[scpubgit/stemmatology.git] / base / t / text_tradition_collation.t

#!/usr/bin/perl -w

use strict;
use Test::More 'no_plan';
$| = 1;


# =begin testing
{
use Text::Tradition;
use TryCatch;

my $cxfile = 't/data/Collatex-16.xml';
my $t = Text::Tradition->new( 
    'name'  => 'inline', 
    'input' => 'CollateX',
    'file'  => $cxfile,
    );
my $c = $t->collation;

my $rno = scalar $c->readings;
# Split n21 ('unto') for testing purposes
my $new_r = $c->add_reading( { 'id' => 'n21p0', 'text' => 'un', 'join_next' => 1 } );
my $old_r = $c->reading( 'n21' );
$old_r->alter_text( 'to' );
$c->del_path( 'n20', 'n21', 'A' );
$c->add_path( 'n20', 'n21p0', 'A' );
$c->add_path( 'n21p0', 'n21', 'A' );
$c->add_relationship( 'n21', 'n22', { type => 'collated', scope => 'local' } );
$c->flatten_ranks();
ok( $c->reading( 'n21p0' ), "New reading exists" );
is( scalar $c->readings, $rno, "Reading add offset by flatten_ranks" );

# Combine n3 and n4 ( with his )
$c->merge_readings( 'n3', 'n4', 1 );
ok( !$c->reading('n4'), "Reading n4 is gone" );
is( $c->reading('n3')->text, 'with his', "Reading n3 has both words" );

# Collapse n9 and n10 ( rood / root )
$c->merge_readings( 'n9', 'n10' );
ok( !$c->reading('n10'), "Reading n10 is gone" );
is( $c->reading('n9')->text, 'rood', "Reading n9 has an unchanged word" );

# Try to combine n21 and n21p0. This should break.
my $remaining = $c->reading('n21');
$remaining ||= $c->reading('n22');  # one of these should still exist
try {
	$c->merge_readings( 'n21p0', $remaining, 1 );
	ok( 0, "Bad reading merge changed the graph" );
} catch( Text::Tradition::Error $e ) {
	like( $e->message, qr/neither concatenated nor collated/, "Expected exception from bad concatenation" );
} catch {
	ok( 0, "Unexpected error on bad reading merge: $@" );
}

try {
	$c->calculate_ranks();
	ok( 1, "Graph is still evidently whole" );
} catch( Text::Tradition::Error $e ) {
	ok( 0, "Caught a rank exception: " . $e->message );
}
}


# =begin testing
{
use Test::Warn;
use Text::Tradition;
use TryCatch;

my $t;
warnings_exist {
	$t = Text::Tradition->new( 'input' => 'Self', 'file' => 't/data/legendfrag.xml' );
} [qr/Cannot set relationship on a meta reading/],
	"Got expected relationship drop warning on parse";

my $c = $t->collation;
# Force the transitive propagation of all existing relationships.
$c->relations->propagate_all_relationships();

my %rdg_ids;
map { $rdg_ids{$_} = 1 } $c->readings;
$c->merge_related( 'orthographic' );
is( scalar( $c->readings ), keys( %rdg_ids ) - 9, 
	"Successfully collapsed orthographic variation" );
map { $rdg_ids{$_} = undef } qw/ r13.3 r11.4 r8.5 r8.2 r7.7 r7.5 r7.4 r7.3 r7.1 /;
foreach my $rid ( keys %rdg_ids ) {
	my $exp = $rdg_ids{$rid};
	is( !$c->reading( $rid ), !$exp, "Reading $rid correctly " . 
		( $exp ? "retained" : "removed" ) );
}
ok( $c->linear, "Graph is still linear" );
try {
	$c->calculate_ranks; # This should succeed
	ok( 1, "Can still calculate ranks on the new graph" );
} catch {
	ok( 0, "Rank calculation on merged graph failed: $@" );
}

# Now add some transpositions
$c->add_relationship( 'r8.4', 'r10.4', { type => 'transposition' } );
$c->merge_related( 'transposition' );
is( scalar( $c->readings ), keys( %rdg_ids ) - 10, 
	"Transposed relationship is merged away" );
ok( !$c->reading('r8.4'), "Correct transposed reading removed" );
ok( !$c->linear, "Graph is no longer linear" );
try {
	$c->calculate_ranks; # This should fail
	ok( 0, "Rank calculation happened on nonlinear graph?!" );
} catch ( Text::Tradition::Error $e ) {
	is( $e->message, 'Cannot calculate ranks on a non-linear graph', 
		"Rank calculation on merged graph threw an error" );
}
}


# =begin testing
{
use Text::Tradition;

my $t = Text::Tradition->new( input => 'CollateX', file => 't/data/Collatex-16.xml' );
my $c = $t->collation;
my $n = scalar $c->readings;
$c->compress_readings();
is( scalar $c->readings, $n - 6, "Compressing readings seems to work" );

# Now put in a join-word and make sure the thing still works.
my $t2 = Text::Tradition->new( input => 'CollateX', file => 't/data/Collatex-16.xml' );
my $c2 = $t2->collation;
# Split n21 ('unto') for testing purposes
my $new_r = $c2->add_reading( { 'id' => 'n21p0', 'text' => 'un', 'join_next' => 1 } );
my $old_r = $c2->reading( 'n21' );
$old_r->alter_text( 'to' );
$c2->del_path( 'n20', 'n21', 'A' );
$c2->add_path( 'n20', 'n21p0', 'A' );
$c2->add_path( 'n21p0', 'n21', 'A' );
$c2->calculate_ranks();
is( scalar $c2->readings, $n + 1, "We have our extra test reading" );
$c2->compress_readings();
is( scalar $c2->readings, $n - 6, "Compressing readings also works with join_next" );
is( $c2->reading( 'n21p0' )->text, 'unto', "The joined word has no space" );
}


