def scrape(workspace):
print "Scraping A&E Waiting Times with workspace {}".format(workspace)
html = requests.get(ROOT)
page = fromstring(html.content)
h3 = hd([h for h in page.cssselect('h3') if h.text_content().strip() == 'Weekly Data and Quarterly Aggregates'])
links = h3.getnext().cssselect('a')
h3 = hd([h for h in page.cssselect('h3') if
h.text_content().strip() == 'Monthly Data'])
links += h3.getnext().cssselect('a')
for l in links:
print l
datasets = []
for l in links:
try:
datasets.extend(scrape_page(l.get("href")))
except:
import traceback
traceback.print_exc()
datasets = filter(lambda x: x is not None, datasets)
print "Processed {} datasets".format(len(datasets))
return datasets
def scrape(workspace):
print "Scraping Delayed Transfer {}".format(workspace)
global DEFAULT_NOTES
html = requests.get(ROOT)
page = fromstring(html.content)
default_notes(page)
h3 = hd([h for h in page.cssselect('h3') if h.text_content().strip() == 'Data'])
links = h3.getnext().cssselect('a')
datasets = []
datasets.extend(scrape_page(links[-1].get("href")))
for l in links:
datasets.extend(scrape_page(l.get("href")))
# Get the annual statistical reports
h3 = hd([h for h in page.cssselect('h3') if h.text_content().strip() == 'Annual Statistical Report'])
links = h3.getnext().cssselect('a')
dataset = {
"resources": [anchor_to_resource(l) for l in links],
"title": "Delayed Transfers of Care - Annual Statistical Reports",
"origin": ROOT,
"notes": DEFAULT_NOTES,
"frequency": "Annually",
"groups": ['delayed_transfer']
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
datasets.append(dataset)
datasets = filter(lambda x: x is not None, datasets)
print "Processed {} datasets".format(len(datasets))
return datasets
def history(page):
link = hd(hd(page.cssselect('#ctlNavigationPastResults')).cssselect('a'))
page = get_dom(urljoin(ROOT, link.get('href')))
div = hd(page.cssselect('.foldout-set'))
links = [a for a in div.cssselect('a')
if 'Detailed Spreadsheets' in a.text_content().strip()]
for link in links:
u = urljoin(ROOT, link.get('href'))
page = get_dom(u)
yield scrape_page(page, u)
def latest(page):
# Find the Latest Data link at "http://www.nhsstaffsurveys.com/" and scrape
# that page.
link = hd(hd(page.cssselect('#ctlNavigationLatestResults')).cssselect('a'))
page = get_dom(urljoin(ROOT, link.get('href')))
h3 = hd([h for h in page.cssselect('h3') if h.text_content().strip() == "Detailed spreadsheets"])
latest_link = hd(h3.getparent().getnext().cssselect('a'))
u = urljoin(ROOT, latest_link.get('href'))
page = get_dom(u)
return scrape_page(page, u)
def history(page):
link = hd(hd(page.cssselect('#ctlNavigationPastResults')).cssselect('a'))
page = get_dom(urljoin(ROOT, link.get('href')))
div = hd(page.cssselect('.foldout-set'))
links = [
a for a in div.cssselect('a')
if 'Detailed Spreadsheets' in a.text_content().strip()
]
for link in links:
u = urljoin(ROOT, link.get('href'))
page = get_dom(u)
yield scrape_page(page, u)
def scrape(workspace):
print "Scraping CWT with workspace {}".format(workspace)
datasets = []
bases = [
'http://www.england.nhs.uk/statistics/statistical-work-areas/cancer-waiting-times/provider-based-cancer-waiting-times-statistics/',
'http://www.england.nhs.uk/statistics/statistical-work-areas/cancer-waiting-times/commissioner-based-cancer-waiting-times-statistics/'
]
targets = []
for base in bases:
html = requests.get(base)
page = fromstring(html.content)
h3 = hd([h for h in page.cssselect('h3') if h.text_content().strip().lower() == 'latest statistics'])
links = [a.get('href') for a in h3.getnext().cssselect('a')]
for l in links:
print l
targets += links
for t in targets:
datasets.append(scrape_commissioner_page(t))
# datasets.extend(commissioner_based())
# datasets.extend(default_cwt())
datasets = filter(lambda x: x is not None, datasets)
return datasets
def latest(page):
