84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
|
# extract first title parts
title = rx_t.search(row["title"])
if title:
title = title.group(0).replace(" ", "%20")
# do a google search
html = ahttp.ajax("http://www.google.de/search?hl=de&q="+title, None)
# find first URL hit
url = rx_u.search(html)
if url:
row["homepage"] = ahttp.fix_url(url.group(1))
pass
#-----------------
|
|
|
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
|
# extract first title parts
title = rx_t.search(row["title"])
if title:
title = title.group(0).replace(" ", "%20")
# do a google search
html = ahttp.get("http://www.google.de/search?hl=de&q="+title, params={}, ajax=1)
# find first URL hit
url = rx_u.search(html)
if url:
row["homepage"] = ahttp.fix_url(url.group(1))
pass
#-----------------
|