# =begin testing
{
use Test::More::UTF8;
use Text::Tradition;
use TryCatch;

my $st = Text::Tradition->new( 'input' => 'Self', 'file' => 't/data/collatecorr.xml' );
is( ref( $st ), 'Text::Tradition', "Got a tradition from test file" );
ok( $st->has_witness('Ba96'), "Tradition has the affected witness" );

my $sc = $st->collation;
my $numr = 17;
ok( $sc->reading('n131'), "Tradition has the affected reading" );
is( scalar( $sc->readings ), $numr, "There are $numr readings in the graph" );
is( $sc->end->rank, 14, "There are fourteen ranks in the graph" );

# Detach the erroneously collated reading
my( $newr, @del_rdgs ) = $sc->duplicate_reading( 'n131', 'Ba96' );
ok( $newr, "New reading was created" );
ok( $sc->reading('n131_0'), "Detached the bad collation with a new reading" );
is( scalar( $sc->readings ), $numr + 1, "A reading was added to the graph" );
is( $sc->end->rank, 10, "There are now only ten ranks in the graph" );
my $csucc = $sc->common_successor( 'n131', 'n131_0' );
is( $csucc->id, 'n136', "Found correct common successor to duped reading" ); 

# Check that the bad transposition is gone
is( scalar @del_rdgs, 1, "Deleted reading was returned by API call" );
is( $sc->get_relationship( 'n130', 'n135' ), undef, "Bad transposition relationship is gone" );

# The collation should not be fixed
my @pairs = $sc->identical_readings();
is( scalar @pairs, 0, "Not re-collated yet" );
# Fix the collation
ok( $sc->merge_readings( 'n124', 'n131_0' ), "Collated the readings correctly" );
@pairs = $sc->identical_readings( start => 'n124', end => $csucc->id );
is( scalar @pairs, 3, "Found three more identical readings" );
is( $sc->end->rank, 11, "The ranks shifted appropriately" );
$sc->flatten_ranks();
is( scalar( $sc->readings ), $numr - 3, "Now we are collated correctly" );

# Check that we can't "duplicate" a reading with no wits or with all wits
try {
	my( $badr, @del_rdgs ) = $sc->duplicate_reading( 'n124' );
	ok( 0, "Reading duplication without witnesses throws an error" );
} catch( Text::Tradition::Error $e ) {
	like( $e->message, qr/Must specify one or more witnesses/, 
		"Reading duplication without witnesses throws the expected error" );
} catch {
	ok( 0, "Reading duplication without witnesses threw the wrong error" );
}

try {
	my( $badr, @del_rdgs ) = $sc->duplicate_reading( 'n124', 'Ba96', 'Mü11475' );
	ok( 0, "Reading duplication with all witnesses throws an error" );
} catch( Text::Tradition::Error $e ) {
	like( $e->message, qr/Cannot join all witnesses/, 
		"Reading duplication with all witnesses throws the expected error" );
} catch {
	ok( 0, "Reading duplication with all witnesses threw the wrong error" );
}

try {
	$sc->calculate_ranks();
	ok( 1, "Graph is still evidently whole" );
} catch( Text::Tradition::Error $e ) {
	ok( 0, "Caught a rank exception: " . $e->message );
}
}


# =begin testing
{
use JSON qw/ from_json /;
use Text::Tradition;

my $t = Text::Tradition->new( 
	'input' => 'Self',
	'file' => 't/data/florilegium_graphml.xml' );
my $c = $t->collation;
	
# Make a connection so we can test rank preservation
$c->add_relationship( 'w91', 'w92', { type => 'grammatical' } );

# Create an adjacency list of the whole thing; test the output.
my $adj_whole = from_json( $c->as_adjacency_list() );
is( scalar @$adj_whole, scalar $c->readings(), 
	"Same number of nodes in graph and adjacency list" );
my @adj_whole_edges;
map { push( @adj_whole_edges, @{$_->{adjacent}} ) } @$adj_whole;
is( scalar @adj_whole_edges, scalar $c->sequence->edges,
	"Same number of edges in graph and adjacency list" );
# Find the reading whose rank should be preserved
my( $test_rdg ) = grep { $_->{id} eq 'w89' } @$adj_whole;
my( $test_edge ) = grep { $_->{id} eq 'w92' } @{$test_rdg->{adjacent}};
is( $test_edge->{minlen}, 2, "Rank of test reading is preserved" );

# Now create an adjacency list of just a portion. w76 to w122
my $adj_part = from_json( $c->as_adjacency_list(
	{ from => $c->reading('w76')->rank,
	  to   => $c->reading('w122')->rank }));
is( scalar @$adj_part, 48, "Correct number of nodes in partial graph" );
my @adj_part_edges;
map { push( @adj_part_edges, @{$_->{adjacent}} ) } @$adj_part;
is( scalar @adj_part_edges, 58,
	"Same number of edges in partial graph and adjacency list" );
# Check for consistency
my %part_nodes;
map { $part_nodes{$_->{id}} = 1 } @$adj_part;
foreach my $edge ( @adj_part_edges ) {
	my $testid = $edge->{id};
	ok( $part_nodes{$testid}, "ID $testid referenced in edge is given as node" );
}
}


