langbrain/Corbia spider.md

50 lines
1.3 KiB
Markdown

In order to index the web properly, corbia spider uses domain based templates.
## Domain declaration
For each domain, a custom python function will be used in order to extract the relevant information and store it as JSON before further processing.
New domains are added to `Domain` class:
```python
class Domain:
def __init__(self, domain, html, url):
self.domain = domain
self.soup = BeautifulSoup(html, 'html.parser')
self.url = url
self.scrapers = {
"www.azlyrics.com" : self.scrape_azlyrics,
"www.monde-diplomatique.fr" : self.scrape_diplo,
"www.amnesty.org" : self.scrape_amnesty,
"www.vindefrance.com" : self.scrape_vdf,
"www.tasteofcinema.com" : self.scrape_taste_of_cinema,
"www.blast-info.fr" : self.scrape_blast,
"www.byredo.com" : self.scrape_byredo
}
```
Then, the function is added to the class as `def scrape_domain(self)`.
## Domain based scrapping function
The function must **ALWAYS** return a JSON with a `"success"` key set to either `True` or `False`.
To make sure the spider doesn't stop working, the scrapping function must be wrapped inside a `try` / `except` block:
```python
def scrape_domain(self):
try:
data = {
"success" : True
}
return json.dumps(data, indent=4, ensure_ascii=False)
except Exception as e:
return json.dumps({"success": False})
```