canon/internal/importer/importer_test.go

package importer

import (
	"fmt"
	"reflect"
	"testing"

	"github.com/cdnutter/canon/internal/providers"
)

func TestParseText(t *testing.T) {
	got, err := Parse([]byte("1. Heat (1995)\n2) 2001: A Space Odyssey\n\n4 - Alien [1979]\n"), "list.txt")
	if err != nil {
		t.Fatal(err)
	}
	want := []Row{{Place: 1, Title: "Heat", Year: "1995"}, {Place: 2, Title: "2001: A Space Odyssey"}, {Place: 4, Title: "Alien", Year: "1979"}}
	if !reflect.DeepEqual(got, want) {
		t.Fatalf("got %+v", got)
	}
	// Not every line is numbered, so nothing is stripped; line numbers are
	// the places, and a blank line leaves one empty.
	got, _ = Parse([]byte("2001: A Space Odyssey\nHeat\n\n1917\n"), "x.txt")
	if got[0].Title != "2001: A Space Odyssey" || got[2].Title != "1917" || got[2].Place != 4 {
		t.Fatalf("got %+v", got)
	}
	// Numbers win over line numbers when every line has one.
	got, _ = Parse([]byte("1. Super Mario Sunshine\n38. Super Mario World\n"), "games.txt")
	if got[0].Place != 1 || got[1].Place != 38 || got[1].Title != "Super Mario World" {
		t.Fatalf("got %+v", got)
	}
	// A repeated number takes the next free place, keeping the file's order.
	got, _ = Parse([]byte("3. A\n3. B\n1. C\n"), "x.txt")
	if fmt.Sprint(got) != fmt.Sprint([]Row{{Place: 1, Title: "C"}, {Place: 3, Title: "A"}, {Place: 4, Title: "B"}}) {
		t.Fatalf("got %+v", got)
	}
}

func TestParseCSV(t *testing.T) {
	csv := "\xef\xbb\xbfRank,Title,Year,Director,Why\n2,Alien,1979,Ridley Scott,\"Tense, perfect\"\n1,Heat,1995,Michael Mann,\n"
	got, err := Parse([]byte(csv), "films.csv")
	if err != nil {
		t.Fatal(err)
	}
	want := []Row{
		{Place: 1, Title: "Heat", Year: "1995", Subtitle: "Michael Mann"},
		{Place: 2, Title: "Alien", Year: "1979", Subtitle: "Ridley Scott", Note: "Tense, perfect"},
	}
	if !reflect.DeepEqual(got, want) {
		t.Fatalf("got %+v", got)
	}
	// Canon's own export: one column, no header, a blank line for an empty
	// place.
	got, _ = Parse([]byte("\xef\xbb\xbfHeat\r\n\r\n\"Crouching Tiger, Hidden Dragon\"\r\n"), "films.csv")
	if len(got) != 2 || got[1].Title != "Crouching Tiger, Hidden Dragon" || got[1].Place != 3 {
		t.Fatalf("got %+v", got)
	}
	// With a header, the first row after it is #1.
	got, _ = Parse([]byte("Title\nHeat\n\nAlien\n"), "x.csv")
	if got[0].Place != 1 || got[1].Place != 3 {
		t.Fatalf("got %+v", got)
	}
	// Headerless rank,title,year.
	got, _ = Parse([]byte("1,Heat,1995\n5,Alien,1979\n"), "x.csv")
	if got[1] != (Row{Place: 5, Title: "Alien", Year: "1979"}) {
		t.Fatalf("got %+v", got)
	}
	if _, err := Parse([]byte("\n \n"), "x.txt"); err != ErrEmpty {
		t.Fatalf("empty file: %v", err)
	}
}

func res(title, year, sub string) providers.Result {
	return providers.Result{Provider: "t", ProviderID: title + year + sub, Title: title, Year: year, Subtitle: sub}
}

func TestRank(t *testing.T) {
	cases := []struct {
		name     string
		row      Row
		category string
		results  []providers.Result
		status   string
		top      string
	}{
		{"exact and alone", Row{Title: "Alien"}, "movie", []providers.Result{res("Alien", "1979", ""), res("Aliens", "1986", "")}, Matched, "Alien1979"},
		{"remake, no year", Row{Title: "Heat"}, "movie", []providers.Result{res("Heat", "1995", ""), res("Heat", "1986", "")}, Review, "Heat1995"},
		{"remake, year given", Row{Title: "Heat", Year: "1986"}, "movie", []providers.Result{res("Heat", "1995", ""), res("Heat", "1986", "")}, Matched, "Heat1986"},
		{"loose title", Row{Title: "Star Wars"}, "movie", []providers.Result{res("Star Wars: A New Hope", "1977", "")}, Review, "Star Wars: A New Hope1977"},
		{"punctuation and accents", Row{Title: "amelie"}, "movie", []providers.Result{res("Amélie", "2001", "")}, Matched, "Amélie2001"},
		{"wikipedia qualifier", Row{Title: "The Matrix", Year: "1999"}, "custom", []providers.Result{res("The Matrix (1999 film)", "1999", "")}, Matched, "The Matrix (1999 film)1999"},
		{"artist - title", Row{Title: "Radiohead - Creep"}, "song", []providers.Result{res("Creep", "1992", "TLC"), res("Creep", "1992", "Radiohead")}, Matched, "Creep1992Radiohead"},
		{"title, no artist", Row{Title: "Creep"}, "song", []providers.Result{res("Creep", "1992", "TLC"), res("Creep", "1992", "Radiohead")}, Review, "Creep1992TLC"},
		{"same album twice", Row{Title: "OK Computer"}, "album", []providers.Result{res("OK Computer", "1997", "Radiohead"), res("OK Computer", "1997", "Radiohead")}, Matched, "OK Computer1997Radiohead"},
		{"wrong year", Row{Title: "Alien", Year: "1990"}, "movie", []providers.Result{res("Alien", "1979", "")}, Review, "Alien1979"},
		{"exact beats a bracketed version", Row{Title: "Radiohead - Creep"}, "song", []providers.Result{res("Creep (Acoustic)", "", "Radiohead"), res("Creep", "", "Radiohead")}, Matched, "CreepRadiohead"},
		{"only bracketed versions", Row{Title: "Hey Jude"}, "song", []providers.Result{res("Hey Jude (Remastered 2015)", "", "The Beatles")}, Matched, "Hey Jude (Remastered 2015)The Beatles"},
		{"year beats no year", Row{Title: "The Matrix", Year: "1999"}, "custom", []providers.Result{res("The Matrix", "", "American media franchise"), res("The Matrix", "1999", "Film by the Wachowskis")}, Matched, "The Matrix1999Film by the Wachowskis"},
		{"artist confirmed beats a literal title", Row{Title: "Radiohead - Creep"}, "song", []providers.Result{res("Radiohead - Creep", "", "Klangsberg"), res("Creep", "", "Radiohead")}, Matched, "CreepRadiohead"},
		{"custom needs confirming", Row{Title: "Heat"}, "custom", []providers.Result{res("Heat", "", "Type of energy transfer")}, Review, "HeatType of energy transfer"},
		{"nothing", Row{Title: "zzz"}, "movie", nil, None, ""},
	}
	for _, c := range cases {
		status, scored := Rank(c.row, c.category, c.results)
		if status != c.status {
			t.Errorf("%s: status %s, want %s", c.name, status, c.status)
		}
		if len(scored) > 0 && scored[0].ProviderID != c.top {
			t.Errorf("%s: top %s, want %s", c.name, scored[0].ProviderID, c.top)
		}
	}
}