[Ruby-DBI-devel-CVS] CVS: src/build DBI-VERSIONS,NONE,1.1 Makefile,NONE,1.1 USER,NONE,1.1 cl2html.sh,NONE,1.1 create_changelog.rb,NONE,1.1 cvs2cl,NONE,1.1

[Michael N. <mne...@us...>]

Update of /cvsroot/ruby-dbi/src/build
In directory usw-pr-cvs1:/tmp/cvs-serv13192/build

Added Files:
	DBI-VERSIONS Makefile USER cl2html.sh create_changelog.rb 
	cvs2cl 
Log Message:
initial import


--- NEW FILE: DBI-VERSIONS ---
# All times are in UCT
dbi-0-0-17
dbi-0-0-16	2002-07-03 20:10 UTC
dbi-0-0-15	2002-05-21 21:59 UTC
dbi-0-0-14	2002-05-14 18:16 UTC
dbi-0-0-13	2002-04-16 21:59 UTC
dbi-0-0-12	2001-12-28 13:08 UTC
dbi-0-0-11	2001-11-14 13:12 UTC
dbi-0-0-10	2001-10-22 16:09 UTC
dbi-0-0-9	2001-09-07 13:40 UTC
dbi-0-0-8	2001-07-28 12:08 UTC
dbi-0-0-7	2001-07-11 21:59 UTC
dbi-0-0-6	2001-07-06 19:42 UTC
dbi-0-0-5	2001-06-08 20:31 UTC

--- NEW FILE: Makefile ---
all: ChangeLog.html

ChangeLog:
	ruby create_changelog.rb

ChangeLog.html: ChangeLog
	./cl2html.sh

clean:
	rm ChangeLog ChangeLog.html

--- NEW FILE: USER ---
mneumann:'Michael Neumann <mne...@fa...>'
michael:'Michael Neumann <mne...@fa...>'
rainer:'Rainer Perl'

--- NEW FILE: cl2html.sh ---
#!/bin/sh

VIM="/usr/local/bin/gvim -f"

${VIM} +"syn on" +"set nonumber" +"run! syntax/2html.vim" +"w! ChangeLog.html" +"q!" +"q!" ChangeLog

--- NEW FILE: create_changelog.rb ---
require "date"

Dir.chdir ".."  # This script should be executed from directory src/ (of ruby-dbi cvs repository)

# modify if needed
FILE = "ChangeLog"
DBI_VERSIONS = "./build/DBI-VERSIONS"
USER = "./build/USER"


lines = File.readlines(DBI_VERSIONS)

START_DATE = "1980-01-01 00:00:00"    # a date before all other dates

v = []
lines.each {|l| 
  l.strip!
  next if l.empty? or l[0] == ?#
  v << l.split(/\s+/, 2) 
}

f = File.new(FILE, "w+")

for i in 0..(v.size-1)
  to = v[i] 
  from = v[i+1] || ["none-tag", START_DATE] 

  if to[1] != nil
    f << "\n\n"
    f << "===============================================================================\n"
    f << " Tag #{to[0]} (#{to[1]})\n"
    f << "===============================================================================\n"
    f << "\n\n"
  end
  to[1] ||= (Date.today + 3).to_s  # a day in the future
  f << `./build/cvs2cl -g -d:ext:mne...@cv...:/cvsroot/ruby-dbi -l "-d '#{from[1]}<#{to[1]}'" --stdout --prune --utc -U #{USER} --no-wrap -R '^\s*[*]\s*'`
end

f.close

--- NEW FILE: cvs2cl ---
#!/bin/sh
exec perl -w -x $0 ${1+"$@"} # -*- mode: perl; perl-indent-level: 2; -*-
#!perl -w

##############################################################
###                                                        ###
### cvs2cl.pl: produce ChangeLog(s) from `cvs log` output. ###
###                                                        ###
##############################################################

## $Revision: 1.1 $
## $Date: 2002/10/02 18:04:09 $
## $Author: mneumann $
##
##   (C) 1999 Karl Fogel <kf...@re...>, under the GNU GPL.
## 
##   (Extensively hacked on by Melissa O'Neill <on...@cs...>.)
##
## cvs2cl.pl is free software; you can redistribute it and/or modify
[...1839 lines suppressed...]
happen.  Interesting.

Anyway, rather than fix this in Text::Wrap, we might as well write a
new wrap() which has the following much-needed features:

* initial indentation, like current Text::Wrap()
* subsequent line indentation, like current Text::Wrap()
* user chooses among: force-break long words, leave them alone, or die()?
* preserve existing indentation: chopped chunks from an indented line
  are indented by same (like this line, not counting the asterisk!)
* optional list of things to preserve on line starts, default ">"

Note that the last two are essentially the same concept, so unify in
implementation and give a good interface to controlling them.

And how about:

Optionally, when encounter a line pre-indented by same as previous
line, then strip the newline and refill, but indent by the same.
Yeah...