Changeset 3142

Timestamp:

2002-06-10T18:14:45+12:00 (22 years ago)

Author:

jrm21

Message:

1) We can't use "Date" for the year metadata, as greenstone assumes Date is YYYYMMDD, so we now use "Year" instead.

2) Tidied up the single-entry-per-line code, as it was missing some entries.

3) bibtex allows 3 letter month abbreviations, so we know expand those.

4) some more tidying up of latex (and bibtex) commands.

File:

• trunk/gsdl/perllib/plugins/BibTexPlug.pm (modified) (6 diffs)

trunk/gsdl/perllib/plugins/BibTexPlug.pm

@@ -105 +105 @@
          'title', 'Title',
          'volume', 'Volume',
-         'year', 'Date',
+         'year', 'Year', # Can't use "Date" as this implies DDDDMMYY!
 
          'keywords', 'Keywords',
@@ -119 +119 @@
     $verbosity = 0 unless $verbosity;
 
-    my $lines=$$textref;
-
-    # Make sure the text has exactly one entry per line
-
-    $lines  =~ s/^\s*(\@[^,]+,)\s*\n/$1=====/; #splitting key in entry
-    $lines =~ s/([\"\}]\s*,)\s*\n/$1=====/g; #splitting by comma, followed by \n (assuming end of lines are " or })
-    $lines =~ s/(\d+\s*\,)\s*\n/$1=====/g; #for the case, when we have number entry without closing "
-    $lines =~ s/\n\s*\n/%%%%%/g; #this was simply added in order to allow to process newline inside  quoted strings, 
-                                 #that continues for several lines
-    $lines =~ s/\s+/ /g;
-    $lines =~ s/\s*=====\s*/\n/g;
-
-    my @all_lines = split(/\n+/, $lines);
+    # Make sure the text has exactly one entry per line -
+    # append line to previous if it doesn't start with "  <key> = "
+
+    my @input_lines=split('\n', $$textref);
+    my @all_lines;
+    my $entry_line=shift @input_lines;
+    foreach my $input_line (@input_lines) {
+    if ($input_line =~ m/^\s*\w+\s*=\s*/) {
+        # this is a new key
+        push(@all_lines, $entry_line);
+        $entry_line=$input_line;
+    } else {
+        # this is a continuation of previous line
+        $entry_line .= $input_line;
+    }
+    
+    }
+    # add final line, removing trailing '}'
+    $entry_line =~ s/\}\s*$//;
+    push(@all_lines, $entry_line);
+    push(@all_lines, "}");
+
+
 
     # Read and process each line in the bib file.
@@ -189 +199 @@
     $value = $metadata{$entryname};
 
+    if ($name =~ /^Month/) {
+        $value=expand_month($value);
+    }
     # Add the various fields as metadata    
     my $html_value = &text_into_html($value);
@@ -338 +351 @@
 
 
+sub expand_month {
+    my $text=shift;
+
+    # bibtex style files expand abbreviations for months.
+    # Entries can contain more than one month (eg ' month = jun # "-" # aug, ')
+    $text =~ s/jan/_textmonth01_/g;
+    $text =~ s/feb/_textmonth02_/g;
+    $text =~ s/mar/_textmonth03_/g;
+    $text =~ s/apr/_textmonth04_/g;
+    $text =~ s/may/_textmonth05_/g;
+    $text =~ s/jun/_textmonth06_/g;
+    $text =~ s/jul/_textmonth07_/g;
+    $text =~ s/aug/_textmonth08_/g;
+    $text =~ s/sep/_textmonth09_/g;
+    $text =~ s/oct/_textmonth10_/g;
+    $text =~ s/nov/_textmonth11_/g;
+    $text =~ s/dec/_textmonth12_/g;
+
+    return $text;
+}
 
 
@@ -416 +449 @@
      '"Y' => chr(0xc3).chr(0xb8),
      # tilde
+     '~A' => chr(0xc3).chr(0x83),
+     '~N' => chr(0xc3).chr(0x91),
+     '~O' => chr(0xc3).chr(0x95),
+     '~a' => chr(0xc3).chr(0xa3),
+     '~n' => chr(0xc3).chr(0xb1),
+     '~o' => chr(0xc3).chr(0xb5),
      # caron - handled specially
 #      ',s' => chr(0xc5).chr(0xa1),
@@ -520 +559 @@
       # only do the change if immediately followed by a space, }, {, or \
       for $latex_code (keys %special_utf8_chars) {
-      $text =~ s/\\${latex_code}([\\\s\{\}])/$special_utf8_chars{$latex_code}$1/g;
+      $text =~ s/\\${latex_code}([\\\s{}])/$special_utf8_chars{$latex_code}$1/g;
       }
     }
-
     # remove latex commands
-    $text =~ s@\\\w+{([^}]*)}@$1@g;
-
+    $text =~ s@\\noopsort{[^}]+\}@@g;
+    $text =~ s@\\\w+{([^}]*)}@$1@g; # all other commands
+    
     # remove latex groupings { } (but not \{ or \} )
-    $text =~ s/([^\\])\{/$1/g;
-    $text =~ s/([^\\])\}/$1/g;
-    $text =~ s/^{//; # remove { if first char
-
+    while ($text =~ s/([^\\])[\{\}]/$1/g) {;}
+    $text =~ s/^\{//; # remove { if first char
+    
     # maths mode $...$ - this is not interpreted in any way at the moment...
     $text =~ s@\$(.*)\$@$1@g;
-
+    
+    # latex characters
+    # spaces - nobr space (~), opt break (\-), append ("#" - bibtex only)
+    $text =~ s/([^\\])~+/$1/g; # non-breaking space  "~"
+    # optional break "\-"
+    if ($text =~ m/\#/) { # concat macros (bibtex)
+    # the non-macro bits have quotes around them - we just remove quotes
+    $text =~ s/[\"\#]//g;
+    }
     # quoted { } chars
     $text =~ s@\\{@{@g;