# =begin testing
{
use Text::Tradition;
use TryCatch;

my $READINGS = 311;
my $PATHS = 361;

my $datafile = 't/data/florilegium_tei_ps.xml';
my $tradition = Text::Tradition->new( 'input' => 'TEI',
                                      'name' => 'test0',
                                      'file' => $datafile,
                                      'linear' => 1 );

ok( $tradition, "Got a tradition object" );
is( scalar $tradition->witnesses, 13, "Found all witnesses" );
ok( $tradition->collation, "Tradition has a collation" );

my $c = $tradition->collation;
is( scalar $c->readings, $READINGS, "Collation has all readings" );
is( scalar $c->paths, $PATHS, "Collation has all paths" );
is( scalar $c->relationships, 0, "Collation has all relationships" );

# Add a few relationships
$c->add_relationship( 'w123', 'w125', { 'type' => 'collated' } );
$c->add_relationship( 'w193', 'w196', { 'type' => 'collated' } );
$c->add_relationship( 'w257', 'w262', { 'type' => 'transposition', 
					  'is_significant' => 'yes' } );

# Now write it to GraphML and parse it again.

my $graphml = $c->as_graphml;
my $st = Text::Tradition->new( 'input' => 'Self', 'string' => $graphml );
is( scalar $st->collation->readings, $READINGS, "Reparsed collation has all readings" );
is( scalar $st->collation->paths, $PATHS, "Reparsed collation has all paths" );
is( scalar $st->collation->relationships, 3, "Reparsed collation has new relationships" );
my $sigrel = $st->collation->get_relationship( 'w257', 'w262' );
is( $sigrel->is_significant, 'yes', "Ternary attribute value was restored" );

# Now add a stemma, write to GraphML, and look at the output.
SKIP: {
	skip "Analysis module not present", 3 unless $tradition->can( 'add_stemma' );
	my $stemma = $tradition->add_stemma( 'dotfile' => 't/data/florilegium.dot' );
	is( ref( $stemma ), 'Text::Tradition::Stemma', "Parsed dotfile into stemma" );
	is( $tradition->stemmata, 1, "Tradition now has the stemma" );
	$graphml = $c->as_graphml;
	like( $graphml, qr/digraph/, "Digraph declaration exists in GraphML" );
}
}


# =begin testing
{
use Text::Tradition;
use Text::CSV;

my $READINGS = 311;
my $PATHS = 361;
my $WITS = 13;
my $WITAC = 4;

my $datafile = 't/data/florilegium_tei_ps.xml';
my $tradition = Text::Tradition->new( 'input' => 'TEI',
                                      'name' => 'test0',
                                      'file' => $datafile,
                                      'linear' => 1 );

my $c = $tradition->collation;
# Export the thing to CSV
my $csvstr = $c->as_csv();
# Count the columns
my $csv = Text::CSV->new({ sep_char => ',', binary => 1 });
my @lines = split(/\n/, $csvstr );
ok( $csv->parse( $lines[0] ), "Successfully parsed first line of CSV" );
is( scalar( $csv->fields ), $WITS + $WITAC, "CSV has correct number of witness columns" );
my @q_ac = grep { $_ eq 'Q'.$c->ac_label } $csv->fields;
ok( @q_ac, "Found a layered witness" );

my $t2 = Text::Tradition->new( input => 'Tabular',
							   name => 'test2',
							   string => $csvstr,
							   sep_char => ',' );
is( scalar $t2->collation->readings, $READINGS, "Reparsed CSV collation has all readings" );
is( scalar $t2->collation->paths, $PATHS, "Reparsed CSV collation has all paths" );

# Now do it with TSV
my $tsvstr = $c->as_tsv();
my $t3 = Text::Tradition->new( input => 'Tabular',
							   name => 'test3',
							   string => $tsvstr,
							   sep_char => "\t" );
is( scalar $t3->collation->readings, $READINGS, "Reparsed TSV collation has all readings" );
is( scalar $t3->collation->paths, $PATHS, "Reparsed TSV collation has all paths" );

my $table = $c->alignment_table;
my $noaccsv = $c->as_csv({ noac => 1 });
my @noaclines = split(/\n/, $noaccsv );
ok( $csv->parse( $noaclines[0] ), "Successfully parsed first line of no-ac CSV" );
is( scalar( $csv->fields ), $WITS, "CSV has correct number of witness columns" );
is( $c->alignment_table, $table, "Request for CSV did not alter the alignment table" );

my $safecsv = $c->as_csv({ safe_ac => 1});
my @safelines = split(/\n/, $safecsv );
ok( $csv->parse( $safelines[0] ), "Successfully parsed first line of safe CSV" );
is( scalar( $csv->fields ), $WITS + $WITAC, "CSV has correct number of witness columns" );
@q_ac = grep { $_ eq 'Q__L' } $csv->fields;
ok( @q_ac, "Found a sanitized layered witness" );
is( $c->alignment_table, $table, "Request for CSV did not alter the alignment table" );

# Test relationship collapse
$c->add_relationship( $c->readings_at_rank( 37 ), { type => 'spelling' } );
$c->add_relationship( $c->readings_at_rank( 60 ), { type => 'spelling' } );

my $mergedtsv = $c->as_tsv({mergetypes => [ 'spelling', 'orthographic' ] });
my $t4 = Text::Tradition->new( input => 'Tabular',
							   name => 'test4',
							   string => $mergedtsv,
							   sep_char => "\t" );
is( scalar $t4->collation->readings, $READINGS - 2, "Reparsed TSV merge collation has fewer readings" );
is( scalar $t4->collation->paths, $PATHS - 4, "Reparsed TSV merge collation has fewer paths" );

