#!/usr/bin/perl -w

use strict;
use Getopt::Long;
use File::Basename;

my($pdfbasepath) = $ENV{'HOME'} . "/public_html/PDFDOCS/";

my($datafilename);
my($verbose) = 0;
my($fromdate, $uptodate) = (0, 999999);
my($mincool)=0;
my($maxmonth);
(undef,undef,undef,undef,undef,$maxmonth,undef,undef,undef) = localtime(time);
$maxmonth += 1901;              # now it's the next year
$maxmonth *= 100;               # now it has two zeros
$maxmonth += 12;                # now it's yyyymm

GetOptions('datafilename=s' => \$datafilename,
           'verbose+' => \$verbose,
           'mincool=i' => \$mincool,
           'from=i' => \$fromdate,
           'upto=i' => \$uptodate);
die "$0: Incorrect commandline args.
Usage:
  $0 --data filename
     [--verbose]
     [--mincool number]
     [--from yyyymm] [--upto yyyymm]

By default, all entries are taken. If you specify an integer mincool,
then entries with inferior coolness are skipped. If you specify either
a starting or an ending date (or both), then entries that don't fit are
screened out.

Say '--verbose' (or -v) multiple times for more verbosity."
    if (!defined $datafilename);

# datafilename could be ans-res.hrb or ../ans-res.hrb or ~ajayshah/PERSONAL/RESUME/ans-res.hrb
my($basename) = &basename($datafilename, ".hrb");

print "Parsed commandline --

  datafile             $datafilename
    -> basename        $basename
  verbosity            $verbose
  mincool              $mincool
  from                 $fromdate
  upto                 $uptodate\n\n" if ($verbose>2);

my($name, $url, $email, $bibfile, $bst, @order);
my(%subjects, %types);
my(@entries);

my(%usedpdf);
for (`ls $pdfbasepath`) {
    chop $_;
    $usedpdf{$_}=0;
}

# Let me explain the data structures.
#
# The hashes 'subjects' and 'types' are just shortname to fullname mappings.
# E.g. you might have $subjects{'hpc'} = 'High performance computing'.
#      and you might have $types{'pub'} = 'Publications'.
#
# The ORDER in which the sections appear in the generated resume is
# supplied as part of basics - it is in @order.
#
# The @entries table stores the actual information. It is indexed by
# the entry number, so the entries in the .hrb file come into this as
# elem 0, 1, 2, ...
# The columns in this file are hash table entries, where the following
# information is stored:
#           date, from, upto, cool, content, type, pdf, key, subject

$/="\n\n";
my($i);                         # just a scratch counter
my($n)=0;
open(F, $datafilename) or die "$0: Data file $datafilename not found.\n";
print "Opened $datafilename...\n" if ($verbose > 0);
while (<F>) {
    s/\#.*?\n//gm;              # deal with comments (ungreedy)
    s/\#.*$//m;                 # comment that ends the file  :-)
    s/\n/ /gm;                  # remove carriage returns
    s/\s+/ /g;                  # remove surplus whitespace
    next if /^\s*$/;            # Do we have nothing after all?

    my(@f) = split /;/;         # If !nothing, it has to be ';' delimited
    die "$0: Error in $datafilename --
             malformed paragraph at " . join(" ", @f) . "\n"
             if ($#f == 0);

    my($taken)=0;
    if ($f[0] eq "basics") {
        print "Processing basics...\n" if ($verbose > 2);
        for ($i=1; $i<$#f; $i++) {
            # These are strings of the form 'key = value'
            my(@words) = split(/=/, $f[$i], 2);
            $words[0] =~ s/^\s+//;
            $words[0] =~ s/\s+$//;
            $words[1] =~ s/^\s+//;
            $words[1] =~ s/\s+$//;
            die "$0: Error in $datafilename --
                     Illegal entry : $f[$i]\n"
                     if ((length($words[0]) eq 0)
                         ||
                         (length($words[1]) eq 0));
            my($entrytaken)=0;
            if ($words[0] eq 'name') {
                $name = $words[1];                $entrytaken=1;
            }
            if ($words[0] eq 'url') {
                $url = $words[1];                 $entrytaken=1;
            }
            if ($words[0] eq 'email') {
                $email = $words[1];               $entrytaken=1;
            }
            if ($words[0] eq 'bibfile') {
                $bibfile = $words[1];             $entrytaken=1;
            }
            if ($words[0] eq 'bst') {
                $bst = $words[1];                 $entrytaken=1;
            }
            if ($words[0] eq 'order') {
                @order = split /,/, $words[1];
                my($j);
                for ($j=0; $j<=$#order; $j++) {
                    $order[$j] =~ s/^\s+//;
                    $order[$j] =~ s/\s$//;
                }                                 $entrytaken=1;
            }
            die "$0: Dunno what you mean by $f[$i]\n" if (!$entrytaken);
        }
        $taken=1;
    }

