def scrape_bill(self, chamber, session):
url = "ftp://www.arkleg.state.ar.us/SessionInformation/LegislativeMeasures.txt"
page = csv.reader(get_utf_16_ftp_content(url).splitlines(),
delimiter="|")
for row in page:
bill_chamber = {"H": "lower", "S": "upper"}[row[0]]
if bill_chamber != chamber:
continue
bill_id = "%s%s %s" % (row[0], row[1], row[2])
type_spec = re.match(r"(H|S)([A-Z]+)\s", bill_id).group(2)
bill_type = {
"B": "bill",
"R": "resolution",
"JR": "joint resolution",
"CR": "concurrent resolution",
"MR": "memorial",
"CMR": "concurrent memorial",
}[type_spec]
if row[-1] != self.slug:
continue
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=row[3],
classification=bill_type,
)
bill.add_source(url)
primary = row[11]
if not primary:
primary = row[12]
if primary:
bill.add_sponsorship(
primary,
classification="primary",
entity_type="person",
primary=True,
)
version_url = ("ftp://www.arkleg.state.ar.us/Bills/"
"%s/Public/Searchable/%s.pdf" %
(self.slug, bill_id.replace(" ", "")))
bill.add_version_link(bill_id,
version_url,
media_type="application/pdf")
yield from self.scrape_bill_page(bill)
self.bills[bill_id] = bill
def scrape_prefiles(self, session):
url = 'https://www.legis.iowa.gov/legislation/billTracking/prefiledBills'
page = lxml.html.fromstring(self.get(url).content)
page.make_links_absolute(url)
for row in page.xpath('//table[contains(@class, "sortable")]/tr[td]'):
title = row.xpath('td[2]/a/text()')[0].strip()
url = row.xpath('td[2]/a/@href')[0]
bill_id = self.extract_doc_id(title)
bill = Bill(
bill_id,
legislative_session=session,
chamber='legislature',
title=title,
classification='proposed bill',
)
if (row.xpath('td[3]/a')):
document_url = row.xpath('td[3]/a/@href')[0]
if '.docx' in document_url:
media_type = 'application/vnd.openxmlformats-officedocument.wordprocessingml.document'
elif '.pdf' in document_url:
media_type = 'application/pdf'
bill.add_document_link(
note="Backround Statement",
url=document_url,
media_type=media_type
)
bill.add_version_link(
note="Prefiled",
url=url,
media_type="application/pdf"
)
bill.add_source(url)
yield bill
def scrape_bill(self, chamber, session, bill_id):
bill_num = bill_id.split()[1]
url = "%s/GetLegislation?biennium=%s&billNumber" "=%s" % (
self._base_url,
self.biennium,
bill_num,
)
page = self.get(url)
page = lxml.etree.fromstring(page.content)
page = xpath(page, "//wa:Legislation")[0]
xml_chamber = xpath(page, "string(wa:OriginalAgency)")
chamber = self._chamber_map[xml_chamber]
title = xpath(page, "string(wa:LongDescription)")
bill_type = xpath(
page, "string(wa:ShortLegislationType/wa:LongLegislationType)"
)
bill_type = bill_type.lower()
if bill_type == "gubernatorial appointment":
return
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=[bill_type],
)
fake_source = (
"http://apps.leg.wa.gov/billinfo/"
"summary.aspx?bill=%s&year=%s" % (bill_num, session[0:4])
)
bill.add_source(fake_source)
try:
for version in self.versions[bill_id]:
bill.add_version_link(
note=version["note"],
url=version["url"],
media_type=version["media_type"],
)
except KeyError:
self.warning("No versions were found for {}".format(bill_id))
try:
for document in self.documents[bill_num]:
bill.add_document_link(
note=document["note"],
url=document["url"],
media_type=document["media_type"],
)
except KeyError:
pass
self.scrape_sponsors(bill)
self.scrape_actions(bill, chamber, fake_source)
self.scrape_hearings(bill, bill_num)
yield from self.scrape_votes(bill)
bill.subject = list(set(self._subjects[bill_id]))
yield bill
def scrape_bill(self, session, history_url):
history_xml = self.get(history_url).text
root = etree.fromstring(history_xml)
bill_title = root.findtext("caption")
if bill_title is None or "Bill does not exist" in history_xml:
self.warning("Bill does not appear to exist")
return
bill_id = " ".join(root.attrib["bill"].split(" ")[1:])
chamber = self.CHAMBERS[bill_id[0]]
if bill_id[1] == "B":
bill_type = ["bill"]
elif bill_id[1] == "R":
bill_type = ["resolution"]
elif bill_id[1:3] == "CR":
bill_type = ["concurrent resolution"]
elif bill_id[1:3] == "JR":
bill_type = ["joint resolution"]
else:
raise ScrapeError("Invalid bill_id: %s" % bill_id)
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=bill_title,
classification=bill_type,
)
bill.add_source(history_url)
bill_id_for_url = bill_id.replace(" ", "")
bill.add_source(
f"https://capitol.texas.gov/BillLookup/History.aspx?LegSess={session}&Bill={bill_id_for_url}"
)
for subject in root.iterfind("subjects/subject"):
bill.add_subject(subject.text.strip())
for version in root.iterfind(
"billtext/docTypes/bill/versions/version"):
if not version:
continue
note = version.find("versionDescription").text
html_url = version.find("WebHTMLURL").text
bill.add_version_link(note=note,
url=html_url,
media_type="text/html")
pdf_url = version.find("WebPDFURL").text
bill.add_version_link(note=note,
url=pdf_url,
media_type="application/pdf")
for analysis in root.iterfind(
"billtext/docTypes/analysis/versions/version"):
if not analysis:
continue
description = analysis.find("versionDescription").text
html_url = analysis.find("WebHTMLURL").text
bill.add_document_link(
note="Analysis ({})".format(description),
url=html_url,
media_type="text/html",
)
for fiscal_note in root.iterfind(
"billtext/docTypes/fiscalNote/versions/version"):
if not fiscal_note:
continue
description = fiscal_note.find("versionDescription").text
html_url = fiscal_note.find("WebHTMLURL").text
bill.add_document_link(
note="Fiscal Note ({})".format(description),
url=html_url,
media_type="text/html",
)
witnesses = [x for x in self.witnesses if x[0] == bill_id]
for witness in witnesses:
bill.add_document_link(
note="Witness List ({})".format(
self.NAME_SLUGS[witness[1][-5]]),
url=witness[1],
media_type="text/html",
)
for action in root.findall("actions/action"):
act_date = datetime.datetime.strptime(action.findtext("date"),
"%m/%d/%Y").date()
action_number = action.find("actionNumber").text
actor = {
"H": "lower",
"S": "upper",
"E": "executive"
}[action_number[0]]
desc = action.findtext("description").strip()
if desc == "Scheduled for public hearing on . . .":
self.warning("Skipping public hearing action with no date")
continue
atype = _categorize_action(desc)
act = bill.add_action(
action.findtext("description"),
act_date,
chamber=actor,
classification=atype,
)
if atype and "referral-committee" in atype:
repls = ["Referred to", "Recommended to be sent to "]
ctty = desc
for r in repls:
ctty = ctty.replace(r, "").strip()
act.add_related_entity(name=ctty, entity_type="organization")
for author in root.findtext("authors").split(" | "):
if author != "":
bill.add_sponsorship(author,
classification="primary",
entity_type="person",
primary=True)
for coauthor in root.findtext("coauthors").split(" | "):
if coauthor != "":
bill.add_sponsorship(
coauthor,
classification="cosponsor",
entity_type="person",
primary=False,
)
for sponsor in root.findtext("sponsors").split(" | "):
if sponsor != "":
bill.add_sponsorship(
sponsor,
classification="primary",
entity_type="person",
primary=True,
)
for cosponsor in root.findtext("cosponsors").split(" | "):
if cosponsor != "":
bill.add_sponsorship(
cosponsor,
classification="cosponsor",
entity_type="person",
primary=False,
)
if root.findtext("companions"):
self._get_companion(bill)
yield bill
def scrape_bill(self, session, history_url):
history_xml = self.get(history_url).text
root = etree.fromstring(history_xml)
bill_title = root.findtext("caption")
if bill_title is None or "Bill does not exist" in history_xml:
self.warning("Bill does not appear to exist")
return
bill_id = " ".join(root.attrib["bill"].split(" ")[1:])
chamber = self.CHAMBERS[bill_id[0]]
if bill_id[1] == "B":
bill_type = ["bill"]
elif bill_id[1] == "R":
bill_type = ["resolution"]
elif bill_id[1:3] == "CR":
bill_type = ["concurrent resolution"]
elif bill_id[1:3] == "JR":
bill_type = ["joint resolution"]
else:
raise ScrapeError("Invalid bill_id: %s" % bill_id)
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=bill_title,
classification=bill_type,
)
bill.add_source(history_url)
for subject in root.iterfind("subjects/subject"):
bill.add_subject(subject.text.strip())
versions = [x for x in self.versions if x[0] == bill_id]
for version in versions:
bill.add_version_link(
note=self.NAME_SLUGS[version[1][-5]],
url=version[1],
media_type="text/html",
)
analyses = [x for x in self.analyses if x[0] == bill_id]
for analysis in analyses:
bill.add_document_link(
note="Analysis ({})".format(self.NAME_SLUGS[analysis[1][-5]]),
url=analysis[1],
media_type="text/html",
)
fiscal_notes = [x for x in self.fiscal_notes if x[0] == bill_id]
for fiscal_note in fiscal_notes:
bill.add_document_link(
note="Fiscal Note ({})".format(
self.NAME_SLUGS[fiscal_note[1][-5]]),
url=fiscal_note[1],
media_type="text/html",
)
witnesses = [x for x in self.witnesses if x[0] == bill_id]
for witness in witnesses:
bill.add_document_link(
note="Witness List ({})".format(
self.NAME_SLUGS[witness[1][-5]]),
url=witness[1],
media_type="text/html",
)
for action in root.findall("actions/action"):
act_date = datetime.datetime.strptime(action.findtext("date"),
"%m/%d/%Y").date()
action_number = action.find("actionNumber").text
actor = {
"H": "lower",
"S": "upper",
"E": "executive"
}[action_number[0]]
desc = action.findtext("description").strip()
if desc == "Scheduled for public hearing on . . .":
self.warning("Skipping public hearing action with no date")
continue
introduced = False
if desc == "Amended":
atype = "amendment-passage"
elif desc == "Amendment(s) offered":
atype = "amendment-introduction"
elif desc == "Amendment amended":
atype = "amendment-amendment"
elif desc == "Amendment withdrawn":
atype = "amendment-withdrawal"
elif desc == "Passed" or desc == "Adopted":
atype = "passage"
elif re.match(r"^Received (by|from) the", desc):
if "Secretary of the Senate" not in desc:
atype = "introduction"
else:
atype = "filing"
elif desc.startswith("Sent to the Governor"):
# But what if it gets lost in the mail?
atype = "executive-receipt"
elif desc.startswith("Signed by the Governor"):
atype = "executive-signature"
elif desc.startswith("Effective on"):
atype = "became-law"
elif desc == "Vetoed by the Governor":
atype = "executive-veto"
elif desc == "Read first time":
atype = ["introduction", "reading-1"]
introduced = True
elif desc == "Read & adopted":
atype = ["passage"]
if not introduced:
introduced = True
atype.append("introduction")
elif desc == "Passed as amended":
atype = "passage"
elif desc.startswith("Referred to") or desc.startswith(
"Recommended to be sent to "):
atype = "referral-committee"
elif desc == "Reported favorably w/o amendment(s)":
atype = "committee-passage"
elif desc == "Filed":
atype = "filing"
elif desc == "Read 3rd time":
atype = "reading-3"
elif desc == "Read 2nd time":
atype = "reading-2"
elif desc.startswith("Reported favorably"):
atype = "committee-passage-favorable"
else:
atype = None
act = bill.add_action(
action.findtext("description"),
act_date,
chamber=actor,
classification=atype,
)
if atype and "referral-committee" in atype:
repls = ["Referred to", "Recommended to be sent to "]
ctty = desc
for r in repls:
ctty = ctty.replace(r, "").strip()
act.add_related_entity(name=ctty, entity_type="organization")
for author in root.findtext("authors").split(" | "):
if author != "":
bill.add_sponsorship(author,
classification="primary",
entity_type="person",
primary=True)
for coauthor in root.findtext("coauthors").split(" | "):
if coauthor != "":
bill.add_sponsorship(
coauthor,
classification="cosponsor",
entity_type="person",
primary=False,
)
for sponsor in root.findtext("sponsors").split(" | "):
if sponsor != "":
bill.add_sponsorship(
sponsor,
classification="primary",
entity_type="person",
primary=True,
)
for cosponsor in root.findtext("cosponsors").split(" | "):
if cosponsor != "":
bill.add_sponsorship(
cosponsor,
classification="cosponsor",
entity_type="person",
primary=False,
)
if root.findtext("companions"):
self._get_companion(bill)
yield bill
def scrape_bill(self, chamber, session, bill_id, url):
try:
page = lxml.html.fromstring(self.get(url).text)
except scrapelib.HTTPError as e:
self.warning("error (%s) fetching %s, skipping" % (e, url))
return
title = page.xpath(
"string(//span[contains(@id, 'PlaceHolder1_txtST')])").strip()
if not title:
self.warning("blank bill on %s - skipping", url)
return
if "JR" in bill_id:
bill_type = ["joint resolution"]
elif "CR" in bill_id:
bill_type = ["concurrent resolution"]
elif "R" in bill_id:
bill_type = ["resolution"]
else:
bill_type = ["bill"]
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=bill_type,
)
bill.add_source(url)
bill.subject = self.subject_map[bill_id]
for link in page.xpath("//a[contains(@id, 'Auth')]"):
name = link.xpath("string()").strip()
if "author not found" in name.lower():
continue
if ":" in name:
raise Exception(name)
if "otherAuth" in link.attrib["id"]:
bill.add_sponsorship(
name,
classification="cosponsor",
entity_type="person",
primary=False,
)
else:
bill.add_sponsorship(name,
classification="primary",
entity_type="person",
primary=True)
act_table = page.xpath("//table[contains(@id, 'Actions')]")[0]
for tr in act_table.xpath("tr")[2:]:
action = tr.xpath("string(td[1])").strip()
if not action or action == "None":
continue
date = tr.xpath("string(td[3])").strip()
date = datetime.datetime.strptime(date, "%m/%d/%Y").date()
actor = tr.xpath("string(td[4])").strip()
if actor == "H":
actor = "lower"
elif actor == "S":
actor = "upper"
attrs = self.categorizer.categorize(action)
related_entities = []
for item in attrs["committees"]:
related_entities.append({"type": "committee", "name": item})
for item in attrs["legislators"]:
related_entities.append({"type": "legislator", "name": item})
bill.add_action(
description=action,
date=date.strftime("%Y-%m-%d"),
chamber=actor,
classification=attrs["classification"],
related_entities=related_entities,
)
version_table = page.xpath("//table[contains(@id, 'Versions')]")[0]
# Keep track of already seen versions to prevent processing duplicates.
version_urls = []
for link in version_table.xpath(".//a[contains(@href, '.PDF')]"):
version_url = link.attrib["href"]
if version_url in version_urls:
self.warning("Skipping duplicate version URL.")
continue
else:
version_urls.append(version_url)
name = link.text.strip()
if re.search("COMMITTEE REPORTS|SCHEDULED CCR", version_url,
re.IGNORECASE):
bill.add_document_link(note=name,
url=version_url,
media_type="application/pdf")
continue
bill.add_version_link(note=name,
url=version_url,
media_type="application/pdf")
self.scrape_amendments(bill, page)
for link in page.xpath(".//a[contains(@href, '_VOTES')]"):
if "HT_" not in link.attrib["href"]:
yield from self.scrape_votes(
bill, self.urlescape(link.attrib["href"]))
# # If the bill has no actions and no versions, it's a bogus bill on
# # their website, which appears to happen occasionally. Skip.
has_no_title = bill.title == "Short Title Not Found."