# Test non-ASCII sigla
my $t5 = Text::Tradition->new( input => 'Tabular',
							   name => 'nonascii',
							   file => 't/data/armexample.xlsx',
							   excel => 'xlsx' );
my $awittsv = $t5->collation->as_tsv({ noac => 1, ascii => 1 });
my @awitlines = split( /\n/, $awittsv );
like( $awitlines[0], qr/_A_5315622/, "Found ASCII sigil variant in TSV" );
}


# =begin testing
{
use Text::Tradition;

my $cxfile = 't/data/Collatex-16.xml';
my $t = Text::Tradition->new( 
    'name'  => 'inline', 
    'input' => 'CollateX',
    'file'  => $cxfile,
    );
my $c = $t->collation;

# Make an svg
my $table = $c->alignment_table;
ok( $c->has_cached_table, "Alignment table was cached" );
is( $c->alignment_table, $table, "Cached table returned upon second call" );
$c->calculate_ranks;
is( $c->alignment_table, $table, "Cached table retained with no rank change" );
$c->add_relationship( 'n13', 'n23', { type => 'repetition' } );
is( $c->alignment_table, $table, "Alignment table unchanged after non-colo relationship add" );
$c->add_relationship( 'n24', 'n23', { type => 'spelling' } );
isnt( $c->alignment_table, $table, "Alignment table changed after colo relationship add" );
}


# =begin testing
{
use Text::Tradition;

my $cxfile = 't/data/Collatex-16.xml';
my $t = Text::Tradition->new( 
    'name'  => 'inline', 
    'input' => 'CollateX',
    'file'  => $cxfile,
    );
my $c = $t->collation;

my @common = $c->calculate_common_readings();
is( scalar @common, 8, "Found correct number of common readings" );
my @marked = sort $c->common_readings();
is( scalar @common, 8, "All common readings got marked as such" );
my @expected = qw/ n1 n11 n16 n19 n20 n5 n6 n7 /;
is_deeply( \@marked, \@expected, "Found correct list of common readings" );
}


# =begin testing
{
use Text::Tradition;

my $cxfile = 't/data/Collatex-16.xml';
my $t = Text::Tradition->new( 
    'name'  => 'inline', 
    'input' => 'CollateX',
    'file'  => $cxfile,
    );
my $c = $t->collation;

is( $c->common_predecessor( 'n24', 'n23' )->id, 
    'n20', "Found correct common predecessor" );
is( $c->common_successor( 'n24', 'n23' )->id, 
    '__END__', "Found correct common successor" );

is( $c->common_predecessor( 'n19', 'n17' )->id, 
    'n16', "Found correct common predecessor for readings on same path" );
is( $c->common_successor( 'n21', 'n10' )->id, 
    '__END__', "Found correct common successor for readings on same path" );
}


