def scrape_votes_old(self, bill, billname, session):
vote_url = ("http://archives.legislature.state.oh.us/bills.cfm?ID=" +
session + "_" + billname)
page = self.get(vote_url).text
page = lxml.html.fromstring(page)
for jlink in page.xpath("//a[contains(@href, 'JournalText')]"):
date = self._tz.localize(
datetime.datetime.strptime(jlink.text, "%m/%d/%Y")).date()
date = "{:%Y-%m-%d}".format(date)
details = jlink.xpath("string(../../../td[2])")
chamber = details.split(" - ")[0]
if chamber == "House":
chamber = "lower"
elif chamber == "Senate":
chamber = "upper"
else:
raise ScrapeError("Bad chamber: %s" % chamber)
motion = details.split(" - ")[1].split("\n")[0].strip()
vote_row = jlink.xpath("../../..")[0].getnext()
yea_div = vote_row.xpath("td/font/div[contains(@id, 'Yea')]")[0]
yeas = []
for td in yea_div.xpath("table/tr/td"):
name = td.xpath("string()")
if name:
yeas.append(name)
no_div = vote_row.xpath("td/font/div[contains(@id, 'Nay')]")[0]
nays = []
for td in no_div.xpath("table/tr/td"):
name = td.xpath("string()")
if name:
nays.append(name)
yes_count = len(yeas)
no_count = len(nays)
vote = VoteEvent(
chamber=chamber,
start_date=date,
motion_text=motion,
result="pass" if yes_count > no_count else "fail",
bill=bill,
classification="passage",
)
for yes in yeas:
vote.yes(yes)
for no in nays:
vote.no(no)
vote.add_source(vote_url)
yield vote
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, title, url):
page = self.lxmlize(url)
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(url)
version_rows = page.xpath(
'//div[@id="ctl00_ContentPlaceHolder1_ctl00_BillVersions"]'
+ "/section/table/tbody/tr"
)
assert len(version_rows) > 0
for row in version_rows:
(date,) = row.xpath('./td[@data-title="Date"]/text()')
date = date.strip()
date = datetime.datetime.strptime(date, "%m/%d/%Y").date()
(html_note,) = row.xpath('./td[@data-title="HTML"]/a/text()')
(html_link,) = row.xpath('./td[@data-title="HTML"]/a/@href')
(pdf_note,) = row.xpath('./td[@data-title="PDF"]/a/text()')
(pdf_link,) = row.xpath('./td[@data-title="PDF"]/a/@href')
assert html_note == pdf_note
note = html_note
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",
)
sponsor_links = page.xpath(
'//div[@id="ctl00_ContentPlaceHolder1_ctl00_BillDetail"]'
+ '/label[contains(text(), "Sponsors:")]'
+ "/following-sibling::div[1]/p/a"
)
for link in sponsor_links:
if link.attrib["href"].startswith("https://sdlegislature.gov/Legislators/"):
sponsor_type = "person"
elif link.attrib["href"].startswith(
"https://sdlegislature.gov/Legislative_Session/Committees"
):
sponsor_type = "organization"
else:
raise ScrapeError(
"Found unexpected sponsor, URL: " + link.attrib["href"]
)
bill.add_sponsorship(
link.text,
classification="primary",
primary=True,
entity_type=sponsor_type,
)
actor = chamber
use_row = False
for row in page.xpath("//table[contains(@id, 'tblBillActions')]//tr"):
# Some tables have null rows, that are just `<tr></tr>`
# Eg: sdlegislature.gov/Legislative_Session/Bills/Bill.aspx?Bill=1005&Session=2018
if row.text_content() == "":
self.debug("Skipping action table row that is completely empty")
continue
if "Date" in row.text_content() and "Action" in row.text_content():
use_row = True
continue
elif not use_row:
continue
action = row.xpath("string(td[2])").strip()
atypes = []
if action.startswith("First read"):
atypes.append("introduction")
atypes.append("reading-1")
if re.match(r"Signed by (?:the\s)*Governor", action, re.IGNORECASE):
atypes.append("executive-signature")
actor = "executive"
match = re.match(r"(.*) Do Pass( Amended)?, (Passed|Failed)", action)
if match:
if match.group(1) in ["Senate", "House of Representatives"]:
first = ""
else:
first = "committee-"
if match.group(3).lower() == "passed":
second = "passage"
elif match.group(3).lower() == "failed":
second = "failure"
atypes.append("%s%s" % (first, second))
if "referred to" in action.lower():
atypes.append("referral-committee")
if "Motion to amend, Passed Amendment" in action:
atypes.append("amendment-introduction")
atypes.append("amendment-passage")
if row.xpath('td[2]/a[contains(@href,"Amendment.aspx")]'):
amd = row.xpath('td[2]/a[contains(@href,"Amendment.aspx")]')[0]
version_name = amd.xpath("string(.)")