if has_no_title:
# If there's no title, this is an empty page. Skip!
return
else:
# Otherwise, save the bills.
yield bill
def _parse_senate_billpage(self, bill_url, year):
bill_page = self.lxmlize(bill_url)
# get all the info needed to record the bill
# TODO probably still needs to be fixed
bill_id = bill_page.xpath('//*[@id="lblBillNum"]')[0].text_content()
bill_title = bill_page.xpath(
'//*[@id="lblBillTitle"]')[0].text_content()
bill_desc = bill_page.xpath(
'//*[@id="lblBriefDesc"]')[0].text_content()
# bill_lr = bill_page.xpath('//*[@id="lblLRNum"]')[0].text_content()
bill_type = "bill"
triplet = bill_id[:3]
if triplet in bill_types:
bill_type = bill_types[triplet]
subs = []
bid = bill_id.replace(" ", "")
if bid in self._subjects:
subs = self._subjects[bid]
self.info("With subjects for this bill")
self.info(bid)
if bid == "XXXXXX":
self.info("Skipping Junk Bill")
return
bill = Bill(
bill_id,
title=bill_desc,
chamber="upper",
legislative_session=self._session_id,
classification=bill_type,
)
bill.subject = subs
bill.add_abstract(bill_desc, note="abstract")
bill.add_source(bill_url)
if bill_title:
bill.add_title(bill_title)
# Get the primary sponsor
try:
sponsor = bill_page.xpath('//a[@id="hlSponsor"]')[0]
except IndexError:
sponsor = bill_page.xpath('//span[@id="lSponsor"]')[0]
bill_sponsor = sponsor.text_content()
# bill_sponsor_link = sponsor.attrib.get('href')
bill.add_sponsorship(bill_sponsor,
entity_type="person",
classification="primary",
primary=True)
# cosponsors show up on their own page, if they exist
cosponsor_tag = bill_page.xpath('//a[@id="hlCoSponsors"]')
if len(cosponsor_tag) > 0 and cosponsor_tag[0].attrib.get("href"):
self._parse_senate_cosponsors(bill,
cosponsor_tag[0].attrib["href"])
# get the actions
action_url = bill_page.xpath('//a[@id="hlAllActions"]')
if len(action_url) > 0:
action_url = action_url[0].attrib["href"]
self._parse_senate_actions(bill, action_url)
# stored on a separate page
versions_url = bill_page.xpath('//a[@id="hlFullBillText"]')
if len(versions_url) > 0 and versions_url[0].attrib.get("href"):
self._parse_senate_bill_versions(bill,
versions_url[0].attrib["href"])
amendment_links = bill_page.xpath(
'//a[contains(@href,"ShowAmendment.asp")]')
for link in amendment_links:
link_text = link.xpath("string(.)").strip()
if "adopted" in link_text.lower():
link_url = link.xpath("@href")[0]
bill.add_version_link(
link_text,
link_url,
media_type="application/pdf",
on_duplicate="ignore",
)
yield bill
def scrape_bill(self, chamber, session, bill_id):
# there will be a space in bill_id if we're doing a one-off bill scrape
# convert HB 102 into H102
if " " in bill_id:
bill_id = bill_id[0] + bill_id.split(" ")[-1]
# if chamber comes in as House/Senate convert to lower/upper
if chamber == "Senate":
chamber = "upper"
elif chamber == "House":
chamber = "lower"
bill_detail_url = (
"http://www.ncleg.net/gascripts/"
"BillLookUp/BillLookUp.pl?Session=%s&BillID=%s&votesToView=all"
) % (session, bill_id)
# parse the bill data page, finding the latest html text
data = self.get(bill_detail_url).text
doc = lxml.html.fromstring(data)
doc.make_links_absolute(bill_detail_url)
title_div_txt = doc.xpath('//div[contains(@class, "h2")]/text()')[0]
if "Joint Resolution" in title_div_txt:
bill_type = "joint resolution"
bill_id = bill_id[0] + "JR " + bill_id[1:]
elif "Resolution" in title_div_txt:
bill_type = "resolution"
bill_id = bill_id[0] + "R " + bill_id[1:]
elif "Bill" in title_div_txt:
bill_type = "bill"
bill_id = bill_id[0] + "B " + bill_id[1:]
bill_title = doc.xpath("//main//div[@class='col-12'][1]")[0]
bill_title = bill_title.text_content().strip()
# For special cases where bill title is blank, a new title is created using Bill ID
if not bill_title:
bill_title = bill_id.replace(" ", "")
bill = Bill(
bill_id,
legislative_session=session,
title=bill_title,
chamber=chamber,
classification=bill_type,
)
bill.add_source(bill_detail_url)
# skip first PDF link (duplicate link to cur version)
if chamber == "lower":
link_xpath = '//a[contains(@href, "/Bills/House/PDF/")]'
else:
link_xpath = '//a[contains(@href, "/Bills/Senate/PDF/")]'
for vlink in doc.xpath(link_xpath)[1:]:
# get the name from the PDF link...
version_name = vlink.text.replace("\xa0", " ")
version_url = vlink.attrib["href"]
media_type = "text/html"
if version_url.lower().endswith(".pdf"):
media_type = "application/pdf"
bill.add_version_link(version_name,
version_url,
media_type=media_type,
on_duplicate="ignore")
# rows with a 'adopted' in the text and an amendment link, skip failed amds
for row in doc.xpath(
'//div[@class="card-body"]/div[contains(., "Adopted")'
' and contains(@class,"row")]//a[@title="Amendment"]'):
version_url = row.xpath("@href")[0]
version_name = row.xpath("string(.)").strip()
bill.add_version_link(
version_name,
version_url,
media_type="application/pdf",
on_duplicate="ignore",
)
# sponsors
spon_row = doc.xpath(
'//div[contains(text(), "Sponsors")]/following-sibling::div')[0]
# first sponsors are primary, until we see (Primary)
spon_type = "primary"
spon_lines = spon_row.text_content().replace("\r\n",
";").replace("\n", ";")
for leg in spon_lines.split(";"):
name = leg.replace("\xa0", " ").strip()
if name.startswith("(Primary)") or name.endswith("(Primary)"):
name = name.replace("(Primary)", "").strip()
spon_type = "cosponsor"
if not name:
continue
bill.add_sponsorship(
name,
classification=spon_type,
entity_type="person",
primary=(spon_type == "primary"),
)
# keywords
kw_row = doc.xpath(
'//div[contains(text(), "Keywords:")]/following-sibling::div')[0]
for subject in kw_row.text_content().split(", "):
bill.add_subject(subject)
# actions
action_tr_xpath = ('//h6[contains(text(), "History")]'
'/ancestor::div[contains(@class, "gray-card")]'
'//div[contains(@class, "card-body")]'
'/div[@class="row"]')
# skip two header rows
for row in doc.xpath(action_tr_xpath):
cols = row.xpath("div")
act_date = cols[1].text
actor = cols[3].text or ""
# if text is blank, try diving in
action = (cols[5].text
or "").strip() or cols[5].text_content().strip()
if act_date is None:
search_action_date = action.split()
for act in search_action_date:
try:
if "/" in act:
# try:
act_date = dt.datetime.strptime(
act, "%m/%d/%Y").strftime("%Y-%m-%d")
except KeyError:
raise Exception("No Action Date Provided")
else:
act_date = dt.datetime.strptime(
act_date, "%m/%d/%Y").strftime("%Y-%m-%d")
if actor == "Senate":
actor = "upper"
elif actor == "House":
actor = "lower"
else:
actor = "executive"
for pattern, atype in self._action_classifiers.items():
if action.startswith(pattern):
break
else:
atype = None
if act_date is not None:
bill.add_action(action,
act_date,
chamber=actor,
classification=atype)
# TODO: Fix vote scraper
for row in doc.xpath("//h6[@id='vote-header']"):
yield from self.scrape_votes(bill, doc)
# For archived votes
if session in ["1997", "1999"]:
yield from self.add_archived_votes(bill, bill_id)
yield bill
def bill_info(self, bill_link, session, main_url):
bill_page = self.lxmlize(bill_link)
long_title = self.get_node(
bill_page, '//div[@class="main-content"]//h2').text.split()
bill_number = long_title[0]
title = ""
for x in range(2, len(long_title)):
title += long_title[x] + " "
title = title[0:-1]
if not title:
self.error("no title, skipping %s", bill_number)
return
bill_type = "resolution" if "LR" in bill_number else "bill"
bill = Bill(bill_number, session, title, classification=bill_type)
bill.add_source(main_url)
bill.add_source(bill_link)
introduced_by = self.get_node(
bill_page,
"//body/div[3]/div[2]/div[2]/div/div[3]/div[1]/ul/li[1]/a[1]/text()",
)
if not introduced_by:
introduced_by = self.get_node(
bill_page,
"//body/div[3]/div[2]/div[2]/div/div[2]/div[1]/ul/li[1]/text()",
)
introduced_by = introduced_by.split("Introduced By:")[1].strip()
introduced_by = introduced_by.strip()
bill.add_sponsorship(
name=introduced_by,
entity_type="person",
primary=True,
classification="primary",
)
action_nodes = self.get_nodes(
bill_page, '//div[@class="main-content"]/div[5]//table/tbody/tr')
for action_node in action_nodes:
date = self.get_node(action_node, "./td[1]").text
date = datetime.strptime(date, "%b %d, %Y")
# The action node may have an anchor element within it, so
# we grab all the text within.
action = self.get_node(action_node, "./td[2]").text_content()
if "Governor" in action:
actor = "executive"
elif "Speaker" in action:
actor = "legislature"
else:
actor = "legislature"
action_type = self.action_types(action)
bill.add_action(
action,
date.strftime("%Y-%m-%d"),
chamber=actor,
classification=action_type,
)
# Grabs bill version documents.
version_links = self.get_nodes(
bill_page, "/html/body/div[3]/div[2]/div[2]/div/"
"div[3]/div[2]/ul/li/a")
for version_link in version_links:
version_name = version_link.text
version_url = version_link.attrib["href"]
# replace Current w/ session number
version_url = version_url.replace("Current", session)
bill.add_version_link(version_name,
version_url,
media_type="application/pdf")
soi = self.get_nodes(bill_page,
".//a[contains(text(), 'Statement of Intent')]")
if soi:
bill.add_document_link("Statement of Intent",
soi[0].get("href"),
media_type="application/pdf")
comstmt = self.get_nodes(
bill_page, ".//a[contains(text(), 'Committee Statement')]")
if comstmt:
bill.add_document_link(
"Committee Statement",
comstmt[0].get("href"),
media_type="application/pdf",
)
fn = self.get_nodes(bill_page, ".//a[contains(text(), 'Fiscal Note')]")
if fn:
bill.add_document_link("Fiscal Note",
fn[0].get("href"),
media_type="application/pdf")
# Adds any documents related to amendments.
amendment_links = self.get_nodes(
bill_page, ".//div[contains(@class, 'amend-link')]/a")
for amendment_link in amendment_links:
amendment_name = amendment_link.text
amendment_url = amendment_link.attrib["href"]
# skip over transcripts
if "/AM/" not in amendment_url:
continue
bill.add_document_link(amendment_name,
amendment_url,
media_type="application/pdf")
yield bill
yield from self.scrape_votes(bill, bill_page, actor)
def scrape_bill(self, session, bill_url):
page = self.get(bill_url).text
page = lxml.html.fromstring(page)
page.make_links_absolute(bill_url)
try:
bill_id = page.xpath('//span[@id="lblBillNumber"]/a[1]')[0].text
except IndexError:
self.logger.warning("Something is wrong with bill page, skipping.")
return
secondary_bill_id = page.xpath('//span[@id="lblCompNumber"]/a[1]')
# checking if there is a matching bill
if secondary_bill_id:
secondary_bill_id = secondary_bill_id[0].text
# swap ids if * is in secondary_bill_id
if "*" in secondary_bill_id:
bill_id, secondary_bill_id = secondary_bill_id, bill_id
secondary_bill_id = secondary_bill_id.strip()
secondary_bill_id = secondary_bill_id.replace(" ", " ")
bill_id = bill_id.replace("*", "").replace(" ", " ").strip()
if "B" in bill_id:
bill_type = "bill"
elif "JR" in bill_id:
bill_type = "joint resolution"
elif "R" in bill_id:
bill_type = "resolution"
primary_chamber = "lower" if "H" in bill_id else "upper"
# secondary_chamber = 'upper' if primary_chamber == 'lower' else 'lower'
title = page.xpath("//span[@id='lblAbstract']")[0].text
if title is None:
msg = "%s detail page was missing title info."
self.logger.warning(msg % bill_id)
return
# bill subject
subject_pos = title.find("-")
subjects = [s.strip() for s in title[:subject_pos - 1].split(",")]
subjects = filter(None, subjects)
bill = Bill(
bill_id,
legislative_session=session,
chamber=primary_chamber,
title=title,
classification=bill_type,
)
for subject in subjects:
bill.add_subject(subject)
if secondary_bill_id:
bill.add_identifier(secondary_bill_id)
if page.xpath('//span[@id="lblCompNumber"]/a'):
companion_id = (page.xpath('//span[@id="lblCompNumber"]/a')
[0].text_content().strip())
bill.add_related_bill(
identifier=companion_id,
legislative_session=session,
relation_type="companion",
)
bill.add_source(bill_url)
# Primary Sponsor
sponsor = (page.xpath("//span[@id='lblBillPrimeSponsor']")
[0].text_content().split("by")[-1])
sponsor = sponsor.replace("*", "").strip()
if sponsor:
bill.add_sponsorship(sponsor,
classification="primary",
entity_type="person",
primary=True)
# bill text
btext = page.xpath("//span[@id='lblBillNumber']/a")[0]
bill.add_version_link("Current Version",
btext.get("href"),
media_type="application/pdf")
# documents
summary = page.xpath('//a[contains(@href, "BillSummaryArchive")]')
if summary:
bill.add_document_link("Summary", summary[0].get("href"))
fiscal = page.xpath('//span[@id="lblFiscalNote"]//a')
if fiscal:
bill.add_document_link("Fiscal Note", fiscal[0].get("href"))
amendments = page.xpath('//a[contains(@href, "/Amend/")]')
for amendment in amendments:
bill.add_version_link(
"Amendment " + amendment.text,
amendment.get("href"),
media_type="application/pdf",
)
# amendment notes in image with alt text describing doc inside <a>
amend_fns = page.xpath('//img[contains(@alt, "Fiscal Memo")]')
for afn in amend_fns:
bill.add_document_link(afn.get("alt"),
afn.getparent().get("href"),
on_duplicate="ignore")
# actions
atable = page.xpath("//table[@id='gvBillActionHistory']")[0]
actions_from_table(bill, atable)
# if there is a matching bill
if secondary_bill_id:
# secondary sponsor
secondary_sponsor = (
page.xpath("//span[@id='lblCompPrimeSponsor']")
[0].text_content().split("by")[-1])
secondary_sponsor = (secondary_sponsor.replace("*", "").replace(
")", "").strip())
# Skip black-name sponsors.
if secondary_sponsor:
bill.add_sponsorship(
secondary_sponsor,
classification="primary",
entity_type="person",
primary=True,
)
# secondary actions
if page.xpath("//table[@id='gvCoActionHistory']"):
cotable = page.xpath("//table[@id='gvCoActionHistory']")[0]
actions_from_table(bill, cotable)
# votes
yield from self.scrape_vote_events(bill, page, bill_url)
bill.actions.sort(key=lambda a: a["date"])
yield bill
def scrape_bill_page(self, chamber, session, bill_url, bill_abbreviation):
page = self.lxmlize(bill_url)
author = self.get_one_xpath(
page, "//a[@id='ctl00_PageBody_LinkAuthor']/text()")
def sbp(x):
return self.scrape_bare_page(
page.xpath("//a[contains(text(), '%s')]" %
(x))[0].attrib["href"])
authors = [x.text for x in sbp("Authors")]
try:
digests = sbp("Digests")
except IndexError:
digests = []
try:
versions = sbp("Text")
except IndexError:
versions = []
try:
amendments = sbp("Amendments")
except IndexError:
amendments = []
title = page.xpath(
"//span[@id='ctl00_PageBody_LabelShortTitle']/text()")[0]
title = title.replace("\u00a0\u00a0", " ")
actions = page.xpath("//div[@id='ctl00_PageBody_PanelBillInfo']/"
"/table[@style='font-size:small']/tr")
bill_id = page.xpath(
"//span[@id='ctl00_PageBody_LabelBillID']/text()")[0]
bill_type = self._bill_types[bill_abbreviation[1:]]
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=bill_type,
)
bill.add_source(bill_url)
authors.remove(author)
bill.add_sponsorship(author,
classification="primary",
entity_type="person",
primary=True)
for author in authors:
bill.add_sponsorship(author,
classification="cosponsor",
entity_type="person",
primary=False)
for digest in digests:
bill.add_document_link(
note=digest.text,
url=digest.attrib["href"],
media_type="application/pdf",
)
for version in versions:
bill.add_version_link(
note=version.text,
url=version.attrib["href"],
media_type="application/pdf",
)
for amendment in amendments:
if "href" in amendment.attrib:
bill.add_version_link(
note=amendment.text,
url=amendment.attrib["href"],
media_type="application/pdf",
)
flags = {
"prefiled": ["filing"],
"referred to the committee": ["referral-committee"],
"sent to the house": ["passage"],
"ordered returned to the house": ["passage"],
"ordered to the senate": ["passage"],
"signed by the governor": ["executive-signature"],
"sent to the governor": ["executive-receipt"],
}
try:
votes_link = page.xpath("//a[text() = 'Votes']")[0]
yield from self.scrape_votes(bill, votes_link.attrib["href"])
except IndexError:
# Some bills don't have any votes
pass
for action in actions:
date, chamber, page, text = [x.text for x in action.xpath(".//td")]
session_year = self.jurisdiction.legislative_sessions[-1][
"start_date"][0:4]
# Session is April -> June. Prefiles look like they're in
# January at earliest.
date += "/{}".format(session_year)
date = dt.datetime.strptime(date, "%m/%d/%Y")
chamber = self._chambers[chamber]
cat = []
for flag in flags:
if flag in text.lower():
cat += flags[flag]
bill.add_action(
description=text,
date=date.strftime("%Y-%m-%d"),
chamber=chamber,
classification=cat,
)
yield bill
def scrape_bill_type(
self,
chamber,
session,
bill_type,
type_abbr,
committee_abbr_regex=get_committee_name_regex(),
):
bills = (self.session.query(CABill).filter_by(
session_year=session).filter_by(measure_type=type_abbr))
archive_year = int(session[0:4])
not_archive_year = archive_year >= 2009
for bill in bills:
bill_session = session
if bill.session_num != "0":
bill_session += " Special Session %s" % bill.session_num
bill_id = bill.short_bill_id
if bill_id.strip() == "SB77" and session == "20052006":
continue
fsbill = Bill(bill_id, bill_session, title="", chamber=chamber)
if (bill_id.startswith("S")
and chamber == "lower") or (bill_id.startswith("A")
and chamber == "upper"):
print("!!!! BAD ID/CHAMBER PAIR !!!!", bill)
continue
# Construct a fake source url
source_url = ("http://leginfo.legislature.ca.gov/faces/"
"billNavClient.xhtml?bill_id=%s") % bill.bill_id
fsbill.add_source(source_url)
fsbill.add_version_link(bill_id,
source_url,
media_type="text/html")
title = ""
type_ = ["bill"]
subject = ""
all_titles = set()
summary = ""
# Get digest test (aka "summary") from latest version.
if bill.versions and not_archive_year:
version = bill.versions[-1]
nsmap = version.xml.nsmap
xpath = "//caml:DigestText/xhtml:p"
els = version.xml.xpath(xpath, namespaces=nsmap)
chunks = []
for el in els:
t = etree_text_content(el)
t = re.sub(r"\s+", " ", t)
t = re.sub(r"\)(\S)", lambda m: ") %s" % m.group(1), t)
chunks.append(t)
summary = "\n\n".join(chunks)
for version in bill.versions:
if not version.bill_xml:
continue
version_date = self._tz.localize(
version.bill_version_action_date)
# create a version name to match the state's format
# 02/06/17 - Enrolled
version_date_human = version_date.strftime("%m/%d/%y")
version_name = "{} - {}".format(version_date_human,
version.bill_version_action)
version_base = "https://leginfo.legislature.ca.gov/faces"
version_url_pdf = "{}/billPdf.xhtml?bill_id={}&version={}".format(
version_base, version.bill_id, version.bill_version_id)
fsbill.add_version_link(
version_name,
version_url_pdf,
media_type="application/pdf",
date=version_date.date(),
)
# CA is inconsistent in that some bills have a short title
# that is longer, more descriptive than title.
if bill.measure_type in ("AB", "SB"):
impact_clause = clean_title(version.title)
title = clean_title(version.short_title)
else:
impact_clause = None
if len(version.title) < len(
version.short_title) and not version.title.lower(
).startswith("an act"):
title = clean_title(version.short_title)
else:
title = clean_title(version.title)
if title:
all_titles.add(title)
type_ = [bill_type]
if version.appropriation == "Yes":
type_.append("appropriation")
tags = []
if version.fiscal_committee == "Yes":
tags.append("fiscal committee")
if version.local_program == "Yes":
tags.append("local program")
if version.urgency == "Yes":
tags.append("urgency")
if version.taxlevy == "Yes":
tags.append("tax levy")
if version.subject:
subject = clean_title(version.subject)
if not title:
self.warning("Couldn't find title for %s, skipping" % bill_id)
continue
fsbill.title = title
if summary:
fsbill.add_abstract(summary, note="summary")
fsbill.classification = type_
fsbill.subject = [subject] if subject else []
fsbill.extras["impact_clause"] = impact_clause
fsbill.extras["tags"] = tags
# We don't want the current title in alternate_titles
all_titles.remove(title)
for title in all_titles:
fsbill.add_title(title)
for author in version.authors:
fsbill.add_sponsorship(
author.name,
classification=SPONSOR_TYPES[author.contribution],
primary=author.primary_author_flg == "Y",
entity_type="person",
)
# fsbill.sponsorships[-1]['extras'] = {'official_type': author.contribution}
seen_actions = set()
for action in bill.actions:
if not action.action:
# NULL action text seems to be an error on CA's part,
# unless it has some meaning I'm missing
continue
actor = action.actor or chamber
actor = actor.strip()
match = re.match(r"(Assembly|Senate)($| \(Floor)", actor)
if match:
actor = {
"Assembly": "lower",
"Senate": "upper"
}[match.group(1)]
elif actor.startswith("Governor"):
actor = "executive"
else:
def replacer(matchobj):
if matchobj:
return {
"Assembly": "lower",
"Senate": "upper"
}[matchobj.group()]
else:
return matchobj.group()
actor = re.sub(r"^(Assembly|Senate)", replacer, actor)
type_ = []
act_str = action.action
act_str = re.sub(r"\s+", " ", act_str)
attrs = self.categorizer.categorize(act_str)
# Add in the committee strings of the related committees, if any.
kwargs = attrs
matched_abbrs = committee_abbr_regex.findall(action.action)
if re.search(r"Com[s]?. on",
action.action) and not matched_abbrs:
msg = "Failed to extract committee abbr from %r."