# Find the Latest Data link at "http://www.nhsstaffsurveys.com/" and scrape
# that page.
link = hd(hd(page.cssselect('#ctlNavigationLatestResults')).cssselect('a'))
page = get_dom(urljoin(ROOT, link.get('href')))
h3 = hd([
h for h in page.cssselect('h3')
if h.text_content().strip() == "Detailed spreadsheets"
])
latest_link = hd(h3.getparent().getnext().cssselect('a'))
u = urljoin(ROOT, latest_link.get('href'))
page = get_dom(u)
return scrape_page(page, u)
def get_groups(self):
if self.groups:
return self.groups
if 'Improving Access to Psychological Therapies' in self.dataset['title']:
self.groups.append('IAPT')
if 'Hospital Episode Statistics' in self.dataset['title']:
self.groups.append('HES')
if 'SHMI' in self.dataset['title']:
self.groups.append('SHMI')
# Check indicator specific data....
firsturl = hd([s['url'] for s in self.dataset.get('sources',[]) if s['description'] == 'Indicator specification'])
if firsturl:
if 'Clinical Commissioning Group Indicators' in firsturl:
self.groups.append('CCGOIS')
if 'Outcomes Framework' in firsturl:
self.groups.append('NHSOF')
for a in self.dataset.get('sources', []):
if 'Quality and Outcomes Framework' in a['description']:
self.groups.append('QOF')
self.groups = list(set(self.groups))
if self.groups:
print "***" * 20
print "Curated into a group {}".format(self.groups)
print "***" * 20
return self.groups
def scrape(workspace):
print "Scraping Child Immunisation with workspace {}".format(workspace)
html = requests.get(ROOT).content
page = fromstring(html)
div = page.cssselect('.center')[0]
links = div.cssselect('a')[3:]
h3 = hd([
h for h in div.cssselect('h3')
if h.text_content().strip() == "Background"
])
desc = h3.getnext().text_content()
dataset = {
"title": "Child Immunisation",
"notes": to_markdown(fix_bad_unicode(unicode(desc))),
"coverage_start_date": "",
"coverage_end_date": "",
"resources": [],
"frequency": "Quarterly",
"origin": ROOT,
"tags": ["immunisation", "children"],
"groups": ['child_immunisation']
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
earliest_quarter, earliest_year = 4, 9999
latest_quarter, latest_year = 1, 2000
for l in links:
y, q = get_quarter_and_year(l.text_content().strip())
if y < earliest_year:
earliest_year = y
if q < earliest_quarter:
earliest_quarter = q
if y > latest_year:
latest_year = y
if latest_quarter > q:
latest_quarter = q
dataset["resources"].append(anchor_to_resource(l))
if earliest_quarter == 4:
earliest_year += 1
if latest_quarter == 4:
latest_year += 1
s, e = QUARTERS[earliest_quarter]
dataset["coverage_start_date"] = "{}-{}-01".format(earliest_year,
str(s).zfill(2))
s, e = QUARTERS[latest_quarter]
_, last_day = calendar.monthrange(latest_year, s - 1)
dataset["coverage_end_date"] = "{}-{}-{}".format(earliest_year,
str(s - 1).zfill(2),
last_day)
return [dataset]
def scrape_page(page, url):
dataset = {
"title":
"NHS {}".format(hd(page.cssselect('h1')).text_content().strip()),
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
dataset["origin"] = url
dataset["tags"] = ["staff survey"]
year = re.match('.*(\d{4}).*', dataset['title']).groups()[0]
dataset["coverage_start_date"] = "{}-01-01".format(year)
dataset["coverage_end_date"] = "{}-12-31".format(year)
desc_node = page.cssselect('div.column-content p')
if desc_node:
dataset["notes"] = hd(desc_node).text_content()
else:
dataset["notes"] = "Results for the Staff Survey {year} can be seen below. "\
