def project(self,response):
## Get the project info proper.
projectObject=ProjectItem()
projectObject['author']=User.objects.get(pk=self.user_id)
projectObject['title']=response.selector.xpath('//*[contains(@class,\'thing-header-data\')]/h1/text()').extract()[0].strip()
projectObject['tags'] = response.selector.xpath("//*[contains(@class,\'thing-info-content thing-detail-tags-container\')]/div/a/text()").extract()
yield projectObject
## get special text files. (readme, instructions, license)
import html2text
h2t = html2text.HTML2Text()
#Get the reame file, do stuff to it.
readme = h2t.handle(response.selector.xpath("//*[@id = 'description']").extract()[0].strip())
import unicodedata
readmeItem=fileObjectItem()
readmeItem["name"]="README.md"
readmeItem["parent"]=projectObject['SID']
readmeItem["filename"]=u""+unicodedata.normalize('NFKD',readme).encode('ascii','ignore')
readmeItem['isReadme'] = True
yield readmeItem
#projectObject['readme'] = u""+unicodedata.normalize('NFKD',readme).encode('ascii','ignore')
#also a markdown file I guess we'd want.
try:
instructions = u""+h2t.handle(response.selector.xpath("//*[@id = 'instructions']").extract()[0].strip()).encode('ascii','ignore')
instructionItem=fileObjectItem()
instructionItem["name"]="Instructions.md"
instructionItem["parent"]=projectObject['SID']
instructionItem["filename"]=instructions
yield instructionItem
except IndexError:
pass
#print("xpath to get the instructions IndexError'd")
## now, because the format of the license on thingi is always the same, we can pull this off.
## but I expect it is rather fragile.
licenseurl =response.selector.xpath("//*[contains(@class,\'license-text\')]/a/@href")[2].extract().strip()
licensetext = response.selector.xpath("//*[contains(@class,\'license-text\')]/a/text()")[1].extract().strip()
licenceItem=fileObjectItem()
licenceItem["name"]="License.md"
licenceItem["parent"]=projectObject['SID']
licenceItem["filename"]="["+licensetext+"]("+licenseurl+")"
yield licenceItem
## get all the projects image and file objects
filelist = response.selector.xpath('//*[contains(@class,\'thing-file\')]/a/@href')
for i in filelist:
yield scrapy.http.Request(url=urlparse.urljoin(response.url, i.extract()), callback=self.item, meta={'parent':projectObject['SID']})
#Grab only raw images.
imagelist = response.selector.xpath('//*[contains(@class,\'thing-gallery-thumbs\')]/div[@data-track-action="viewThumb"][@data-thingiview-url=""]/@data-large-url')
for i in imagelist:
yield scrapy.http.Request(dont_filter=True, url=urlparse.urljoin(response.url, i.extract()), callback=self.item, meta={'parent':projectObject['SID']})
def item(self,response):
item=fileObjectItem()
## warning stupid preasent here.
# splitting and grabing from urlparse for filename may not be best.
item['name']=urlparse.urlparse(response.url)[2].split("/")[-1]
item['name']=item['name'].replace("_display_large","")
item['parent'] = response.meta['parent']
item['filename']=response.body
yield(item)
