remove </?unstructured_citation> earlier, so ending punctuation on ur…

…ls can be recognized git-svn-id: svn://tug.org/tugboat/trunk@617 7237770a-693a-0410-b4d3-c34f48dbc3f6
TeXUsersGroup · Nov 16, 2024 · c408745 · c408745
1 parent bf0f670
commit c408745
Show file tree

Hide file tree

Showing 2 changed files with 16 additions and 8 deletions.
diff --git a/capsules/ChangeLog b/capsules/ChangeLog
@@ -1,3 +1,9 @@
+2024-11-16  Karl Berry  <karl@freefriends.org>
+
+	* cr-landing-bbl-abs (read_crossref_xml): remove
+	</?unstructured_citation> early, so our check for [., ] at the end
+	of a url is effective.
+
 2024-11-15  Karl Berry  <karl@tug.org>
 
 	* tb140capsule.txt (mittelbach-tagging24.pdf): primary category

diff --git a/capsules/cr-landing-bbl-abs b/capsules/cr-landing-bbl-abs
@@ -17,8 +17,8 @@ Usage: $0 [CROSSREF.xml] [CROSSREF_WORKING_DIR]
 
 In CROSSREF_WORKING_DIR/*.html (Crossref landing) files,
 1) replace bibliography placeholders with the actual bbl text
-   taken from CROSSREF.xml, and also
-2) replace abstract placeholds with the abstract texts taken from
+   taken from CROSSREF.xml, including making urls live, and also
+2) replace abstract placeholders with the abstract texts taken from
    CROSSREF_WORKING_DIR/ARTICLE.abs.
 END_USAGE
     exit ($ARGV[0] ne "--help");
@@ -121,12 +121,16 @@ sub update_landing_bbl {
           my $newcite = "";
           for my $word (split (" ", $cite)) {
             #warn "checking $word\n";
-            my ($w1,$w2);
-
+            # remove the crossref element.
+            $word =~ s,</?unstructured_citation>,,g;
+
             # try to detect http(s) urls, with or without a protocol,
             # possibly followed by a period or comma.
             my ($prot,$urlbase) = "";
 
+            # used conditionally below to save capture groups.
+            my ($w1,$w2);
+
             # protocol is specified.
             if ($word =~ m!^https?://!) {
               ($prot,$urlbase) = ($word =~ m!^(https?)://(\S+)!);
@@ -170,14 +174,12 @@ sub update_landing_bbl {
               my $url = "$prot://$urlbase";
               $newcite .=  qq!<a href="$url">$urlbase</a>!;
             } else {
-              # As a second horrible special case, insert a line break
-              # before bullets (used in tb136andre bbl).
+              # Special case: insert a line break (in both html source
+              # and output) before bullets (used in tb136andre bbl).
               $newcite .= "\n<br>" if $word =~ /^&#x2022;/;
               $newcite .= $word;
             }
           }
-          # remove the crossref element.
-          $newcite =~ s,</?unstructured_citation>,,g;
           print $temp_fh "<li class=\"tugbbl\">$newcite</li>\n";
         }
         print $temp_fh "</ol>\n";