Simple Word Processing With Enscript

The Software

  $EDITOR
  enscript
  a postscript viewer
  pr (only to help with tabular data)

  A list of the software sources and example files can be
  found at the end of this page.


Introduction

  Enscript will convert plain text files into PS
  (PostScript) or other formats, such as  HTML and RTF, and
  can also send files directly to a printer.  It is much
  more advanced than plain lpr or lp, and also supports
  special filters for highlighting source code.

  It is useful for printing letters, or technical files, and
  supports using mixed fonts, such as italic and bold fonts
  for emphasis.

  You may find it best to use real tabs in documents, set to
  a small tabstop, rather than spaces.  This makes it a bit
  easier to align things like addresses to the right margin
  without them wrapping off the edge to the next line.  In
  vim you would need to `set noexpandtab'.

  Note: I've cut some long lines with a `\' for readability.

Method

 The steps in enscript printing:

  1. Edit a plain text file
  2. Use enscript with `-o file.ps' to format and output text to PS
  3. Preview output in a PostScript viewer
  4. Repeat 1-3 until satisfied with the output
  5. Repeat the enscript command omitting the `-o file.ps'

  The last command will send the PS output to the printer.
  The output can rival that of any GUI word processor.  The
  source documents are plain text, so they will be much
  easier to use normal file operations like `grep' and
  `find' etc. on.


Print Previewing

  Any PostScript viewer will do.

  Viewers for the framebuffer console:

    fimgs/fim
    fbgs/fbi

  For X Windows:

    gv
    evince
    (probably others)

  Fim is `fbi improved'.  The strange diagonal scrolling
  that sometimes occurs in fbgs/fbi has been fixed.  It also
  supports scripting and a lot of customisation.  When
  viewing a file with fimgs, it is converted to png images
  in a temporary directory using `ghostscript', and the
  images are viewed in `fim'.  The `fimgs' application
  itself is a shell script.


Options

  I won't go into all of enscript's options here - the
  manual is very detailed.  But here are a few of the most
  common and useful options for word processing.


 General Options

    -e[char], --escapes[=char]
      Allow escape codes.  This allows us to set italic or
      bold fonts, among other things.  Vim refuses to add a
      ^@ via a key map, but ^_ (-e\037) works fine.
      Default: ^@

    -f font, --font=name
      Main body font name and size.
      Default: Courier10 or Courier7

    -i num, --indent=num
      Indent each line by num units of c (centimeters),
      i (inches), l (characters), or p (PS points.)
      Default: l

    -M name, --media=name
      Media - e.g. A4, letter etc.
      Default: letter

    --margins=left:right:top:bottom
      Margins in PS points (usually 72 DPI.)
      Default: none.

    -o file, -p file, --output=file
      Output to file.
      Default: send to printer

    -P name, --printer=name
      Printer name to use.  When adding printers in CUPS
      it's helpful to use short, simple names to avoid too
      much typing here - e.g. "Photosmart" or "BJ200".
      Default: default printer.

    -T num, --tabsize=num
      Tabsize to use.
      Default: 8

    -w lang, --language=lang
      Output language:

      PostScript, HTML, OverStrike (for line printers and
      `less'), RTF, ANSI (ANSI terminal control codes.)
      Default: PostScript

    --word-wrap

    -X enc, --encoding=enc
      Input encoding.  The manual has a long list.
      Default: latin1/88591.

    -B, --no-header
      Do not print page headers.


  Options that may be useful for technical documents or source code:

    -G, --fancy-header[=name]
      Print a fancy header (very customisable)

    -F name, --header-font=name
      Font for headers.

    -C[start_line], --line-numbers[=start_line]

    -E[lang], --highlight[=lang]
      Source code highlighter.

      List of highlight filters included with enscript
      1.6.6:

      ada asm awk bash c changelog cpp csh delphi diff diffs
      diffu elisp f90 fortran fortran_pp haskell html idl
      inf java javascript ksh m4 mail makefile matlab nroff
      objc outline pascal perl postscript pyrex python rfc
      scheme sh skill sql states synopsys tcl tcsh tex vba
      verilog vhdl vrml wmlscript zsh

    -b header_string, --header=header_string
      Use text string as header.  Can use %Format escapes
      described in the manual.  E.G.:

      enscript --header='$n %W Page $% of $=' *.c

    --color[=bool]

    --footer=footer_string
      Similar to --header.

