Fixed adress parsing and page extraction#3
Conversation
torfsen
left a comment
torfsen
left a comment
There was a problem hiding this comment.
Nice work! There are some minor things that I'd like to see fixed or don't understand, otherwise this is great. However, I still find some error messages in my logs when I run this:
Traceback (most recent call last):
File "./scrape.py", line 409, in <module>
get_new_listings(db)
File "./scrape.py", line 352, in get_new_listings
listings = extract_listings(page)
File "./scrape.py", line 159, in extract_listings
street_span = entry.find('div', class_='result-list-entry__address').find('span').contents[0]
AttributeError: 'NoneType' object has no attribute 'contents'
Looks like we either need to check whether that element we're looking for is really there (in case it's optional) or we need to fix the way we're looking for it (in case it's always there but our selector sometimes fails).
| @@ -47,7 +47,7 @@ | |||
| # Immobilienscout24 URLs for listings in Karlsruhe | |||
| BASE_URL = 'http://www.immobilienscout24.de/Suche/S-T/Wohnung-Miete/Baden-Wuerttemberg/Karlsruhe' | |||
| PAGE_URL = 'http://www.immobilienscout24.de/Suche/S-T/P-%d/Wohnung-Miete/Baden-Wuerttemberg/Karlsruhe?pagerReporting=true' | |||
There was a problem hiding this comment.
I guess these URLs need to be changed as well? The best way would be if one could set just the city using a single variable, perhaps we can use a generic search URL (if one exists)?
| dd = dl.find('dd') | ||
| content = unicode(dd.string).strip() | ||
| if content.endswith('€'): | ||
| if content.endswith(' €'): |
There was a problem hiding this comment.
The new check covers less cases then the old (it needs an additional space). Why is that necessary?
| if content.endswith(' €'): | ||
| rent = parse_german_float(content.split()[0]) | ||
| elif content.endswith('m²'): | ||
| elif content.endswith(' m²'): |
| 'rent': rent, | ||
| 'area': area, | ||
| } | ||
| print(listings) |
There was a problem hiding this comment.
Right now the scraper doesn't output anything to STDOUT, all output goes into the database or into the log file. I'd like to keep it that way for the moment.
| logger.info("Fetching page %d" % page_index) | ||
| page = get_page(page_index) | ||
| num_pages = num_pages or extract_number_of_pages(page) | ||
| num_pages = extract_number_of_pages(page) |
| BASE_URL = 'http://www.immobilienscout24.de/Suche/S-T/Wohnung-Miete/Baden-Wuerttemberg/Karlsruhe' | ||
| PAGE_URL = 'http://www.immobilienscout24.de/Suche/S-T/P-%d/Wohnung-Miete/Baden-Wuerttemberg/Karlsruhe?pagerReporting=true' | ||
|
|
||
| CITY = 'Karlsruhe' |
There was a problem hiding this comment.
And can we please make this a command line parameter while we're at it?
2 participants
No description provided.