version_url = amd.xpath("@href")[0]
if "htm" in version_url:
mimetype = "text/html"
elif "pdf" in version_url:
mimetype = "application/pdf"
bill.add_version_link(
version_name,
version_url,
media_type=mimetype,
on_duplicate="ignore",
)
if "Veto override, Passed" in action:
atypes.append("veto-override-passage")
elif "Veto override, Failed" in action:
atypes.append("veto-override-failure")
if "Delivered to the Governor" in action:
atypes.append("executive-receipt")
match = re.match("First read in (Senate|House)", action)
if match:
if match.group(1) == "Senate":
actor = "upper"
else:
actor = "lower"
date = row.xpath("string(td[1])").strip()
match = re.match(r"\d{2}/\d{2}/\d{4}", date)
if not match:
self.warning("Bad date: %s" % date)
continue
date = datetime.datetime.strptime(date, "%m/%d/%Y").date()
for link in row.xpath("td[2]/a[contains(@href, 'RollCall')]"):
yield from self.scrape_vote(bill, date, link.attrib["href"])
if action:
bill.add_action(action, date, chamber=actor, classification=atypes)
for link in page.xpath("//a[contains(@href, 'Keyword')]"):
bill.add_subject(link.text.strip())
yield bill
def scrape_vote(self, bill, date, url):
page = self.get(url).text
page = lxml.html.fromstring(page)
header = page.xpath("string(//h3[contains(@id, 'hdVote')])")
if "No Bill Action" in header:
self.warning("bad vote header -- skipping")
return
location = header.split(", ")[1]
if location.startswith("House"):
chamber = "lower"
elif location.startswith("Senate"):
chamber = "upper"
elif location.startswith("Joint"):
chamber = "legislature"
else:
raise ScrapeError("Bad chamber: %s" % location)
motion = ", ".join(header.split(", ")[2:]).strip()
if motion:
# If we can't detect a motion, skip this vote
yes_count = int(page.xpath("string(//span[contains(@id, 'tdAyes')])"))
no_count = int(page.xpath("string(//span[contains(@id, 'tdNays')])"))
excused_count = int(
page.xpath("string(//span[contains(@id, 'tdExcused')])")
)
absent_count = int(page.xpath("string(//span[contains(@id, 'tdAbsent')])"))
passed = yes_count > no_count
if motion.startswith("Do Pass"):
type = "passage"
elif motion == "Concurred in amendments":
type = "amendment"
elif motion == "Veto override":
type = "veto_override"
else:
type = "other"
vote = VoteEvent(
chamber=chamber,
start_date=date,
motion_text=motion,
result="pass" if passed else "fail",
classification=type,
bill=bill,
)
# The vote page URL has a unique ID
# However, some votes are "consent calendar" events,
# and relate to the passage of _multiple_ bills
# These can't be modeled yet in Pupa, but for now we can
# append a bill ID to the URL that forms the `pupa_id`
# https://github.com/opencivicdata/pupa/issues/308
vote.pupa_id = "{}#{}".format(url, bill.identifier.replace(" ", ""))
vote.add_source(url)
vote.set_count("yes", yes_count)
vote.set_count("no", no_count)
vote.set_count("excused", excused_count)
vote.set_count("absent", absent_count)
for td in page.xpath("//table[@id='tblVoteTotals']/tbody/tr/td"):
option_or_person = td.text.strip()
if option_or_person in ("Aye", "Yea"):
vote.yes(td.getprevious().text.strip())
elif option_or_person == "Nay":
vote.no(td.getprevious().text.strip())
elif option_or_person == "Excused":
vote.vote("excused", td.getprevious().text.strip())
elif option_or_person == "Absent":
vote.vote("absent", td.getprevious().text.strip())
yield vote