1;
Commit	Line	Data
0e47f4f6	1	#!/usr/bin/perl -w
	2
	3	use strict;
	4	use Test::More 'no_plan';
	5	$\| = 1;
	6
	7
	8
	9	# =begin testing
	10	{
	11	use Text::Tradition;
58568d5c	12	use TryCatch;
0e47f4f6	13
4e483aa5	14	my $cxfile = 't/data/Collatex-16.xml';
	15	my $t = Text::Tradition->new(
	16	'name' => 'inline',
	17	'input' => 'CollateX',
	18	'file' => $cxfile,
	19	);
	20	my $c = $t->collation;
	21
	22	my $rno = scalar $c->readings;
3c234eb6	23	# Split n21 ('unto') for testing purposes
4e483aa5	24	my $new_r = $c->add_reading( { 'id' => 'n21p0', 'text' => 'un', 'join_next' => 1 } );
	25	my $old_r = $c->reading( 'n21' );
	26	$old_r->alter_text( 'to' );
	27	$c->del_path( 'n20', 'n21', 'A' );
	28	$c->add_path( 'n20', 'n21p0', 'A' );
	29	$c->add_path( 'n21p0', 'n21', 'A' );
7a0956c1	30	$c->add_relationship( 'n21', 'n22', { type => 'collated', scope => 'local' } );
4e483aa5	31	$c->flatten_ranks();
	32	ok( $c->reading( 'n21p0' ), "New reading exists" );
	33	is( scalar $c->readings, $rno, "Reading add offset by flatten_ranks" );
	34
679f17e1	35	# Combine n3 and n4 ( with his )
4e483aa5	36	$c->merge_readings( 'n3', 'n4', 1 );
	37	ok( !$c->reading('n4'), "Reading n4 is gone" );
	38	is( $c->reading('n3')->text, 'with his', "Reading n3 has both words" );
	39
679f17e1	40	# Collapse n9 and n10 ( rood / root )
	41	$c->merge_readings( 'n9', 'n10' );
	42	ok( !$c->reading('n10'), "Reading n10 is gone" );
	43	is( $c->reading('n9')->text, 'rood', "Reading n9 has an unchanged word" );
4e483aa5	44
58568d5c	45	# Try to combine n21 and n21p0. This should break.
4e483aa5	46	my $remaining = $c->reading('n21');
4e483aa5	47	$remaining \|\|= $c->reading('n22'); # one of these should still exist
58568d5c	48	try {
	49	$c->merge_readings( 'n21p0', $remaining, 1 );
	50	ok( 0, "Bad reading merge changed the graph" );
	51	} catch( Text::Tradition::Error $e ) {
	52	like( $e->message, qr/neither concatenated nor collated/, "Expected exception from bad concatenation" );
	53	} catch {
	54	ok( 0, "Unexpected error on bad reading merge: $@" );
	55	}
	56
	57	try {
	58	$c->calculate_ranks();
	59	ok( 1, "Graph is still evidently whole" );
	60	} catch( Text::Tradition::Error $e ) {
	61	ok( 0, "Caught a rank exception: " . $e->message );
	62	}
4e483aa5	63	}
	64
	65
	66
	67	# =begin testing
	68	{
58568d5c	69	use Test::Warn;
8d9494a8	70	use Text::Tradition;
	71	use TryCatch;
	72
58568d5c	73	my $t;
	74	warnings_exist {
	75	$t = Text::Tradition->new( 'input' => 'Self', 'file' => 't/data/legendfrag.xml' );
	76	} [qr/Cannot set relationship on a meta reading/],
	77	"Got expected relationship drop warning on parse";
	78
8d9494a8	79	my $c = $t->collation;
a512d0e6	80	# Force the transitive propagation of all existing relationships.
a512d0e6	81	$c->relations->propagate_all_relationships();
8d9494a8	82
	83	my %rdg_ids;
	84	map { $rdg_ids{$_} = 1 } $c->readings;
	85	$c->merge_related( 'orthographic' );
751ee528	86	is( scalar( $c->readings ), keys( %rdg_ids ) - 9,
8d9494a8	87	"Successfully collapsed orthographic variation" );
751ee528	88	map { $rdg_ids{$_} = undef } qw/ r13.3 r11.4 r8.5 r8.2 r7.7 r7.5 r7.4 r7.3 r7.1 /;
8d9494a8	89	foreach my $rid ( keys %rdg_ids ) {
	90	my $exp = $rdg_ids{$rid};
	91	is( !$c->reading( $rid ), !$exp, "Reading $rid correctly " .
	92	( $exp ? "retained" : "removed" ) );
	93	}
	94	ok( $c->linear, "Graph is still linear" );
	95	try {
	96	$c->calculate_ranks; # This should succeed
	97	ok( 1, "Can still calculate ranks on the new graph" );
	98	} catch {
	99	ok( 0, "Rank calculation on merged graph failed: $@" );
	100	}
	101
	102	# Now add some transpositions
	103	$c->add_relationship( 'r8.4', 'r10.4', { type => 'transposition' } );
	104	$c->merge_related( 'transposition' );
751ee528	105	is( scalar( $c->readings ), keys( %rdg_ids ) - 10,
8d9494a8	106	"Transposed relationship is merged away" );
	107	ok( !$c->reading('r8.4'), "Correct transposed reading removed" );
	108	ok( !$c->linear, "Graph is no longer linear" );
	109	try {
	110	$c->calculate_ranks; # This should fail
	111	ok( 0, "Rank calculation happened on nonlinear graph?!" );
	112	} catch ( Text::Tradition::Error $e ) {
	113	is( $e->message, 'Cannot calculate ranks on a non-linear graph',
	114	"Rank calculation on merged graph threw an error" );
	115	}
	116	}
	117
	118
	119
	120	# =begin testing
	121	{
c67190dd	122	use Text::Tradition;
c67190dd	123
2e818703	124	my $t = Text::Tradition->new( input => 'CollateX', file => 't/data/Collatex-16.xml' );
c67190dd	125	my $c = $t->collation;
	126	my $n = scalar $c->readings;
	127	$c->compress_readings();
	128	is( scalar $c->readings, $n - 6, "Compressing readings seems to work" );
	129
	130	# Now put in a join-word and make sure the thing still works.
2e818703	131	my $t2 = Text::Tradition->new( input => 'CollateX', file => 't/data/Collatex-16.xml' );
