diff --git a/FourmiCrawler/sources/ChemSpider.py b/FourmiCrawler/sources/ChemSpider.py index 2fcd07c..254c1a5 100644 --- a/FourmiCrawler/sources/ChemSpider.py +++ b/FourmiCrawler/sources/ChemSpider.py @@ -1,9 +1,12 @@ -from source import Source +import re + from scrapy import log from scrapy.http import Request from scrapy.selector import Selector + +from source import Source from FourmiCrawler.items import Result -import re + # [TODO] - Maybe clean up usage of '.extract()[0]', because of possible IndexError exception. diff --git a/FourmiCrawler/sources/NIST.py b/FourmiCrawler/sources/NIST.py index 0b75b17..2fe5966 100644 --- a/FourmiCrawler/sources/NIST.py +++ b/FourmiCrawler/sources/NIST.py @@ -1,9 +1,12 @@ -from source import Source +import re + from scrapy import log from scrapy.http import Request from scrapy.selector import Selector + +from source import Source from FourmiCrawler.items import Result -import re + # [TODO]: values can be '128.', perhaps remove the dot in that case? # [TODO]: properties have references and comments which do not exist in the diff --git a/FourmiCrawler/sources/WikipediaParser.py b/FourmiCrawler/sources/WikipediaParser.py index cb7d0b9..c4f7a0f 100644 --- a/FourmiCrawler/sources/WikipediaParser.py +++ b/FourmiCrawler/sources/WikipediaParser.py @@ -1,9 +1,11 @@ +import re + from scrapy.http import Request from scrapy import log -from source import Source from scrapy.selector import Selector + +from source import Source from FourmiCrawler.items import Result -import re class WikipediaParser(Source):