Changeset 1734

Timestamp:

2000-12-01T16:36:33+13:00 (23 years ago)

Author:

jrm21

Message:

For postscript, fall back to some simple text extraction if ps2ascii isn't
found. (This should be portable as it is perl). It won't be formatted though,
so currently is only useful for indexing - users will have to view the
postscript for now...

File:

• trunk/gsdl/bin/script/gsConvert.pl (modified) (16 diffs)

trunk/gsdl/bin/script/gsConvert.pl

@@ -84 +84 @@
     # Deduce filenames
     my ($tailname,$dirname,$suffix) 
-    = File::Basename::fileparse($input_filename,'\.[^\.]+');
+    = File::Basename::fileparse($input_filename,'\..+');
     my $output_filestem = &util::filename_cat($dirname,"$tailname");
 
@@ -102 +102 @@
     } 
     elsif ($input_type eq "doc") {
+    print STDERR "I recognise this to be a Word document...\n"; # remove
     print &convertDOC($input_filename, $output_filestem, $output_type);
     print "\n";
@@ -148 +149 @@
     my $realtype = &find_docfile_type($input_filename);
 
-    if ($realtype eq "word678") {
+    print STDERR "The real type of this Word document is $realtype\n"; # remove
+
+    if ($realtype eq "word6" || $realtype eq "word7" || $realtype eq "word8") {
+    print STDERR "I recognise this to be a word678 document...\n"; # remove
     return &convertWord678($input_filename, $output_filestem, $output_type);
     } elsif ($realtype eq "rtf") {
@@ -166 +170 @@
     # Attempt specialised conversion to HTML
     if (!$output_type || ($output_type =~ /html/i)) {
+    print STDERR "I am about to call doc_to_html...\n";
     $success = &doc_to_html($input_filename, $output_filestem);
     if ($success) {
@@ -278 +283 @@
     ($input_filename) = @_;
     
+    open(TMP, ">temp.txt");
+    binmode(TMP);
     open(CHK, "<$input_filename");
+    binmode(CHK);
     my $line = "";
     my $first = 1;
@@ -285 +293 @@
     
     $line = $_;
-
+    print TMP "$line\n\n";
     if ($first) {
         # check to see if this is an rtf file
@@ -294 +302 @@
     }
     
-    # is theis a word 6/7/8 document?
-    if ($line =~ /Word\.Document\.[678]/) {
+    # is this is a word 6/7/8 document?
+    if ($line =~ /Word\.Document\.([678])/) {
         close(CHK);
-        return "word678";
+        return "word$1";
     }
 
@@ -309 +317 @@
 
 
-# Specific type-to-type cponversions
+# Specific type-to-type conversions
 #
 # Each of the following functions attempts to convert a document from 
@@ -320 +328 @@
 
 sub doc_to_html {
+    print STDERR "/;-DG I am in doc_to_html...\n"; # remove
     ($input_filename, $output_filestem) = @_;
 
-    # formulate the command
-    my $wv_home = &util::filename_cat($ENV{'GSDLHOME'}, "packages", "unix", "wv");
-    my $wv_conf = &util::filename_cat($wv_home, "lib", "wv", "wvHtml.xml");
-    my $wvWare = &util::filename_cat($wv_home, "bin", "wvWare");
+    my $wvWare = "";
+    my $wv_conf = "";
+    if ($ENV{'GSDLOS'} =~ /^windows$/i) {
+    $wvWare = "$ENV{'GSDLHOME'}\\bin\\windows\\wvWare.exe";
+    $wv_conf = "$ENV{'GSDLHOME'}\\bin\\windows\\wvHtml.xml";
+
+    } else {
+    # formulate the command
+    my $wv_home = &util::filename_cat($ENV{'GSDLHOME'}, "packages", "unix", "wv");
+    $wv_conf = &util::filename_cat($wv_home, "lib", "wv", "wvHtml.xml");
+    $wvWare = &util::filename_cat($wv_home, "bin", "wvWare");
+    }
+    print STDERR "I am about to test if your file exists...\n";
     return 0 unless (-e "$wvWare");
     $cmd = "";
@@ -332 +350 @@
     $cmd .= " \"$input_filename\" > \"$output_filestem.html\" 2>\"$output_filestem.err\"";
 
