This is reformat.pl in view mode; [Download] [Up]
#!/usr/bin/perl -w
#
# Written by: Sean Ahern (ahern@llnl.gov)
#
# this is a reimplementation, as a filter, of the vile paragraph reformatter.
#
# USAGE: reformat [-w width] [prefix]
# It reads stdin and outputs a reformatted paragraph on stdout.
#
# The vile macro I use to run this is:
#
# ; Format a paragraph
# 1 store-macro
# set-variable %savcol $curcol
# set-variable %savline $curline
# ~if &equal $char &ascii ">"
# set-variable %pre @"Prefix string: "
# ~else
# set-variable %pre ""
# ~endif
# ~force filter-til forward-paragraph &cat "reformat -w78 " %pre
# ~force %savline goto-line
# ~force %savcol goto-column
# ~endm
# bind-key execute-macro-1 ^P
#
# It does the standard stuff of filtering a paragraph, but it will also prompt
# you for a prefix string if your cursor is on a ">". I probably should put in
# math to calculate the width from the wrapmargin or the fill-col, rather than
# hardcoding "78", but I'll leave that as an exercise for the reader.
#
# Here it is... Enjoy!
require("getopt.pl");
$debug = 0;
print "Debugging messages turned on.\n" if ($debug == 1);
$endings = "'\\\"\\\`\\)\\]}"; # Extra slashes because this string is used in
# regular expressions.
# Get the width. If none specified, default to 72.
$opt_w = 72;
&Getopt("w:");
$save_width = $opt_w;
# If we have an argument, use it, otherwise "".
if (scalar(@ARGV) >= 1)
{
$name = shift(@ARGV);
} else
{
$name = "";
}
# Get the indents from a line. If there is a special character sequence (#,
# >, *, or /*) then return that sequence and the second indent. Otherwise
# just return the first indent and nothing for the characters and second
# indent.
sub get_indent
{
local($line) = @_;
local(@ret);
$ret[0] = "";
$ret[1] = "";
$ret[2] = "";
# if there is a special character at the front of the line...
if ($line =~ m=^\s*([#>\*]+|/\*)=)
{
$line =~ m=^(\s*)([#\*]+|/\*|(\s*>)+)(\s*)=;
$ret[0] = $1;
$ret[1] = $2;
$ret[2] = $4;
} else
{
$line =~ m=^(\s*)=;
$ret[0] = $1;
}
return(@ret);
}
$have_name = 0 if ($name =~ /^\s*$/);
$have_name = 1 if ($name !~ /^\s*$/);
# At the start, we haven't printed anything out.
$length = 0;
$printed_word = 0;
$first_word = 1;
$indent_characters = "";
$C_comment = 0;
$output_line = "";
# Read in all of the input
@lines = <STDIN>;
# Check to see if we already have something of the form "name>" on the first
# line.
if ($lines[0] =~ /^\s*(\w+)>/)
{
$name = $1;
$have_name = 1;
for($i=0;$i<scalar(@lines);$i++)
{
$lines[$i] =~ s/$name>/>/;
}
}
# Process the first line for spaces and indents.
($first_indent[0],$indent_characters,$first_indent[1]) = &get_indent($lines[0]);
print "First indent 0: \"$first_indent[0]\"\n" if ($debug == 1);
print "First indent 1: \"$first_indent[1]\"\n" if ($debug == 1);
if ($indent_characters !~ /^\s*$/)
{
print "We have an indent of \"$indent_characters\".\n" if ($debug == 1);
$have_indent = 1;
} else
{
$have_indent = 0;
}
# If we do have a name set up the indent to be a ">".
if ($have_name == 1)
{
$indent_characters = ">";
$indent_characters = "$name$indent_characters";
$have_indent = 1;
}
# Set up the width for use on the first line. We'll reset it for later
# lines.
$width = $save_width;
# Adjust the width to accomodate the indent and the spaces.
$save_width -= length($indent_characters);
# If we have more than one line...
if (scalar(@lines) > 1)
{
# Process the second line for spaces.
($second_indent[0],$foobar,$second_indent[1]) = &get_indent($lines[1]);
$second_indent[0] .= " " if (length($foobar) == 1);
$second_indent[0] .= " " if (length($foobar) == 2);
$foobar = "";
} else
{
$second_indent[0] = $first_indent[0];
$second_indent[1] = $first_indent[1];
}
# If we have a newline for the second indent, use the first indent.
$second_indent[0] = $first_indent[0] if ($second_indent[0] =~ /\n/);
$second_indent[1] = $first_indent[1] if ($second_indent[1] =~ /\n/);
print "Second indent 0: \"$second_indent[0]\"\n" if ($debug == 1);
print "Second indent 1: \"$second_indent[1]\"\n" if ($debug == 1);
