html-get 2.13.5 → 2.13.6
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/package.json +2 -2
- package/src/auto-domains.json +24 -24
- package/src/index.js +3 -3
package/package.json
CHANGED
|
@@ -2,7 +2,7 @@
|
|
|
2
2
|
"name": "html-get",
|
|
3
3
|
"description": "Get the HTML from any website, using prerendering when is necessary.",
|
|
4
4
|
"homepage": "https://nicedoc.com/microlinkhq/html-get",
|
|
5
|
-
"version": "2.13.
|
|
5
|
+
"version": "2.13.6",
|
|
6
6
|
"main": "src/index.js",
|
|
7
7
|
"bin": {
|
|
8
8
|
"html-get": "bin/index.js"
|
|
@@ -29,6 +29,7 @@
|
|
|
29
29
|
"request"
|
|
30
30
|
],
|
|
31
31
|
"dependencies": {
|
|
32
|
+
"@kikobeats/time-span": "~1.0.2",
|
|
32
33
|
"@metascraper/helpers": "~5.37.1",
|
|
33
34
|
"cheerio": "~1.0.0-rc.12",
|
|
34
35
|
"css-url-regex": "~4.0.0",
|
|
@@ -43,7 +44,6 @@
|
|
|
43
44
|
"p-cancelable": "~2.1.0",
|
|
44
45
|
"p-retry": "~4.6.0",
|
|
45
46
|
"replace-string": "~3.1.0",
|
|
46
|
-
"time-span": "~4.0.0",
|
|
47
47
|
"top-sites": "~1.1.169",
|
|
48
48
|
"write-json-file": "~4.3.0"
|
|
49
49
|
},
|
package/src/auto-domains.json
CHANGED
|
@@ -2,25 +2,25 @@
|
|
|
2
2
|
[
|
|
3
3
|
[
|
|
4
4
|
"domainWithoutSuffix",
|
|
5
|
-
"
|
|
5
|
+
"youtube"
|
|
6
6
|
]
|
|
7
7
|
],
|
|
8
8
|
[
|
|
9
9
|
[
|
|
10
10
|
"domainWithoutSuffix",
|
|
11
|
-
"
|
|
11
|
+
"google"
|
|
12
12
|
]
|
|
13
13
|
],
|
|
14
14
|
[
|
|
15
15
|
[
|
|
16
16
|
"domainWithoutSuffix",
|
|
17
|
-
"
|
|
17
|
+
"microsoft"
|
|
18
18
|
]
|
|
19
19
|
],
|
|
20
20
|
[
|
|
21
21
|
[
|
|
22
22
|
"domainWithoutSuffix",
|
|
23
|
-
"
|
|
23
|
+
"wordpress"
|
|
24
24
|
]
|
|
25
25
|
],
|
|
26
26
|
[
|
|
@@ -44,13 +44,13 @@
|
|
|
44
44
|
[
|
|
45
45
|
[
|
|
46
46
|
"domainWithoutSuffix",
|
|
47
|
-
"
|
|
47
|
+
"github"
|
|
48
48
|
]
|
|
49
49
|
],
|
|
50
50
|
[
|
|
51
51
|
[
|
|
52
52
|
"domainWithoutSuffix",
|
|
53
|
-
"
|
|
53
|
+
"vimeo"
|
|
54
54
|
]
|
|
55
55
|
],
|
|
56
56
|
[
|
|
@@ -62,7 +62,7 @@
|
|
|
62
62
|
[
|
|
63
63
|
[
|
|
64
64
|
"domainWithoutSuffix",
|
|
65
|
-
"
|
|
65
|
+
"bbc"
|
|
66
66
|
]
|
|
67
67
|
],
|
|
68
68
|
[
|
|
@@ -80,43 +80,43 @@
|
|
|
80
80
|
[
|
|
81
81
|
[
|
|
82
82
|
"domainWithoutSuffix",
|
|
83
|
-
"
|
|
83
|
+
"nytimes"
|
|
84
84
|
]
|
|
85
85
|
],
|
|
86
86
|
[
|
|
87
87
|
[
|
|
88
88
|
"domainWithoutSuffix",
|
|
89
|
-
"
|
|
89
|
+
"huffingtonpost"
|
|
90
90
|
]
|
|
91
91
|
],
|
|
92
92
|
[
|
|
93
93
|
[
|
|
94
94
|
"domainWithoutSuffix",
|
|
95
|
-
"
|
|
95
|
+
"soundcloud"
|
|
96
96
|
]
|
|
97
97
|
],
|
|
98
98
|
[
|
|
99
99
|
[
|
|
100
100
|
"domainWithoutSuffix",
|
|
101
|
-
"
|
|
101
|
+
"spotify"
|
|
102
102
|
]
|
|
103
103
|
],
|
|
104
104
|
[
|
|
105
105
|
[
|
|
106
