Bug 9886: ParseLetter removes too often punctuation characters from fields

This patch makes ParseLetter somewhat more restrictive in removing
punctuation characters from the end of a table field.
Based on the assumption that we want to remove punctuation from fields in
biblio and biblioitems (like ISBD).
ParseLetter should not remove e.g. a parenthesis in itemcallnumber, but still
removes e.g. a colon (:) at the end of a title.
Removed an unneeded global and lookahead from the regex.

Test plan:
1) Add a colon (:) to the end of a title.
2) Add a colon to the end of item copynumber.
3) Place a hold on that item. Check it in. Confirm hold.
4) Check the email or print notice generated. There should be no colon at the
end of the title, but the colon in the copynumber should still be there.

Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Katrin Fischer <Katrin.Fischer.83@web.de>
I compared checkout notices with lots of different fields before
and after applying the patch. For example the ) at the end of a
field in branches is now longer removed. Other fields looked ok
before and after.
Passes all tests and QA script.
Signed-off-by: Jared Camins-Esakov <jcamins@cpbibliography.com>

diff --git a/C4/Letters.pm b/C4/Letters.pm
index fdbb473..b8e2722 100644 (file)
--- a/C4/Letters.pm
+++ b/C4/Letters.pm
@@ -620,7 +620,11 @@ sub _parseletter {
     while ( my ($field, $val) = each %$values ) {
         my $replacetablefield = "<<$table.$field>>";
         my $replacefield = "<<$field>>";
-        $val =~ s/\p{P}(?=$)//g if $val;
+        $val =~ s/\p{P}$// if $val && $table=~/biblio/;
+            #BZ 9886: Assuming that we want to eliminate ISBD punctuation here
+            #Therefore adding the test on biblio. This includes biblioitems,
+            #but excludes items. Removed unneeded global and lookahead.
+
         my $replacedby   = defined ($val) ? $val : '';
         ($letter->{title}  ) and do {
             $letter->{title}   =~ s/$replacetablefield/$replacedby/g;