# If we have a newline for the first indent, just use the second indent.
$first_indent[0] = $second_indent[0] if ($first_indent[0] =~ /\n/);
$first_indent[1] = $second_indent[1] if ($first_indent[1] =~ /\n/);
if (length($indent_characters) != 0)
{
# Adjust the first indent if we have less of an indent on the second
# line.
if (length($first_indent[0]) > length($second_indent[0]))
{
$first_indent[1] = substr($first_indent[0],length($second_indent[0]));
$first_indent[0] = $second_indent[0];
}
# Adjust the second indent.
if (length($second_indent[0]) >=
(length($first_indent[0]) +
length($indent_characters) -
length($name)
)
)
{
$test = substr($second_indent[0],
length($first_indent[0])+
length($indent_characters)-
length($name));
if (length($test) != 0)
{
$second_indent[1] = $test;
}
}
$second_indent[0] = $first_indent[0];
}
# Print the indent for the first line. It's different from the rest, most
# likely.
if ($have_indent == 1)
{
$output_line .= "$first_indent[0]$indent_characters$first_indent[1]";
$length = length($first_indent[0])+length($indent_characters)+
length($first_indent[1]);
} else
{
$output_line .= "$first_indent[0]";
$length = length($first_indent[0]);
}
# Here we go
for($count = 0; defined($lines[$count]); $count++)
{
$_ = $lines[$count];
# Get rid of the $indent_characters if they exist at the front of the
# string.
if ($have_indent == 1)
{
if ($have_name == 1)
{
$temp_indent = $indent_characters;
$temp_indent =~ s/^$name//;
s/^\s*($temp_indent ?)*\s*//;
} else
{
if (($C_comment == 0) || (! m|^\s*\*/|))
{
$temp_indent = $indent_characters;
$temp_indent =~ s/\*/\\\*/;
$temp_indent =~ s/^\s*//;
s/^\s*($temp_indent ?)*\s*//;
}
}
}
# Get rid of any excess white space in all cases.
s/^\s*//;
s/\s*$//;
# Split the line into words by whitespace.
for $word (split(/\s+/))
{
# If printing out the current word would push us over the $width, do
# a line break.
if ($length+length($word) > $width)
{
# Print out the line so far, after processing.
$output_line =~ s/\s*$//;
print "$output_line\n";
# Reset the line.
$output_line = "";
$length = 0;
$first_word = 1;
$width = $save_width;
# If we don't have an indent, preserve the second indent level.
$output_line .= "$second_indent[0]$indent_characters$second_indent[1]";
$length += length($second_indent[0])+
length($indent_characters)+
length($second_indent[1]);
}
# It's safe to print out the next word and adjust the $length.
if ($first_word != 1)
{
$output_line .= " ";
$length++;
}
$output_line .= "$word ";
$printed_word = 1;
$length += length($word)+1;
# If the word ends in a period, question mark, or exclamation point,
# we might be ending a sentence. Check for common abbreviations to
# see. If we *are* ending a sentence, we need to add another space
# to the output. We also might be right after a colon.
if ($word =~ /[\.\?!:][$endings]*$/)
{
# Check for standard abbs.
if ( ($word !~ /\bMr\.[$endings]*$/) &&
($word !~ /\bMrs\.[$endings]*$/) &&
($word !~ /\bMs\.[$endings]*$/) &&
($word !~ /\bCo\.[$endings]*$/) &&
($word !~ /\bLtd\.[$endings]*$/) &&
($word !~ /\bltd\.[$endings]*$/) &&
($word !~ /\bp\.m\.[$endings]*$/) &&
($word !~ /\bP\.M\.[$endings]*$/) &&
($word !~ /\ba\.m\.[$endings]*$/) &&
($word !~ /\bA\.M\.[$endings]*$/) &&
($word !~ /\bi\.e\.[$endings]*$/)
)
{
$output_line .= " ";
$length++;
}
}
}
# If we have a "/*" indent, change it to be " *".
if ($indent_characters eq "/\*")
{
$C_comment = 1;
$indent_characters = " *";
# If we don't have a second indent 1, make one.
if (length($second_indent[1]) == 0)
{
$second_indent[1] = " ";
}
}
# If we got to the end of the first input line without printing anything,
# do a line break.
if ($printed_word == 0)
{
# Print out the line so far, after processing.
$output_line =~ s/\s*$//;
print "$output_line\n";
# Reset the line.
$output_line = "";
$length = 0;
$width = $save_width;
# If we don't have an indent, preserve the second indent level.
$output_line .= "$second_indent[0]$indent_characters$second_indent[1]";
$length += length($second_indent[0])+
length($indent_characters)+
length($second_indent[1]);
}
}
# Print out the last line, after processing.
$output_line =~ s/\s*$//;
print "$output_line\n";
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Netfuture.ch.