self.logger.warning(msg % action.action)
if matched_abbrs:
committees = []
for abbr in matched_abbrs:
try:
name = self.committee_abbr_to_name(chamber, abbr)
committees.append(name)
except KeyError:
msg = ("Mapping contains no committee name for "
"abbreviation %r. Action text was %r.")
args = (abbr, action.action)
self.warning(msg % args)
committees = filter(None, committees)
kwargs["committees"] = committees
code = re.search(r"C[SXZ]\d+", actor)
if code is not None:
code = code.group()
kwargs["actor_info"] = {"committee_code": code}
if not_archive_year:
assert len(list(committees)) == len(matched_abbrs)
for committee, abbr in zip(committees, matched_abbrs):
act_str = act_str.replace("Coms. on ", "")
act_str = act_str.replace("Com. on " + abbr, committee)
act_str = act_str.replace(abbr, committee)
if not act_str.endswith("."):
act_str = act_str + "."
# Determine which chamber the action originated from.
changed = False
for committee_chamber in ["upper", "lower", "legislature"]:
if actor.startswith(committee_chamber):
actor = committee_chamber
changed = True
break
if not changed:
actor = "legislature"
if actor != action.actor:
actor_info = kwargs.get("actor_info", {})
actor_info["details"] = action.actor
kwargs["actor_info"] = actor_info
# Add strings for related legislators, if any.
rgx = r"(?:senator|assembly[mwp][^ .,:;]+)\s+[^ .,:;]+"
legislators = re.findall(rgx, action.action, re.I)
if legislators:
kwargs["legislators"] = legislators
date = action.action_date
date = self._tz.localize(date)
date = date.date()
if (actor, act_str, date) in seen_actions:
continue
kwargs.update(self.categorizer.categorize(act_str))
action = fsbill.add_action(
act_str,
date.strftime("%Y-%m-%d"),
chamber=actor,
classification=kwargs["classification"],
)
for committee in kwargs.get("committees", []):
action.add_related_entity(committee,
entity_type="organization")
seen_actions.add((actor, act_str, date))
source_url = (
"http://leginfo.legislature.ca.gov/faces/billVotesClient.xhtml?"
)
source_url += f"bill_id={session}{bill.session_num}{fsbill.identifier}"
# Votes for non archived years
if archive_year > 2009:
for vote_num, vote in enumerate(bill.votes):
if vote.vote_result == "(PASS)":
result = True
else:
result = False
if not vote.location:
continue
full_loc = vote.location.description
first_part = full_loc.split(" ")[0].lower()
if first_part in ["asm", "assembly"]:
vote_chamber = "lower"
# vote_location = ' '.join(full_loc.split(' ')[1:])
elif first_part.startswith("sen"):
vote_chamber = "upper"
# vote_location = ' '.join(full_loc.split(' ')[1:])
else:
# raise ScrapeError("Bad location: %s" % full_loc) # To uncomment
continue
if vote.motion:
motion = vote.motion.motion_text or ""
else:
motion = ""
if "Third Reading" in motion or "3rd Reading" in motion:
vtype = "passage"
elif "Do Pass" in motion:
vtype = "passage"
else:
vtype = "other"
motion = motion.strip()
motion = re.compile(r"(\w+)( Extraordinary)? Session$",
re.IGNORECASE).sub("", motion)
motion = re.compile(r"^(Senate|Assembly) ",
re.IGNORECASE).sub("", motion)
motion = re.sub(r"^(SCR|SJR|SB|AB|AJR|ACR)\s?\d+ \w+\.? ",
"", motion)
motion = re.sub(r" \(\w+\)$", "", motion)
motion = re.sub(r"(SCR|SB|AB|AJR|ACR)\s?\d+ \w+\.?$", "",
motion)
motion = re.sub(
r"(SCR|SJR|SB|AB|AJR|ACR)\s?\d+ \w+\.? "
r"Urgency Clause$",
"(Urgency Clause)",
motion,
)
motion = re.sub(r"\s+", " ", motion)
if not motion:
self.warning("Got blank motion on vote for %s" %
bill_id)
continue
# XXX this is responsible for all the CA 'committee' votes, not
# sure if that's a feature or bug, so I'm leaving it as is...
# vote_classification = chamber if (vote_location == 'Floor') else 'committee'
# org = {
# 'name': vote_location,
# 'classification': vote_classification
# }
fsvote = VoteEvent(
motion_text=motion,
start_date=self._tz.localize(vote.vote_date_time),
result="pass" if result else "fail",
classification=vtype,
# organization=org,
chamber=vote_chamber,
bill=fsbill,
)
fsvote.extras = {"threshold": vote.threshold}
fsvote.add_source(source_url)
fsvote.pupa_id = source_url + "#" + str(vote_num)
rc = {"yes": [], "no": [], "other": []}
for record in vote.votes:
if record.vote_code == "AYE":
rc["yes"].append(record.legislator_name)
elif record.vote_code.startswith("NO"):
rc["no"].append(record.legislator_name)
else:
rc["other"].append(record.legislator_name)
# Handle duplicate votes
for key in rc.keys():
rc[key] = list(set(rc[key]))
for key, voters in rc.items():
for voter in voters:
fsvote.vote(key, voter)
# Set counts by summed votes for accuracy
fsvote.set_count(key, len(voters))
yield fsvote
if len(bill.votes) > 0 and archive_year <= 2009:
vote_page_url = (
"http://leginfo.legislature.ca.gov/faces/billVotesClient.xhtml?"
)
vote_page_url += (
f"bill_id={session}{bill.session_num}{fsbill.identifier}")
# parse the bill data page, finding the latest html text
data = self.get(vote_page_url).content
doc = html.fromstring(data)
doc.make_links_absolute(vote_page_url)
num_of_votes = len(doc.xpath("//div[@class='status']"))
for vote_section in range(1, num_of_votes + 1):
lines = doc.xpath(
f"//div[@class='status'][{vote_section}]//div[@class='statusRow']"
)
date, result, motion, vtype, location = "", "", "", "", ""
votes = {}
for line in lines:
line = line.text_content().split()
if line[0] == "Date":
date = line[1]
date = datetime.datetime.strptime(date, "%m/%d/%y")
date = self._tz.localize(date)
elif line[0] == "Result":
result = "pass" if "PASS" in line[1] else "fail"
elif line[0] == "Motion":
motion = " ".join(line[1:])
elif line[0] == "Location":
location = " ".join(line[1:])
elif len(line) > 1:
if line[0] == "Ayes" and line[1] != "Count":
votes["yes"] = line[1:]
elif line[0] == "Noes" and line[1] != "Count":
votes["no"] = line[1:]
elif line[0] == "NVR" and line[1] != "Count":
votes["not voting"] = line[1:]
# Determine chamber based on location
first_part = location.split(" ")[0].lower()
vote_chamber = ""
if first_part in ["asm", "assembly"]:
vote_chamber = "lower"
elif first_part.startswith("sen"):
vote_chamber = "upper"
if "Third Reading" in motion or "3rd Reading" in motion:
vtype = "passage"
elif "Do Pass" in motion:
vtype = "passage"
else:
vtype = "other"
if len(motion) > 0:
fsvote = VoteEvent(
motion_text=motion,
start_date=date,
result=result,
classification=vtype,
chamber=vote_chamber,
bill=fsbill,
)
fsvote.add_source(vote_page_url)
fsvote.pupa_id = vote_page_url + "#" + str(
vote_section)
for how_voted, voters in votes.items():
for voter in voters:
voter = voter.replace(",", "")
fsvote.vote(how_voted, voter)
yield fsvote
yield fsbill
self.session.expire_all()
def scrape_bill(self, chamber, session, bill_id):
# try and get bill for the first year of the session biennium
url = "http://legislature.mi.gov/doc.aspx?%s-%s" % (
session[:4],
bill_id.replace(" ", "-"),
)
html = self.get(url).text
# Otherwise, try second year of the session biennium
if (
"Page Not Found" in html
or "The bill you are looking for is not available yet" in html
):
url = "http://legislature.mi.gov/doc.aspx?%s-%s" % (
session[-4:],
bill_id.replace(" ", "-"),
)
html = self.get(url).text
if (
"Page Not Found" in html
or "The bill you are looking for is not available yet" in html
):
self.warning("Cannot open bill page for {}; skipping".format(bill_id))
return
doc = lxml.html.fromstring(html)
doc.make_links_absolute("http://legislature.mi.gov")
title = doc.xpath('//span[@id="frg_billstatus_ObjectSubject"]')[
0
].text_content()
# get B/R/JR/CR part and look up bill type
bill_type = bill_types[bill_id.split(" ")[0][1:]]
bill = Bill(bill_id, session, title, chamber=chamber, classification=bill_type)
bill.add_source(url)
# sponsors
sponsors = doc.xpath('//span[@id="frg_billstatus_SponsorList"]/a')
for sponsor in sponsors:
name = sponsor.text.replace(u"\xa0", " ")
# sometimes district gets added as a link
if name.isnumeric():
continue
if len(sponsors) > 1:
classification = (
"primary"
if sponsor.tail and "primary" in sponsor.tail
else "cosponsor"
)
else:
classification = "primary"
bill.add_sponsorship(
name=name.strip(),
chamber=chamber,
entity_type="person",
primary=classification == "primary",
classification=classification,
)
bill.subject = doc.xpath('//span[@id="frg_billstatus_CategoryList"]/a/text()')
# actions (skip header)
for row in doc.xpath('//table[@id="frg_billstatus_HistoriesGridView"]/tr')[1:]:
tds = row.xpath("td") # date, journal link, action
date = tds[0].text_content()
journal = tds[1].text_content()
action = tds[2].text_content()
try:
date = TIMEZONE.localize(datetime.datetime.strptime(date, "%m/%d/%y"))
except ValueError:
try:
date = TIMEZONE.localize(
datetime.datetime.strptime(date, "%m/%d/%Y")
)
except ValueError:
self.warning(
"{} has action with invalid date. Skipping Action".format(
bill_id
)
)
continue
# use journal for actor
# then fall back to upper/lower case
# Journal entries are often posted with 'Expected Soon' as the cite,
# then changed to the journal entry.
if "SJ" in journal.upper():
actor = "upper"
elif "HJ" in journal.upper():
actor = "lower"
elif action.split()[0].islower():
actor = "lower"
elif action.split()[0].isupper():
actor = "upper"
else:
actor = "legislature"
classification = categorize_action(action)
bill.add_action(action, date, chamber=actor, classification=classification)
# check if action mentions a sub
submatch = re.search(
r"WITH SUBSTITUTE\s+([\w\-\d]+)", action, re.IGNORECASE
)
if submatch and tds[2].xpath("a"):
version_url = tds[2].xpath("a/@href")[0]
version_name = tds[2].xpath("a/text()")[0].strip()
version_name = "Substitute {}".format(version_name)
self.info("Found Substitute {}".format(version_url))
if version_url.lower().endswith(".pdf"):
mimetype = "application/pdf"
elif version_url.lower().endswith(".htm"):
mimetype = "text/html"
bill.add_version_link(version_name, version_url, media_type=mimetype)
# check if action mentions a vote
rcmatch = re.search(r"Roll Call # (\d+)", action, re.IGNORECASE)
if rcmatch:
rc_num = rcmatch.groups()[0]
# in format mileg.aspx?page=getobject&objectname=2011-SJ-02-10-011
journal_link = tds[1].xpath("a/@href")
if journal_link:
objectname = journal_link[0].rsplit("=", 1)[-1]
chamber_name = {"upper": "Senate", "lower": "House"}[actor]
vote_url = BASE_URL + "/documents/%s/Journal/%s/htm/%s.htm" % (
session,
chamber_name,
objectname,
)
results = self.parse_roll_call(vote_url, rc_num, session)
if results is not None:
vote_passed = len(results["yes"]) > len(results["no"])
vote = VoteEvent(
start_date=date,
chamber=actor,
bill=bill,
motion_text=action,
result="pass" if vote_passed else "fail",
classification="passage",
)
# check the expected counts vs actual
count = re.search(r"YEAS (\d+)", action, re.IGNORECASE)
count = int(count.groups()[0]) if count else 0
if count != len(results["yes"]):
self.warning(
"vote count mismatch for %s %s, %d != %d"
% (bill_id, action, count, len(results["yes"]))
)
count = re.search(r"NAYS (\d+)", action, re.IGNORECASE)
count = int(count.groups()[0]) if count else 0
if count != len(results["no"]):
self.warning(
"vote count mismatch for %s %s, %d != %d"
% (bill_id, action, count, len(results["no"]))
)
vote.set_count("yes", len(results["yes"]))
vote.set_count("no", len(results["no"]))
vote.set_count("other", len(results["other"]))
possible_vote_results = ["yes", "no", "other"]
for pvr in possible_vote_results:
for name in results[pvr]:
if session == "2017-2018":
names = name.split("\t")
for n in names:
vote.vote(pvr, name.strip())
else:
# Prevents voter names like "House Bill No. 4451, entitled" and other sentences
if len(name.split()) < 5:
vote.vote(pvr, name.strip())
vote.add_source(vote_url)
yield vote
else:
self.warning("missing journal link for %s %s" % (bill_id, journal))
# versions
for row in doc.xpath('//table[@id="frg_billstatus_DocumentGridTable"]/tr'):
parsed = self.parse_doc_row(row)
if parsed:
name, url = parsed
if url.endswith(".pdf"):
mimetype = "application/pdf"
elif url.endswith(".htm"):
mimetype = "text/html"
bill.add_version_link(name, url, media_type=mimetype)
# documents
for row in doc.xpath('//table[@id="frg_billstatus_HlaTable"]/tr'):
document = self.parse_doc_row(row)
if document:
name, url = document
bill.add_document_link(name, url)
for row in doc.xpath('//table[@id="frg_billstatus_SfaTable"]/tr'):
document = self.parse_doc_row(row)
if document:
name, url = document
bill.add_document_link(name, url)
yield bill
def scrape_bill(self, chamber, session, bill_id, url):
page = self.lxmlize(url)
(header, ) = page.xpath('//h3[@class="heading"]/text()')
title = header.replace(bill_id, "").strip()
if ".B. " in bill_id:
bill_type = "bill"
elif bill_id.startswith("H.R. ") or bill_id.startswith("S.R. "):
bill_type = "resolution"
elif ".C.R. " in bill_id:
bill_type = "concurrent resolution"
elif ".J.R. " in bill_id:
bill_type = "joint resolution"
for flag in SUB_BLACKLIST:
if flag in bill_id:
bill_id = bill_id.replace(flag, " ")
bill_id = re.sub(r"\s+", " ", bill_id).strip().replace(".", "")
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=bill_type,
)
bill.add_source(url)
primary_info = page.xpath('//div[@id="billsponsordiv"]')
for info in primary_info:
try:
(title, name) = [
x.strip() for x in info.xpath(".//text()") if x.strip()
]
except ValueError:
self.warning(
"Could not find sponsor's name for {}".format(bill_id))
continue
assert title == "Bill Sponsor:"
name = name.replace("Sen. ", "").replace("Rep. ", "")
bill.add_sponsorship(name,
classification="primary",
entity_type="person",
primary=True)
floor_info = page.xpath('//div[@id="floorsponsordiv"]//text()')
floor_info = [x.strip() for x in floor_info if x.strip()]
if len(floor_info) in (0, 1):
# This indicates that no floor sponsor was found
pass
elif len(floor_info) == 2:
assert floor_info[0] == "Floor Sponsor:"
floor_sponsor = floor_info[1].replace("Sen. ",
"").replace("Rep. ", "")
bill.add_sponsorship(
floor_sponsor,
classification="cosponsor",
entity_type="person",
primary=False,
)
else:
self.warning("Unexpected floor sponsor HTML found")
versions = page.xpath(
'//b[text()="Bill Text"]/following-sibling::ul/li/'
'a[text() and not(text()=" ")]')
for version in versions:
# sometimes the href is on the following <a> tag and the tag we
# have has an onclick
url = version.get("href")
if not url:
url = version.xpath("following-sibling::a[1]/@href")[0]
bill.add_version_link(version.xpath("text()")[0].strip(),
url,
media_type="application/pdf")
for related in page.xpath(
'//b[text()="Related Documents "]/following-sibling::ul/li/'
'a[contains(@class,"nlink")]'):
href = related.xpath("@href")[0]
if ".fn.pdf" in href:
bill.add_document_link("Fiscal Note",
href,
media_type="application/pdf")
else:
text = related.xpath("text()")[0]
bill.add_document_link(text,
href,
media_type="application/pdf")
subjects = []
for link in page.xpath("//a[contains(@href, 'RelatedBill')]"):
subjects.append(link.text.strip())
bill.subject = subjects
if page.xpath('//div[@id="billStatus"]//table'):
status_table = page.xpath('//div[@id="billStatus"]//table')[0]
yield from self.parse_status(bill, status_table, chamber)
yield bill
def scrape(self, session=None, chambers=None):
# Bills endpoint can sometimes take a very long time to load
self.timeout = 300
if not session:
session = self.latest_session()
self.info("no session, using %s", session)
if int(session) < 128:
raise AssertionError("No data for period {}".format(session))
elif int(session) < 131:
# they changed their data format starting in 131st and added
# an undocumented API
yield from self.old_scrape(session)
else:
chamber_dict = {
"Senate": "upper",
"House": "lower",
"House of Representatives": "lower",
"house": "lower",
"senate": "upper",
}
# so presumably not everything passes, but we haven't
# seen anything not pass yet, so we'll need to wait
# till it fails and get the right language in here
vote_results = {
"approved": True,
"passed": True,
"adopted": True,
"true": True,
"false": False,
"failed": False,
True: True,
False: False,
}
action_dict = {
"ref_ctte_100": "referral-committee",
"intro_100": "introduction",