+    print STDERR "$cmd\n"; #remove
+
     # execute the command
+    print STDERR system($cmd);
+    print STDERR "\n";
     if (system($cmd)>0)
     {
@@ -352 +374 @@
     }
     }
+    print STDERR "/;-DG I am leaving doc_to_html...\n";
     return 0;
 }
@@ -488 +511 @@
     my $cmd = "ps2ascii \"$input_filename\" > \"$output_filestem.text\"";
     $cmd .= " 2> $output_filestem.err";
-    
     if (system($cmd)>0)
     {
@@ -494 +516 @@
     &util::rm("$output_filestem.text") if (-e "$output_filestem.text");
     &util::rm("$output_filestem.err") if (-e "$output_filestem.err");
-    return 0;
-    }
-
+
+    # Fine then. We'll just do a lousy job by ourselves...
+    # Based on code nicked from:
+    # http://snark.ptc.spbu.ru/mail-archives/lout/brown/msg00003.html
+    # 
+    print STDERR "Attempting to strip text from postscript.\n";
+    my $errorcode=0;
+    open (IN, "$input_filename") 
+        ||  ($errorcode=1, warn "Couldn't read file: $!");
+    open (OUT, ">$output_filestem.text") 
+        ||  ($errorcode=1, warn "Couldn't write file: $!");
+    if ($errorcode) {print STDERR "errors\n";return 0;}
+
+    my $in_a_sentence=0;
+    while (<IN>) {
+        if (/^[^\(\)]+$/ && !$in_a_sentence) {next ;} # no brackets in line
+        # attempt to add whitespace between different lines...
+        s/F.?\(/\( /g; # this might break up some other words though...
+        ### remove all postscript control data
+        if (!$in_a_sentence) {
+        s/^[^\(\)]*?\(//;}  # rm start of line up to first open bracket
+        s/\\\(/\{/g;s/\\\)/\}/g ; # change quoted braces
+        s/\)([^\(\)])*?\(//g ;   # close bracket up to next open unquoted bracket
+        if (s/\)[^\(\)]*?$//g)  # last close bracket to end of line
+        {$in_a_sentence=0;chomp;}     
+        if (s/\\$//) # if line is a continuation
+        {$in_a_sentence=1;chomp;} 
+        s/^$//g ;             # remove empty lines
+        ### ligatures have special characters...
+        s/\\214/fi/g;
+        s/\\215/fl/g;
+        print OUT "$_";
+    }
+    close IN; close OUT;
+    }
     &util::rm("$output_filestem.err") if (-e "$output_filestem.err");
     return 1;
@@ -506 +560 @@
 
 sub any_to_html {
+    print STDERR "/;-Dg I am in any_to_html!\n";
     ($input_filename, $output_filestem) = @_;
 
@@ -518 +573 @@
 <META HTTP-EQUIV="Content-Type" CONTENT="text/html">
 <META NAME="GENERATOR" CONTENT="Greenstone any_to_html">
-</head><body>\n\n';
+</head><body>';
+    print HTML "\n\n";
+
     while (<TEXT>) { 
     print HTML "<p> ", $_;
     
     }
-    print HTML "\n</body></html>]\n";
+    print HTML "\n</body></html>\n";
 
     &util::rm("$output_filestem.text") if (-e "$output_filestem.text");
+    print STDERR "/;-Dg I am getting out of  any_to_html!\n";
     return 1;
 }
@@ -535 +593 @@
     ($input_filename, $output_filestem) = @_;
 
+    #open(TEMP, ">temp.txt");
     open(IN, "<$input_filename");
+    binmode(IN);
     open(OUT, ">$output_filestem.text");
 
     my ($line);
+    my $dgcount = 0;
     while (<IN>) { 
     $line = $_;
-    
+
     # delete anything that isn't a printable character
+    #print TEMP $line;
     $line =~ s/[^\040-\176]+/\n/sg;
 
     # delete any string less than 10 characters long
-    $line =~ s/^[^\n]{0,9}$/\n/mg;
-    while ($line =~ /^[^\n]{1,9}$/m) {
-        $line =~ s/^[^\n]{0,9}$/\n/mg;
+    $line =~ s/^.{0,9}$/\n/mg;
+    while ($line =~ /^.{1,9}$/m) {
+        $line =~ s/^.{0,9}$/\n/mg;
         $line =~ s/\n+/\n/sg;
     }