    if ($f[0] eq "mappings") {
        print "Processing mappings...\n" if ($verbose > 2);
        my($i);
        for ($i=1; $i<$#f; $i++) {
            # These are strings of the form 'subject risk -> Risk management'
            #                                0       1    2  3
            my(@words) = split(' ', $f[$i], 4);
            die "$0: Error in $datafilename --
                     Illegal mappings string : $f[$i]\n"
                     if ((($words[0] ne 'type') && ($words[0] ne 'subject'))
                         ||
                         ($words[2] ne '->')
                         ||
                         (length($words[3]) eq 0));
            if ($words[0] eq 'subject') {
                $subjects{$words[1]} = $words[3];
            }
            if ($words[0] eq 'type') {
                $types{$words[1]} = $words[3];
            }
        }
        $taken=1;
    }

    if ($f[0] =~ /date/) {
        print "Processing data...\n" if ($verbose > 2);
        my($i);
        for ($i=0; $i<$#f; $i++) {
            # These are strings of the form 'key = value'
            #                                0     1
            my(@words) = split(/=/, $f[$i], 2);
            $words[0] =~ s/^\s+//;
            $words[0] =~ s/\s+$//;
            $words[1] =~ s/^\s+//;
            $words[1] =~ s/\s+$//;
            die "$0: Error in $datafilename --
                     Illegal entry : $f[$i]\n"
                     if ((length($words[0]) eq 0)
                         ||
                         (length($words[1]) eq 0));

            my($entrytaken)=0;
            if ($words[0] eq 'date') {
                my($dateditem, $daterangeitem, $date, $from, $upto);
                &processdate($words[1], \$dateditem, \$daterangeitem,
                             \$date, \$from, \$upto);
                die "$0: You can't be both dated and daterange (at $f[$i])\n"
                    if ($dateditem && $daterangeitem);
                if ($dateditem) {
                    $entries[$n]{'date'} = $date;
                } else {
                    $entries[$n]{'from'} = $from;
                    $entries[$n]{'upto'} = $upto;
                }
                $entrytaken=1;
            }
            if ($words[0] eq 'cool') {
                $entries[$n]{'cool'} = $words[1];        $entrytaken=1;
            }
            if ($words[0] eq 'content') {
                $entries[$n]{'content'} = $words[1];     $entrytaken=1;
            }
            if ($words[0] eq 'type') {
                die "$0: Type $words[1] was not defined earlier.\n"
                    if (!defined $types{$words[1]});
                $entries[$n]{'type'} = $words[1];        $entrytaken=1;
            }
            if ($words[0] eq 'pdf') {
                $entries[$n]{'pdf'} = $words[1];         $entrytaken=1;
                die "$0: File " . " " . $entries[$n]{'pdf'} . " is required but is missing.\n"
                    if (! -f $pdfbasepath . '/' . $entries[$n]{'pdf'});
                $usedpdf{$words[1]} = 1;
            }
            if ($words[0] eq 'key') {
                $entries[$n]{'key'} = $words[1];         $entrytaken=1;
            }
            if ($words[0] eq 'subject') {
                die "$0: Subject $words[1] was not defined earlier.\n"
                    if (!defined $subjects{$words[1]});
                $entries[$n]{'subject'} = $words[1];     $entrytaken=1;
            }
            die "$0: I did not know what to do with $f[$i]\n"
                if (!$entrytaken);
        }
        $taken=1;
        $entries[$n]{'taken'}=0;
        $n++;
    }

    die "$0: I just don't know what to do with " . join(" ", @f) . "\n"
        if (!$taken);
}
print "Finished reading $datafilename.\n" if ($verbose > 0);
close(F);

