test cases for the naco_normalize stored function

Note that running the tests would require sticking in
the connection parameters for an Evergreen database; check
comments for some musings on how to do this better.

Signed-off-by: Galen Charlton <gmc@esilibrary.com>

git-svn-id: svn://svn.open-ils.org/ILS/trunk@18865 dcc99617-32d9-48b4-a31d-7c20da2025e4

diff --git a/Open-ILS/tests/naco_normalize.t b/Open-ILS/tests/naco_normalize.t
new file mode 100644 (file)
index 0000000..b43f344
--- /dev/null
+++ b/Open-ILS/tests/naco_normalize.t
@@ -0,0 +1,83 @@
+use strict;
+use warnings;
+use utf8;
+
+use Test::More;
+use Unicode::Normalize;
+use DBI;
+
+# This could be made better in at least one of two ways (or both);
+# 1. put PL/Perl code that doesn't require a database into external
+#    modules so that test frameworks can get at it more easily
+# 2. Build a test harness that knows how to find an Evergreen
+#    database to use for non-destructive testing.  Of course, there
+#    can be a chicken-and-egg problem here; also, a complete test
+#    suite would need to be able to do *destructive* testing, from
+#    which we'd presumably want to protect production databases.
+
+# Database connection parameters
+my $db_driver = 'Pg';
+my $db_host   = 'evergreen';
+my $db_port   = '5432';
+my $db_name   = 'evergreen';
+my $db_user   = 'evergreen';
+my $db_pw     = 'evergreen';
+my $dsn       = "dbi:" . $db_driver . ":dbname=" . $db_name .';host=' . $db_host . ';port=' . $db_port;
+
+my $dbh = DBI->connect($dsn, $db_user, $db_pw, {AutoCommit => 1, pg_enable_utf8 => 1, PrintError => 0});
+if (!defined($dbh)) {
+    plan skip_all => "Failed to connect to database: $DBI::errstr";
+}
+
+binmode STDOUT, ':utf8';
+binmode STDERR, ':utf8';
+
+my @test_cases = (
+    [ 'abc', 'abc', 'regular text' ],
+    [ 'ABC', 'abc', 'regular text' ],
+    [ 'åbçdéñœöîøæÇıÂÅÍÎÏÔÔÒÚÆŒè', 'abcdenoeoioaeciaaiiiooouaeoee', 'European diacritics' ],
+    [ '“‘„«quotes»’”', 'quotes', 'special quotes' ],
+    [ '\98abc\9c def', 'def', 'special non-filing characters designation' ],
+    [ '\9cabcdef', 'abcdef', 'unpaired start of string' ],
+    [ 'ß', 'ss', 'sharp S (eszett)' ],
+    [ 'flfiff', 'flfiff', 'ligatures' ],
+    [ 'ƠơƯư²IJij', 'oouu2ijij', 'NFKD applied correctly' ],
+    [ 'ÆØÞæðøþĐđıŁłŒœʻʼℓ', 'aeothaedothddilloeoel', 'part 3.6' ],
+    [ 'Ð', 'd', 'uppercase eth (missing from 3.6?)' ],
+    [ 'ıİ', 'ii', 'Turkish I' ],
+    [ '[book\'s cover]', 'books cover', 'square brackets and apostrophe' ],
+    [ '  grue   food ', 'grue food', 'trim spaces' ],
+    # note addition of NFKD() to transform expected output
+    [ '한국어 조선말', NFKD('한국어 조선말'), 'Korean text' ],
+    [ '普通話 / 普通话', '普通話 普通话', 'Chinese text' ],
+    [ 'العربية', 'العربية', 'Arabic text' ],
+    [ 'ქართული ენა', 'ქართული ენა', 'Georgian text' ],
+    [ 'русский язык', 'русскии язык', 'Russian text' ],
+    [ "\r\npa\tper\f", 'paper', 'other whitespace' ],
+    [ '#1: ∃ C++, @ home & abroad', '#1 c++ @ home & abroad', 'other punctuation' ],
+    [ '٠١٢٣٤٥', '012345', 'other decimal digits' ],
+    [ '²³¹', '231', 'superscript numbers' ],
+    [ '♭©®♯', '♭ ♯', 'other symbols' ],
+);
+
+my $sth1 = $dbh->prepare_cached('SELECT public.naco_normalize(?)');
+my $sth2 = $dbh->prepare_cached('SELECT public.naco_normalize(?, ?)');
+sub naco_normalize_wrapper {
+    my ($str, $sf) = @_;
+    if (defined $sf) {
+        $sth2->execute($str, $sf);
+        return $sth2->fetchrow_array;
+    } else {
+        $sth1->execute($str);
+        return $sth1->fetchrow_array;
+    }
+}
+
+foreach my $case (@test_cases) {
+    is(naco_normalize_wrapper($case->[0]), $case->[1], $case->[2]);    
+}
+
+is(naco_normalize_wrapper('Smith, Jane. Poet, painter, and author', 'a'), 'smith, jane poet painter and author', 
+    'retain first comma');
+
+done_testing;