This package include modules for findng links in a webpage and its children.
In the main module
find_links_by_extension links are found using two sub-modules and then added together:
- Using Google Search Results (
Since we can specify which types of files we are looking for when we search in Google, this methos scrapes these results. But this method is not complete:
- Google search works based on crawlers, and sometimes they don't index properly. For example this webpage has three pdf files at the moment (Aug 7 2018), but when we use google search to find them it finds only two although the files were uploaded 4 years ago.
- It doesn't work with some websites. For example this webpage has three pdf files but google cannot find any.
- If many requests are sent in a short period of time, Google blocks access and asks for CAPTCHA solving.
- Using a direct method of finding all urls in the given page and following those links if they are refering to children pages and seach recursively (
While this method does not miss any files in pages that it gets to (in contrast to method 1 which sometimes do), it may not find all the files because:
- Some webpages in the domain may be isolated i.e. there is no link to them in the parent pages. For these cases method 1 above works.
- In rare cases the link to a file of type xyz may not have .xyz in the link (example). In these cases method 2 cannot detect the file (because it only relies on the extesion appearing in the links), but method 1 detects correctly in these cases.
So the two methods complete each other's gaps.
$ pip install web-scraper
Note: This package relies on
html5lib libraries. So You need to have them installed:
$ pip install requests bs4 html5lib
>>> import web_scraper as ws >>> print(ws.find_links_by_extension("https://www.sfu.ca", 'pdf'))
get_links_directlyto find links to files that have the given extension in the given webpage and its children.
url: The webpage's url
extension: The file extension. Do not include dot.
maxNumberOfLinksToFind: Upper bound for number of links to find (default 30).
sorted: Whether to sort the output list.
See the source file.
See the source file.