# Warn him about PDF files not used
for (keys %usedpdf) {
    print "Your pdf file $_ exists but is not used.\n"
           if ($usedpdf{$_} == 0);
}

# Validation rules for basics.
die "$0: You did not tell me the bst file.\n" if (!defined $bst);
die "$0: You did not tell me the bib file.\n" if (!defined $bibfile);
die "$0: You did not tell me your name.\n"    if (!defined $name);
die "$0: You did not tell me your URL.\n"     if (!defined $url);
die "$0: You did not tell me your email address.\n" if (!defined $email);

if ($verbose >= 2) {
    print "The basics:\n";
    print "  name    = $name
  url     = $url
  email   = $email
  bibfile = $bibfile
  bst     = $bst
  order   = ", join(",", @order), "\n\n";

    # Throw in a sanity check here.
    for (@order) {
        die "$0: Don't know what you mean by $_ (in 'order').\n"
            if ((!defined $subjects{$_}) && (!defined $types{$_}));
    }

    print "The subjects mapping table is:\n";
    for (keys %subjects) {
        print "  $_ -> $subjects{$_}\n";
    }
    print "\nThe types mapping table is:\n";
    for (keys %types) {
        print "  $_ -> $types{$_}\n";
    }
    print "\nThe table of $n entries is:\n";
    my($i);
    for ($i=0; $i<$n; $i++) {
        print "$i. ";
        if (defined $entries[$i]{'date'}) {
            print 'date: ', $entries[$i]{'date'}, "; ";
        }
        if (defined $entries[$i]{'from'}) {
            print 'from: ', $entries[$i]{'from'}, "; ";
        }
        if (defined $entries[$i]{'upto'}) {
            print 'upto: ', $entries[$i]{'upto'}, "; ";
        }
        if (defined $entries[$i]{'cool'}) {
            print 'cool: ', $entries[$i]{'cool'}, "; ";
        }
        if (defined $entries[$i]{'content'}) {
            print 'content: ', $entries[$i]{'content'}, "; ";
        }
        if (defined $entries[$i]{'type'}) {
            print 'type: ', $entries[$i]{'type'}, "; ";
        }
        if (defined $entries[$i]{'pdf'}) {
            print 'pdf: ', $entries[$i]{'pdf'}, "; ";
        }
        if (defined $entries[$i]{'key'}) {
            print 'key: ', $entries[$i]{'key'}, "; ";
        }
        if (defined $entries[$i]{'subject'}) {
            print 'subject: ', $entries[$i]{'subject'}, "; ";
        }
        print "\n";
    }
}

# Now we get on with resume generation...
print "Starting to write ./$basename.tex...\n" if ($verbose>0);
&cleantex();
open(F, ">$basename.tex") or die "$0: Could not write ./$basename.tex\n";
print F <<'EOF';
\documentclass[11pt,a4paper]{article}
\usepackage{parskip}
\usepackage{bibentry}
\input resume-config

% ------------------------------------------------------------------
% Put username on front page nicely
% Arg 1 - your name    (e.g. 'Ajay Shah')
% Arg 2 - homepage URL (e.g. 'http://www.igidr.ac.in/\string~ajayshah')
%                            just saying blah/~ajayshah won't work!
% Arg 3 - your email address
\newcommand{\username}[3]{
  \hfill\begin{tabular}{r}
        \textbf{\LARGE \myhref{#2}{#1}} \\
        \hline
        \texttt{#3} \\
        \texttt{#2} \\
        \end{tabular}
}
% ------------------------------------------------------------------