"We have released detailed spreadsheets showing key finding "\
"and question level information for each trust who took part "\
"in the {year} survey.".format(year=year)
dataset['notes'] = to_markdown(dataset['notes'])
dataset["resources"] = []
boxes = page.cssselect('.document-box')
for box in boxes:
a = box.cssselect('a')[0]
resource = anchor_to_resource(a)
resource['description'] = box.cssselect('h4')[0].text_content().strip()
resource['url'] = urljoin(ROOT, resource['url'])
dataset["resources"].append(resource)
key = hd([
a for a in page.cssselect('a')
if a.text_content().strip() == 'Click here'
])
if key is not None:
resource = anchor_to_resource(key)
resource['description'] = "Key Findings"
resource['url'] = urljoin(ROOT, resource['url'])
dataset["resources"].append(resource)
return dataset
def scrape(workspace):
print "Scraping Child Immunisation with workspace {}".format(workspace)
html = requests.get(ROOT).content
page = fromstring(html)
div = page.cssselect('.center')[0]
links = div.cssselect('a')[3:]
h3 = hd([h for h in div.cssselect('h3') if h.text_content().strip() == "Background"])
desc = h3.getnext().text_content()
dataset = {
"title": "Child Immunisation",
"notes": to_markdown(fix_bad_unicode(unicode(desc))),
"coverage_start_date": "",
"coverage_end_date": "",
"resources": [],
"frequency": "Quarterly",
"origin": ROOT,
"tags": ["immunisation", "children"],
"groups": ['child_immunisation']
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
earliest_quarter, earliest_year = 4, 9999
latest_quarter, latest_year = 1, 2000
for l in links:
y, q = get_quarter_and_year(l.text_content().strip())
if y < earliest_year:
earliest_year = y
if q < earliest_quarter:
earliest_quarter = q
if y > latest_year:
latest_year = y
if latest_quarter > q:
latest_quarter = q
dataset["resources"].append(anchor_to_resource(l))
if earliest_quarter == 4:
earliest_year += 1
if latest_quarter == 4:
latest_year += 1
s, e = QUARTERS[earliest_quarter]
dataset["coverage_start_date"] = "{}-{}-01".format(earliest_year, str(s).zfill(2))
s, e = QUARTERS[latest_quarter]
_, last_day = calendar.monthrange(latest_year, s-1)
dataset["coverage_end_date"] = "{}-{}-{}".format(earliest_year, str(s-1).zfill(2), last_day)
return [dataset]
def scrape_page(page, url):
dataset = {
"title": "NHS {}".format(hd(page.cssselect('h1')).text_content().strip()),
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
dataset["origin"] = url
dataset["tags"] = ["staff survey"]
year = re.match('.*(\d{4}).*', dataset['title']).groups()[0]
dataset["coverage_start_date"] = "{}-01-01".format(year)
dataset["coverage_end_date"] = "{}-12-31".format(year)
desc_node = page.cssselect('div.column-content p')
if desc_node:
dataset["notes"] = hd(desc_node).text_content()
else:
dataset["notes"] = "Results for the Staff Survey {year} can be seen below. "\
"We have released detailed spreadsheets showing key finding "\
"and question level information for each trust who took part "\
"in the {year} survey.".format(year=year)
dataset['notes'] = to_markdown(dataset['notes'])
dataset["resources"] = []
boxes = page.cssselect('.document-box')
for box in boxes:
a = box.cssselect('a')[0]
resource = anchor_to_resource(a)
resource['description'] = box.cssselect('h4')[0].text_content().strip()
resource['url'] = urljoin(ROOT, resource['url'])
dataset["resources"].append(resource)