  You can also add the most used settings to ~/.enscriptrc.


Aliasing Commands

  That's a fair number of options to remember, but we can
  use aliases to save typing and to avoid making typos.

  Proportional font, suitable for most correspondence:

  alias ep='enscript -e\037 -BM A4 --word-wrap -fTimes-Roman12 \
    --margins=50:50:50:50 -T2'

  Fixed-width font with a fancy header, suitable for source:

  alias ep-mono='enscript -e\037 -GM A4 --word-wrap -fCourier12 \
    --margins=50:50:50:50 -FCourier12 -T2'

  Now all we need to do is `ep letter1.txt -o out.ps', but
  see the vim key mappings below.  Extra options can be added
  to aliases on-the-fly.

Encodings

  Enscript doesn't support UTF-8, so one of the supported
  ISO-8859 or other encodings will need to be used.

  This means we need to find a workaround for some
  characters like UTF-8 bullets, and e.g. use something like
  a `middle dot' instead.

  Supported encodings from the man page:

  88591, latin1 ISO-8859-1 (ISO Latin1) (enscript's default
  encoding).
  88592, latin2 ISO-8859-2 (ISO Latin2)
  88593, latin3 ISO-8859-3 (ISO Latin3)
  88594, latin4 ISO-8859-4 (ISO Latin4)
  88595, cyrillic ISO-8859-5 (ISO Cyrillic)
  88597, greek ISO-8859-7 (ISO Greek)
  88599, latin5 ISO-8859-9 (ISO Latin5)
  885910, latin6 ISO-8859-10 (ISO Latin6)
  ascii   7-bit ascii
  asciifise, asciifi, asciise 7-bit ascii with some
    scandinavian (Finland, Sweden) extensions
  asciidkno, asciidk, asciino 7-bit ascii with some
    scandinavian (Denmark, Norway) extensions
  ibmpc, pc, dos IBM PC charset
  mac     Mac charset
  vms     VMS multinational charset
  hp8     HP Roman-8 charset
  koi8    Adobe Standard Cyrillic Font KOI8 charset
  ps, PS  PostScript font's default encoding
  pslatin1, ISOLatin1Encoding
    PostScript interpreter's `ISOLatin1Encoding'

  There is an application called `paps' that does a similar
  task to enscript and supports UTF-8, however it doesn't
  seem to support changing fonts within documents, nor any
  of the advanced features of enscript.  It would be useful
  to create similar aliases or editor binds as I use here
  with it, if a need for UTF-8 support arises.


Editing

  The key is a good editor.  Make sure that it can edit
  plain text and has the ability to enter control/escape
  codes.  I use vim pretty much all the time so I've
  included some of the settings and binds I use.  I have
  briefly used emacs, but wouldn't know how to do the same
  in that.

  We can get a good idea of the end result when previewing
  by limiting the columns in vim to something roughly
  corresponding to the hard copy.  You need to adjust by
  trial and error until it looks right.  It helps to not
  break lines, so we use `textwidth=0' and `linebreak',
  which displays long lines as wrapped, but doesn't insert
  newline characters:

  " Stop lines breaking and limit the terminal width to 68
  " characters.
  set columns=68
  set textwidth=0
  set linebreak
  set nolist

  " No line numbering and small tab stops
  set nonumber
  set tabstop=2
  set softtabstop=2
  set shiftwidth=2

  " Spell-checking:
  hi SpellBad ctermfg=red cterm=Bold
  setlocal spell


Using Emphasis

  Escape codes in enscript begin with a `NUL', which are a
  `^@' (literal Ctrl-@) in vim, or `\0' in C, a zero byte.
  In vim this is done by pressing Ctrl-v in insert mode,
  then pressing Ctrl-@.  However, it seems that these can't
  be entered automatically, e.g. via a keymap, as they get
  transposed to newlines, but we can change `^@' to
  something else and use enscript's `-e' flag to use that
  instead.

