Micro Web Crawler in PHP & Manticore
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

222 lines
5.4 KiB

{
"manticore":
{
"server":
{
"host":"127.0.0.1",
"port":9308
},
"index":
{
"document":{
"name":"yo_gemini",
"settings":
{
"morphology":"stem_cz,stem_enru",
"index_exact_words":1,
"html_strip":1,
"min_word_len":3,
"min_prefix_len":3
}
}
}
},
"gui":
{
"pagination":
{
"limit":20
},
"search":
{
"index":
{
"request":
{
"url":{
"enabled":false,
"regex":"/^gemini:\\/\\/.*/ui"
}
}
},
"extended":
{
"enabled":false
},
"options":
{
"field_weights":
{
"url":100,
"title":200,
"description":300,
"keywords":400,
"body":500
}
},
"highlight":
{
"fields":
[
"url",
"h1",
"h2",
"h3",
"body"
],
"options":
{
"around":5,
"limit":140
}
}
},
"index":
{
"enabled":true
},
"captcha":
{
"length":5,
"phrase":"0123456789abcdef",
"background":
{
"r":46,
"g":52,
"b":54
}
}
},
"cli":
{
"document":
{
"crawl":
{
"debug":
{
"level":
{
"notice":true,
"warning":true,
"error":true
}
},
"network":
{
"check":
{
"enabled":false,
"timeout":5,
"socket":
{
"1.1.1.1":80
}
}
},
"connection":
{
"timeout":3,
"length":10485760
},
"queue":
{
"limit":1,
"delay":1
},
"url":
{
"external":true,
"regex":"/^gemini:\\/\\/.*/ui",
"skip":
{
"stripos":
[
"?"
]
}
},
"snap":
{
"enabled":true,
"timeout":2592000
}
}
}
},
"snap":
{
"storage":
{
"tmp":{
"directory":"storage/tmp/snap"
},
"local":{
"enabled":false,
"directory":"storage/snap",
"size":
{
"max":1048576
},
"meta":
{
"stripos":
[
"text/gemini",
"image/"
]
},
"url":
{
"stripos":
[
"gemini://"
]
}
},
"remote":
{
"ftp":
[
{
"enabled":false,
"connection":
{
"port":21,
"host":"",
"username":"",
"password":"",
"directory":"/snap/yo",
"timeout":30,
"passive":true,
"attempts":
{
"limit":0,
"delay":60
}
},
"size":
{
"max":1048576
},
"meta":
{
"stripos":
[
"text/gemini",
"image/"
]
},
"url":
{
"stripos":
[
"gemini://"
]
}
}
]
}
}
}
}