% ------------------------------------------------------------------
% Setup new style for section/subsection titles
% Stolen from `The Latex Companion'
\makeatletter
\renewcommand\section{\@startsection {section}{1}{\z@}%
                              {-3.5ex \@plus -1ex \@minus -.2ex}%
                              {2.3ex \@plus.2ex}%
                              {\normalfont\large\bfseries\sffamily}}
\renewcommand\subsection{\@startsection{subsection}{2}{\z@}%
                                {-3.25ex\@plus -1ex \@minus -.2ex}%
                                {1.5ex \@plus .2ex}%
                                {\normalfont\large\sffamily}}
\renewcommand\subsubsection{\@startsection{subsubsection}{3}{\z@}%
                                {-3.25ex\@plus -1ex \@minus -.2ex}%
                                {1.5ex \@plus .2ex}%
                                {\normalfont\sffamily}}
\makeatother
% ------------------------------------------------------------------

\begin{document}
\sloppy
EOF

printf F "\\bibliographystyle\{%s\}
\\nobibliography\{%s\}
\\username\{%s\}
         \{%s\}
         \{%s\}\n",
    $bst, $bibfile, $name, $url, $email;

my($key);
for $key (@order) {
    my($utilised)=0;
    # Create the resume content for $key...
    # Is it a type or is it a subject?
    if (defined $types{$key}) {
        # First check - do we have any content, or is this section blank?
        my($count)=0;
        for ($i=0; $i<$n; $i++) {
            $count++
                if ((defined $entries[$i]{'type'})
                    && ($entries[$i]{'type'} eq $key)
                    && ($entries[$i]{'taken'} == 0));
        }
        if ($count > 0) {
            printf F "\n\n\n\\section*{%s}\n", $types{$key};
            for ($i=0; $i<$n; $i++) {
                if ((defined $entries[$i]{'type'})
                      && ($entries[$i]{'type'} eq $key)
                      && ($entries[$i]{'taken'} == 0)) {
                    &emit_tex(\*F, \@entries, $i,
                              $fromdate, $uptodate, $mincool);
                    $entries[$i]{'taken'}=1;
                }
            }
        } else {
            printf F "\n\n\n%% Skipped %s since there was nothing\n", $key;
        }
        $utilised=1;
    }
    if (defined $subjects{$key}) {
        # First check - do we have any content, or is this section blank?
        my($count)=0;
        for ($i=0; $i<$n; $i++) {
            $count++
                if ((defined $entries[$i]{'subject'})
                    && ($entries[$i]{'subject'} eq $key)
                    && ($entries[$i]{'taken'} == 0));
        }
        if ($count > 0) {
            printf F "\n\n\n\\section*{%s}\n", $subjects{$key};
            for ($i=0; $i<$n; $i++) {
                if ((defined $entries[$i]{'subject'})
                      && ($entries[$i]{'subject'} eq $key)
                      && ($entries[$i]{'taken'} == 0)) {
                    &emit_tex(\*F, \@entries, $i,
                              $fromdate, $uptodate, $mincool);
                    $entries[$i]{'taken'}=1;
                }
            }
        } else {
            printf F "\n\n\n%% Skipped %s since there was nothing\n", $key;
        }
        $utilised=1;
    }
    die "$0: You have \"$key\" in `order', which is
              neither a subject nor a type.\n"
              if (!$utilised);
}
print F <<'EOF';
\end{document}
EOF

print "Finished making ./$basename.tex\n" if ($verbose>0);
close(F);

sleep(5);

# Time to put ink on paper...
# PDF -- 
print "\n\n\n\nProducing pdf...\n";
open(F, ">resume-config.tex")
    or die "$0: Could not write ./resume-config.tex\n";
print F <<'EOF';
% Switch into Times Roman for use with pdflatex
% \usepackage{times}
% Zap out the cute macros used in resume-config-html
\newcommand{\pdffile}[1]{~}
\newcommand{\myhref}[2]{#2}
EOF
close(F);
system("pdflatex $basename.tex && bibtex $basename && pdflatex $basename.tex && pdflatex $basename.tex");
&cleantex();

# Text --
print "\n\n\n\nProducing text...\n";
open(F, ">resume-config.tex")
    or die "$0: Could not write ./resume-config.tex\n";
print F <<'EOF';
% Use tex4ht if you want text
\usepackage[html]{tex4ht}
\newcommand{\myhref}[2]{#2}
\newcommand{\pdffile}[1]{~}
EOF
close(F);
system("latex $basename.tex && bibtex $basename && latex $basename.tex && latex $basename.tex && tex4ht $basename.tex && lynx -dump $basename.html > $basename.text");
&cleantex();
system("rm resume-config.tex");