key = hd([a for a in page.cssselect('a') if a.text_content().strip() == 'Click here'])
if key is not None:
resource = anchor_to_resource(key)
resource['description'] = "Key Findings"
resource['url'] = urljoin(ROOT, resource['url'])
dataset["resources"].append(resource)
return dataset
def scrape(workspace):
print "Scraping Delayed Transfer {}".format(workspace)
global DEFAULT_NOTES
html = requests.get(ROOT)
page = fromstring(html.content)
default_notes(page)
h3 = hd([
h for h in page.cssselect('h3') if h.text_content().strip() == 'Data'
])
links = h3.getnext().cssselect('a')
datasets = []
datasets.extend(scrape_page(links[-1].get("href")))
for l in links:
datasets.extend(scrape_page(l.get("href")))
# Get the annual statistical reports
h3 = hd([
h for h in page.cssselect('h3')
if h.text_content().strip() == 'Annual Statistical Report'
])
links = h3.getnext().cssselect('a')
dataset = {
"resources": [anchor_to_resource(l) for l in links],
"title": "Delayed Transfers of Care - Annual Statistical Reports",
"origin": ROOT,
"notes": DEFAULT_NOTES,
"frequency": "Annually",
"groups": ['delayed_transfer']
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
datasets.append(dataset)
datasets = filter(lambda x: x is not None, datasets)
print "Processed {} datasets".format(len(datasets))
return datasets
def scrape(workspace):
print "Scraping Diagnostic Imaging Data with workspace {}".format(workspace)
datasets = []
page = get_dom(ROOT)
datasets.append(scrape_page(ROOT, title="Diagnostic Imaging Dataset - Previous versions"))
h3 = hd([h for h in page.cssselect('h3') if h.text_content().strip() == "Data"])
for a in h3.getnext().cssselect('a'):
datasets.append(scrape_page(a.get('href')))
datasets = filter(lambda x: x is not None, datasets)
return datasets
def scrape(workspace):
print "Scraping Critical Care Capacity {}".format(workspace)
html = requests.get(ROOT)
page = fromstring(html.content)
default_notes(page)
h3 = hd([h for h in page.cssselect('h3') if h.text_content().strip() == 'Latest Data'])
links = h3.getnext().cssselect('a')
datasets = []
datasets.extend(scrape_page(links[-1].get("href")))
for l in links:
datasets.extend(scrape_page(l.get("href")))
datasets = filter(lambda x: x is not None, datasets)
print "Processed {} datasets".format(len(datasets))
return datasets
def scrape(workspace):
print "Scraping Diagnostic Imaging Data with workspace {}".format(
workspace)
datasets = []
page = get_dom(ROOT)
datasets.append(
scrape_page(ROOT,
title="Diagnostic Imaging Dataset - Previous versions"))
h3 = hd([
h for h in page.cssselect('h3') if h.text_content().strip() == "Data"
])
for a in h3.getnext().cssselect('a'):
datasets.append(scrape_page(a.get('href')))
datasets = filter(lambda x: x is not None, datasets)
return datasets
def default_notes(page):
""" Some pages don't have a description. If we have no DEFAULT_NOTES then
see if we can find them on the current page for the use in later pages """
global DEFAULT_NOTES
if DEFAULT_NOTES:
return
print "Getting default notes"
p = hd([h for h in page.cssselect('h3') if h.text_content().strip() == 'Background'])
if p is None:
return
desc = []
while True:
p = p.getnext()
if p.tag not in ['p', 'ul']:
break
s = tostring(p)
s = s.replace('&', '&')
desc.append(s)
DEFAULT_NOTES = to_markdown("".join(desc))
def default_notes(page):
""" Some pages don't have a description. If we have no DEFAULT_NOTES then
