tanpengsccd/AV_Data_Capture
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update javdb.py

Browse Source
This commit is contained in:
mo_yy 2020-01-10 18:25:35 +08:00 committed by GitHub
parent 9c71e5976a
commit 3545705283
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 4 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
javdb.py
View File

@ -105,6 +105,7 @@ def getOutline(htmlcode):
def main(number): def main(number):
number = number.upper()
try: try:
a = get_html('https://javdb.com/search?q=' + number + '&f=all').replace(u'\xa0', u' ') a = get_html('https://javdb.com/search?q=' + number + '&f=all').replace(u'\xa0', u' ')
html = etree.fromstring(a, etree.HTMLParser()) # //table/tr[1]/td[1]/text() html = etree.fromstring(a, etree.HTMLParser()) # //table/tr[1]/td[1]/text()
@ -114,7 +115,7 @@ def main(number):
'actor': getActor(b), 'actor': getActor(b),
'title': getTitle(b).replace("\\n", '').replace(' ', '').replace(getActor(a), '').replace(getNum(a), 'title': getTitle(b).replace("\\n", '').replace(' ', '').replace(getActor(a), '').replace(getNum(a),
'').replace( '').replace(
'无码', '').replace('有码', '').lstrip(' ').replace(number, ''), '无码', '').replace('有码', '').lstrip(' ').replace(number,''),
'studio': getStudio(b), 'studio': getStudio(b),
'outline': getOutline(b), 'outline': getOutline(b),
'runtime': getRuntime(b), 'runtime': getRuntime(b),
@ -140,13 +141,14 @@ def main(number):
a = get_html('https://javdb.com/search?q=' + number + '&f=all').replace(u'\xa0', u' ') a = get_html('https://javdb.com/search?q=' + number + '&f=all').replace(u'\xa0', u' ')
html = etree.fromstring(a, etree.HTMLParser()) # //table/tr[1]/td[1]/text() html = etree.fromstring(a, etree.HTMLParser()) # //table/tr[1]/td[1]/text()
result1 = html.xpath('//*[@id="videos"]/div/div/a/@href')[0] result1 = html.xpath('//*[@id="videos"]/div/div/a/@href')[0]
print(html.xpath('//*[@id="videos"]/div/div/a/@href'))
b = get_html('https://javdb.com' + result1).replace(u'\xa0', u' ') b = get_html('https://javdb.com' + result1).replace(u'\xa0', u' ')
dic = { dic = {
'actor': getActor(b), 'actor': getActor(b),
'title': getTitle(b).replace("\\n", '').replace(' ', '').replace(getActor(a), '').replace( 'title': getTitle(b).replace("\\n", '').replace(' ', '').replace(getActor(a), '').replace(
getNum(b), getNum(b),
'').replace( '').replace(
'无码', '').replace('有码', '').lstrip(' ').replace(number, ''), '无码', '').replace('有码', '').lstrip(' ').replace(number,''),
'studio': getStudio(b), 'studio': getStudio(b),
'outline': getOutline(b), 'outline': getOutline(b),
'runtime': getRuntime(b), 'runtime': getRuntime(b),