            ######   #######   ########  ##        ##
           ########  ########  ########  ##        ##
           ##        ##    ##  ##        ##        ##
           #######   #######   ######    ##        ##
                 ##  ##        ##        ##        ##
                 ##  ##        ##        ##        ##
           ########  ##        ########  ########  ########
            ######   ##        ########  ########  ########

    This program checks a file for words that it does not have  in
a lexicon - that is, words which may be misspelt. The command line
format is

[newfile,errorsfile=]inputfile[,mylex[,syslex]]

where errorsfile defaults to inputfile.ERR . Two lexicons are read
and one is produced. If no mylex is specified then the name LEXICON
is used.  Similarly the default for syslex is LB:[1,1]LEXICON.
A new version of mylex is made. Thus you may keep
locally a lexicon of words you use that are not in the system one.
If  either  the  system  or  the  local  lexicon cannot be read, a
warning message is output;  neither conditions are fatal, however.
The program works in three passes. First it reads in the whole text
and then it checks the words. For every word in neither lexicon, a
message is printed, with the prompt 'Is this a word?'.  The  valid
replies are:
 Y : yes, enter it into my lexicon
 L : yes, enter it into my lexicon but with all letters lowercased
 E : this word is eccentric - do not enter it into the  lexicon  or
     the error listing
 carriage return, or any other single letter response :
     this word is erroneous
 CTRL/Z or W: this gives the response 'erroneous' to every  succeeding
     question automatically.
 word: where the word is longer than one character; this  word  is
     the correct answer, and is placed in the errors file.  If a
     new output file was specified (implicitly or explicitly),   a
     third pass is performed on the document,  correcting these
     spelling errors.
 Q : quit immediately. Delete all new files created by this run.

   Care should be taken with capital letters - words placed in the
lexicon with an initial capital or completely in capitals may only
be  so  spelt;  words  in  the  lexicon in lower case may occur in
capitals in the text, however.
   The file LEXICON.SOR is produced and  used  for  sorting; it is
deleted at the end of a normal run.
   The program LEXMERGE can be found in the SPELL utility directory
and may be used to merge lexicons.  The  program  CHECKLEX  can  be
used to check a lexicon.
   An errorsfile is not produced if there were no errors to report,
or the errors file was suppressed.
   A new output is not produced if there were no corrections to do,
or the new output file was suppressed.

Examples
SPE>MY.TXT

MY.ERR will be produced if there were any errors or corrections to do.
MY.TXT (a new version) will be produced if there were any corrections
  to do.

SPE>,MY=MY.TXT

No new output will be produced.

SPE>MY,=MY

No error file will be produced.