# HTML --
print "\n\n\n\nProducing HTML...\n";
open(F, ">resume-config.tex")
    or die "$0: Could not write ./resume-config.tex\n";
print F <<'EOF';
% Use tex4ht if you want HTML
\usepackage[html,2]{tex4ht}
% myhref:
%    Usage: \myhref{URL}{string}
%    In the HTML file, it comes out as <a href="URL">string</a>
%    In the TeX file,  it comes out as string.
\newcommand{\myhref}[2]{\Link[#1]{}{}#2\EndLink}
% pdffile:
%    Usage: \pdffile{a.pdf}
%    In the HTML file, it comes out as <a href="PDFDOCS/a.pdf">[pdf]</a>
%    In the TeX file,  it is zapped out.
\newcommand{\pdffile}[1]{\Link[PDFDOCS/#1]{}{}[pdf]\EndLink}
% FYI - \pdffile{r.pdf}{s} is the same as
%	\htmlembellish{PDFDOCS/r.pdf}{[pdf]}
% or	\pdffilemyanchor{r.pdf}{[pdf]}
EOF
close(F);
system("latex $basename.tex && bibtex $basename && latex $basename.tex && latex $basename.tex && tex4ht $basename.tex");
&cleantex();

# Time to generate one file per subject
my($subject);
for $subject (keys %subjects) {
    open(F, ">$subject.tex") or die;
    print F <<'EOF';
\documentclass[11pt,a4paper]{article}
\usepackage{parskip}
\usepackage{bibentry}
\input resume-config

% ------------------------------------------------------------------
% Setup new style for section/subsection titles
% Stolen from `The Latex Companion'
\makeatletter
\renewcommand\section{\@startsection {section}{1}{\z@}%
                              {-3.5ex \@plus -1ex \@minus -.2ex}%
                              {2.3ex \@plus.2ex}%
                              {\normalfont\large\bfseries\sffamily}}
\renewcommand\subsection{\@startsection{subsection}{2}{\z@}%
                                {-3.25ex\@plus -1ex \@minus -.2ex}%
                                {1.5ex \@plus .2ex}%
                                {\normalfont\large\sffamily}}
\renewcommand\subsubsection{\@startsection{subsubsection}{3}{\z@}%
                                {-3.25ex\@plus -1ex \@minus -.2ex}%
                                {1.5ex \@plus .2ex}%
                                {\normalfont\sffamily}}
\makeatother
% ------------------------------------------------------------------

\begin{document}
\sloppy
EOF
printf F "\\bibliographystyle\{%s\}
\\nobibliography\{%s\}\n",
    $bst, $bibfile;

    printf F "\\section*{Work in %s}\n", $subjects{$subject};

    for ($i=0; $i<$n; $i++) {
        if ((defined $entries[$i]{'subject'})
            && ($entries[$i]{'subject'} eq $subject)) {
            &emit_tex(\*F, \@entries, $i,
                      0, 999999, $mincool);
        }
    }

    printf F "\n\\vspace*{1cm}\n\n";
    printf F "\\HCode{<hr><a href=\"%s\">Return to %s's home page</a>\n<hr>\n<h6 align=right>Generated using <a href=\"http://www.mayin.org/\\string~ajayshah/CODE/hrbuilder-manual.html\">hrbuilder</a>.</h6>\n}\n", $url, $name;
    printf F "\\end{document}\n";
    close(F);

    print "Finished making ./$subject.tex\n" if ($verbose>0);

# HTML --
    print "\nProducing HTML...\n";
    open(F, ">resume-config.tex")
        or die "$0: Could not write ./resume-config.tex\n";
    print F <<'EOF';
% Use tex4ht if you want HTML
\usepackage[html]{tex4ht}
% myhref:
%    Usage: \myhref{URL}{string}
%    In the HTML file, it comes out as <a href="URL">string</a>
%    In the TeX file,  it comes out as string.
\newcommand{\myhref}[2]{\Link[#1]{}{}#2\EndLink}
% pdffile:
%    Usage: \pdffile{a.pdf}
%    In the HTML file, it comes out as <a href="PDFDOCS/a.pdf">[pdf]</a>
%    In the TeX file,  it is zapped out.
\newcommand{\pdffile}[1]{\Link[PDFDOCS/#1]{}{}[pdf]\EndLink}
% FYI - \pdffile{r.pdf}{s} is the same as
%	\htmlembellish{PDFDOCS/r.pdf}{[pdf]}
% or	\pdffilemyanchor{r.pdf}{[pdf]}
EOF
    close(F);
    system("latex $subject.tex && bibtex $subject && latex $subject.tex && latex $subject.tex && tex4ht $subject.tex");
    &cleantex();
}