"intro_101": "introduction",
"pass_300": "passage",
"intro_110": "reading-1",
"refer_210": "referral-committee",
"crpt_301": None,
"crpt_317": None,
"concur_606": "passage",
"pass_301": "passage",
"refer_220": "referral-committee",
"intro_102": ["introduction", "passage"],
"intro_105": ["introduction", "passage"],
"intro_ref_ctte_100": "referral-committee",
"refer_209": None,
"intro_108": ["introduction", "passage"],
"intro_103": ["introduction", "passage"],
"msg_reso_503": "passage",
"intro_107": ["introduction", "passage"],
"imm_consid_360": "passage",
"refer_213": None,
"adopt_reso_100": "passage",
"adopt_reso_110": "passage",
"msg_507": "amendment-passage",
"confer_713": None,
"concur_603": None,
"confer_712": None,
"msg_506": "amendment-failure",
"receive_message_100": "passage",
"motion_920": None,
"concur_611": None,
"confer_735": None,
"third_429": None,
"final_501": None,
"concur_608": None,
"infpass_217": "passage",
}
base_url = "https://search-prod.lis.state.oh.us"
first_page = base_url
first_page += "/solarapi/v1/general_assembly_{session}/".format(
session=session)
legislators = self.get_legislator_ids(first_page)
all_amendments = self.get_other_data_source(
first_page, base_url, "amendments")
all_fiscals = self.get_other_data_source(first_page, base_url,
"fiscals")
all_synopsis = self.get_other_data_source(first_page, base_url,
"synopsiss")
all_analysis = self.get_other_data_source(first_page, base_url,
"analysiss")
for row in self.get_bill_rows(session):
(
spacer,
number_link,
_ga,
title,
primary_sponsor,
status,
spacer,
) = row.xpath("td")
# S.R.No.1 -> SR1
bill_id = number_link.text_content().replace("No.", "")
bill_id = bill_id.replace(".", "").replace(" ", "")
# put one space back in between type and number
bill_id = re.sub(r"([a-zA-Z]+)(\d+)", r"\1 \2", bill_id)
title = title.text_content().strip()
title = re.sub(r"^Title", "", title)
chamber = "lower" if "H" in bill_id else "upper"
classification = "bill" if "B" in bill_id else "resolution"
if not title and session == "134" and bill_id == "HR 35":
# Exception for HR 35 which is a real bill
title = "No title provided"
elif not title:
self.warning(f"no title for {bill_id}, skipping")
continue
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=classification,
)
bill.add_source(number_link.xpath("a/@href")[0])
if (session, bill_id) in BAD_BILLS:
self.logger.warning(
f"Skipping details for known bad bill {bill_id}")
yield bill
continue
# get bill from API
bill_api_url = (
"https://search-prod.lis.state.oh.us/solarapi/v1/"
"general_assembly_{}/{}/{}/".format(
session,
"bills" if "B" in bill_id else "resolutions",
bill_id.lower().replace(" ", ""),
))
data = self.get(bill_api_url, verify=False).json()
if len(data["items"]) == 0:
self.logger.warning(
"Data for bill {bill_id} has empty 'items' array,"
" cannot process related information".format(
bill_id=bill_id.lower().replace(" ", "")))
yield bill
continue
# add title if no short title
if not bill.title:
bill.title = data["items"][0]["longtitle"]
bill.add_title(data["items"][0]["longtitle"], "long title")
# this stuff is version-specific
for version in data["items"]:
version_name = version["version"]
version_link = base_url + version["pdfDownloadLink"]
bill.add_version_link(version_name,
version_link,
media_type="application/pdf")
# we'll use latest bill_version for everything else
bill_version = data["items"][0]
bill.add_source(bill_api_url)
# subjects
for subj in bill_version["subjectindexes"]:
try:
bill.add_subject(subj["primary"])
except KeyError:
pass
try:
secondary_subj = subj["secondary"]
except KeyError:
secondary_subj = ""
if secondary_subj:
bill.add_subject(secondary_subj)
# sponsors
sponsors = bill_version["sponsors"]
for sponsor in sponsors:
sponsor_name = self.get_sponsor_name(sponsor)
bill.add_sponsorship(
sponsor_name,
classification="primary",
entity_type="person",
primary=True,
)
cosponsors = bill_version["cosponsors"]
for sponsor in cosponsors:
sponsor_name = self.get_sponsor_name(sponsor)
bill.add_sponsorship(
sponsor_name,
classification="cosponsor",
entity_type="person",
primary=False,
)
try:
action_doc = self.get(base_url +
bill_version["action"][0]["link"])
except scrapelib.HTTPError:
pass
else:
actions = action_doc.json()
for action in reversed(actions["items"]):
actor = chamber_dict[action["chamber"]]
action_desc = action["description"]
try:
action_type = action_dict[action["actioncode"]]
except KeyError:
self.warning(
"Unknown action {desc} with code {code}."
" Add it to the action_dict"
".".format(desc=action_desc,
code=action["actioncode"]))
action_type = None
date = self._tz.localize(
datetime.datetime.strptime(action["datetime"],
"%Y-%m-%dT%H:%M:%S"))
date = "{:%Y-%m-%d}".format(date)
bill.add_action(action_desc,
date,
chamber=actor,
classification=action_type)
# attach documents gathered earlier
self.add_document(all_amendments, bill_id, "amendment", bill,
base_url)
self.add_document(all_fiscals, bill_id, "fiscal", bill,
base_url)
self.add_document(all_synopsis, bill_id, "synopsis", bill,
base_url)
self.add_document(all_analysis, bill_id, "analysis", bill,
base_url)
# votes
vote_url = base_url + bill_version["votes"][0]["link"]
try:
vote_doc = self.get(vote_url)
except scrapelib.HTTPError:
self.warning(
"Vote page not loading; skipping: {}".format(vote_url))
yield bill
continue
votes = vote_doc.json()
yield from self.process_vote(
votes,
vote_url,
base_url,
bill,
legislators,
chamber_dict,
vote_results,
)
vote_url = base_url
vote_url += bill_version["cmtevotes"][0]["link"]
try:
vote_doc = self.get(vote_url)
except scrapelib.HTTPError:
self.warning(
"Vote page not loading; skipping: {}".format(vote_url))
yield bill
continue
votes = vote_doc.json()
yield from self.process_vote(
votes,
vote_url,
base_url,
bill,
legislators,
chamber_dict,
vote_results,
)
if data["items"][0]["effective_date"]:
effective_date = datetime.datetime.strptime(
data["items"][0]["effective_date"], "%Y-%m-%d")
effective_date = self._tz.localize(effective_date)
# the OH website adds an action that isn't in the action list JSON.
# It looks like:
# Effective 7/6/18
effective_date_oh = "{:%-m/%-d/%y}".format(effective_date)
effective_action = "Effective {}".format(effective_date_oh)
bill.add_action(
effective_action,
effective_date,
chamber="executive",
classification=["became-law"],
)
# we have never seen a veto or a disapprove, but they seem important.
# so we'll check and throw an error if we find one
# life is fragile. so are our scrapers.
if "veto" in bill_version:
veto_url = base_url + bill_version["veto"][0]["link"]
veto_json = self.get(veto_url).json()
if len(veto_json["items"]) > 0:
raise AssertionError("Whoa, a veto! We've never"
" gotten one before."
" Go write some code to deal"
" with it: {}".format(veto_url))
if "disapprove" in bill_version:
disapprove_url = base_url + bill_version["disapprove"][0][
"link"]
disapprove_json = self.get(disapprove_url).json()
if len(disapprove_json["items"]) > 0:
raise AssertionError(
"Whoa, a disapprove! We've never"
" gotten one before."
" Go write some code to deal "
"with it: {}".format(disapprove_url))
yield bill
def scrape_bills(self, chamber_to_scrape, session):
url = (
"http://billstatus.ls.state.ms.us/%s/pdf/all_measures/allmsrs.xml"
% session)
bill_dir_page = self.get(url)
root = lxml.etree.fromstring(bill_dir_page.content)
for mr in root.xpath("//LASTACTION/MSRGROUP"):
bill_id = mr.xpath("string(MEASURE)").replace(" ", "")
if bill_id[0] == "S":
chamber = "upper"
else:
chamber = "lower"
bill_type = {
"B": "bill",
"C": "concurrent resolution",
"R": "resolution",
"N": "nomination",
}[bill_id[1]]
# just skip past bills that are of the wrong chamber
if chamber != chamber_to_scrape:
continue
link = mr.xpath("string(ACTIONLINK)").replace("..", "")
main_doc = mr.xpath("string(MEASURELINK)").replace("../../../", "")
main_doc_url = "http://billstatus.ls.state.ms.us/%s" % main_doc
bill_details_url = "http://billstatus.ls.state.ms.us/%s/pdf%s" % (
session,
link,
)
try:
details_page = self.get(bill_details_url)
except scrapelib.HTTPError:
self.warning(
"Bill page not loading for {}; skipping".format(bill_id))
continue
page = details_page.content
# Some pages have the (invalid) byte 11 sitting around. Just drop
# them out. Might as well.
details_root = lxml.etree.fromstring(page)
title = details_root.xpath("string(//SHORTTITLE)")
longtitle = details_root.xpath("string(//LONGTITLE)")
if title == "":
self.warning(f"No title yet for {bill_id}, skipping")
return
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=bill_type,
)
bill.extras["summary"] = longtitle
bill.add_source(main_doc_url)
# sponsors
main_sponsor = details_root.xpath("string(//P_NAME)").split()
if main_sponsor:
main_sponsor = main_sponsor[0]
main_sponsor_link = details_root.xpath(
"string(//P_LINK)").replace(" ", "_")
main_sponsor_url = ("http://billstatus.ls.state.ms.us/%s/"
"pdf/%s") % (
session,
main_sponsor_link.strip("../"),
)
type = "primary"
bill.add_source(main_sponsor_url)
bill.add_sponsorship(
self.clean_voter_name(main_sponsor),
classification=type,
entity_type="person",
primary=True,
)
for author in details_root.xpath("//AUTHORS/ADDITIONAL"):
leg = author.xpath("string(CO_NAME)").replace(" ", "_")
if leg:
leg_url = ("http://billstatus.ls.state.ms.us/%s/"
"pdf/House_authors/%s.xml") % (session, leg)
type = "cosponsor"
bill.add_source(leg_url)
bill.add_sponsorship(
self.clean_voter_name(leg),
classification=type,
entity_type="person",
primary=False,
)
# Versions
curr_version = details_root.xpath("string(//CURRENT_OTHER"
")").replace("../../../../", "")
if curr_version != "":
curr_version_url = "http://billstatus.ls.state.ms.us/" + curr_version
bill.add_version_link(
"Current version",
curr_version_url,
on_duplicate="ignore",
media_type="text/html",
)
curr_pdf_url = re.sub("html?", "pdf", curr_version_url)
bill.add_version_link(
"Current version",
curr_pdf_url,
on_duplicate="ignore",
media_type="application/pdf",
)
intro_version = details_root.xpath(
"string(//INTRO_OTHER)").replace("../../../../", "")
if intro_version != "":
intro_version_url = "http://billstatus.ls.state.ms.us/" + intro_version
bill.add_version_link(
"As Introduced",
intro_version_url,
on_duplicate="ignore",
media_type="text/html",
)
intro_pdf_url = re.sub("html?", "pdf", intro_version_url)
bill.add_version_link(
"As Introduced",
intro_pdf_url,
on_duplicate="ignore",
media_type="application/pdf",
)
comm_version = details_root.xpath("string(//CMTESUB_OTHER"
")").replace("../../../../", "")
if comm_version.find("documents") != -1:
comm_version_url = "http://billstatus.ls.state.ms.us/" + comm_version
bill.add_version_link(
"Committee Substitute",
comm_version_url,
on_duplicate="ignore",
media_type="text/html",
)
comm_pdf_url = re.sub("html?", "pdf", comm_version_url)
bill.add_version_link(
"Committee Substitute",
comm_pdf_url,
on_duplicate="ignore",
media_type="application/pdf",
)
passed_version = details_root.xpath("string(//PASSED_OTHER"
")").replace(
"../../../../", "")
if passed_version.find("documents") != -1:
passed_version_url = ("http://billstatus.ls.state.ms.us/" +
passed_version)
title = "As Passed the " + chamber
bill.add_version_link(
title,
passed_version_url,
on_duplicate="ignore",
media_type="text/html",
)
passed_pdf_url = re.sub("html?", "pdf", passed_version_url)
bill.add_version_link(
title,
passed_pdf_url,
on_duplicate="ignore",
media_type="application/pdf",
)
asg_version = details_root.xpath("string(//ASG_OTHER)").replace(
"../../../../", "")
if asg_version.find("documents") != -1:
asg_version_url = "http://billstatus.ls.state.ms.us/" + asg_version
bill.add_version_link(
"Approved by the Governor",
asg_version_url,
on_duplicate="ignore",
media_type="text/html",
)
asg_pdf_url = re.sub("html?", "pdf", asg_version_url)
bill.add_version_link(
"Approved by the Governor",
asg_pdf_url,
on_duplicate="ignore",
media_type="application/pdf",
)
# amendments
# ex: http://billstatus.ls.state.ms.us/2018/pdf/history/HB/HB1040.xml
for amd in details_root.xpath("//AMENDMENTS/*"):
if amd.tag == "HAM":
name = amd.xpath("HAM_DESC[1]/text()")[0]
name = append_parens(amd, "HAM_DISP", name)
name = append_parens(amd, "HAM_VDESC", name)
pdf_url = amd.xpath("string(HAM_PDF"
")").replace("../", "")
html_url = amd.xpath("string(HAM_OTHER"
")").replace("../", "")
elif amd.tag == "SAM":
name = amd.xpath("SAM_DESC[1]/text()")[0]
name = append_parens(amd, "SAM_DISP", name)
name = append_parens(amd, "SAM_VDESC", name)
pdf_url = amd.xpath("string(SAM_PDF"
")").replace("../", "")
html_url = amd.xpath("string(SAM_OTHER"
")").replace("../", "")
elif amd.tag == "AMRPT":
name = amd.xpath("AMRPT_DESC[1]/text()")[0]
pdf_url = amd.xpath("string(AMRPT_PDF"
")").replace("../", "")
html_url = amd.xpath("string(AMRPT_OTHER"
")").replace("../", "")
pdf_url = "http://billstatus.ls.state.ms.us/" + pdf_url
html_url = "http://billstatus.ls.state.ms.us/" + html_url
if "adopted" in name.lower(
) or "amendment report" in name.lower():
bill.add_version_link(
name,
pdf_url,
on_duplicate="ignore",
media_type="application/pdf",
)
bill.add_version_link(name,
html_url,
on_duplicate="ignore",
media_type="text/html")
# avoid duplicate votes
seen_votes = set()
# Actions
for action in details_root.xpath("//HISTORY/ACTION"):
# action_num = action.xpath('string(ACT_NUMBER)').strip()
# action_num = int(action_num)
act_vote = action.xpath("string(ACT_VOTE)").replace(
"../../../..", "")
action_desc = action.xpath("string(ACT_DESC)")
date, action_desc = action_desc.split(" ", 1)
date = date + "/" + session[0:4]
date = datetime.strptime(date, "%m/%d/%Y")
if action_desc.startswith("(H)"):
actor = "lower"
action = action_desc[4:]
elif action_desc.startswith("(S)"):
actor = "upper"
action = action_desc[4:]
else:
actor = "executive"
action = action_desc
if "Veto" in action and actor == "executive":
version_path = details_root.xpath("string(//VETO_OTHER)")
version_path = version_path.replace("../../../../", "")
version_url = "http://billstatus.ls.state.ms.us/" + version_path
bill.add_document_link("Veto", version_url)
atype = "other"
for prefix, prefix_type in self._action_types:
if action.startswith(prefix):
atype = prefix_type
break
bill.add_action(
action,
self._tz.localize(date),
chamber=actor,
classification=atype if atype != "other" else None,
)
# use committee names as scraped subjects
subjects = details_root.xpath("//H_NAME/text()")
subjects += details_root.xpath("//S_NAME/text()")
for subject in subjects:
if subject not in bill.subject:
bill.add_subject(subject)
if act_vote:
vote_url = "http://billstatus.ls.state.ms.us%s" % act_vote
if vote_url not in seen_votes:
seen_votes.add(vote_url)
yield from self.scrape_votes(vote_url, action, date,
actor, bill)
bill.add_source(bill_details_url)
yield bill
def scrape(self, session=None):
if not session:
session = self.jurisdiction.legislative_sessions[-1]["identifier"]
self.info("no session specified, using %s", session)
chamber_types = {
"H": "lower",
"S": "upper",
"G": "executive",
"C": "legislature",
}
# pull the current session's details to tell if it's a special
session_details = next(
each for each in self.jurisdiction.legislative_sessions
if each["identifier"] == session)
is_special = False
if ("classification" in session_details
and session_details["classification"] == "special"):
is_special = True
session_id = SESSION_SITE_IDS[session]
self.init_sftp(session_id)
bill_url_base = "https://lis.virginia.gov/cgi-bin/"
if not is_special:
self.load_members()
self.load_sponsors()
self.load_fiscal_notes()
self.load_summaries()
self.load_history()
self.load_votes()
self.load_bills()
if not is_special:
self.load_amendments()
for bill in self._bills:
bill = self._bills[bill][0]
bill_id = bill["bill_id"]
chamber = chamber_types[bill_id[0]]
bill_type = {
"B": "bill",
"J": "joint resolution",
"R": "resolution"
}[bill_id[1]]
b = Bill(
bill_id,
session,
bill["bill_description"],
chamber=chamber,
classification=bill_type,
)
bill_url = bill_url_base + f"legp604.exe?{session_id}+sum+{bill_id}"
b.add_source(bill_url)