c67190dd	132	my $c2 = $t2->collation;
	133	# Split n21 ('unto') for testing purposes
	134	my $new_r = $c2->add_reading( { 'id' => 'n21p0', 'text' => 'un', 'join_next' => 1 } );
	135	my $old_r = $c2->reading( 'n21' );
	136	$old_r->alter_text( 'to' );
	137	$c2->del_path( 'n20', 'n21', 'A' );
	138	$c2->add_path( 'n20', 'n21p0', 'A' );
	139	$c2->add_path( 'n21p0', 'n21', 'A' );
	140	$c2->calculate_ranks();
	141	is( scalar $c2->readings, $n + 1, "We have our extra test reading" );
	142	$c2->compress_readings();
	143	is( scalar $c2->readings, $n - 6, "Compressing readings also works with join_next" );
	144	is( $c2->reading( 'n21p0' )->text, 'unto', "The joined word has no space" );
	145	}
	146
	147
	148
	149	# =begin testing
	150	{
68e48c06	151	use Test::More::UTF8;
4e483aa5	152	use Text::Tradition;
68e48c06	153	use TryCatch;
f97ef19e	154
	155	my $st = Text::Tradition->new( 'input' => 'Self', 'file' => 't/data/collatecorr.xml' );
	156	is( ref( $st ), 'Text::Tradition', "Got a tradition from test file" );
	157	ok( $st->has_witness('Ba96'), "Tradition has the affected witness" );
	158
	159	my $sc = $st->collation;
	160	my $numr = 17;
	161	ok( $sc->reading('n131'), "Tradition has the affected reading" );
	162	is( scalar( $sc->readings ), $numr, "There are $numr readings in the graph" );
	163	is( $sc->end->rank, 14, "There are fourteen ranks in the graph" );
	164
	165	# Detach the erroneously collated reading
2dcb5d11	166	my( $newr, @del_rdgs ) = $sc->duplicate_reading( 'n131', 'Ba96' );
ef73c20a	167	ok( $newr, "New reading was created" );
f97ef19e	168	ok( $sc->reading('n131_0'), "Detached the bad collation with a new reading" );
	169	is( scalar( $sc->readings ), $numr + 1, "A reading was added to the graph" );
	170	is( $sc->end->rank, 10, "There are now only ten ranks in the graph" );
3c234eb6	171	my $csucc = $sc->common_successor( 'n131', 'n131_0' );
3c234eb6	172	is( $csucc->id, 'n136', "Found correct common successor to duped reading" );
f97ef19e	173
f97ef19e	174	# Check that the bad transposition is gone
2dcb5d11	175	is( scalar @del_rdgs, 1, "Deleted reading was returned by API call" );
f97ef19e	176	is( $sc->get_relationship( 'n130', 'n135' ), undef, "Bad transposition relationship is gone" );
f97ef19e	177
e19635f8	178	# The collation should not be fixed
	179	my @pairs = $sc->identical_readings();
	180	is( scalar @pairs, 0, "Not re-collated yet" );
f97ef19e	181	# Fix the collation
3c234eb6	182	ok( $sc->merge_readings( 'n124', 'n131_0' ), "Collated the readings correctly" );
e19635f8	183	@pairs = $sc->identical_readings( start => 'n124', end => $csucc->id );
3c234eb6	184	is( scalar @pairs, 3, "Found three more identical readings" );
e19635f8	185	is( $sc->end->rank, 11, "The ranks shifted appropriately" );
3c234eb6	186	$sc->flatten_ranks();
f97ef19e	187	is( scalar( $sc->readings ), $numr - 3, "Now we are collated correctly" );
68e48c06	188
	189	# Check that we can't "duplicate" a reading with no wits or with all wits
	190	try {
	191	my( $badr, @del_rdgs ) = $sc->duplicate_reading( 'n124' );
	192	ok( 0, "Reading duplication without witnesses throws an error" );
	193	} catch( Text::Tradition::Error $e ) {
	194	like( $e->message, qr/Must specify one or more witnesses/,
	195	"Reading duplication without witnesses throws the expected error" );
	196	} catch {
	197	ok( 0, "Reading duplication without witnesses threw the wrong error" );
	198	}
	199
	200	try {
	201	my( $badr, @del_rdgs ) = $sc->duplicate_reading( 'n124', 'Ba96', 'Mü11475' );
	202	ok( 0, "Reading duplication with all witnesses throws an error" );
	203	} catch( Text::Tradition::Error $e ) {
	204	like( $e->message, qr/Cannot join all witnesses/,
	205	"Reading duplication with all witnesses throws the expected error" );
	206	} catch {
	207	ok( 0, "Reading duplication with all witnesses threw the wrong error" );
	208	}
58568d5c	209
	210	try {
	211	$sc->calculate_ranks();
	212	ok( 1, "Graph is still evidently whole" );
	213	} catch( Text::Tradition::Error $e ) {
	214	ok( 0, "Caught a rank exception: " . $e->message );
	215	}
f97ef19e	216	}
	217
	218
	219
	220	# =begin testing
	221	{
8a9a8200	222	use JSON qw/ from_json /;
	223	use Text::Tradition;
	224
	225	my $t = Text::Tradition->new(
	226	'input' => 'Self',
	227	'file' => 't/data/florilegium_graphml.xml' );
	228	my $c = $t->collation;
	229
	230	# Make a connection so we can test rank preservation
	231	$c->add_relationship( 'w91', 'w92', { type => 'grammatical' } );
	232
	233	# Create an adjacency list of the whole thing; test the output.
	234	my $adj_whole = from_json( $c->as_adjacency_list() );
	235	is( scalar @$adj_whole, scalar $c->readings(),
	236	"Same number of nodes in graph and adjacency list" );
	237	my @adj_whole_edges;
	238	map { push( @adj_whole_edges, @{$_->{adjacent}} ) } @$adj_whole;