see if we can find them on the current page for the use in later pages """
global DEFAULT_NOTES
if DEFAULT_NOTES:
return
print "Getting default notes"
p = hd([h for h in page.cssselect('h3') if h.text_content().strip() == 'Background'])
if p is None:
return
desc = []
while True:
p = p.getnext()
if p.tag != 'p':
break
s = p.text_content().strip()
s = s.replace('&', '&')
desc.append(s)
DEFAULT_NOTES = to_markdown("".join(desc))
def scrape(workspace):
print "Scraping Critical Care Capacity {}".format(workspace)
html = requests.get(ROOT)
page = fromstring(html.content)
default_notes(page)
h3 = hd([
h for h in page.cssselect('h3')
if h.text_content().strip() == 'Latest Data'
])
links = h3.getnext().cssselect('a')
datasets = []
datasets.extend(scrape_page(links[-1].get("href")))
for l in links:
datasets.extend(scrape_page(l.get("href")))
datasets = filter(lambda x: x is not None, datasets)
print "Processed {} datasets".format(len(datasets))
return datasets
def get_groups(self):
if self.groups:
return self.groups
if 'Improving Access to Psychological Therapies' in self.dataset[
'title']:
self.groups.append('IAPT')
if 'Hospital Episode Statistics' in self.dataset['title']:
self.groups.append('HES')
if 'SHMI' in self.dataset['title']:
self.groups.append('SHMI')
# Check indicator specific data....
firsturl = hd([
s['url'] for s in self.dataset.get('sources', [])
if s['description'] == 'Indicator specification'
])
if firsturl:
if 'Clinical Commissioning Group Indicators' in firsturl:
self.groups.append('CCGOIS')
if 'Outcomes Framework' in firsturl:
self.groups.append('NHSOF')
for a in self.dataset.get('sources', []):
if 'Quality and Outcomes Framework' in a['description']:
self.groups.append('QOF')
self.groups = list(set(self.groups))
if self.groups:
print "***" * 20
print "Curated into a group {}".format(self.groups)
print "***" * 20
return self.groups
def get_page_count():
dom = get_dom(ROOT.format(1))
return int(hd(dom.cssselect('#paging li a.last')).text_content())
def get_page_count():
dom = get_dom(ROOT.format(1))
return int(hd(dom.cssselect('#paging li a.last')).text_content())
def process_monthly(page):
datasets = []
title = "Monthly Hospital Activity Data"
description = "Monthly activity data relating to elective and non-elective inpatient "\
"admissions (FFCEs) and outpatient referrals and attendances for first "\
"consultant outpatient appointments."
headers = page.cssselect('h3,h4')
for h in headers:
text = h.text_content().strip()
if re.match("(\d{4})-(\d{2})", text):
datasets.extend(
process_block(h, _p_strong("Provider based"),
_p_strong("Commissioner based"), title,
description, QUARTERLY))
provider_links, commissioner_links = [], []
h3prev = hd([
h for h in page.cssselect('h3')
if h.text_content().strip().startswith("Previous")
])
p = h3prev.getnext()
while True:
if len(p) == 0:
break
if _p_strong("Provider based")(p):
provider_links = p.getnext().cssselect('a')
if _p_strong("Commissioner based")(p):
commissioner_links = p.getnext().cssselect('a')
p = p.getnext()
for l in provider_links:
m = re.match(".*(\d{4})-\d{2}.*", l.text_content().encode('utf8'))
yr = int(m.groups()[0])
csd = "{}-04-01".format(yr)
ced = "{}-03-31".format(yr + 1)
pdataset = {
"title": "{} - Provider based - {}-{}".format(title, yr, yr + 1),
"notes": description,
"origin": MONTHLY,
"resources": [anchor_to_resource(l)],
"frequency": "Annual",
