System Booster/System Booster.iso/Archives/Converters/TextPort.lha/TextPort/README

Text File | 1993-07-16 | 8.5 KB | 196 lines

TEXT FILE PORTABILITY AIDS    6/6/92 public domain
--------------------------    Alex Matulich, Unicorn Research Corp
                              4621 N. Landmark Drive
                              Orlando, FL 32817-1235
                              [phone]
                              [e-mail]

This archive contains three programs plus C source code for each:
   AddCR    -- converts unix or AmigaDOS text files to MS-DOS
   StripCR  -- converts MS-DOS text files to unix or AmigaDOS
   StripHR  -- selectively strips hard returns from a text file.
   Reformat -- reformats a text file using a different line length

Typing the name of each command on the command line will display its
function and syntax.

These 4 utilities are helpful when transferring text files between
dissimilar operating systems or when importing text files into word
processing (WP) or desktop-publishing (DTP) software, or uploading
text into BBS message editors.

IMPORTANT:  Normally, you specify an input file to process, and an output
file to write.  However, all these utilities are capable of processing a
file "in place" if you do not specify an output file.  This practice is
not recommended because (1) it is MUCH slower, and (2) your original file
may be destroyed if something goes wrong.  Be warned!


Explanation (i.e. why bother?)
------------------------------
   The end-of-line (EOL) code used in MS-DOS text files is different
from that used in other operating systems such as unix or AmigaDOS.
MS-DOS uses two characters at the end of each text line: a carriage
return (CR) plus a linefeed (LF).  Unix uses only one character as the
EOL code, LF.

   The programs StripCR and AddCR will convert EOL codes from CR+LF to
LF and back again.  StripCR strips out the CR characters from a text
file, leaving only LF characters at the end of each line.  AddCR looks
for LF characters in a file, and inserts a CR in front of those that are
not already preceded by CR, thus ensuring that all LF's will be
converted to CR+LF.

   Now that you have your text file all nice and compatible with your
operating system, you probably want to import it into WP or DTP
software, right?  Before doing so, run the text file through the StripHR
utility.  Why, you ask?  Well, all WP or DTP packages have the ability
to import text.  Invariably, however, the imported text will not fit
properly within the margins you wish to use, which means that you will
have to reformat everything in your WP program manually.  Many software
packages give you the option to strip out the hard returns (EOL codes)
in your text file as it is being imported, so that the software can
format it to your margins by itself.  Trouble is, often ALL, or too
many, of the hard returns are stripped out, and this messes up paragraph
breaks and section titles, so you still have to go through your text
and reformat everything manually.

   StripHR is more selective about removing hard returns.  It removes
all hard returns except the ones that precede a paragraph indent or a
blank line.  This means that the paragraph formatting in your text file
will be preserved.  A normal paragraph which is flush to the left will
have all hard returns removed within it, but there will still be a hard
return at the end of the paragraph.  Also, indented blocks of text
(where each line in the block starts with a space or a tab) will be left
alone, as will lines that are intentionally blank.  After running your
text file through StripHR, just import it straight into your WP or DTP
software, without stripping any more hard returns.

   Reformat will take a text file and reformat it to fit within a
specified line length.  This is useful if you have to upload a text file
into an online editor that has its own idea about where the wrap margin
is located.  Two situations can occur:  The lines in your text file
might be a little too long, and the editor will split each line, usually
leaving single words on lines by themselves.  Or, if you have a
"continuous" text file (say, the output of StripHR), you might be using
an online editor that doesn't insert carriage returns into the text as
you upload it, resulting in long one-line paragraphs that wrap around
and around the screen.  Running a text file through Reformat first
should take care of most problems, though some minor post-editing may be
necessary.

   StripHR and Reformat leave indented text alone.  So if you have a
textfile with a "left margin" throughout, these utilities won't have any
effect.  If I feel like it, I'll modify these utilities to handle this
special case, but it's not likely to be soon, because I'm satisfied.


Usage
-----

Each of these utilities will give you an explanation on its use when you
enter its name by itself on the command line.  More detailed information
is below.


----- AddCR -----

To add CR characters to a text file, type:

    AddCR file             (operates on the file "in place" - avoid this!)
 or AddCR infile outfile   (infile left alone, new version is outfile)
 Example:  addcr unixfile.txt msdos.txt

If there are existing CR+LF codes, AddCR will leave them alone, instead
of converting them to CR+CR+LF codes, which are meaningless.


----- StripCR -----

To strip CR characters from a text file, type:

    StripCR infile         (operates in place)
 or StripCR infile outfile (infile left alone, stripped version is
                            saved in outfile)
 Example:  stripcr msdos.txt unixfile.txt

When stripping is done in place, the resulting file size is equal to or
shorter than the original (because characters are being removed).  The
new replaced file will have its end flagged with a ctrl-Z character,
followed by some useless text from the old version of the file.  Most
MS-DOS software will ignore anything after the ctrl-Z.  You can delete
anything after the ctrl-Z with a text editor if you wish.


----- StripHR -----

To strip out hard returns from a file, leaving paragraphs intact, type:

    StripHR infile           (operates in place)
 or StripHR infile outfile   (creates a new modified file)
 Example:  striphr article.txt article.wp

Here's the decision logic that determines which hard returns will be
kept and which will be thrown out:  A hard return is discarded unless
(1) it is followed by "white space" (a space, tab, or blank line), or
(2) it is preceded by another hard return.
These rules ensure that paragraphs and blank lines will be kept intact.


----- Reformat -----

To reformat a text file to new line lengths, the command syntax is:

    Reformat [-r] [-lN] [-tM] infile [outfile]
 Example:  Reformat -l70 -t4 article.wp article.txt

Parameters for Reformat:
   -r  = Re-wrap lines that happen to be shorter than the specified
         length.  The default is to leave shorter lines alone.  Either
         setting may be a compromise, depending on what you want.
   -lN = Set maximum line length to N characters.  Default is 72.
   -tM = Assume that tabs are M characters wide.  Default assumes 8.

Reformat can also be used as a functional equivalent to StripHR.  You
could delete StripHR altogether, and set up a command alias to use
Reformat, something like:

   alias StripHR = Reformat -r -l20000

This will work exactly like StripHR, assuming each paragraphs is less
than 20,000 characters long!  (StripHR itself has no limitation, and
and it's a bit faster as well.)

-----

>>> IMPORTANT! >>>  Be sure your input text file is compatible with
your operating system, or the output of Reformat/StripHR could be trash!
If you're running MS-DOS, make sure that the EOL codes are the CR+LF
type. To be safe, specify both an input and output file when using
StripHR or Reformat; this will preserve your original file in case
something goes awry, or if you just don't like the result.


Changes from the previous release
---------------------------------

All utilities are now capable of operating on files in place, instead of
just the stripping utilities.  This practice is not recommended, however,
because it is slower and you risk destroying your original file if
something goes wrong.  But if you have little disk space to spare, this
method might be the best way.

In-place operations used to be excruciatingly slow.  Now the in-place
mode is significantly faster.  Instead of reading & writing one character
at a time, a 16K buffer is used to hold the output so that reads will be
smoother.

The Reformat utility was added to the collection.

Online help is now more informative.

StripHR used to lose the last character in a file.  No more.

Endless-loop condition on buffer-emptying process in AddCR fixed.