Transcript of Bernstein hearing is now available

We received the transcript of October 20's oral hearing from the court
stenographer.  It's up on the EFF Web site at:

http://www.eff.org/pub/Legal/Cases/Bernstein_v_DoS/Legal/951020_hearing.transcript

This is from Dan Bernstein's case, which is trying to get the crypto
export laws invalidated as unconstitutional.  This particular hearing
is about the government's motion to throw the case out of court
because the courts don't have jurisdiction to decide the issue.  Some
of it is deep legalese, and some of it is quite readable.

There's still no news from the judge on what her decision is, or when
she will decide.

	John Gilmore

> John Gilmore <gnu@toad.com> writes:
> We received the transcript of October 20's oral hearing from the court
> stenographer.  It's up on the EFF Web site at:

> http://www.eff.org/pub/Legal/Cases/Bernstein_v_DoS/Legal/951020_hearing.transcript

I find it hard to read things in all upper case.  Here's a slowish 95%
hack to make it more legible.  I imagine there's a 5-line way to do it
twenty times faster for 99% success, but what the heck...

	Jim Gillogly
	Highday, 12 Blotmath S.R. 1995, 19:47

-----------------------------------------------------------------------
#!/usr/bin/perl
# delegal: quick-n-dirty case conversion for Bernstein transcript
# 2 Nov 95, Gillogly

@propers =
( "daniel", "dan", "marilyn", "hall", "patel", "california", "bernstein",
  "united", "department", "state", "cohn", "coppolino", "national",
  "security", "agency", "steefel", "levitt", "weiss", "court", "i",
  "mc", "glashan", "sarrall", "lee", "tien", "ed", "ross", "susan", "arnold",
  "justice", "anthony", "mandel", "bazarov", "appeals", "helme", "webster",
  "states", "dorfmont", "constitution", "constitutional", "doe", "schechter",
  "snuffle", "june", "lowell", "cj", "edler", "olc", "mr", "ninth",
  "circuit", "judge", "ritchie", "english", "dr", "freedman", "o", "brien",
  "pentagon", "cubby", "compuserve", "golden", "gate", "san", "francisco",

);

$INDENT_UPALL = 13;     # If indented deeper than this, upcase each word
$INDENT_UPCOLON = 6;    # If indented with a colon and these spaces, upcase
$INDENT_PARA = 10;      # If indented this deep, upcase first word
$INDENT_SENT = 2;       # Pick up sentence starts

while ($proper = pop(@propers))
{
	($first, $rest) = ($proper =~ /^(.)(.*)$/);
	$first =~ tr/a-z/A-Z/;
	$caps{$proper} = $first . $rest;
}


while (<>)
{
	tr/A-Z/a-z/;            # Downcase everything
	s/u\.s\./U.S./g;        # special case
	s/d\.c\./D.C./g;        # special case
	s/([^a-z])nsa([^a-z])/$1NSA$2/g;  # Another one
	s/([^a-z])itar([^a-z])/$1ITAR$2/g;  # Another one

	# Upcase known proper names
	while (($proper, $cap) = each(%caps))
	{
		($first, $rest) = ($proper =~ /^(.)(.*)$/);
		s/([^a-z])$proper([^a-z])/$1$cap$2/g;
	}
	# If it's indented deeply, upcase each word
	if (/ {$INDENT_UPALL}/ || /: {$INDENT_UPCOLON}/)
	{
		while (($low) = /[^a-zA-Z]([a-z])/)
		{
			$low =~ tr/a-z/A-Z/;
			s/([^a-zA-Z])[a-z]/$1$low/;
		}
	}
	# Upcase middle initials
	while (($init) = / ([a-z])\./)
	{
		$init =~ tr/a-z/A-Z/;
		s/ [a-z]\./ $init\./;
	}
	# Upcase paragraphs
	if (($init) = / {$INDENT_PARA}([a-z])/)
	{
		$init =~ tr/a-z/A-Z/;
		s/( {$INDENT_PARA})[a-z]/$1$init/;
	}
	# Sentences
	($num, $_) = /^([ \d]*)([^ \d].*)$/; # Simplify
	while (($init) = /[^ ] {$INDENT_SENT}([a-z])/)
	{
		$init =~ tr/a-z/A-Z/;
		s/( {$INDENT_SENT})[a-z]/$1$init/;
	}
	$_ = $num . $_ . "\n";


	print $_;
}
-----------------------------------------------------------------------

In article <199511021951.LAA03152@mycroft.rand.org>,
Jim Gillogly <jim@acm.org> wrote:
> 	s/([^a-z])nsa([^a-z])/$1NSA$2/g;  # Another one

It'd be faster to use `s/\bnsa\b/NSA/g' for these constructs.
-- 
Shields.