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ New Edition for the 1990's © Oxford University Press 1964, 1976, 1982, 1990 ISBN 0-19-861200-1 The Bible! The Shorter Oxford English Dictionary (two volumes) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ © Oxford University Press 1973 ISBN 0-19-861126-9 The Last Word and final arbiter! Harrap's English Spelling Rules ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ © G. Terry Page 1991 ISBN 0 245-60144-9 Great little pocket-sized authoritative volume. A Word about Americanisms ~~~~~~~~~~~~~~~~~~~~~~~~~ There is some confusion and contention over the use of the suffix `ise' or `ize'. Some maintain that the former is the only correct English usage and that the latter is an Americanisation. In fact both are quite legal in UK English; here is an extract from the Concise Oxford: The form -ize has been in use in English since the 16th c.; it is widely used in American English, but is not an American- ism. The alternative -ise (reflecting a French influence) is in common use, esp. in British English, and is obligatory in certain cases...... Because of this the supplied dictionary (MAIN044.DCT) uses the -ise form exclusively. If you wish to change to the -ize form, use the DED programme only; as some words will not generate correctly (eg. Realize will not make Realism) if you use a global replacement option in a word-processor. ************************************************************ DED & the Dictionary Tools ~~~~~~~~~~~~~~~~~~~~~~~~~~ ************************************************************ DED the main dictionary editor ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Usage: DED [dictionary.file] <flags> <pattern> DED will load any ASCII file but the best format is a file that contains single words; each on a single line, and with no spaces. DED produces effects upon the words by a system of suffix flagging. The suffix flags are the characters of the alphabet and the numerals 1-9. A full description of which flags produce what would take volumes to write as many flags double up to produce different effects depending on the word, an example will serve here and you will soon learn which flags do what as you use the editor. Here are a few examples: Example Word: register Flag G produces Registering Flag S produces Registers Flag 1 produces Registered Example Word: Geography Flag Y produces Geographic Flag H produces Geographical Flag 5 produces Geographically Flag 6 produces Geographer Example Word: Interregnum Flag S produces Interregna Flags in the dictionary are separated from the root word by a forward slash `/'. You can see the large saving in file size, apart from the limited amount of typing that needs to be done. DED Flags ~~~~~~~~~ a for ADD This concatenates a full list of all 32 suffix flags to every word in the dictionary file. BEWARE! This can (correction - this WILL) suddenly make small files into very large ones at a stroke. If your memory or disk space is limited I suggest you work on small subsets of your word list at a time. Several tools are provided to help you do this. n for NEXT Goes directly to the first line in the file that has a full list of suffix flags. s for SEARCH Searches for the first occurrence of the pattern set in <pattern>. DED uses the UNIX style wildcards in it's pattern matching, that is '*' is equivalent to `#?' and matches any number of characters, `?' as usual matches a single character. The suffix flags are effectively ignored in searching, So: DED MAIN044.DCT s format goes straight to the word `format' in the file MAIN044.DCT DED NEW.WORDS s *ise/ goes straight to the first word ending in `ise' in the file NEW.WORDS. Note the use of the divider '/' to denote the actual end of a word. After the first match has been found pressing RETURN will search out the next match. DED NEW.WORDS s *s??le/ will find possible, missile, fusable etc. DED NEW.WORDS s */*Y finds all words with the `Y' flag set. Function Keys ~~~~~~~~~~~~~ F1 Save and Exit. F3 Abandon & Exit. F5 Delete whole line - CAUTION! Once deleted the line is not recoverable without aborting (F3). F6 Deletes all suffixes flags but leaves the '/' divider. F7 Save & Continue. F8 Go to he first line of the file. F9 Adds a full list of suffix flags regardless of what flags have already been set. (Tip: F9 followed by ESC for a quick look see). F10 Sets a new search pattern from anywhere in the file. ESC Restores a partially edited line (not a deleted line). Prefix Cutting ~~~~~~~~~~~~~~ As mentioned earlier DSpell uses a system of prefix cutting when checking words, this means words such as `fix' will also pass `prefix', `ground' will pass `underground', `head' will pass `overhead' etc. A small list of the most common cuttable prefixes is shown on the right side of the display when you click `Prefixes'; clicking again restores the gadgets. Refer to earlier in this document for the full list. Although DSpell passes plurals and `ly' forms `on the fly' you should still generate them in DED for the benefit of the other programmes. Double Words ~~~~~~~~~~~~ Don't worry if a word generates more than once in the dictionary, all doubles are checked for and ignored by the data builder DASH described next. Quitting DED ~~~~~~~~~~~~ The F3 (abandon) and F1 (save & exit) keys are not usable while DED is searching or loading. The close window gadget, although hidden is actually there and will abort without saving at any time. ************************************************************ DASH & DASH3 The Data Builders ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Usage: DASH [dictionary.file] Usage: DASH3 [dictionary.file] DASH & DASH2 process the main dictionary file and produce a file called DSPELL.DATA. This is the dictionary fully expanded, sorted and tagged in the manner required for DSpell to read. DSpell will not run with any other file so do not rename it. DASH3 is the latest implementation and writes only those words to disk that DSpell will require. DASH is included for anyone with RAM enough and patience enough to load the full expanded dictionary. Both DASH & DASH3 can take