  The man page has a list of escape format codes near the
  end, (see `man enscript | less -p "^SPECIAL ESCAPES"'),
  and can also inject PostScript code.  But we are most
  interested in the code to change the font.  The syntax is
  `^@font{NEWFONTNAME}'.

  To use an italic or bold font we simply have to switch to
  that font before the word(s) we want emphasised, and then
  switch back to the default font afterwards.  E.G.:

  This is ^@font{Times-Italic12}italic^@font{default} text.

  (NOTE: These commands and binds may not display correctly
  in my gopherhole yet.  Compare with the real text.vim
  linked below.)

  We can make some helper mappings in vim to save typing:

  " normal/command mode font maps
  nmap <buffer> ,ii i^_font{Times-Italic12}<Esc>ea^_font{default}<Esc>
  nmap <buffer> ,bb i^_font{Times-Bold12}<Esc>ea^_font{default}<Esc>
  nmap <buffer> ,bi i^_font{Times-BoldItalic12}<Esc>ea^_font{default}<Esc>
  nmap <buffer> ,cc i^_font{Courier11}<Esc>ea^_font{default}<Esc>

  " visual select mode font maps
  vmap <buffer> ,ii <Esc>`>a^_font{default}<Esc>`<i^_font{Times-Italic12}<Esc>
  vmap <buffer> ,bb <Esc>`>a^_font{default}<Esc>`<i^_font{Times-Bold12}<Esc>
  vmap <buffer> ,bi <Esc>`>a^_font{default}<Esc>`<i^_font{Times-BoldItalic12}<Esc>
  vmap <buffer> ,cc <Esc>`>a^_font{default}<Esc>`<i^_font{Courier11}<Esc>

  For the nmap maps, place the cursor on the first letter of
  a word, press comma followed by `bb' for bold, or `ii' for
  italic etc.  The vmap visual mode maps work by selecting
  text in visual mode and then using the same comma key
  combinations.

  Some maps not really connected with enscript, but might be
  useful, that quote words or selected text with backticks
  and single quotes:

  nmap <buffer> ,`` i`<Esc>ea'<Esc>
  vmap <buffer> ,`` <Esc>`>a'<Esc>`<i`<Esc>

  " Insert dates
  nmap ,dd "=strftime("%A, %-d %B %Y")<CR>p
  " Output: Saturday, 4 January 2020

  nmap ,du "=strftime("%a %_d %b %T UTC %Y")<CR>p
  " Output: Sat  4 Jan 17:28:31 UTC 2020

  nmap ,dt "=strftime("%a %-d %b %T %Z %Y")<CR>p
  " Output: Sat 4 Jan 17:28:31 GMT 2020

  This will insert a tab and a round bullet at the start of
  the current line:

  nmap <buffer> ,tt I<Tab>·<Tab><Esc>

  Since enscript doesn't support the UTF-8 bullet symbols,
  this is actually a `middle dot' from ISO-8559-1.  It may
  look like a small dot in the terminal, but it's much
  larger when rendered in PostScript or printed out:

  char dec col/row oct hex  description
  [·]  183  11/07  267  B7  MIDDLE DOT

More here:

http://www.columbia.edu/kermit/latin1.html and
http://www.columbia.edu/kermit/latin9.html
and a good reference and list:
https://en.wikipedia.org/wiki/ISO/IEC_8859

  Map <leader> + <Tab> to `make' enscript:

  nmap <buffer> <leader><tab> :silent make!<CR>

  Either `makeprg' or `&mp' need to be set for `make' to
  work:

  let &mp = 'enscript -e\037 -MA4 --word-wrap --margins=50:50:50:50 \
  -T4 -o %:r.ps; fimgs -r 240 %:r.ps'

  The above sets the output file name to the same as the
  edited text file name, but changes the extension to .ps,
  and then loads the file in fimgs.  The -r 240 flag is a
  resolution.


Vim settings in an Autogroup

  An example autogroup that sets up a make command and binds for
  italic and bold.  It's easier to add these commands to a
  ~/.vim/after/ftplugin/text.vim though, which I've linked to
  below.