"coverage_start_date": csd,
"coverage_end_date": ced,
"groups": ['hospital_activity']
}
pdataset["name"] = slugify.slugify(pdataset["title"]).lower()
datasets.append(pdataset)
for l in commissioner_links:
m = re.match(".*(\d{4})-\d{2}.*", l.text_content().encode('utf8'))
yr = int(m.groups()[0])
csd = "{}-04-01".format(yr)
ced = "{}-03-31".format(yr + 1)
cdataset = {
"title": "{} - Provider based - {}-{}".format(title, yr, yr + 1),
"notes": description,
"origin": MONTHLY,
"resources": [anchor_to_resource(l)],
"frequency": "Annual",
"coverage_start_date": csd,
"coverage_end_date": ced,
"groups": ['hospital_activity']
}
cdataset["name"] = slugify.slugify(cdataset["title"]).lower()
datasets.append(cdataset)
time_series_links = [
a for a in page.cssselect('a') if 'Timeseries' in a.get('href')
]
dataset = {
"title": "{} - Time Series".format(title),
"notes": description,
"origin": MONTHLY,
"resources": [anchor_to_resource(a) for a in time_series_links]
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
datasets.append(dataset)
return datasets
def process_monthly(page):
datasets = []
title = "Monthly Hospital Activity Data"
description = "Monthly activity data relating to elective and non-elective inpatient "\
"admissions (FFCEs) and outpatient referrals and attendances for first "\
"consultant outpatient appointments."
headers = page.cssselect('h3,h4')
for h in headers:
text = h.text_content().strip()
if re.match("(\d{4})-(\d{2})", text):
datasets.extend(process_block(h, _p_strong("Provider based"),
_p_strong("Commissioner based"),title, description,QUARTERLY))
provider_links, commissioner_links = [], []
h3prev = hd([h for h in page.cssselect('h3') if h.text_content().strip().startswith("Previous")])
p = h3prev.getnext()
while True:
if len(p) == 0:
break
if _p_strong("Provider based")(p):
provider_links = p.getnext().cssselect('a')
if _p_strong("Commissioner based")(p):
commissioner_links = p.getnext().cssselect('a')
p = p.getnext()
for l in provider_links:
m = re.match(".*(\d{4})-\d{2}.*", l.text_content().encode('utf8'))
yr = int(m.groups()[0])
csd = "{}-04-01".format(yr)
ced = "{}-03-31".format(yr+1)
pdataset = {
"title": "{} - Provider based - {}-{}".format(title, yr, yr+1),
"notes": description,
"origin": MONTHLY,
"resources": [anchor_to_resource(l)],
"frequency": "Annual",
"coverage_start_date": csd,
"coverage_end_date": ced,
"groups": ['hospital_activity']
}
pdataset["name"] = slugify.slugify(pdataset["title"]).lower()
datasets.append(pdataset)
for l in commissioner_links:
m = re.match(".*(\d{4})-\d{2}.*", l.text_content().encode('utf8'))
yr = int(m.groups()[0])
csd = "{}-04-01".format(yr)
ced = "{}-03-31".format(yr+1)
cdataset = {
"title": "{} - Provider based - {}-{}".format(title, yr, yr+1),
"notes": description,
"origin": MONTHLY,
"resources": [anchor_to_resource(l)],
"frequency": "Annual",
"coverage_start_date": csd,
"coverage_end_date": ced,
"groups": ['hospital_activity']
}
cdataset["name"] = slugify.slugify(cdataset["title"]).lower()
datasets.append(cdataset)
time_series_links = [a for a in page.cssselect('a') if 'Timeseries' in a.get('href')]
dataset = {
"title": "{} - Time Series".format(title),
"notes": description,
"origin": MONTHLY,
"resources": [anchor_to_resource(a) for a in time_series_links]
}
dataset["name"] = slugify.slugify(dataset["title"]).lower()
datasets.append(dataset)
return datasets