# Long Bill ID needs to have 6 characters to work with vote urls, sponsors, and summaries.
# Fill in blanks with 0s
long_bill_id = bill_id
if len(bill_id) == 3:
long_bill_id = bill_id[0:2] + "000" + bill_id[-1]
elif len(bill_id) == 4:
long_bill_id = bill_id[0:2] + "00" + bill_id[-2:]
elif len(bill_id) == 5:
long_bill_id = bill_id[0:2] + "0" + bill_id[-3:]
# Sponsors
if long_bill_id not in self._sponsors:
if "patron_name" in bill and bill["patron_name"].strip() != "":
b.add_sponsorship(
bill["patron_name"],
classification="primary",
entity_type="person",
primary=True,
)
for spon in self._sponsors[long_bill_id]:
if spon["member_name"].strip() == "":
continue
sponsor_type = spon["patron_type"]
if sponsor_type.endswith("Chief Patron"):
sponsor_type = "primary"
else:
sponsor_type = "cosponsor"
b.add_sponsorship(
spon["member_name"],
classification=sponsor_type,
entity_type="person",
primary=sponsor_type == "primary",
)
# Summary
summary_texts = self._summaries[long_bill_id]
for sum_text in summary_texts:
b.add_abstract(sum_text["summary_text"],
sum_text["summary_type"])
# Amendment docs
amendments = self._amendments[bill_id]
for amend in amendments:
doc_link = (
bill_url_base +
f"legp604.exe?{session_id}+amd+{amend['txt_docid']}")
b.add_document_link("Amendment: " + amend["txt_docid"],
doc_link,
media_type="text/html")
# fiscal notes
for fn in self._fiscal_notes[long_bill_id]:
doc_link = bill_url_base + f"legp604.exe?{session_id}+oth+{fn['refid']}"
b.add_document_link(
"Fiscal Impact Statement: " + fn["refid"],
doc_link.replace(".PDF", "+PDF"),
media_type="application/pdf",
)
# actions with 8-digit number followed by D are version titles too
doc_actions = defaultdict(list)
# History and then votes
for hist in self._history[bill_id]:
action = hist["history_description"]
action_date = hist["history_date"]
date = datetime.datetime.strptime(action_date,
"%m/%d/%y").date()
chamber = chamber_types[action[0]]
vote_id = hist["history_refid"]
cleaned_action = action[2:]
if re.findall(r"\d{8}D", cleaned_action):
doc_actions[action_date].append(cleaned_action)
# categorize actions
for pattern, atype in ACTION_CLASSIFIERS:
if re.match(pattern, cleaned_action):
break
else:
atype = None
if atype != SKIP:
b.add_action(cleaned_action,
date,
chamber=chamber,
classification=atype)
if len(vote_id) > 0:
total_yes = 0
total_no = 0
total_not_voting = 0
total_abstain = 0
for v in self._votes[vote_id]:
if v["vote_result"] == "yes":
total_yes += 1
elif v["vote_result"] == "no":
total_no += 1
elif v["vote_result"] == "not voting":
total_not_voting += 1
elif v["vote_result"] == "abstain":
total_abstain += 1
vote = VoteEvent(
identifier=vote_id,
start_date=date,
chamber=chamber,
motion_text=cleaned_action,
result="pass" if total_yes > total_no else "fail",
classification="passage",
bill=b,
)
vote.set_count("yes", total_yes)
vote.set_count("no", total_no)
vote.set_count("not voting", total_not_voting)
vote.set_count("abstain", total_abstain)
vote_url = (
bill_url_base +
f"legp604.exe?{session_id}+vot+{vote_id}+{long_bill_id}"
)
vote.add_source(vote_url)
for v in self._votes[vote_id]:
vote.vote(v["vote_result"], v["member_id"])
yield vote
# Versions
for version in bill["text_docs"]:
# Checks if abbr is blank as not every bill has multiple versions
if version["doc_abbr"]:
version_url = (
bill_url_base +
f"legp604.exe?{session_id}+ful+{version['doc_abbr']}")
version_date = datetime.datetime.strptime(
version["doc_date"], "%m/%d/%y").date()
# version text will default to abbreviation provided in CSV
# but if there is an unambiguous action from that date with
# a version, we'll use that as the document title
version_text = version["doc_abbr"]
if len(doc_actions[version["doc_date"]]) == 1:
version_text = doc_actions[version["doc_date"]][0]
b.add_version_link(
version_text,
version_url,
date=version_date,
media_type="text/html",
on_duplicate="ignore",
)
yield b
def scrape_bill(self, chamber, session, bill_id, short_title=None):
"""
Scrapes documents, actions, vote counts and votes for
bills from the 2009 session and above.
"""
url = BILL_URL % (session, bill_id.replace(" ", ""))
bill_page = self.get(url).text
html = lxml.html.fromstring(bill_page)
html.make_links_absolute(
"https://legislature.idaho.gov/legislation/%s/" % session)
bill_tables = html.xpath('//table[contains(@class, "bill-table")]')
title = bill_tables[1].text_content().strip()
bill_type = get_bill_type(bill_id)
bill = Bill(
legislative_session=session,
chamber=chamber,
identifier=bill_id,
title=title,
classification=bill_type,
)
bill.add_source(url)
for subject in self._subjects[bill_id.replace(" ", "")]:
bill.add_subject(subject)
if short_title and title.lower() != short_title.lower():
bill.add_title(short_title, "short title")
# documents
doc_links = html.xpath('//div[contains(@class,"insert-page")]//a')
for link in doc_links:
name = link.text_content().strip()
href = link.get("href")
if "Engrossment" in name or "Bill Text" in name or "Amendment" in name:
bill.add_version_link(note=name,
url=href,
media_type="application/pdf")
else:
bill.add_document_link(note=name,
url=href,
media_type="application/pdf")
def _split(string):
return re.split(r"\w+[,|AND]\s+", string)
# sponsors range from a committee to one legislator to a group of legs
sponsor_lists = bill_tables[0].text_content().split("by")
if len(sponsor_lists) > 1:
for sponsors in sponsor_lists[1:]:
if "COMMITTEE" in sponsors.upper():
bill.add_sponsorship(
name=sponsors.strip(),
entity_type="organization",
primary=True,
classification="primary",
)
else:
for person in _split(sponsors):
person = person.strip()
if person != "":
bill.add_sponsorship(
classification="primary",
name=person,
entity_type="person",
primary=True,
)
actor = chamber
last_date = None
# if a bill has passed a chamber or been 'received from'
# then the next committee passage is in the opposite chamber
has_moved_chambers = False
for row in bill_tables[2]:
# lots of empty rows
if len(row) == 1:
continue
_, date, action, _ = [x.text_content().strip() for x in row]
if date:
last_date = date
else:
date = last_date
date = datetime.datetime.strptime(date + "/" + session[0:4],
"%m/%d/%Y").strftime("%Y-%m-%d")
if action.startswith("House"):
actor = "lower"
elif action.startswith("Senate"):
actor = "upper"
# votes
if "AYES" in action or "NAYS" in action:
yield from self.parse_vote(actor, date, row[2], session,
bill_id, chamber, url)
# bill.add_vote_event(vote)
# some td's text is seperated by br elements
if len(row[2]):
action = "".join(row[2].itertext())
action = action.replace("\xa0", " ").strip()
atype = get_action(actor, action)
if atype and "passage" in atype:
has_moved_chambers = True
if atype and "committee-passage" in atype and has_moved_chambers:
actor = _OTHER_CHAMBERS[actor]
bill.add_action(action, date, chamber=actor, classification=atype)
# after voice vote/roll call and some actions the bill is sent
# 'to House' or 'to Senate'
if "to House" in action:
actor = "lower"
elif "to Senate" in action:
actor = "upper"
yield bill
def scrape_bill_list(self, url):
bill_list = self._get_bill_list(url)
for bill_info in bill_list:
(bill_id, ) = bill_info.xpath("td[1]/font/input/@value")
(sponsor, ) = bill_info.xpath("td[2]/font/input/@value")
(subject, ) = bill_info.xpath("td[3]//text()")
subject = subject.strip()
chamber = self.CHAMBERS[bill_id[0]]
if "B" in bill_id:
bill_type = "bill"
elif "JR" in bill_id:
bill_type = "joint resolution"
elif "R" in bill_id:
bill_type = "resolution"
else:
raise AssertionError(
"Unknown bill type for bill '{}'".format(bill_id))
bill = Bill(
bill_id,
legislative_session=self.session,
chamber=chamber,
title="",
classification=bill_type,
)
if subject:
bill.subject = [subject]
if sponsor:
bill.add_sponsorship(
name=sponsor,
entity_type="person",
classification="primary",
primary=True,
)
bill.add_source(url)
bill_url = ("http://alisondb.legislature.state.al.us/Alison/"
"SESSBillStatusResult.aspx?BILL={}".format(bill_id))
bill.add_source(bill_url)
bill_html = self._get_bill_response(bill_url)
if bill_html is None:
self.warning(
"Bill {} has no webpage, and will be skipped".format(
bill_id))
continue
bill_doc = lxml.html.fromstring(bill_html)
if bill_doc.xpath(
'//span[@id="ContentPlaceHolder1_lblShotTitle"]'):
title = (bill_doc.xpath(
'//span[@id="ContentPlaceHolder1_lblShotTitle"]')
[0].text_content().strip())
if not title:
title = "[No title given by state]"
bill.title = title
session = "2022FS" if self.session == "2022s1" else self.session
version_url_base = (
"http://alisondb.legislature.state.al.us/ALISON/"
"SearchableInstruments/{0}/PrintFiles/{1}-".format(
session, bill_id))
versions = bill_doc.xpath(
'//table[@class="box_versions"]/tr/td[2]/font/text()')
for version in versions:
name = version
if version == "Introduced":
version_url = version_url_base + "int.pdf"
elif version == "Engrossed":
version_url = version_url_base + "eng.pdf"
elif version == "Enrolled":
version_url = version_url_base + "enr.pdf"
else:
raise NotImplementedError(
"Unknown version type found: '{}'".format(name))
bill.add_version_link(
name,
version_url,
media_type="application/pdf",
on_duplicate="ignore",
)
# Fiscal notes exist, but I can't figure out how to build their URL
fiscal_notes = bill_doc.xpath(
'//table[@class="box_fiscalnote"]')[1:]
for fiscal_note in fiscal_notes:
pass
# Budget Isolation Resolutions are handled as extra actions/votes
birs = bill_doc.xpath(
'//div[@class="box_bir"]//table//table/tr')[1:]
for bir in birs:
bir_action = bir.xpath("td[1]")[0].text_content().strip()
# Sometimes ALISON's database puts another bill's
# actions into the BIR action list; ignore these
if bill_id not in bir_action:
self.warning(
"BIR action found ({}) ".format(bir_action) +
"that doesn't match the bill ID ({})".format(bill_id))
continue
bir_date = datetime.datetime.strptime(
bir.xpath("td[2]/font/text()")[0], self.DATE_FORMAT)
bir_type = bir.xpath("td[1]/font/text()")[0].split(" ")[0]
bir_chamber = self.CHAMBERS[bir_type[0]]
bir_text = "{0}: {1}".format(
bir_type,
bir.xpath("td[3]/font/text()")[0].strip())
bill.add_action(
bir_text,
TIMEZONE.localize(bir_date),
chamber=bir_chamber,
classification="other",
)
try:
(bir_vote_id, ) = bir.xpath("td[4]/font/input/@value")
except ValueError:
bir_vote_id = ""
bir_vote_id = bir_vote_id.strip()
if bir_vote_id.startswith("Roll "):
bir_vote_id = bir_vote_id.split(" ")[-1]
yield from self.scrape_vote(
bill=bill,
vote_chamber=bir_type[0],
bill_id="{0}%20for%20{1}".format(bir_type, bill_id),
vote_id=bir_vote_id,
vote_date=TIMEZONE.localize(bir_date),
action_text=bir_text,
)
actions = bill_doc.xpath(
'//table[@id="ContentPlaceHolder1_gvHistory"]/tr')[1:]
action_date = None
for action in actions:
# If actions occur on the same day, only one date will exist
if (action.xpath("td[1]/font/text()")[0].encode(
"ascii", "ignore").strip()):
action_date = datetime.datetime.strptime(
action.xpath("td[1]/font/text()")[0], self.DATE_FORMAT)
(action_chamber, ) = action.xpath("td[2]/font/text()")
possible_amendment = action.xpath("td[3]/font/u/text()")
if (len(possible_amendment) > 0
and not possible_amendment[0].strip() == ""):
(amendment, ) = possible_amendment
else:
amendment = None
(action_text, ) = action.xpath("td[4]/font/text()")
action_type = _categorize_action(action_text)
# check for occasional extra last row
if not action_chamber.strip():
continue
# The committee cell is just an abbreviation, so get its name
actor = self.CHAMBERS[action_chamber]
try:
action_committee = (re.search(
r".*? referred to the .*? committee on (.*?)$",
action_text).group(1).strip())
except AttributeError:
action_committee = ""
if action_date is not None and action_text.strip():
act = bill.add_action(
action_text,
TIMEZONE.localize(action_date),
chamber=actor,
classification=action_type,
)
if action_committee:
act.add_related_entity(action_committee,
entity_type="organization")
try:
vote_button = action.xpath("td[9]//text()")[0].strip()
except IndexError:
vote_button = ""
if vote_button.startswith("Roll "):
vote_id = vote_button.split(" ")[-1]
yield from self.scrape_vote(
bill=bill,
vote_chamber=action_chamber,
bill_id=bill_id,
vote_id=vote_id,
vote_date=TIMEZONE.localize(action_date),
action_text=action_text,
)
if amendment:
session = "2021FS" if self.session == "2021s1" else self.session
amend_url = (
"http://alisondb.legislature.state.al.us/ALISON/"
"SearchableInstruments/{0}/PrintFiles/{1}.pdf".format(
session, amendment))
amend_name = "Amd/Sub {}".format(amendment)
bill.add_version_link(
amend_name,
amend_url,
media_type="application/pdf",
on_duplicate="ignore",
)
yield bill
def scrape_bill(self, chamber, session, session_id, bill_id, url):
sidebar = lxml.html.fromstring(self.get(url).text)
sidebar.make_links_absolute("https://www.legis.iowa.gov")
hist_url = (f"https://www.legis.iowa.gov/legislation/billTracking/"
f"billHistory?billName={bill_id}&ga={session_id}")
req_session = requests.Session()
req = requests.get(hist_url)
if req.status_code == 500:
self.warning("500 error on {}, skipping".format(hist_url))
return
page = lxml.html.fromstring(req.text)
page.make_links_absolute("https://www.legis.iowa.gov")
title = page.xpath('string(//div[@id="content"]/div[@class='
'"divideVert"]/div/div[4]/div[2])').strip()
if title == "":
# Sometimes the title is moved, see
# https://www.legis.iowa.gov/legislation/billTracking/billHistory?billName=SF%20139&ga=88
title = page.xpath('string(//div[@id="content"]/div[@class='
'"divideVert"]/div[4]/div[2])').strip()
if title == "":
self.warning("URL: %s gives us an *EMPTY* bill. Aborting." %
url)
return
if title.lower().startswith("in"):
title = page.xpath("string(//table[2]/tr[3])").strip()
if "HR" in bill_id or "SR" in bill_id:
bill_type = ["resolution"]
elif "HJR" in bill_id or "SJR" in bill_id:
bill_type = ["joint resolution"]
elif "HCR" in bill_id or "SCR" in bill_id:
bill_type = ["concurrent resolution"]
else:
bill_type = ["bill"]
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=bill_type,
)
bill.add_source(hist_url)
# base url for text version (version_abbrev, session_id, bill_id)
version_html_url_template = (
"https://www.legis.iowa.gov/docs/"
"publications/LG{}/{}/attachments/{}.html")
version_pdf_url_template = ("https://www.legis.iowa.gov/docs/"
"publications/LG{}/{}/{}.pdf")
# get pieces of version_link
vpieces = sidebar.xpath('//select[@id="billVersions"]/option')
if vpieces:
for version in vpieces:
version_name = version.text
version_abbrev = version.xpath("string(@value)")
# Get HTML document of bill version.
version_html_url = version_html_url_template.format(
version_abbrev.upper(), session_id,
bill_id.replace(" ", ""))
bill.add_version_link(note=version_name,
url=version_html_url,
media_type="text/html")