	239	is( scalar @adj_whole_edges, scalar $c->sequence->edges,
	240	"Same number of edges in graph and adjacency list" );
	241	# Find the reading whose rank should be preserved
	242	my( $test_rdg ) = grep { $_->{id} eq 'w89' } @$adj_whole;
	243	my( $test_edge ) = grep { $_->{id} eq 'w92' } @{$test_rdg->{adjacent}};
	244	is( $test_edge->{minlen}, 2, "Rank of test reading is preserved" );
	245
	246	# Now create an adjacency list of just a portion. w76 to w122
	247	my $adj_part = from_json( $c->as_adjacency_list(
	248	{ from => $c->reading('w76')->rank,
	249	to => $c->reading('w122')->rank }));
	250	is( scalar @$adj_part, 48, "Correct number of nodes in partial graph" );
	251	my @adj_part_edges;
	252	map { push( @adj_part_edges, @{$_->{adjacent}} ) } @$adj_part;
	253	is( scalar @adj_part_edges, 58,
	254	"Same number of edges in partial graph and adjacency list" );
	255	# Check for consistency
	256	my %part_nodes;
	257	map { $part_nodes{$_->{id}} = 1 } @$adj_part;
	258	foreach my $edge ( @adj_part_edges ) {
	259	my $testid = $edge->{id};
	260	ok( $part_nodes{$testid}, "ID $testid referenced in edge is given as node" );
	261	}
	262	}
	263
	264
	265
	266	# =begin testing
	267	{
f97ef19e	268	use Text::Tradition;
951ddfe8	269	use TryCatch;
4e483aa5	270
56eefa04	271	my $READINGS = 311;
	272	my $PATHS = 361;
	273
	274	my $datafile = 't/data/florilegium_tei_ps.xml';
	275	my $tradition = Text::Tradition->new( 'input' => 'TEI',
	276	'name' => 'test0',
	277	'file' => $datafile,
	278	'linear' => 1 );
	279
	280	ok( $tradition, "Got a tradition object" );
	281	is( scalar $tradition->witnesses, 13, "Found all witnesses" );
	282	ok( $tradition->collation, "Tradition has a collation" );
	283
	284	my $c = $tradition->collation;
	285	is( scalar $c->readings, $READINGS, "Collation has all readings" );
	286	is( scalar $c->paths, $PATHS, "Collation has all paths" );
	287	is( scalar $c->relationships, 0, "Collation has all relationships" );
	288
	289	# Add a few relationships
	290	$c->add_relationship( 'w123', 'w125', { 'type' => 'collated' } );
	291	$c->add_relationship( 'w193', 'w196', { 'type' => 'collated' } );
b71e7ea8	292	$c->add_relationship( 'w257', 'w262', { 'type' => 'transposition',
b71e7ea8	293	'is_significant' => 'yes' } );
56eefa04	294
	295	# Now write it to GraphML and parse it again.
	296
	297	my $graphml = $c->as_graphml;
	298	my $st = Text::Tradition->new( 'input' => 'Self', 'string' => $graphml );
	299	is( scalar $st->collation->readings, $READINGS, "Reparsed collation has all readings" );
	300	is( scalar $st->collation->paths, $PATHS, "Reparsed collation has all paths" );
	301	is( scalar $st->collation->relationships, 3, "Reparsed collation has new relationships" );
b71e7ea8	302	my $sigrel = $st->collation->get_relationship( 'w257', 'w262' );
b71e7ea8	303	is( $sigrel->is_significant, 'yes', "Ternary attribute value was restored" );
2a812726	304
9fef629b	305	# Now add a stemma, write to GraphML, and look at the output.
951ddfe8	306	SKIP: {
37bf09f4	307	skip "Analysis module not present", 3 unless $tradition->can( 'add_stemma' );
951ddfe8	308	my $stemma = $tradition->add_stemma( 'dotfile' => 't/data/florilegium.dot' );
	309	is( ref( $stemma ), 'Text::Tradition::Stemma', "Parsed dotfile into stemma" );
	310	is( $tradition->stemmata, 1, "Tradition now has the stemma" );
	311	$graphml = $c->as_graphml;
	312	like( $graphml, qr/digraph/, "Digraph declaration exists in GraphML" );
	313	}
56eefa04	314	}
	315
	316
	317
	318	# =begin testing
	319	{
16203db5	320	use Text::Tradition;
34ca808b	321	use Text::CSV;
16203db5	322
	323	my $READINGS = 311;
	324	my $PATHS = 361;
34ca808b	325	my $WITS = 13;
34ca808b	326	my $WITAC = 4;
16203db5	327
	328	my $datafile = 't/data/florilegium_tei_ps.xml';
	329	my $tradition = Text::Tradition->new( 'input' => 'TEI',
	330	'name' => 'test0',
	331	'file' => $datafile,
	332	'linear' => 1 );
	333
	334	my $c = $tradition->collation;
	335	# Export the thing to CSV
	336	my $csvstr = $c->as_csv();
34ca808b	337	# Count the columns
	338	my $csv = Text::CSV->new({ sep_char => ',', binary => 1 });
	339	my @lines = split(/\n/, $csvstr );
	340	ok( $csv->parse( $lines[0] ), "Successfully parsed first line of CSV" );
	341	is( scalar( $csv->fields ), $WITS + $WITAC, "CSV has correct number of witness columns" );
cbc8e08f	342	my @q_ac = grep { $_ eq 'Q'.$c->ac_label } $csv->fields;
	343	ok( @q_ac, "Found a layered witness" );
	344
16203db5	345	my $t2 = Text::Tradition->new( input => 'Tabular',
	346	name => 'test2',
	347	string => $csvstr,
	348	sep_char => ',' );
	349	is( scalar $t2->collation->readings, $READINGS, "Reparsed CSV collation has all readings" );
	350	is( scalar $t2->collation->paths, $PATHS, "Reparsed CSV collation has all paths" );
	351
	352	# Now do it with TSV
	353	my $tsvstr = $c->as_tsv();
