1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331

package main

import (
	"math/bits"
	"net/http"
	"net/http/httptest"
	"net/url"
	"path/filepath"
	"slices"
	"strings"
	"testing"
)

func TestNormaliseUrl(t *testing.T) {
	for in, want := range map[string]string{
		"https://www.Example.org/":         "example.org",
		"http://example.org/blog/":         "example.org/blog",
		"https://example.org/x?a=1":        "example.org/x?a=1",
		"  https://sub.example.org/a/b/  ": "sub.example.org/a/b",
	} {
		if got := normaliseUrl(in); got != want {
			t.Errorf("normaliseUrl(%q) = %q, want %q", in, got, want)
		}
	}
}

func TestUrlSimhash(t *testing.T) {
	d := func(a, b string) int { return bits.OnesCount64(urlSimhash(a) ^ urlSimhash(b)) }
	if d("https://example.org/", "http://www.example.org") != 0 {
		t.Error("scheme, www. and trailing slash should not count")
	}
	near := d("https://www.youtube.com/channel/UCabcdefghijk", "https://www.youtube.com/channel/UCzyxwvutsrqp")
	far := d("https://www.youtube.com/channel/UCabcdefghijk", "https://blog.example.org/")
	if near >= far {
		t.Errorf("two YouTube channels are %d bits apart, YouTube and a blog %d", near, far)
	}
}

func TestFarthestPoints(t *testing.T) {
	// Three clusters, of 50, 5 and 1, far apart from each other.
	var hashes []uint64
	for i := range 50 {
		hashes = append(hashes, uint64(i&3)) // a few bits from 0
	}
	for i := range 5 {
		hashes = append(hashes, ^uint64(0)^uint64(i&1))
	}
	hashes = append(hashes, 0x00000000ffffffff)

	picks, dists := farthestPoints(hashes, 3)
	if len(picks) != 3 {
		t.Fatalf("%d picks, want 3", len(picks))
	}
	cluster := func(i int) int {
		switch {
		case i < 50:
			return 0
		case i < 55:
			return 1
		}
		return 2
	}
	var got []int
	for _, p := range picks {
		got = append(got, cluster(p))
	}
	slices.Sort(got)
	if !slices.Equal(got, []int{0, 1, 2}) {
		t.Errorf("picked from clusters %v, want one from each", got)
	}
	if dists[0] != 64 || dists[1] < dists[2] {
		t.Errorf("distances %v should start at 64 and not grow", dists)
	}

	// Asking for more than there are gives all of them, each once.
	picks, _ = farthestPoints(hashes, 1000)
	slices.Sort(picks)
	if n := len(picks); n != len(hashes) || len(slices.Compact(picks)) != n {
		t.Errorf("%d picks of %d, or some twice", n, len(hashes))
	}
}

// The steps against a database, with a fake site.
func TestSampleAndFetch(t *testing.T) {
	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
		if r.Header.Get("User-Agent") != userAgent {
			t.Errorf("User-Agent %q", r.Header.Get("User-Agent"))
		}
		switch r.URL.Path {
		case "/big":
			w.Write([]byte(strings.Repeat("x", fetchLimit+10)))
		case "/moved":
			http.Redirect(w, r, "/a", http.StatusFound)
		default:
			w.Header().Set("Content-Type", "text/html")
			w.Write([]byte("<html><head><title>" + r.URL.Path + "</title></head></html>"))
		}
	}))
	defer srv.Close()

	db, err := openDB(filepath.Join(t.TempDir(), "s.db"))
	if err != nil {
		t.Fatal(err)
	}
	defer db.Close()
	if err = storeListingPage(db, 0, []entry{
		{Handle: "a", SiteUrl: srv.URL + "/a"},
		{Handle: "a2", SiteUrl: srv.URL + "/a/"}, // the same site as a
		{Handle: "big", SiteUrl: srv.URL + "/big"},
		{Handle: "moved", SiteUrl: srv.URL + "/moved"},
		{Handle: "down", SiteUrl: "http://127.0.0.1:1/"},
	}); err != nil {
		t.Fatal(err)
	}