# Get PDF document of bill version.
version_pdf_url = version_pdf_url_template.format(
version_abbrev.upper(), session_id,
bill_id.replace(" ", ""))
if "Marked Up" in version_name:
version_pdf_url = sidebar.xpath(
"//iframe[@id='bbContextDoc']/@src")[0]
bill.add_version_link(note=version_name,
url=version_pdf_url,
media_type="application/pdf")
sponsors_str = page.xpath('string(//div[@id="content"]/div[@class='
'"divideVert"]/div/div[4]/div[1])').strip()
if re.search("^By ", sponsors_str):
sponsors = re.split(",| and ", sponsors_str.split("By ")[1])
# for some bills sponsors listed in different format
else:
sponsors = re.findall(r"[\w-]+(?:, [A-Z]\.)?(?:,|(?: and)|\.$)",
sponsors_str)
for sponsor in sponsors:
sponsor = sponsor.replace(" and", "").strip(" .,")
# a few sponsors get mangled by our regex
sponsor = {
"Means": "Ways & Means",
"Iowa": "Economic Growth/Rebuild Iowa",
"Safety": "Public Safety",
"Resources": "Human Resources",
"Affairs": "Veterans Affairs",
"Protection": "Environmental Protection",
"Government": "State Government",
"Boef": "De Boef",
}.get(sponsor, sponsor)
if sponsor[0].islower():
# SSBs catch cruft in it ('charges', 'overpayments')
# https://sunlight.atlassian.net/browse/DATA-286
continue
bill.add_sponsorship(
name=sponsor,
classification="primary",
entity_type="person",
primary=True,
)
for tr in page.xpath(
"//table[contains(@class, 'billActionTable')][1]/tbody/tr"):
date = tr.xpath("string(td[contains(text(), ', 20')])").strip()
if date.startswith("***"):
continue
elif "No history is recorded at this time." in date:
return
if date == "":
continue
date = datetime.datetime.strptime(date, "%B %d, %Y").date()
action = tr.xpath("string(td[3])").strip()
action = re.sub(r"\s+", " ", action)
# Capture any amendment links.
links = [
link
for link in [version["links"] for version in bill.versions]
]
version_urls = [
link["url"] for link in [i for sub in links for i in sub]
]
if "amendment" in action.lower():
for anchor in tr.xpath(".//a[1]"):
if "-" in anchor.text:
# https://www.legis.iowa.gov/docs/publications/AMDI/88/S3071.pdf
amd_pattern = "https://www.legis.iowa.gov/docs/publications/AMDI/{}/{}.pdf"
amd_id = anchor.text.replace("-", "").strip()
amd_url = amd_pattern.format(session_id, amd_id)
amd_name = "Amendment {}".format(anchor.text.strip())
if amd_url not in version_urls:
bill.add_version_link(note=amd_name,
url=amd_url,
media_type="application/pdf")
version_urls.append(amd_url)
else:
self.info(
"Already Added {}, skipping".format(amd_url))
if "S.J." in action or "SCS" in action:
actor = "upper"
elif "H.J." in action or "HCS" in action:
actor = "lower"
else:
actor = "legislature"
action = re.sub(r"(H|S)\.J\.\s+\d+\.$", "", action).strip()
if action.startswith("Introduced"):
atype = ["introduction"]
if ", referred to" in action:
atype.append("referral-committee")
elif action.startswith("Read first time"):
atype = "reading-1"
elif action.startswith("Referred to"):
atype = "referral-committee"
elif action.startswith("Sent to Governor"):
atype = "executive-receipt"
elif action.startswith("Reported Signed by Governor"):
atype = "executive-signature"
elif action.startswith("Signed by Governor"):
atype = "executive-signature"
elif action.startswith("Vetoed by Governor"):
atype = "executive-veto"
elif action.startswith("Item veto"):
atype = "executive-veto-line-item"
elif re.match(r"Passed (House|Senate)", action):
atype = "passage"
elif re.match(r"Amendment (S|H)-\d+ filed", action):
atype = ["amendment-introduction"]
if ", adopted" in action:
atype.append("amendment-passage")
elif re.match(r"Amendment (S|H)-\d+( as amended,)? adopted",
action):
atype = "amendment-passage"
elif re.match(r"Amendment (S|N)-\d+ lost", action):
atype = "amendment-failure"
elif action.startswith("Resolution filed"):
atype = "introduction"
elif action.startswith("Resolution adopted"):
atype = "passage"
elif action.startswith("Committee report") and action.endswith(
"passage."):
atype = "committee-passage"
elif action.startswith("Withdrawn"):
atype = "withdrawal"
else:
atype = None
if action.strip() == "":
continue
if re.search(r"END OF \d+ ACTIONS", action):
continue
if "$history" not in action:
bill.add_action(description=action,
date=date,
chamber=actor,
classification=atype)
self.scrape_subjects(bill, bill_id, session, req_session)
yield bill
def scrape_bills(self, chamber, session, subjects):
idex = bill_start_numbers(session)[chamber]
FROM = "ctl00$rilinContent$txtBillFrom"
TO = "ctl00$rilinContent$txtBillTo"
YEAR = "ctl00$rilinContent$cbYear"
blocks = "FOO" # Ugh.
while len(blocks) > 0:
default_headers = get_default_headers(SEARCH_URL)
default_headers[FROM] = idex
default_headers[TO] = idex + MAXQUERY
default_headers[YEAR] = session
idex += MAXQUERY
blocks = self.parse_results_page(
self.post(SEARCH_URL, data=default_headers).text)
blocks = blocks[1:-1]
blocks = self.digest_results_page(blocks)
for block in blocks:
bill = blocks[block]
subs = []
try:
subs = subjects[bill["bill_id"]]
except KeyError:
pass
title = bill["title"][len("ENTITLED, "):]
billid = bill["bill_id"]
try:
subs = subjects[bill["bill_id"]]
except KeyError:
subs = []
for b in BILL_NAME_TRANSLATIONS:
if billid[:len(b)] == b:
billid = (BILL_NAME_TRANSLATIONS[b] +
billid[len(b) + 1:].split()[0])
b = Bill(
billid,
title=title,
chamber=chamber,
legislative_session=session,
classification=self.get_type_by_name(bill["bill_id"]),
)
b.subject = subs
# keep bill ID around
self._bill_id_by_type[(chamber,
re.findall(r"\d+", billid)[0])] = billid
self.process_actions(bill["actions"], b)
sponsors = bill["sponsors"][len("BY"):].strip()
sponsors = sponsors.split(",")
sponsors = [s.strip() for s in sponsors]
for href in bill["bill_id_hrefs"]:
b.add_version_link(href.text,
href.attrib["href"],
media_type="application/pdf")
for sponsor in sponsors:
b.add_sponsorship(
sponsor,
entity_type="person",
classification="primary",
primary=True,
)
b.add_source(SEARCH_URL)
yield b
def scrape_bills(self, session, year_abr):
# Main Bill information
main_bill_csv = self.to_csv("MAINBILL.TXT")
# keep a dictionary of bills (mapping bill_id to Bill obj)
bill_dict = {}
for rec in main_bill_csv:
bill_type = rec["BillType"].strip()
bill_number = int(rec["BillNumber"])
bill_id = bill_type + str(bill_number)
title = rec["Synopsis"]
if bill_type[0] == "A":
chamber = "lower"
else:
chamber = "upper"
# some bills have a blank title.. just skip it
if not title:
continue
bill = Bill(
bill_id,
title=title,
chamber=chamber,
legislative_session=session,
classification=self._bill_types[bill_type[1:]],
)
if rec["IdenticalBillNumber"].strip():
bill.add_related_bill(
rec["IdenticalBillNumber"].split()[0],
legislative_session=session,
relation_type="companion",
)
# TODO: last session info is in there too
bill_dict[bill_id] = bill
# Sponsors
bill_sponsors_csv = self.to_csv("BILLSPON.TXT")
for rec in bill_sponsors_csv:
bill_type = rec["BillType"].strip()
bill_number = int(rec["BillNumber"])
bill_id = bill_type + str(bill_number)
if bill_id not in bill_dict:
self.warning("unknown bill %s in sponsor database" % bill_id)
continue
bill = bill_dict[bill_id]
name = rec["Sponsor"]
sponsor_type = rec["Type"]
if sponsor_type == "P":
sponsor_type = "primary"
else:
sponsor_type = "cosponsor"
bill.add_sponsorship(
name,
classification=sponsor_type,
entity_type="person",
primary=sponsor_type == "primary",
)
# Documents
bill_document_csv = self.to_csv("BILLWP.TXT")
for rec in bill_document_csv:
bill_type = rec["BillType"].strip()
bill_number = int(rec["BillNumber"])
bill_id = bill_type + str(bill_number)
if bill_id not in bill_dict:
self.warning("unknown bill %s in document database" % bill_id)
continue
bill = bill_dict[bill_id]
document = rec["Document"]
document = document.split("\\")
document = document[-2] + "/" + document[-1]
htm_url = "https://www.njleg.state.nj.us/Bills/{}/{}".format(
year_abr, document.replace(".DOC", ".HTM"))
pdf_url = "https://www.njleg.state.nj.us/Bills/{}/{}".format(
year_abr, document.replace(".DOC", ".PDF"))
# name document based _doctype
try:
doc_name = self._doctypes[rec["DocType"]]
except KeyError:
raise Exception("unknown doctype %s on %s" %
(rec["DocType"], bill_id))
if rec["Comment"]:
doc_name += " " + rec["Comment"]
# Clean links.
if htm_url.endswith("HTMX"):
htm_url = re.sub("X$", "", htm_url)
if pdf_url.endswith("PDFX"):
pdf_url = re.sub("X$", "", pdf_url)
if rec["DocType"] in self._version_types:
if htm_url.lower().endswith("htm"):
mimetype = "text/html"
elif htm_url.lower().endswith("wpd"):
mimetype = "application/vnd.wordperfect"
try:
bill.add_version_link(doc_name,
htm_url,
media_type=mimetype)
bill.add_version_link(doc_name,
pdf_url,
media_type="application/pdf")
except ValueError:
self.warning(
"Couldn't find a document for bill {}".format(bill_id))
pass
else:
bill.add_document_link(doc_name, htm_url)
# Votes
next_year = int(year_abr) + 1
vote_info_list = [
"A%s" % year_abr,
"A%s" % next_year,
"S%s" % year_abr,
"S%s" % next_year,
"CA%s-%s" % (year_abr, next_year),
"CS%s-%s" % (year_abr, next_year),
]
# keep votes clean globally, a few votes show up in multiple files
votes = {}
for filename in vote_info_list:
s_vote_url = f"https://www.njleg.state.nj.us/votes/{filename}.zip"
try:
s_vote_zip, resp = self.urlretrieve(s_vote_url)
except scrapelib.HTTPError:
self.warning("could not find %s" % s_vote_url)
continue
zippedfile = zipfile.ZipFile(s_vote_zip)
for vfile in ["%s.txt" % (filename), "%sEnd.txt" % (filename)]:
try:
vote_file = io.TextIOWrapper(zippedfile.open(vfile, "r"),
encoding="latin-1")
except KeyError:
#
# Right, so, 2011 we have an "End" file with more
# vote data than was in the original dump.
#
self.warning("No such file: %s" % (vfile))
continue
vdict_file = csv.DictReader(vote_file)
if filename.startswith("A") or filename.startswith("CA"):
chamber = "lower"
else:
chamber = "upper"
if filename.startswith("C"):
vote_file_type = "committee"
else:
vote_file_type = "chamber"
for rec in vdict_file:
if vote_file_type == "chamber":
bill_id = rec["Bill"].strip()
leg = rec["Full_Name"]
date = rec["Session_Date"]
action = rec["Action"]
leg_vote = rec["Legislator_Vote"]
vote_parts = (bill_id, chamber, action)
else:
bill_id = "%s%s" % (rec["Bill_Type"],
rec["Bill_Number"])
leg = rec["Name"]
# drop time portion
date = rec["Agenda_Date"].split()[0]
# make motion readable
action = self._com_vote_motions[rec["BillAction"]]
# first char (Y/N) use [0:1] to ignore ''
leg_vote = rec["LegislatorVote"][0:1]
committee = rec["Committee_House"]
vote_parts = (bill_id, chamber, action, committee)
date = datetime.strptime(date, "%m/%d/%Y")
vote_id = "_".join(vote_parts).replace(" ", "_")
if vote_id not in votes:
votes[vote_id] = VoteEvent(
start_date=TIMEZONE.localize(date),
chamber=chamber,
motion_text=action,
classification="passage",
result=None,
bill=bill_dict[bill_id],
)
votes[vote_id].dedupe_key = vote_id
if leg_vote == "Y":
votes[vote_id].vote("yes", leg)
elif leg_vote == "N":
votes[vote_id].vote("no", leg)
else:
votes[vote_id].vote("other", leg)
# remove temp file
os.remove(s_vote_zip)
# Counts yes/no/other votes and saves overall vote
for vote in votes.values():
counts = collections.defaultdict(int)
for count in vote.votes:
counts[count["option"]] += 1
vote.set_count("yes", counts["yes"])
vote.set_count("no", counts["no"])
vote.set_count("other", counts["other"])
# Veto override.
if vote.motion_text == "OVERRIDE":
# Per the NJ leg's glossary, a veto override requires
# 2/3ds of each chamber. 27 in the senate, 54 in the house.
# http://www.njleg.state.nj.us/legislativepub/glossary.asp
if "lower" in vote.bill:
vote.result = "pass" if counts["yes"] >= 54 else "fail"
elif "upper" in vote.bill:
vote.result = "pass" if counts["yes"] >= 27 else "fail"
else:
# Regular vote.
vote.result = "pass" if counts["yes"] > counts[
"no"] else "fail"
vote.add_source("http://www.njleg.state.nj.us/downloads.asp")
yield vote
# Actions
bill_action_csv = self.to_csv("BILLHIST.TXT")
actor_map = {"A": "lower", "G": "executive", "S": "upper"}
for rec in bill_action_csv:
bill_type = rec["BillType"].strip()
bill_number = int(rec["BillNumber"])
bill_id = bill_type + str(bill_number)
if bill_id not in bill_dict:
self.warning("unknown bill %s in action database" % bill_id)
continue
bill = bill_dict[bill_id]
action = rec["Action"]
date = rec["DateAction"]
date = dateutil.parser.parse(date)
actor = actor_map[rec["House"]]
comment = rec["Comment"]
action, atype = self.categorize_action(action, bill_id)
if comment:
action += " " + comment
bill.add_action(
action,
date=TIMEZONE.localize(date),
classification=atype,
chamber=actor,
)
# Subjects
subject_csv = self.to_csv("BILLSUBJ.TXT")
for rec in subject_csv:
bill_id = rec["BillType"].strip() + str(int(rec["BillNumber"]))
if bill_id not in bill_dict:
self.warning("unknown bill %s in subject database" % bill_id)
continue
bill = bill_dict.get(bill_id)
if bill:
bill.subject.append(rec["SubjectKey"])
else:
self.warning("invalid bill id in BillSubj: %s" % bill_id)
phony_bill_count = 0
# save all bills at the end
for bill in bill_dict.values():
# add sources
if not bill.actions and not bill.versions:
self.warning("probable phony bill detected %s",
bill.identifier)
phony_bill_count += 1
else:
bill.add_source("http://www.njleg.state.nj.us/downloads.asp")
yield bill
if phony_bill_count:
self.warning("%s total phony bills detected", phony_bill_count)
def _parse_house_bill(self, url, session):
# using the print page makes the page simpler, and also *drastically* smaller
# (8k rather than 100k)
url = re.sub("billsummary", "billsummaryprn", url)
url = "%s/%s" % (self._house_base_url, url)
# the URL is an iframed version now, so swap in for the actual bill page
url = url.replace("Bill.aspx", "BillContent.aspx")
url = url.replace("&code=R", "&code=R&style=new")
# http://www.house.mo.gov/Bill.aspx?bill=HB26&year=2017&code=R
# http://www.house.mo.gov/BillContent.aspx?bill=HB26&year=2017&code=R&style=new
bill_page = self.get(url).text
bill_page = lxml.html.fromstring(bill_page)
bill_page.make_links_absolute(url)
bill_id = bill_page.xpath('//*[@class="entry-title"]/div')
if len(bill_id) == 0:
self.info("WARNING: bill summary page is blank! (%s)" % url)
self._bad_urls.append(url)
return
bill_id = bill_id[0].text_content()
bill_id = clean_text(bill_id)
bill_desc = bill_page.xpath(
'//*[@class="BillDescription"]')[0].text_content()
bill_desc = clean_text(bill_desc)
table_rows = bill_page.xpath("//table/tr")
# if there is a cosponsor all the rows are pushed down one for the extra row
# for the cosponsor:
cosponsorOffset = 0
if table_rows[2][0].text_content().strip() == "Co-Sponsor:":
cosponsorOffset = 1
lr_label_tag = table_rows[3 + cosponsorOffset]
assert lr_label_tag[0].text_content().strip() == "LR Number:"
# bill_lr = lr_label_tag[1].text_content()
lastActionOffset = 0
if (table_rows[4 + cosponsorOffset][0].text_content().strip() ==
"Governor Action:"):
lastActionOffset = 1
official_title_tag = table_rows[5 + cosponsorOffset + lastActionOffset]
assert official_title_tag[0].text_content().strip() == "Bill String:"
official_title = official_title_tag[1].text_content()