several minutes to process a file so patience is required. Both programmes report the number of words written to disk. ************************************************************* The Dictionary Tools ~~~~~~~~~~~~~~~~~~~~ Five tools are provided to help in the maintenance of the dictionary and in processing large files on low memory, they are: DCount ~~~~~~ Usage: DCount [dictionary.file] A quick programme that counts all root words and flags in the dictionary to give you a fair idea of overall size; DCount does not check for doubles. DChop ~~~~~ Usage: DChop [in-file] [out-file] [number of lines] Suppose you have a file called NEW.WORDS containing 7,000 new additions to your dictionary (quite possible!), DED would take a long time to load such a file and if the `a' option is used to add all the flags, its size will be enormous . This programme will `chop up' the file into more manageable sections. Dchop NEW.WORDS LIST 500 produces 14 files; LIST1, LIST2, LIST3 etc. of 500 words each, much easier to work with. DExpand ~~~~~~~ Usage: DExpand [dictionary.file] [out-file] [pattern] This will expand all the words in the dictionary that match the pattern. All flags are expanded so do not include them in the pattern. FlagExp ~~~~~~~ Usage: FlagExp [dictionary.file] [out-file] [pattern] [flag] Expands words from the dictionary that match the pattern but will only expand the SINGLE flag in the argument, for example: FlagExp MAIN044.DCT TEMP *ise/ G will expand enterprising, realising etc. to the file TEMP. Note the space between the `/' and the `G'. DoubCut ~~~~~~~ Usage: DoubCut [anyfile] Cuts occurrences of double words from any SORTED file. In DSpell dictionary files it will cut doubles regardless of differences in the suffix flags (works on root words only). NOTE: this will only work correctly on sorted files. All doubles that are cut are saved to a file called DOUBLES.LST. RootStrip ~~~~~~~~~ Usage: RootStrip [dictionary.file] [out-file] Cuts all the suffix flags and dividers out of a dictionary file and leaves you with a clean list of root words in the out-file. The AmigaDOS commands JOIN and SORT are also required for a typical dictionary editing session, here is a blow-by-blow account of such a session: After DSpell has alerted you to the fact that you have 150 words ready to process- 1 Rename DSPELL.UDCT as NEW.WORDS (or anything). 2 Use ED to create a new, empty user dictionary ED DSPELL.UDCT ED will produce a window with the legend `Creating New File' in it, press ESCAPE `x' RETURN in that order to save the new, empty, user dictionary. 3 Run DED with the `a' option on the file NEW.WORDS 4 Spend the next year processing the 30,000 words you've always meant to see to, but have left to accumulate. 5 Manually back-up your current dictionary or copy it to a spare floppy disk (after all accidents can happen). 6 Use AmigaDOS to join the dictionaries together JOIN MAIN044.DCT NEW.WORDS as TEMP 7 Delete MAIN044.DCT (hence the back-up). 8 Use AmigaDOS to sort the new file SORT from TEMP to MAIN044.DCT This can take a few minutes (the ARP sort and join commands are faster and more efficient than AmigaDOS, they are public domain - so check them out!). 9 DoubCut MAIN044.DCT, print or use MORE to examine the contents of the file DOUBLES.LST that this produces. 10 Use DED in search mode on MAIN044.DCT to check the cut doubles against the dictionary for flag settings. 11 Dash3 MAIN044.DCT and away you go! ************************************************************* The File Tools ~~~~~~~~~~~~~~ Three tools are provided that work on files Report ~~~~~~ Syntax: Report filename Runs through a text file looking for non standard characters, escape sequences, binary characters etc. and reports if any of these are present reports the following: File contains Tabs! File contains Non-Standard Line Ends! File contains Hyphenated Words! File contains Binary Characters! File contains Escape Sequences! File contains WordStar Line Ends! Or File Free from Special Characters! Strip ~~~~~ Syntax: Strip [text-file] (-s -dw -jhw -cur) This little prog will process a file ready for DSpell, stripping out unwanted code, expanding tabs to spaces and modifying line endings. It will also produce other effects depending on the flag settings. You should run the programme with the -s option to check that it is actually doing what you want it to. On a large file, once your satisfied abort with Control-C and re-run without the -s flag, this is MUCH FASTER. Flags: -s Strip usually runs with no screen output, however this flag will write everything thats going on to the console. -dw Strips all white space at the beginning of lines. -jhw Joins hyphenated words (best used in combination with the -dw flag). -cur Cuts all unnecessary returns (allows three max) Note that the flags may produce different effects if used in combination, feel free to experiment as Strip does not touch your original file but creates a new file with the extension STP. ************************************************************ I hope you enjoy DSpell and find it useful. Please address any suggestions for improvements or bug fixes to: Rod Young 24 Furzedown Crescent West Leigh Havant Hants PO9 5SD All the programmes were written and compiled using SAS/C version 5.10 (formerly Lattice) on an Amiga 2000 with 3 Meg of precious Memory. I am currently using a dictionary of 56,500+ words of which DASH3 writes only 39,000 to disk. Cop Out! ~~~~~~~~ I've tried to make these programmes as fool-proof as possible, every care has been taken to ensure data accuracy, remember always work with, and on copies. Should DSpell's little cursor, in its manic haste, decide to break out of the window boundaries and crash through the side of your new 1084S monitor, career across your desk top, upsetting that cold cup of coffee into your very expensive laser printer causing all the fuses in your house/office to suddenly overload sending massive voltages through the mains, and your A3000 blows up spectacularly causing your hard disk to go spinning through the plate glass and knock out a passing policeman - then I am not to blame!.