	var log strings.Builder
	if err = sample(db, 10, &log); err != nil {
		t.Fatal(err)
	}
	var n int
	db.QueryRow(`SELECT COUNT(*) FROM sample`).Scan(&n)
	if n != 4 {
		t.Errorf("%d sampled, want 4 (a2 is the same site as a)", n)
	}
	if err = sample(db, 10, &log); err == nil {
		t.Error("a second sample should be refused")
	}

	if err = fetch(db, 2, &log); err != nil {
		t.Fatal(err)
	}
	type res struct {
		status    *int
		finalUrl  *string
		size      int
		truncated *bool
		errText   *string
	}
	got := map[string]res{}
	rows, err := db.Query(`SELECT handle, status, final_url, length(body), truncated, error FROM fetch`)
	if err != nil {
		t.Fatal(err)
	}
	for rows.Next() {
		var h string
		var r res
		var size *int
		if err = rows.Scan(&h, &r.status, &r.finalUrl, &size, &r.truncated, &r.errText); err != nil {
			t.Fatal(err)
		}
		if size != nil {
			r.size = *size
		}
		got[h] = r
	}
	rows.Close()
	if len(got) != 4 {
		t.Fatalf("%d fetched, want 4: %v", len(got), got)
	}
	if r := got["big"]; r.size != fetchLimit || r.truncated == nil || !*r.truncated {
		t.Errorf("big: %d bytes, truncated %v", r.size, r.truncated)
	}
	if r := got["moved"]; r.finalUrl == nil || *r.finalUrl != srv.URL+"/a" || *r.status != 200 {
		t.Errorf("moved: %+v", r)
	}
	if r := got["down"]; r.errText == nil || r.status != nil {
		t.Errorf("down: %+v", r)
	}

	// Fetching again fetches nothing.
	log.Reset()
	if err = fetch(db, 2, &log); err != nil {
		t.Fatal(err)
	}
	if !strings.HasPrefix(log.String(), "0 pages to fetch") {
		t.Errorf("second fetch: %q", log.String())
	}
}

func TestRules(t *testing.T) {
	page := `<html><head>
		<link rel="alternate" type="application/atom+xml" href="https://other.example/atom">
		<link rel="alternate" type="application/rss+xml" href="/rss">
		<link rel="Alternate home" type="application/feed+json; charset=utf-8" href="/json">
		<link rel="alternate" type="application/rss+xml" href="/comments/feed">
		</head><body>
		<link rel="alternate" type="application/rss+xml" href="/late">
		<a href="/feed/">RSS</a><a href="/about">About</a>
		</body></html>`
	s := scanPage([]byte(page))
	if !s.headEnded || !s.bodyReached || len(s.links) != 5 || s.links[4].inHead {
		t.Fatalf("scan: %+v", s)
	}
	if len(s.anchors) != 1 || s.anchors[0].href != "/feed/" {
		t.Errorf("anchors: %+v", s.anchors)
	}
	var head []scannedLink
	for _, l := range s.links {
		if l.inHead {
			head = append(head, l)
		}
	}
	base, _ := url.Parse("https://blog.example/")
	for name, c := range map[string]struct{ got, want string }{
		"upstream":  {upstreamRule(head), "/rss"},
		"first":     {firstFeedRule(head), "https://other.example/atom"},
		"same host": {sameHostRule(head, base), "/rss"},
	} {
		if c.got != c.want {
			t.Errorf("%s rule: %q, want %q", name, c.got, c.want)
		}
	}
	if got := feedKey(base, "/feed/?utm_source=x#top"); got != "blog.example/feed" {
		t.Errorf("feedKey: %q", got)
	}
}