# could substitute the description for the name,
# but keeping it separate for now.
bill_type = "bill"
triplet = bill_id[:3]
if triplet in bill_types:
bill_type = bill_types[triplet]
bill_number = int(bill_id[3:].strip())
else:
bill_number = int(bill_id[3:])
subs = []
bid = bill_id.replace(" ", "")
if bid in self._subjects:
subs = self._subjects[bid]
self.info("With subjects for this bill")
self.info(bid)
if bill_desc == "":
if bill_number <= 20:
# blank bill titles early in session are approp. bills
bill_desc = "Appropriations Bill"
else:
self.error("Blank title. Skipping. {} / {} / {}".format(
bill_id, bill_desc, official_title))
return
bill = Bill(
bill_id,
chamber="lower",
title=bill_desc,
legislative_session=self._session_id,
classification=bill_type,
)
bill.subject = subs
bill.add_title(official_title, note="official")
bill.add_source(url)
bill_sponsor = clean_text(table_rows[0][1].text_content())
# try:
# bill_sponsor_link = table_rows[0][1][0].attrib['href']
# except IndexError:
# return
bill.add_sponsorship(bill_sponsor,
entity_type="person",
classification="primary",
primary=True)
# check for cosponsors
(sponsors_url,
) = bill_page.xpath("//a[contains(@href, 'CoSponsors.aspx')]/@href")
self._parse_cosponsors_from_bill(bill, sponsors_url)
# actions_link_tag = bill_page.xpath('//div[@class="Sections"]/a')[0]
# actions_link = '%s/%s' % (self._house_base_url,actions_link_tag.attrib['href'])
# actions_link = re.sub("content", "print", actions_link)
(actions_link,
) = bill_page.xpath("//a[contains(@href, 'BillActions.aspx')]/@href")
yield from self._parse_house_actions(bill, actions_link)
# get bill versions
doc_tags = bill_page.xpath('//div[@class="BillDocuments"][1]/span')
for doc_tag in reversed(doc_tags):
doc = clean_text(doc_tag.text_content())
text_url = "%s%s" % (self._house_base_url,
doc_tag[0].attrib["href"])
bill.add_document_link(doc, text_url, media_type="text/html")
# get bill versions
version_tags = bill_page.xpath('//div[@class="BillDocuments"][2]/span')
for version_tag in reversed(version_tags):
version = clean_text(version_tag.text_content())
for vurl in version_tag.xpath(".//a"):
if vurl.text == "PDF":
mimetype = "application/pdf"
else:
mimetype = "text/html"
bill.add_version_link(
version,
vurl.attrib["href"],
media_type=mimetype,
on_duplicate="ignore",
)
# house bill versions
# everything between the row containing "Bill Text" in an h2 and the next div.DocHeaderRow
version_rows = bill_page.xpath(
'//div[h2[contains(text(),"Bill Text")]]/'
'following-sibling::div[contains(@class,"DocRow") '
'and count(preceding-sibling::div[contains(@class,"DocHeaderRow")])=1]'
)
for row in version_rows:
# some rows are just broken links, not real versions
if row.xpath('.//div[contains(@class,"textType")]/a/@href'):
version = row.xpath(
'.//div[contains(@class,"textType")]/a/text()')[0].strip()
path = row.xpath(
'.//div[contains(@class,"textType")]/a/@href')[0].strip()
if ".pdf" in path:
mimetype = "application/pdf"
else:
mimetype = "text/html"
bill.add_version_link(version,
path,
media_type=mimetype,
on_duplicate="ignore")
# house bill summaries
# everything between the row containing "Bill Summary" in an h2
# and the next div.DocHeaderRow
summary_rows = bill_page.xpath(
'//div[h2[contains(text(),"Bill Summary")]]/'
'following-sibling::div[contains(@class,"DocRow") '
'and count(following-sibling::div[contains(@class,"DocHeaderRow")])=1]'
)
# if there are no amedments, we need a different xpath for summaries
if not summary_rows:
summary_rows = bill_page.xpath(
'//div[h2[contains(text(),"Bill Summary")]]/'
'following-sibling::div[contains(@class,"DocRow")]')
for row in reversed(summary_rows):
version = row.xpath(
'.//div[contains(@class,"textType")]/a/text()')[0].strip()
if version:
path = row.xpath(
'.//div[contains(@class,"textType")]/a/@href')[0].strip()
summary_name = "Bill Summary ({})".format(version)
if ".pdf" in path:
mimetype = "application/pdf"
else:
mimetype = "text/html"
bill.add_document_link(summary_name,
path,
media_type=mimetype,
on_duplicate="ignore")
# house bill amendments
amendment_rows = bill_page.xpath(
'//div[h2[contains(text(),"Amendment")]]/'
'following-sibling::div[contains(@class,"DocRow")]')
for row in reversed(amendment_rows):
version = row.xpath(
'.//div[contains(@class,"DocInfoCell")]/a[1]/text()')[0].strip(
)
path = row.xpath(
'.//div[contains(@class,"DocInfoCell")]/a[1]/@href')[0].strip(
)
summary_name = "Amendment {}".format(version)
defeated_icon = row.xpath('.//img[contains(@title,"Defeated")]')
if defeated_icon:
summary_name = "{} (Defeated)".format(summary_name)
adopted_icon = row.xpath('.//img[contains(@title,"Adopted")]')
if adopted_icon:
summary_name = "{} (Adopted)".format(summary_name)
distributed_icon = row.xpath(
'.//img[contains(@title,"Distributed")]')
if distributed_icon:
summary_name = "{} (Distributed)".format(summary_name)
if ".pdf" in path:
mimetype = "application/pdf"
else:
mimetype = "text/html"
bill.add_version_link(summary_name,
path,
media_type=mimetype,
on_duplicate="ignore")
yield bill
def scrape_bill(self, row, session):
bill_id = row["LegislationDisplayCode"]
amendment = None
substitute = None
if bill_id.count(" ") > 1:
if " w/ " in bill_id:
self.info("Found amended bill `{}`".format(bill_id))
bill_id, amendment = bill_id.split(" w/ ")
if " -" in bill_id:
self.info("Found amended bill `{}`".format(bill_id))
bill_id, amendment = bill_id.split(" -")
# A bill can _both_ be amended and be substituted
if " for " in bill_id:
self.info(
"Found substitute to use instead: `{}`".format(bill_id))
substitute, bill_id = bill_id.split(" for ")
if amendment is None and substitute is None:
raise ValueError("unknown bill_id format: " + bill_id)
bill_type = self.classify_bill(bill_id)
chamber = "upper" if bill_id.startswith("S") else "lower"
bill = Bill(
identifier=bill_id,
legislative_session=session,
chamber=chamber,
title=row["LongTitle"],
classification=bill_type,
)
if row["Synopsis"]:
bill.add_abstract(row["Synopsis"], "synopsis")
if row["ShortTitle"]:
bill.add_title(row["ShortTitle"], "short title")
if row["SponsorPersonId"]:
self.add_sponsor_by_legislator_id(bill, row["SponsorPersonId"],
"primary")
if substitute:
bill.extras["substitute"] = substitute
if amendment:
bill.extras["amendment"] = amendment
# TODO: Is there a way get additional sponsors and cosponsors, and versions/fns via API?
html_url = "https://legis.delaware.gov/BillDetail?LegislationId={}".format(
row["LegislationId"])
bill.add_source(html_url, note="text/html")
html = self.lxmlize(html_url)
additional_sponsors = html.xpath(
'//label[text()="Additional Sponsor(s):"]'
"/following-sibling::div/a/@href")
for sponsor_url in additional_sponsors:
sponsor_id = sponsor_url.replace(
"https://legis.delaware.gov/LegislatorDetail?"
"personId=", "")
self.add_sponsor_by_legislator_id(bill, sponsor_id, "primary")
cosponsors = html.xpath('//label[text()="Co-Sponsor(s):"]/'
"following-sibling::div/a/@href")
for sponsor_url in cosponsors:
sponsor_id = sponsor_url.replace(
"https://legis.delaware.gov/LegislatorDetail?"
"personId=", "")
self.add_sponsor_by_legislator_id(bill, sponsor_id, "cosponsor")
versions = html.xpath(
'//label[text()="Original Text:"]/following-sibling::div/a/@href')
for version_url in versions:
media_type = self.mime_from_link(version_url)
version_name = "Bill Text"
bill.add_version_link(version_name,
version_url,
media_type=media_type)
fiscals = html.xpath('//div[contains(@class,"fiscalNote")]/a/@href')
for fiscal in fiscals:
self.scrape_fiscal_note(bill, fiscal)
self.scrape_actions(bill, row["LegislationId"])
if row["HasAmendments"] is True:
self.scrape_amendments(bill, row["LegislationId"])
yield from self.scrape_votes(bill, row["LegislationId"], session)
yield bill
def scrape(self, session=None):
HTML_TAGS_RE = r"<.*?>"
if session is None:
session = self.latest_session()
year_slug = self.jurisdiction.get_year_slug(session)
# Load all bills and resolutions via the private API
bills_url = "http://legislature.vermont.gov/bill/loadBillsReleased/{}/".format(
year_slug)
bills_json = self.get(bills_url).text
bills = json.loads(bills_json)["data"] or []
bills_url = "http://legislature.vermont.gov/bill/loadBillsIntroduced/{}/".format(
year_slug)
bills_json = self.get(bills_url).text
bills.extend(json.loads(bills_json)["data"] or [])
resolutions_url = "http://legislature.vermont.gov/bill/loadAllResolutionsByChamber/{}/both".format(
year_slug)
resolutions_json = self.get(resolutions_url).text
bills.extend(json.loads(resolutions_json)["data"] or [])
# Parse the information from each bill
for info in bills:
# Strip whitespace from strings
info = {k: v.strip() for k, v in info.items()}
# Identify the bill type and chamber
if info["BillNumber"].startswith("J.R.H."):
bill_type = "joint resolution"
bill_chamber = "lower"
elif info["BillNumber"].startswith("J.R.S."):
bill_type = "joint resolution"
bill_chamber = "upper"
elif info["BillNumber"].startswith("H.C.R."):
bill_type = "concurrent resolution"
bill_chamber = "lower"
elif info["BillNumber"].startswith("S.C.R."):
bill_type = "concurrent resolution"
bill_chamber = "upper"
elif info["BillNumber"].startswith("H.R."):
bill_type = "resolution"
bill_chamber = "lower"
elif info["BillNumber"].startswith("S.R."):
bill_type = "resolution"
bill_chamber = "upper"
elif info["BillNumber"].startswith("PR."):
bill_type = "constitutional amendment"
if info["Body"] == "H":
bill_chamber = "lower"
elif info["Body"] == "S":
bill_chamber = "upper"
else:
raise AssertionError("Amendment not tied to chamber")
elif info["BillNumber"].startswith("H."):
bill_type = "bill"
bill_chamber = "lower"
elif info["BillNumber"].startswith("S."):
bill_type = "bill"
bill_chamber = "upper"
else:
raise AssertionError("Unknown bill type found: '{}'".format(
info["BillNumber"]))
bill_id_original_format = (info["BillNumber"].replace(".",
"").replace(
" ", ""))
bill_id = bill_id_original_format
# put one space back in between type and number
bill_id = re.sub(r"([a-zA-Z]+)(\d+)", r"\1 \2", bill_id)
# Create the bill using its basic information
bill = Bill(
identifier=bill_id,
legislative_session=session,
chamber=bill_chamber,
title=info["Title"],
classification=bill_type,
)
if "resolution" in bill_type:
bill.add_source(resolutions_url)
else:
bill.add_source(bills_url)
# Load the bill's information page to access its metadata
bill_url = "http://legislature.vermont.gov/bill/status/{0}/{1}".format(
year_slug, info["BillNumber"])
doc = self.lxmlize(bill_url)
bill.add_source(bill_url)
# Capture sponsors
sponsors = doc.xpath(
'//dl[@class="summary-table"]/dt[text()="Sponsor(s)"]/'
"following-sibling::dd[1]/ul/li")
sponsor_type = "primary"
for sponsor in sponsors:
if sponsor.xpath("span/text()") == ["Additional Sponsors"]:
sponsor_type = "cosponsor"
continue
sponsor_name = (sponsor.xpath("a/text()")[0].replace(
"Rep.", "").replace("Sen.", "").strip())
if sponsor_name and not (sponsor_name[:5] == "Less"
and len(sponsor_name) == 5):
bill.add_sponsorship(
name=sponsor_name,
classification=sponsor_type,
entity_type="person",
primary=(sponsor_type == "primary"),
)
# Capture bill text versions
# Warning: There's a TODO in VT's source code saying 'move this to where it used to be'
# so leave in the old and new positions
versions = doc.xpath(
'//dl[@class="summary-table"]/dt[text()="Bill/Resolution Text"]/'
"following-sibling::dd[1]/ul/li/a |"
'//ul[@class="bill-path"]//a')
for version in versions:
if version.xpath("text()"):
bill.add_version_link(
note=version.xpath("text()")[0],
url=version.xpath("@href")[0].replace(" ", "%20"),
media_type="application/pdf",
)
# Identify the internal bill ID, used for actions and votes
# If there is no internal bill ID, then it has no extra information
try:
internal_bill_id = re.search(
r'"bill/loadBillDetailedStatus/.+?/(\d+)"',
lxml.etree.tostring(doc).decode("utf-8"),
).group(1)
except AttributeError:
self.warning("Bill {} appears to have no activity".format(
info["BillNumber"]))
yield bill
continue
# Capture actions
actions_url = "http://legislature.vermont.gov/bill/loadBillDetailedStatus/{0}/{1}".format(
year_slug, internal_bill_id)
actions_json = self.get(actions_url)
# Checks if page actually has json posted
if "json" in actions_json.headers.get("Content-Type"):
actions = json.loads(actions_json.text)["data"]
# Checks to see if any data is actually there
if actions == "":
continue
else:
continue
bill.add_source(actions_url)
chambers_passed = set()
for action in actions:
action = {k: v for k, v in action.items() if v is not None}
if "Signed by Governor" in action["FullStatus"]:
actor = "executive"
elif action["ChamberCode"] == "H":
actor = "lower"
elif action["ChamberCode"] == "S":
actor = "upper"
else:
raise AssertionError("Unknown actor for bill action")
# Categorize action
if "Signed by Governor" in action["FullStatus"]:
# assert chambers_passed == set("HS")
action_type = "executive-signature"
elif "Vetoed by the Governor" in action["FullStatus"]:
action_type = "executive-veto"
elif ("Read first time" in action["FullStatus"]
or "Read 1st time" in action["FullStatus"]):
action_type = "introduction"
elif "Reported favorably" in action["FullStatus"]:
action_type = "committee-passage-favorable"
elif actor == "lower" and any(
x.lower().startswith("aspassed")
for x in action["keywords"].split(";")):
action_type = "passage"
chambers_passed.add("H")
elif actor == "upper" and any(
x.lower().startswith(" aspassed")
or x.lower().startswith("aspassed")
for x in action["keywords"].split(";")):
action_type = "passage"
chambers_passed.add("S")
else:
action_type = None
# Manual fix for data error in
# https://legislature.vermont.gov/bill/status/2020/H.511
action["StatusDate"] = action["StatusDate"].replace(
"/0209", "/2019")
# Manual fix for data error in
# https://legislature.vermont.gov/bill/status/2020/H.754
if bill_id == "H 754" and session == "2019-2020":
action["StatusDate"] = action["StatusDate"].replace(
"/0202", "/2020")
# https://legislature.vermont.gov/bill/status/2020/H.942
if bill_id == "H 942" and session == "2019-2020":
action["StatusDate"] = action["StatusDate"].replace(
"/0200", "/2020")
action_date = datetime.datetime.strftime(
datetime.datetime.strptime(action["StatusDate"],
"%m/%d/%Y"),
"%Y-%m-%d",
)
# strftime doesn't always pad year value (%Y) (https://bugs.python.org/issue32195)
# and sometimes this state has typos in year part of the StatusDate value
# which can cause validation errors, so fix leading zeroes if they are missing
if action_date.find("-") < 4:
action_date = ("0" *
(4 - action_date.find("-"))) + action_date
bill.add_action(
description=re.sub(HTML_TAGS_RE, "", action["FullStatus"]),
date=action_date,
chamber=actor,
classification=action_type,
)
# Capture votes
votes_url = "http://legislature.vermont.gov/bill/loadBillRollCalls/{0}/{1}".format(
year_slug, internal_bill_id)
votes_json = self.get(votes_url).text
votes = json.loads(votes_json)["data"]
bill.add_source(votes_url)
for vote in votes:
roll_call_id = vote["VoteHeaderID"]
roll_call_url = ("http://legislature.vermont.gov/bill/"
"loadBillRollCallDetails/{0}/{1}".format(
year_slug, roll_call_id))
roll_call_json = self.get(roll_call_url).text
roll_call = json.loads(roll_call_json)["data"]
roll_call_yea = []
roll_call_nay = []
roll_call_not_voting = []
for member in roll_call:
(member_name,
_district) = member["MemberName"].split(" of ")
member_name = member_name.strip()
if member["MemberVote"] == "Yea":
roll_call_yea.append(member_name)
elif member["MemberVote"] == "Nay":
roll_call_nay.append(member_name)
else:
roll_call_not_voting.append(member_name)
if ("Passed -- " in vote["FullStatus"]
# seems like we've seen both
or "Governor overridden" in vote["FullStatus"] or
"Governor overriden" in vote["FullStatus"]):
did_pass = True
elif ("Failed -- " in vote["FullStatus"] or
"Veto of the Governor sustained" in vote["FullStatus"]):
did_pass = False
else:
raise AssertionError("Roll call vote result is unclear: " +
vote["FullStatus"])
# Check vote counts
yea_count = int(
re.search(r"Yeas = (\d+)", vote["FullStatus"]).group(1))
nay_count = int(
re.search(r"Nays = (\d+)", vote["FullStatus"]).group(1))
vote_start_date = datetime.datetime.strftime(
datetime.datetime.strptime(vote["StatusDate"], "%m/%d/%Y"),
"%Y-%m-%d",
)
motion_text = re.sub(HTML_TAGS_RE, "",
vote["FullStatus"]).strip()
vote_identifer = (vote["StatusDate"] + "--" + motion_text +
"--" + roll_call_url)
vote_to_add = VoteEvent(
identifier=vote_identifer,
bill=bill,
chamber=("lower"
if vote["ChamberCode"] == "H" else "upper"),
start_date=vote_start_date,
motion_text=motion_text,
result="pass" if did_pass else "fail",
classification="passage",
legislative_session=session,
)
vote_to_add.add_source(roll_call_url)
vote_to_add.set_count("yes", yea_count)
vote_to_add.set_count("no", nay_count)
vote_to_add.set_count("not voting", len(roll_call_not_voting))
for member in roll_call_yea:
vote_to_add.yes(member)
for member in roll_call_nay:
vote_to_add.no(member)
for member in roll_call_not_voting:
vote_to_add.vote("not voting", member)
yield vote_to_add
# Witnesses:
# http://legislature.vermont.gov/bill/loadBillWitnessList/{year_slug}/{internal_bill_id}
witnesses_doc_link_url = "https://legislature.vermont.gov/bill/print/2020/{0}/witnesses".format(
bill_id_original_format)
bill.add_document_link(note="Witness List",
url=witnesses_doc_link_url,
media_type="text/html")
# Conference committee members:
# http://legislature.vermont.gov/bill/loadBillConference/{year_slug}/{bill_number}
conferees_doc_link_url = "https://legislature.vermont.gov/bill/print/2020/{0}/conference".format(
bill_id_original_format)
page = self.lxmlize(conferees_doc_link_url)
no_data = page.xpath('//div[@class="no-data"]/text()')
if not no_data:
bill.add_document_link(
note="Conference Committee Members",
url=conferees_doc_link_url,
media_type="text/html",
)
# Committee meetings:
# http://legislature.vermont.gov/committee/loadHistoryByBill/{year_slug}?LegislationId={internal_bill_id}
meetings_doc_link_url = "https://legislature.vermont.gov/bill/print/2020/{0}/meetings".format(
bill_id_original_format)
bill.add_document_link(
note="Committee Meetings",
url=meetings_doc_link_url,
media_type="text/html",
)
yield bill
def scrape_bill(self, session, bill_id, chamber):
# https://malegislature.gov/Bills/189/SD2739
session_for_url = self.replace_non_digits(session)
bill_url = "https://malegislature.gov/Bills/{}/{}".format(
session_for_url, bill_id)
try:
response = self.get(bill_url)
self.info("GET (with `requests`) - {}".format(bill_url))
except requests.exceptions.RequestException:
self.warning("Server Error on {}".format(bill_url))
return False
html = response.text
page = lxml.html.fromstring(html)
if not page.xpath('//div[contains(@class, "followable")]/h1/text()'):
self.warning("Server Error on {}".format(bill_url))
return False
# The state website will periodically miss a few bills' titles for a few days
# These titles will be extant on the bill list page, but missing on the bill detail page
# The titles are eventually populated under one of two markups
try:
bill_title = page.xpath(
'//div[@id="contentContainer"]/div/div/h2/text()')[0]
except IndexError:
bill_title = None
pass
if bill_title is None:
try:
bill_title = page.xpath(
'//div[contains(@class,"followable")]/h1/text()')[0]
bill_title = bill_title.replace("Bill", "").strip()
except IndexError:
self.warning(
"Couldn't find title for {}; skipping".format(bill_id))
return False
bill_types = ["H", "HD", "S", "SD", "SRes"]
if re.sub("[0-9]", "", bill_id) not in bill_types:
self.warning(
"Unsupported bill type for {}; skipping".format(bill_id))
return False
if "SRes" in bill_id:
bill_id = bill_id.replace("SRes", "SR")
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=bill_title,
classification="bill",
)
bill_summary = None
if page.xpath('//p[@id="pinslip"]/text()'):
bill_summary = page.xpath('//p[@id="pinslip"]/text()')[0]
if bill_summary:
bill.add_abstract(bill_summary, "summary")
bill.add_source(bill_url)
