-
Notifications
You must be signed in to change notification settings - Fork 0
/
crawler.go
214 lines (180 loc) · 4.53 KB
/
crawler.go
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
package wikicrawl
import (
"io"
"net/http"
"net/http/cookiejar"
"net/url"
"strings"
"time"
log "github.com/Sirupsen/logrus"
"golang.org/x/net/html"
)
// Wikimedia namespaces to ignore.
var ignore = []string{
"User:", "User_talk:",
"Help:", "Help_talk:",
"Talk:", "File_talk:", "Category_talk:",
"Create_Article", "Special:",
}
// Results of crawling wiki.
// 1. Visited: List of visited links.
// 2. Broken: List of Broken links.
type CrawlResult struct {
Visited LinkSet
Broken LinkSet
}
// Crawler type holds state and methods for exploring a wiki.
type Crawler struct {
base *url.URL
Client *http.Client
}
// Simple constructor for Crawler type.
func NewCrawler(base Link, session string) *Crawler {
c := new(Crawler)
result, err := url.Parse(base)
if err != nil {
panic(err)
}
c.base = result
jar, _ := cookiejar.New(nil)
cookie := &http.Cookie{
Name: "wikidb2_is__session",
Value: session,
Path: "/",
Domain: c.base.Host,
}
cookies := []*http.Cookie{cookie}
jar.SetCookies(c.base, cookies)
c.Client = &http.Client{
Timeout: time.Second * 10,
Jar: jar,
}
return c
}
// Crawls all valid links that can be found from the initial url.
func (c *Crawler) Crawl(source Link) *CrawlResult {
queue := NewWorkQueue(*c, 1000)
queue.Start(10)
queue.AddWork(source)
queue.Wait()
return queue.Result
}
func (c *Crawler) FollowLink(source Link, queue *WorkQueue) {
// Avoid duplicate visits.
if ok := queue.Result.Visited.Add(source); !ok {
return
}
log.WithFields(log.Fields{"source": source}).Debug("Crawling new url")
resp, err := c.Client.Get(source)
if err != nil {
log.WithFields(log.Fields{
"err": err,
}).Warn("GET returned with error")
queue.Result.Broken.Add(source)
return
}
defer resp.Body.Close()
if resp.StatusCode != 200 {
log.WithFields(log.Fields{
"source": source,
"status": resp.Status,
}).Warn("GET returned with non 200 response")
queue.Result.Broken.Add(source)
return
}
if source != resp.Request.URL.String() {
log.WithFields(log.Fields{
"requested": source,
"redirect": resp.Request.URL,
}).Warn("Redirect detected.")
if ok := queue.Result.Visited.Add(resp.Request.URL.String()); !ok {
return
}
}
for raw := range ParseLinks(resp.Body).Set {
result, err := url.Parse(raw)
if err != nil {
queue.Result.Broken.Add(raw)
continue
}
href := NormalizeUrl(result, c.base)
if c.ValidateLink(href) && !queue.Result.Visited.Contains(href.String()) {
queue.AddWork(href.String())
} else {
log.WithFields(log.Fields{"href": href}).Debug("Skipping link.")
}
}
}
// Validates if link should be followed.
//
// 1. Only crawls internal links.
// 2. Skips trivial Wikimedia namespaces.
func (c *Crawler) ValidateLink(link *url.URL) bool {
if !strings.Contains(link.String(), c.base.String()) {
return false
}
if title := WikiPageTitle(link); len(title) > 0 {
for _, trivial := range ignore {
if strings.HasPrefix(title, trivial) {
return false
}
}
}
return true
}
// Parse WikiMedia page title with namespace.
// WikiMedia short url's not supported.
func WikiPageTitle(link *url.URL) string {
query, err := url.ParseQuery(link.RawQuery)
if err != nil {
panic(err)
}
if title, found := query["title"]; found {
return title[0]
}
return ""
}
// Normalize a url to facilitate comparison.
//
// 1. Resolve url from known base (/relative => http://base/relative)
// 2. Cleanup query by filtering unnecessary parameters
// 3. Remove any URL fragment (#junk)
// 4. Force protocol to match base
// 5. Unify case of host and protocol
func NormalizeUrl(link *url.URL, base *url.URL) *url.URL {
clean := base.ResolveReference(link)
if title := WikiPageTitle(clean); len(title) != 0 {
clean.RawQuery = url.Values{"title": []string{title}}.Encode()
}
clean.Fragment = ""
clean.Scheme = strings.ToLower(base.Scheme)
clean.Host = strings.ToLower(clean.Host)
log.WithFields(log.Fields{
"base": base.String(),
"original": link.String(),
"cleaned": clean.String(),
}).Debug("Normalized URL.")
return clean
}
// Parses HTML and returns a list of all href values found.
func ParseLinks(reader io.Reader) LinkSet {
links := NewLinkSet()
z := html.NewTokenizer(reader)
for {
tokenType := z.Next()
switch {
case tokenType == html.ErrorToken:
return links
case tokenType == html.StartTagToken || tokenType == html.SelfClosingTagToken:
token := z.Token()
if token.Data == "a" {
for _, attr := range token.Attr {
if attr.Key == "href" {
links.Add(attr.Val)
break
}
}
}
}
}
}