106
|
"domainWithoutSuffix",
|
|
107
|
-
"
|
|
107
|
+
"pinterest"
|
|
108
108
|
]
|
|
109
109
|
],
|
|
110
110
|
[
|
|
111
111
|
[
|
|
112
112
|
"domainWithoutSuffix",
|
|
113
|
-
"
|
|
113
|
+
"telegraph"
|
|
114
114
|
]
|
|
115
115
|
],
|
|
116
116
|
[
|
|
117
117
|
[
|
|
118
118
|
"domainWithoutSuffix",
|
|
119
|
-
"
|
|
119
|
+
"techcrunch"
|
|
120
120
|
]
|
|
121
121
|
],
|
|
122
122
|
[
|
|
@@ -128,13 +128,7 @@
|
|
|
128
128
|
[
|
|
129
129
|
[
|
|
130
130
|
"domainWithoutSuffix",
|
|
131
|
-
"
|
|
132
|
-
]
|
|
133
|
-
],
|
|
134
|
-
[
|
|
135
|
-
[
|
|
136
|
-
"domainWithoutSuffix",
|
|
137
|
-
"techcrunch"
|
|
131
|
+
"eventbrite"
|
|
138
132
|
]
|
|
139
133
|
],
|
|
140
134
|
[
|
|
@@ -146,19 +140,19 @@
|
|
|
146
140
|
[
|
|
147
141
|
[
|
|
148
142
|
"domainWithoutSuffix",
|
|
149
|
-
"
|
|
143
|
+
"yelp"
|
|
150
144
|
]
|
|
151
145
|
],
|
|
152
146
|
[
|
|
153
147
|
[
|
|
154
148
|
"domainWithoutSuffix",
|
|
155
|
-
"
|
|
149
|
+
"engadget"
|
|
156
150
|
]
|
|
157
151
|
],
|
|
158
152
|
[
|
|
159
153
|
[
|
|
160
154
|
"domainWithoutSuffix",
|
|
161
|
-
"
|
|
155
|
+
"theverge"
|
|
162
156
|
]
|
|
163
157
|
],
|
|
164
158
|
[
|
|
@@ -203,6 +197,12 @@
|
|
|
203
197
|
"ghost"
|
|
204
198
|
]
|
|
205
199
|
],
|
|
200
|
+
[
|
|
201
|
+
[
|
|
202
|
+
"domainWithoutSuffix",
|
|
203
|
+
"giphy"
|
|
204
|
+
]
|
|
205
|
+
],
|
|
206
206
|
[
|
|
207
207
|
[
|
|
208
208
|
"domainWithoutSuffix",
|
package/src/index.js
CHANGED
|
@@ -2,10 +2,10 @@
|
|
|
2
2
|
|
|
3
3
|
const { parseUrl, isMediaUrl } = require('@metascraper/helpers')
|
|
4
4
|
const debug = require('debug-logfmt')('html-get')
|
|
5
|
+
const timeSpan = require('@kikobeats/time-span')
|
|
5
6
|
const PCancelable = require('p-cancelable')
|
|
6
7
|
const { AbortError } = require('p-retry')
|
|
7
8
|
const htmlEncode = require('html-encode')
|
|
8
|
-
const timeSpan = require('time-span')
|
|
9
9
|
const got = require('got')
|
|
10
10
|
|
|
11
11
|
const autoDomains = require('./auto-domains')
|
|
@@ -203,7 +203,7 @@ module.exports = PCancelable.fn(
|
|
|
203
203
|
const toEncode = htmlEncode(encoding)
|
|
204
204
|
const reqMode = getMode(targetUrl, { prerender })
|
|
205
205
|
|
|
206
|
-
const
|
|
206
|
+
const duration = timeSpan()
|
|
207
207
|
|
|
208
208
|
const promise = getContent(targetUrl, reqMode, {
|
|
209
209
|
getBrowserless,
|
|
@@ -218,7 +218,7 @@ module.exports = PCancelable.fn(
|
|
|
218
218
|
|
|
219
219
|
const { mode, ...payload } = await promise
|
|
220
220
|
|
|
221
|
-
return Object.assign(payload, { stats: { mode, timing:
|
|
221
|
+
return Object.assign(payload, { stats: { mode, timing: duration() } })
|
|
222
222
|
}
|
|
223
223
|
)
|
|
224
224
|
|