crabfarm 0.0.12 → 0.0.13
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/crabfarm/modes/generator.rb +1 -1
- data/lib/crabfarm/templates/Crabfile.erb +67 -1
- data/lib/crabfarm/version.rb +1 -1
- metadata +1 -1
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 8c7539be24f102d8e0e7eb4961b2bb10678cfda2
|
4
|
+
data.tar.gz: 7c5fc40e8a2fa4a5b2474b1b8b34bc5287b88d2f
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: bc1848de3ad1f401834f17d3c7691e6cfa90f17939ec5cde588fc1fc933f25e4ecd6af3b1be43a26d42182bb4f008fbd6d42b68ffe23fcd195b396540bfb72db
|
7
|
+
data.tar.gz: 6b5a780eddbf1ca4cdf53b9d5ababfcd10825badda27f170e51f1e07ea22a12ff5aabea661f35166cf4f153955b9f13a589bbafe556efa845630a77cd2ce7ca1
|
@@ -30,7 +30,7 @@ module Crabfarm
|
|
30
30
|
path(_name, 'spec', 'snapshots', '.gitkeep').render('dot_gitkeep')
|
31
31
|
path(_name, 'spec', 'mementos', '.gitkeep').render('dot_gitkeep')
|
32
32
|
path(_name, 'spec', 'integration', '.gitkeep').render('dot_gitkeep')
|
33
|
-
path(_name, 'logs').render('dot_gitkeep')
|
33
|
+
path(_name, 'logs', '.gitkeep').render('dot_gitkeep')
|
34
34
|
end
|
35
35
|
end
|
36
36
|
|
@@ -1,4 +1,70 @@
|
|
1
|
+
|
2
|
+
# The default crawling dsl to use in states and parsers, can be overriden in each component using the `browser_dsl :dsl` modifier
|
3
|
+
# Available options are :surfer, :watir and :capybara
|
1
4
|
set_browser_dsl :surfer
|
2
|
-
set_driver :phantomjs
|
3
5
|
|
6
|
+
# Change the defaut output builder used in a state to generate the output document.
|
7
|
+
# Available options are :hash, :ostruct, :jbuilder
|
8
|
+
# set_output_builder :hash
|
9
|
+
|
10
|
+
# The path where every crawler log is stored.
|
4
11
|
set_log_path 'logs'
|
12
|
+
|
13
|
+
|
14
|
+
# General driver configuration
|
15
|
+
########################################
|
16
|
+
|
17
|
+
# The selected selenium driver to be using throughout the crawler
|
18
|
+
# Other options (may require instalation of additional libraries): :phantomjs, :remote, :chrome, :firefox
|
19
|
+
set_driver :phantomjs
|
20
|
+
|
21
|
+
# Set the selected webdriver capabilities (check the driver documentation for more details)
|
22
|
+
# set_driver_capabilities
|
23
|
+
|
24
|
+
# Set the browser window width
|
25
|
+
# set_driver_window_width 1280
|
26
|
+
|
27
|
+
# Set the browser window height
|
28
|
+
# set_driver_window_height 800
|
29
|
+
|
30
|
+
# Set the driver proxy address
|
31
|
+
# set_proxy
|
32
|
+
|
33
|
+
|
34
|
+
# Phantom launcher configuration
|
35
|
+
########################################
|
36
|
+
|
37
|
+
# The following parameters only apply if using the :phantomjs driver
|
38
|
+
|
39
|
+
# Make phantom load images when requesting documents
|
40
|
+
# set_phantom_load_images false
|
41
|
+
|
42
|
+
# Set phantom ssl mode, defaults to :any, posible options are sslv3, sslv2, tlsv1 or any
|
43
|
+
# set_phantom_ssl 'any'
|
44
|
+
|
45
|
+
# Set phantomjs executable location, by default crabfarm expects phantomjs to be included in enviroment
|
46
|
+
# set_phantom_bin_path 'phantomjs'
|
47
|
+
|
48
|
+
|
49
|
+
# Remote driver options
|
50
|
+
########################################
|
51
|
+
|
52
|
+
# The following parameters only apply if using the :remote driver
|
53
|
+
|
54
|
+
# Remote driver host
|
55
|
+
# set_driver_host www.myseleniumgrid.com
|
56
|
+
|
57
|
+
# Remote driver port
|
58
|
+
# set_driver_port 8080
|
59
|
+
|
60
|
+
# Remote driver response timeout, in seconds
|
61
|
+
# set_driver_remote_timeout 120
|
62
|
+
|
63
|
+
|
64
|
+
# Crabtrap launcher configuration
|
65
|
+
########################################
|
66
|
+
|
67
|
+
# The following parameters only apply when running the crabtrap proxy on specs
|
68
|
+
|
69
|
+
# Set the crabtrap executable location, by default crabfarm expects crabtrap to be included in enviroment
|
70
|
+
# set_crabtrap_bin_path 'crabtrap'
|
data/lib/crabfarm/version.rb
CHANGED