# https://malegislature.gov/Bills/189/SD2739 has a presenter
# https://malegislature.gov/Bills/189/S2168 no sponsor
# Find the non-blank text of the dt following Sponsor or Presenter,
# including any child link text.
sponsor = page.xpath(
'//dt[text()="Sponsor:" or text()="Presenter:"]/'
"following-sibling::dd/descendant-or-self::*/text()[normalize-space()]"
)
if sponsor:
sponsor = (sponsor[0].replace("*", "").replace("%", "").replace(
"This sponsor is an original petitioner.", "").strip())
bill.add_sponsorship(sponsor,
classification="primary",
primary=True,
entity_type="person")
self.scrape_cosponsors(bill, bill_url)
version = page.xpath(
"//div[contains(@class, 'modalBtnGroup')]/"
"a[contains(text(), 'Download PDF') and not(@disabled)]/@href")
if version:
version_url = "https://malegislature.gov{}".format(version[0])
bill.add_version_link("Bill Text",
version_url,
media_type="application/pdf")
self.scrape_actions(bill, bill_url, session)
yield bill
def scrape(self, session=None):
if not session:
session = self.latest_session()
self.info("no session specified, using %s", session)
for category in self._categories:
leg_listing_url = (
self._API_BASE_URL + f"BulkData/{category['categoryId']}/{session}"
)
resp = requests.post(leg_listing_url, headers=self._headers, verify=False,)
resp.raise_for_status()
leg_listing = resp.json()
for leg in leg_listing:
bill = Bill(
leg["legislationNumber"],
legislative_session=session,
title=leg["title"],
classification=category["name"],
)
bill.add_source(leg_listing_url)
bill_url = (
f"https://lims.dccouncil.us/Legislation/{leg['legislationNumber']}"
)
bill.add_source(bill_url)
if leg['lawNumber']:
bill.extras['lawNumber'] = leg['lawNumber']
# Actions
for hist in leg["legislationHistory"]:
hist_date = datetime.datetime.strptime(
hist["actionDate"], "%b %d, %Y"
)
hist_date = self._TZ.localize(hist_date)
hist_action = hist["actionDescription"]
if hist_action.split()[0] in ["OtherAmendment", "OtherMotion"]:
hist_action = hist_action[5:]
hist_class = self.classify_action(hist_action)
if "mayor" in hist_action.lower():
actor = "executive"
else:
actor = "legislature"
bill.add_action(
hist_action, hist_date, classification=hist_class, chamber=actor
)
# Documents with download links
if hist["downloadURL"] and ("download" in hist["downloadURL"]):
download = hist["downloadURL"]
if not download.startswith("http"):
download = "https://lims.dccouncil.us/" + download
mimetype = (
"application/pdf" if download.endswith("pdf") else None
)
is_version = False
# figure out if it's a version from type/name
possible_version_types = [
"SignedAct",
"Introduction",
"Enrollment",
"Engrossment",
]
for vt in possible_version_types:
if vt.lower() in download.lower():
is_version = True
doc_type = vt
if "amendment" in download.lower():
doc_type = "Amendment"
if is_version:
bill.add_version_link(
doc_type,
download,
media_type=mimetype,
on_duplicate="ignore",
)
else:
bill.add_document_link(
hist["actionDescription"],
download,
media_type=mimetype,
on_duplicate="ignore",
)
# Grabs Legislation details
leg_details_url = (
self._API_BASE_URL
+ f"LegislationDetails/{leg['legislationNumber']}"
)
details_resp = requests.get(
leg_details_url, headers=self._headers, verify=False,
)
details_resp.raise_for_status()
leg_details = details_resp.json()
# Sponsors
for i in leg_details["introducers"]:
name = i["memberName"]
bill.add_sponsorship(
name,
classification="primary",
entity_type="person",
primary=True,
)
# Co-sponsor
if leg_details["coSponsors"]:
for cs in leg_details["coSponsors"]:
name = i["memberName"]
bill.add_sponsorship(
name,
classification="cosponsor",
entity_type="person",
primary=True,
)
# Committee Hearing Doc
for commHearing in leg_details["committeeHearing"]:
if commHearing["hearingRecord"]:
bill.add_document_link(
commHearing["hearingType"],
commHearing["hearingRecord"],
media_type="application/pdf",
on_duplicate="ignore",
)
for committeeMarkup in leg_details["committeeMarkup"]:
if committeeMarkup["committeeReport"]:
bill.add_document_link(
"Committee Markup",
committeeMarkup["committeeReport"],
media_type="application/pdf",
on_duplicate="ignore",
)
# Actions and Votes
if leg_details["actions"]:
# To prevent duplicate votes
vote_ids = []
for act in leg_details["actions"]:
action_name = act["action"]
action_date = datetime.datetime.strptime(
act["actionDate"][:10], "%Y-%m-%d"
)
action_date = self._TZ.localize(action_date)
if action_name.split()[0] == "Other":
action_name = " ".join(action_name.split()[1:])
if "mayor" in action_name.lower():
actor = "executive"
else:
actor = "legislature"
# Documents and Versions
if act["attachment"]:
mimetype = (
"application/pdf"
if act["attachment"].endswith("pdf")
else None
)
is_version = False
# figure out if it's a version from type/name
possible_version_types = [
"SignedAct",
"Introduction",
"Enrollment",
"Engrossment",
]
for vt in possible_version_types:
if vt.lower() in act["attachment"].lower():
is_version = True
doc_type = vt
if "amendment" in act["attachment"].lower():
doc_type = "Amendment"
if is_version:
bill.add_version_link(
doc_type,
act["attachment"],
media_type=mimetype,
on_duplicate="ignore",
)
else:
bill.add_document_link(
doc_type,
act["attachment"],
media_type=mimetype,
on_duplicate="ignore",
)
# Votes
if act["voteDetails"]:
result = act["voteDetails"]["voteResult"]
if result:
status = self._vote_statuses[result.lower()]
id_text = (
str(leg["legislationNumber"])
+ "-"
+ action_name
+ "-"
+ result
)
if id_text not in vote_ids:
vote_ids.append(id_text)
action_class = self.classify_action(action_name)
v = VoteEvent(
identifier=id_text,
chamber=actor,
start_date=action_date,
motion_text=action_name,
result=status,
classification=action_class,
bill=bill,
)
v.add_source(leg_listing_url)
yes_count = (
no_count
) = absent_count = abstain_count = other_count = 0
for leg_vote in act["voteDetails"]["votes"]:
mem_name = leg_vote["councilMember"]
if leg_vote["vote"] == "Yes":
yes_count += 1
v.yes(mem_name)
elif leg_vote["vote"] == "No":
no_count += 1
v.no(mem_name)
elif leg_vote["vote"] == "Absent":
absent_count += 1
v.vote("absent", mem_name)
elif leg_vote["vote"] == "Recused":
v.vote("abstain", mem_name)
abstain_count += 1
elif leg_vote["vote"] == "Present":
v.vote("other", mem_name)
other_count += 1
else:
# Incase anything new pops up
other_count += 1
v.vote("other", mem_name)
v.set_count("yes", yes_count)
v.set_count("no", no_count)
v.set_count("absent", absent_count)
v.set_count("abstain", abstain_count)
v.set_count("other", other_count)
yield v
yield bill
def scrape_details(self, bill_detail_url, session, chamber, bill_id):
"""
Create the Bill and add the information obtained from the provided bill_detail_url.
and then yield the bill object.
:param bill_detail_url:
:param session:
:param chamber:
:param bill_id:
:return:
"""
page = self.get(bill_detail_url).text
if "INVALID BILL NUMBER" in page:
self.warning("INVALID BILL %s" % bill_detail_url)
return
doc = lxml.html.fromstring(page)
doc.make_links_absolute(bill_detail_url)
bill_div = doc.xpath('//div[@style="margin:0 0 40px 0;"]')[0]
bill_type = bill_div.xpath("span/text()")[0]
if "General Bill" in bill_type:
bill_type = "bill"
elif "Concurrent Resolution" in bill_type:
bill_type = "concurrent resolution"
elif "Joint Resolution" in bill_type:
bill_type = "joint resolution"
elif "Resolution" in bill_type:
bill_type = "resolution"
else:
raise ValueError("unknown bill type: %s" % bill_type)
# this is fragile, but less fragile than it was
b = bill_div.xpath('./b[text()="Summary:"]')[0]
bill_summary = b.getnext().tail.strip()
bill = Bill(
bill_id,
legislative_session=session, # session name metadata's `legislative_sessions`
chamber=chamber, # 'upper' or 'lower'
title=bill_summary,
classification=bill_type,
)
subjects = list(self._subjects[bill_id])
for subject in subjects:
bill.add_subject(subject)
# sponsors
for sponsor in doc.xpath('//a[contains(@href, "member.php")]/text()'):
bill.add_sponsorship(
name=sponsor,
classification="primary",
primary=True,
entity_type="person",
)
for sponsor in doc.xpath('//a[contains(@href, "committee.php")]/text()'):
sponsor = sponsor.replace("\xa0", " ").strip()
bill.add_sponsorship(
name=sponsor,
classification="primary",
primary=True,
entity_type="organization",
)
# find versions
version_url = doc.xpath('//a[text()="View full text"]/@href')[0]
version_html = self.get(version_url).text
version_doc = lxml.html.fromstring(version_html)
version_doc.make_links_absolute(version_url)
for version in version_doc.xpath('//a[contains(@href, "/prever/")]'):
# duplicate versions with same date, use first appearance
bill.add_version_link(
note=version.text, # Description of the version from the state;
# eg, 'As introduced', 'Amended', etc.
url=version.get("href"),
on_duplicate="ignore",
media_type="text/html", # Still a MIME type
)
# actions
for row in bill_div.xpath("table/tr"):
date_td, chamber_td, action_td = row.xpath("td")
date = datetime.datetime.strptime(date_td.text, "%m/%d/%y")
action_chamber = {"Senate": "upper", "House": "lower", None: "legislature"}[
chamber_td.text
]
action = action_td.text_content()
action = action.split("(House Journal")[0]
action = action.split("(Senate Journal")[0].strip()
atype = action_type(action)
bill.add_action(
description=action, # Action description, from the state
date=date.strftime("%Y-%m-%d"), # `YYYY-MM-DD` format
chamber=action_chamber, # 'upper' or 'lower'
classification=atype, # Options explained in the next section
)
# votes
vurl = doc.xpath('//a[text()="View Vote History"]/@href')
if vurl:
vurl = vurl[0]
yield from self.scrape_vote_history(bill, vurl)
bill.add_source(bill_detail_url)
yield bill
def scrape_bill(self, chamber, session, bill_id, title, url):
page = self.get(url).json()
api_id = page["BillId"]
if re.match(r"^(S|H)B ", bill_id):
btype = ["bill"]
elif re.match(r"(S|H)C ", bill_id):
btype = ["commemoration"]
elif re.match(r"(S|H)JR ", bill_id):
btype = ["joint resolution"]
elif re.match(r"(S|H)CR ", bill_id):
btype = ["concurrent resolution"]
else:
btype = ["bill"]
bill = Bill(
bill_id,
legislative_session=session,
chamber=chamber,
title=title,
classification=btype,
)
bill.add_source(f"https://sdlegislature.gov/Session/Bill/{api_id}")
bill.add_source(url)
version_rows = page["Documents"]
assert len(version_rows) > 0
for version in version_rows:
date = version["DocumentDate"]
if date:
match = re.match(r"\d{4}-\d{2}-\d{2}", date)
date = datetime.datetime.strptime(match.group(0),
"%Y-%m-%d").date()
html_link = f"https://sdlegislature.gov/Session/Bill/{api_id}/{version['DocumentId']}"
pdf_link = f"https://mylrc.sdlegislature.gov/api/Documents/{version['DocumentId']}.pdf"
note = version["BillVersion"]
bill.add_version_link(
note,
html_link,
date=date,
media_type="text/html",
on_duplicate="ignore",
)
bill.add_version_link(
note,
pdf_link,
date=date,
media_type="application/pdf",
on_duplicate="ignore",
)
else:
self.warning("Version listed but no date or documents")
sponsors = page["BillSponsor"]
if sponsors:
for sponsor in sponsors:
sponsor_type = "person"
member = sponsor["Member"]
# first and last name are available, but UniqueName is the old link text
# could change later?
bill.add_sponsorship(
member["UniqueName"],
classification="primary",
primary=True,
entity_type=sponsor_type,
)
else:
sponsor_type = "organization"
committee_sponsor = re.search(r">(.*)</a>",
page["BillCommitteeSponsor"])[1]
bill.add_sponsorship(
committee_sponsor,
classification="primary",
primary=True,
entity_type=sponsor_type,
)
for keyword in page["Keywords"]:
bill.add_subject(keyword["Keyword"]["Keyword"])
actions_url = f"https://sdlegislature.gov/api/Bills/ActionLog/{api_id}"
yield from self.scrape_action(bill, actions_url, chamber)
yield bill
def _scrape_bill(self, session, bill_data):
details = self._parse_bill_details(bill_data)
if details is None:
return
(
senate_url,
assembly_url,
bill_chamber,
bill_type,
bill_id,
title,
(prefix, number, active_version),
) = details
bill = Bill(
bill_id,
legislative_session=session,
chamber=bill_chamber,
title=title or bill_data["summary"],
classification=bill_type,
)
if bill_data["summary"]:
bill.add_abstract(bill_data["summary"], note="")
bill_active_version = None
if active_version != "":
bill_active_version = bill_data["amendments"]["items"][active_version]
else:
self.warning("No active version for {}".format(bill_id))
# Parse sponsors.
if bill_data["sponsor"] is not None:
if bill_data["sponsor"]["rules"] is True:
bill.add_sponsorship(
"Rules Committee",
entity_type="organization",
classification="primary",
primary=True,
)
elif not bill_data["sponsor"]["budget"]:
primary_sponsor = bill_data["sponsor"]["member"]
bill.add_sponsorship(
primary_sponsor["shortName"],
entity_type="person",
classification="primary",
primary=True,
)
if bill_active_version:
# There *shouldn't* be cosponsors if there is no sponsor.
cosponsors = bill_active_version["coSponsors"]["items"]
for cosponsor in cosponsors:
bill.add_sponsorship(
cosponsor["shortName"],
entity_type="person",
classification="cosponsor",
primary=False,
)
if bill_active_version:
# List companion bill.
same_as = bill_active_version.get("sameAs", {})
# Check whether "sameAs" property is populated with at least one bill.
if same_as["items"]:
# Get companion bill ID.
companion_bill_id = same_as["items"][0]["basePrintNo"]
# Build companion bill session.
start_year = same_as["items"][0]["session"]
end_year = start_year + 1
companion_bill_session = "-".join([str(start_year), str(end_year)])
# Attach companion bill data.
bill.add_related_bill(
companion_bill_id, companion_bill_session, relation_type="companion"
)
# Parse actions.
chamber_map = {"senate": "upper", "assembly": "lower"}
for action in bill_data["actions"]["items"]:
chamber = chamber_map[action["chamber"].lower()]
action_datetime = datetime.datetime.strptime(action["date"], "%Y-%m-%d")
action_date = action_datetime.date()
types, _ = NYBillScraper.categorizer.categorize(action["text"])
bill.add_action(
action["text"],
action_date.strftime("%Y-%m-%d"),
chamber=chamber,
classification=types,
)
# Handling of sources follows. Sources serving either chamber
# maintain duplicate data, so we can see certain bill data
# through either chamber's resources. However, we have to refer
# to a specific chamber's resources if we want to grab certain
# specific information such as vote data.
#
# As such, I'm placing all potential sources in the interest of
# thoroughness. - Andy Lo
# List Open Legislation API endpoint as a source.
api_url = self.api_client.root + self.api_client.resources["bill"].format(
session_year=session, bill_id=bill_id, summary="", detail=""
)
bill.add_source(api_url)
bill.add_source(senate_url)
bill.add_source(assembly_url)
# Chamber-specific processing.
for vote_data in bill_data["votes"]["items"]:
yield self._parse_senate_votes(vote_data, bill, api_url)
yield from self.scrape_assembly_votes(session, bill, assembly_url, bill_id)
# A little strange the way it works out, but the Assembly
# provides the HTML version documents and the Senate provides
# the PDF version documents.
amendments = bill_data["amendments"]["items"]
for key, amendment in amendments.items():
version = amendment["printNo"]
html_url = (
"http://assembly.state.ny.us/leg/?sh=printbill&bn="
"{}&term={}&Text=Y".format(bill_id, self.term_start_year)
)
bill.add_version_link(
version, html_url, on_duplicate="ignore", media_type="text/html"
)
pdf_url = "http://legislation.nysenate.gov/pdf/bills/{}/{}".format(
self.term_start_year, version
)
bill.add_version_link(
version, pdf_url, on_duplicate="ignore", media_type="application/pdf"
)
yield bill