	354	my $t3 = Text::Tradition->new( input => 'Tabular',
	355	name => 'test3',
	356	string => $tsvstr,
	357	sep_char => "\t" );
	358	is( scalar $t3->collation->readings, $READINGS, "Reparsed TSV collation has all readings" );
	359	is( scalar $t3->collation->paths, $PATHS, "Reparsed TSV collation has all paths" );
34ca808b	360
4e64b669	361	my $table = $c->alignment_table;
34ca808b	362	my $noaccsv = $c->as_csv({ noac => 1 });
	363	my @noaclines = split(/\n/, $noaccsv );
	364	ok( $csv->parse( $noaclines[0] ), "Successfully parsed first line of no-ac CSV" );
	365	is( scalar( $csv->fields ), $WITS, "CSV has correct number of witness columns" );
4e64b669	366	is( $c->alignment_table, $table, "Request for CSV did not alter the alignment table" );
cbc8e08f	367
	368	my $safecsv = $c->as_csv({ safe_ac => 1});
	369	my @safelines = split(/\n/, $safecsv );
	370	ok( $csv->parse( $safelines[0] ), "Successfully parsed first line of safe CSV" );
	371	is( scalar( $csv->fields ), $WITS + $WITAC, "CSV has correct number of witness columns" );
	372	@q_ac = grep { $_ eq 'Q__L' } $csv->fields;
	373	ok( @q_ac, "Found a sanitized layered witness" );
	374	is( $c->alignment_table, $table, "Request for CSV did not alter the alignment table" );
de20588d	375
	376	# Test relationship collapse
	377	$c->add_relationship( $c->readings_at_rank( 37 ), { type => 'spelling' } );
	378	$c->add_relationship( $c->readings_at_rank( 60 ), { type => 'spelling' } );
	379
	380	my $mergedtsv = $c->as_tsv({mergetypes => [ 'spelling', 'orthographic' ] });
	381	my $t4 = Text::Tradition->new( input => 'Tabular',
	382	name => 'test4',
	383	string => $mergedtsv,
	384	sep_char => "\t" );
	385	is( scalar $t4->collation->readings, $READINGS - 2, "Reparsed TSV merge collation has fewer readings" );
	386	is( scalar $t4->collation->paths, $PATHS - 4, "Reparsed TSV merge collation has fewer paths" );
53ec2b6d	387
	388	# Test non-ASCII sigla
	389	my $t5 = Text::Tradition->new( input => 'Tabular',
	390	name => 'nonascii',
	391	file => 't/data/armexample.xlsx',
	392	excel => 'xlsx' );
	393	my $awittsv = $t5->collation->as_tsv({ noac => 1, ascii => 1 });
	394	my @awitlines = split( /\n/, $awittsv );
	395	like( $awitlines[0], qr/_A_5315622/, "Found ASCII sigil variant in TSV" );
16203db5	396	}
	397
	398
	399
	400	# =begin testing
	401	{
56eefa04	402	use Text::Tradition;
56eefa04	403
0e47f4f6	404	my $cxfile = 't/data/Collatex-16.xml';
	405	my $t = Text::Tradition->new(
	406	'name' => 'inline',
	407	'input' => 'CollateX',
	408	'file' => $cxfile,
	409	);
	410	my $c = $t->collation;
4633f9e4	411
b365fbae	412	# Make an svg
bfcbcecb	413	my $table = $c->alignment_table;
	414	ok( $c->has_cached_table, "Alignment table was cached" );
	415	is( $c->alignment_table, $table, "Cached table returned upon second call" );
b365fbae	416	$c->calculate_ranks;
bfcbcecb	417	is( $c->alignment_table, $table, "Cached table retained with no rank change" );
864ee4bf	418	$c->add_relationship( 'n13', 'n23', { type => 'repetition' } );
	419	is( $c->alignment_table, $table, "Alignment table unchanged after non-colo relationship add" );
	420	$c->add_relationship( 'n24', 'n23', { type => 'spelling' } );
	421	isnt( $c->alignment_table, $table, "Alignment table changed after colo relationship add" );
b365fbae	422	}
	423
	424
	425
	426	# =begin testing
	427	{
	428	use Text::Tradition;
	429
	430	my $cxfile = 't/data/Collatex-16.xml';
	431	my $t = Text::Tradition->new(
	432	'name' => 'inline',
	433	'input' => 'CollateX',
	434	'file' => $cxfile,
	435	);
	436	my $c = $t->collation;
0e47f4f6	437
d4b75f44	438	my @common = $c->calculate_common_readings();
	439	is( scalar @common, 8, "Found correct number of common readings" );
	440	my @marked = sort $c->common_readings();
	441	is( scalar @common, 8, "All common readings got marked as such" );
679f17e1	442	my @expected = qw/ n1 n11 n16 n19 n20 n5 n6 n7 /;
d4b75f44	443	is_deeply( \@marked, \@expected, "Found correct list of common readings" );
	444	}
	445
	446
	447
	448	# =begin testing
	449	{
	450	use Text::Tradition;
	451
	452	my $cxfile = 't/data/Collatex-16.xml';
	453	my $t = Text::Tradition->new(
	454	'name' => 'inline',
	455	'input' => 'CollateX',
	456	'file' => $cxfile,
	457	);
	458	my $c = $t->collation;
	459
679f17e1	460	is( $c->common_predecessor( 'n24', 'n23' )->id,
0e47f4f6	461	'n20', "Found correct common predecessor" );
679f17e1	462	is( $c->common_successor( 'n24', 'n23' )->id,
10e4b1ac	463	'__END__', "Found correct common successor" );
0e47f4f6	464
4e5a7b2c	465	is( $c->common_predecessor( 'n19', 'n17' )->id,
0e47f4f6	466	'n16', "Found correct common predecessor for readings on same path" );
679f17e1	467	is( $c->common_successor( 'n21', 'n10' )->id,
10e4b1ac	468	'__END__', "Found correct common successor for readings on same path" );
0e47f4f6	469	}
	470
	471
	472
	473
	474	1;