// A page named after its newest post, as apenwarr.ca's is, and its feed,
// through fetch, feeds and analyze.
func TestFeedsAndMeta(t *testing.T) {
	const rss = `<?xml version="1.0"?><rss version="2.0"><channel>
		<title> apenwarr </title><description>apenwarr - NITLog</description>
		<item><title>
		thundersnap 0.01: an undo button for everything
		</title><link>https://example.org/1</link></item>
		<item><title>apenwarr</title><link>https://example.org/2</link></item>
		</channel></rss>`
	var feedAccepts []string
	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
		if r.Header.Get("User-Agent") != userAgent {
			t.Errorf("User-Agent %q", r.Header.Get("User-Agent"))
		}
		switch r.URL.Path {
		case "/post":
			w.Header().Set("Content-Type", "text/html")
			w.Write([]byte(`<html><head><link rel="alternate" type="application/rss+xml" href="/rss?x=1">
				<title>thundersnap 0.01: an undo button for everything - apenwarr</title></head></html>`))
		case "/plain":
			w.Header().Set("Content-Type", "text/html")
			w.Write([]byte(`<html><head><link rel="alternate" type="application/rss+xml" href="/rss2">
				<title>apenwarr</title><meta name="description" content="A log"></head></html>`))
		case "/nofeed":
			w.Header().Set("Content-Type", "text/html")
			w.Write([]byte(`<html><head><title>No feed</title></head></html>`))
		case "/rss", "/rss2":
			feedAccepts = append(feedAccepts, r.Header.Get("Accept"))
			w.Header().Set("Content-Type", "application/rss+xml")
			w.Write([]byte(rss))
		default:
			http.NotFound(w, r)
		}
	}))
	defer srv.Close()

	db, err := openDB(filepath.Join(t.TempDir(), "s.db"))
	if err != nil {
		t.Fatal(err)
	}
	defer db.Close()
	if err = storeListingPage(db, 0, []entry{
		{Handle: "post", SiteUrl: srv.URL + "/post"},
		{Handle: "plain", SiteUrl: srv.URL + "/plain"},
		{Handle: "nofeed", SiteUrl: srv.URL + "/nofeed"},
	}); err != nil {
		t.Fatal(err)
	}
	var log strings.Builder
	for _, step := range []func() error{
		func() error { return sample(db, 10, &log) },
		func() error { return fetch(db, 2, &log) },
		func() error { return feeds(db, 2, &log) },
		func() error { return analyze(db, &log) },
		func() error { return analyze(db, &log) }, // again, over its own rows
	} {
		if err = step(); err != nil {
			t.Fatalf("%v; log:\n%s", err, log.String())
		}
	}
	if len(feedAccepts) != 2 || !strings.HasPrefix(feedAccepts[0], "application/atom+xml") {
		t.Errorf("feed requests asked for %q", feedAccepts)
	}

	type meta struct {
		pageTitle, pageDesc string
		feedUrl, feedTitle  *string
		feedDesc            *string
		pageIsItem          *bool
	}
	got := map[string]meta{}
	rows, err := db.Query(`SELECT handle, page_title, page_desc, feed_url, feed_title, feed_desc, page_is_item FROM meta`)
	if err != nil {
		t.Fatal(err)
	}
	for rows.Next() {
		var h string
		var m meta
		if err = rows.Scan(&h, &m.pageTitle, &m.pageDesc, &m.feedUrl, &m.feedTitle, &m.feedDesc, &m.pageIsItem); err != nil {
			t.Fatal(err)
		}
		got[h] = m
	}
	rows.Close()

	post := got["post"]
	if post.feedUrl == nil || *post.feedUrl != srv.URL+"/rss" {
		t.Errorf("post: feed URL %v, want the query trimmed", post.feedUrl)
	}
	if post.feedTitle == nil || *post.feedTitle != "apenwarr" || post.feedDesc == nil || *post.feedDesc != "apenwarr - NITLog" {
		t.Errorf("post: feed title %v, description %v", post.feedTitle, post.feedDesc)
	}
	if post.pageIsItem == nil || !*post.pageIsItem {
		t.Errorf("post: page title %q not taken for an item's", post.pageTitle)
	}
	// The page is named like the site, and an item is too: that is no post.
	plain := got["plain"]
	if plain.pageIsItem == nil || *plain.pageIsItem || plain.pageDesc != "A log" {
		t.Errorf("plain: %+v", plain)
	}
	if nf, ok := got["nofeed"]; !ok || nf.feedUrl != nil || nf.pageIsItem != nil || nf.pageTitle != "No feed" {
		t.Errorf("nofeed: %+v, present %v", nf, ok)
	}

	// Fetching again fetches nothing.
	log.Reset()
	if err = feeds(db, 2, &log); err != nil {
		t.Fatal(err)
	}
	if !strings.HasPrefix(log.String(), "0 feeds to fetch") {
		t.Errorf("second feeds: %q", log.String())
	}
}