  """""""""""""""""""""""""""""""""""""""""""""""""""""""""""
  augroup text
    au!
    autocmd BufRead *txt
    \ let &mp = 'enscript -e\037 -DCollate -MA4 --word-wrap \
    --margins=50:50:50:50 -T4  -o %:r.ps' |
    \ nnoremap <leader><tab> :silent make!<CR> |
    \ nnoremap ,ii :normal i^_font{Times-Italic12}<Esc>ea^_font{default}<Esc> |
    \ nnoremap ,bb :normal i^_font{Times-Bold12}<Esc>ea^_font{default}<Esc> |
    \ nnoremap ^[[34~ :call Print()<CR>

  " The next lines can change the terminal column width if
  " needed.
  "  autocmd  ShellCmdPost *txt exec "!stty cols 103"
  "  if exists("##ExitPre")
  "    autocmd  ExitPre *txt "!stty cols 114"
  "  endif
  augroup END

  """""""""""""""""""""""""""""""""""""""""""""""""""""""""""


A Template Script

  Templates always makes writing easier, so here's an idea for
  a script that copies a template, adds the date, and then
  starts $EDITOR on it:

  ############################################################
  #!/bin/sh

  set -e

  TEMP="$HOME/letters/templates/letter-template.txt"
  LETTERDATE=$(date +"%A, %d %B %Y")
  DATE=$(date +%Y-%m-%d-%S)
  FNAME="$1-$DATE.txt"

  cp -iv "$TEMP" "$FNAME"
  sed -i "s#@DATE@#$LETTERDATE#g" "$FNAME"
  $EDITOR "$FNAME"
  ############################################################

  This takes the first argument to use as a prefix for the
  file name and adds the date: prefix-YYYY-MM-DD-SS.txt

  Fill the template with your name and address and
  put @DATE@ where the date would normally go.


Headers

  There are a few example header files that come with
  enscript.  They are written in the PostScript language.
  Here is a simple example:

 % -- code follows this line --
 %HeaderHeight: 50
 %Format: pagenumstr Re: Appointment on 19/12/2019, J Smith, ($%/$=)

 /do_header { % print the header
   gsave
     d_header_x d_header_y translate
     0 20 moveto
     0.2 setgray
     pagenumstr show
   grestore
 } def

  To try it, copy the file to ~/.enscript/, use the -G
  flag (instead of -B), or --fancy-header=standard-letter,
  or add the following to ~/.enscriptrc:

  DefaultFancyHeader: standard-letter

  This will print:

  Re: Appointment on 19/12/2019, J Smith, (1/4)

  At the top of each page, where the numbers in parentheses
  are page number/total pages, although the --header flag
  can do simple headers like this inline.  There are example
  header files in /usr/share/enscript/ that show how to go
  further, like add horizontal rules etc.


Tabular Data With PR

  pr can paginate and columnise text, which can then be
  piped to enscript, lp/lpr, or redirected to a new file.

  A Few of The Options

  -COLUMNS
    Number of columns

  -a, --across
    Print across, then down

  -l lines, --length=lines
    Page length
    Default: 66

  -o margin, --indent=MARGIN
    Number of characters to use for margin
    Default: 0

  -w width, --width=WIDTH
    Width of page in characters
    Default: 72

  Data is entered in a single column, which pr will
  columnise.  E.G.:

  % cat cols.txt:

  one
  two
  three
  four
  five
  six

  % pr -a -3 -l 15 -w 35 cols.txt

  Output:

  2019-03-03 12:47  cols.txt   Page 1


  one         two         three
  four        five        six


Software List:

You may have some or all of these available in your distro's
software repo.

enscript homepage:
https://www.gnu.org/software/enscript
fim (fbi improved) homepage:
http://www.nongnu.org/fbi-improved/
fbida (the original fbi) homepage:
https://www.kraxel.org/blog/linux/fbida/
ghostscript homepage:
https://ghostscript.com/
gv homepages?:
http://pages.cs.wisc.edu/~ghost/gv/gv.htm
https://sourceforge.net/projects/gnu-gv/
paps homepage:
https://github.com/dov/paps
pr is part of gnucore Utils:
https://www.gnu.org/software/coreutils/manual/html_node/pr-invocation.html



File List:
NameLast modifiedSize

standard-letter.hdr 2020-01-03 11:45 263
text.vim 2020-01-04 17:53 3.8K

2022-05-27 04:37:42+01:00