# END OF MAIN PROGRAM

# Function cleantex
sub cleantex {
    system("rm -f *.{aux,dvi,log,ps,bbl,blg,toc,out,ent,fff,ttt}");
}

# Function processdate
# Inputs
#    a string like yyyymm, yyyymm- or yyyymm-yyyymm
# Outputs
#    if it's just a date, then dateditem=1 and date is set.
#    if it's a daterange then daterangeitem=1 and fromdate,uptodate are set.
#    if it's yyyymm- then uptodate is set as 999999.
sub processdate {
    my($thestring,
       $pdateditem, $pdaterangeitem,
       $pdate,
       $pfrom, $pupto) = @_;
    if ($thestring =~ /\-/) {
        $$pdaterangeitem=1;
        $$pdateditem=0;
        my(@f) = split /\-/, $thestring;
        $$pfrom = $f[0];
        if (defined $f[1]) {
            $$pupto = $f[1];
        } else {
            $$pupto = 999999;
        }
        die "$0: Illegal month range in $thestring"
            if ($$pupto < $$pfrom);
        for ($$pupto, $$pfrom) {
            if ($_ != 999999) {
                die "$0: Error in date $_.\n"
                    if (($_ < 0) || ($_ > $maxmonth));
            }
        }
    } else {
        $$pdaterangeitem=0;
        $$pdateditem=1;
        $$pdate=$thestring;
        die "$0: Error in date $thestring.\n"
            if (($$pdate != 999999)
                &&
                (($$pdate < 0) || ($$pdate > $maxmonth)));
    }
}

# Function emit_tex
# Inputs
#   a file handle
#   the vector of entries
#   the entry number where we have to write .tex.
#   limiting from date
#   limiting upto date
#   amount of mincool required.
# Outputs
#   the file handle is written into.
#   We variously utilise --
#     date fields: date, from, upto
#     coolness:    cool
#     classifiers: key, subject
#     The goods: {key, pdf} or {content}
sub emit_tex {
    my($fh, $pentries, $i, $fromdate, $uptodate, $mincool) = @_;

    # Test for mincool
    return 0 
        if ((defined $pentries->[$i]{'cool'})
            &&
            ($pentries->[$i]{'cool'} < $mincool));

    # Test if the date fits.
    return 0 
        if ((defined $pentries->[$i]{'date'})
            &&
            (($pentries->[$i]{'date'} < $fromdate)
             ||
             ($pentries->[$i]{'date'} > $uptodate)));
    return 0
        if ((defined $pentries->[$i]{'from'})
            &&
            ($pentries->[$i]{'from'} > $uptodate));
    return 0
        if ((defined $pentries->[$i]{'upto'})
            &&
            ($pentries->[$i]{'upto'} < $fromdate));

    # Emit tex.
    print $fh "\n% Item $i\n";
    if (defined $pentries->[$i]{'key'}) { # it's a pub.
        # In this case, we definitely do a \bibentry, but we'll also
        # take the contents if it's there.
        printf $fh "\\bibentry\{%s\} ", $pentries->[$i]{'key'};
        if (defined $pentries->[$i]{'content'}) {
            printf $fh $pentries->[$i]{'content'};
        }
    } else {
        # In this case, we bloody well need a content.
        if (!defined $pentries->[$i]{'content'}) {
            die "$0: Content not defined for elem $i\n";
        } else {
            printf $fh $pentries->[$i]{'content'};
        }
    }
    if (defined $pentries->[$i]{'pdf'}) { # he gave us a PDF file
        printf $fh " \\pdffile\{%s\}", $pentries->[$i]{'pdf'};
    }
    printf $fh "\n";
}