def scrape_record(case_number):
"""
Scrapes a record once the case has been opened.
:param case_number: The current case's case number.
"""
# Wait for court summary to load
for i in range(settings['connect-thresh']):
try:
WebDriverWait(driver, 5).until(EC.presence_of_element_located((By.ID, 'summaryAccordion')))
except TimeoutException:
if i == settings['connect-thresh'] - 1:
raise RuntimeError('Summary details did not load for case {}.'.format(case_number))
else:
driver.refresh()
# Get relevant page content
summary_table_col1 = driver.find_elements_by_xpath('//*[@id="summaryAccordionCollapse"]/table/tbody/tr/td[1]/dl/dd')
summary_table_col2 = driver.find_elements_by_xpath('//*[@id="summaryAccordionCollapse"]/table/tbody/tr/td[2]/dl/dd')
summary_table_col3 = driver.find_elements_by_xpath('//*[@id="summaryAccordionCollapse"]/table/tbody/tr/td[3]/dl/dd')
# Wait for court dockets to load
for i in range(settings['connect-thresh']):
try:
WebDriverWait(driver, 5).until(EC.presence_of_element_located((By.ID, 'gridDocketsView')))
except TimeoutException:
if i == settings['connect-thresh'] - 1:
raise RuntimeError('Dockets did not load for case {}.'.format(case_number))
else:
driver.refresh()
charges_table = driver.find_elements_by_xpath('//*[@id="gridCharges"]/tbody/tr')
docket_public_defender = driver.find_elements_by_xpath(
"//*[contains(text(), 'COURT APPOINTED ATTORNEY') and contains(text(), 'ASSIGNED')]")
docket_attorney = driver.find_elements_by_xpath("//*[contains(text(), 'DEFENSE') and contains(text(), 'ASSIGNED')]")
docket_pleas = driver.find_elements_by_xpath("//*[contains(text(), 'PLEA OF')]")
docket_attachments = driver.find_elements_by_class_name('casedocketimage')
_id = str(uuid.uuid4())
_state = settings['state-code']
_county = settings['county']
CaseNum = summary_table_col2[1].text.strip()
AgencyReportNum = summary_table_col1[4].text.strip()
ArrestDate = None # Can't be found on this portal
FilingDate = summary_table_col1[2].text.strip()
OffenseDate = None # Can't be found on this portal
DivisionName = summary_table_col3[3].text.strip()
CaseStatus = summary_table_col3[1].text.strip()
if settings['collect-pii']:
# Create list of assigned defense attorney(s)
defense_attorney_text = list(map(lambda x: x.text, docket_attorney))
DefenseAttorney = ScraperUtils.parse_attorneys(defense_attorney_text)
# Create list of assigned public defenders / appointed attorneys
public_defender_text = list(map(lambda x: x.text, docket_public_defender))
PublicDefender = ScraperUtils.parse_attorneys(public_defender_text)
# Get Judge
Judge = summary_table_col1[0].text.strip()
# Download docket attachments.
# Todo(OscarVanL): This could be parallelized to speed up scraping if save-attachments is set to 'all'.
if settings['save-attachments']:
for attachment_link in docket_attachments:
attachment_text = attachment_link.find_element_by_xpath('./../../td[3]').text.strip()
if settings['save-attachments'] == 'filing':
if not ('CITATION FILED' in attachment_text or 'CASE FILED' in attachment_text):
# Attachment is not a filing, don't download it.
continue
ScraperUtils.save_attached_pdf(driver, output_attachments, '{}-{}'.format(case_number, attachment_text),
settings['portal-base'], attachment_link, 20, settings['verbose'])
else:
DefenseAttorney = []
PublicDefender = []
Judge = None
Charges = {}
for charge in charges_table:
charge_details = charge.find_elements_by_tag_name('td')
count = int(charge_details[0].text.strip())
long_desc = charge_details[1].text.strip()
# Statute is contained within brackets
if '(' in long_desc and ')' in long_desc:
statute = long_desc[long_desc.find('(') + 1:long_desc.find(')')]
else:
statute = None
description = long_desc.split('(')[0]
level = charge_details[2].text.strip()
degree = charge_details[3].text.strip()
# plea = charge_details[4].text.strip() # Plea is not filled out on this portal.
disposition = charge_details[5].text.strip()
disposition_date = charge_details[6].text.strip()
offense_date = None # Not shown on this portal
citation_number = None # Not shown on this portal
Charges[count] = Charge(count, statute, description, level, degree, disposition, disposition_date, offense_date,
citation_number, None, None)
# Pleas are not in the 'plea' field, but instead in the dockets.
for plea_element in docket_pleas:
plea_text = plea_element.text.strip()
plea = ScraperUtils.parse_plea_type(plea_text)
plea_date = plea_element.find_element_by_xpath('./../td[2]').text.strip()
plea_number = ScraperUtils.parse_plea_case_numbers(plea_text, list(Charges.keys()))
# If no case number is specified in the plea, then we assume it applies to all charges in the trial.
if len(plea_number) == 0:
for charge in Charges.values():
charge.plea = plea
charge.plea_date = plea_date
else:
# Apply plea to relevant charge count(s).
for count in plea_number:
Charges[count].plea = plea
Charges[count].plea_date = plea_date
ArrestingOfficer = None # Can't be found on this portal
ArrestingOfficerBadgeNumber = None # Can't be found on this portal
profile_link = driver.find_element_by_xpath("//table[@id='gridParties']/tbody/tr/*[contains(text(), 'DEFENDANT')]/../td[2]/div/a").get_attribute(
'href')
# profile_link = driver.find_element_by_xpath('//*[@id="gridParties"]/tbody/tr[1]/td[2]/div[1]/a').get_attribute(
# 'href')
load_page(profile_link, 'Party Details:', settings['verbose'])
Suffix = None
DOB = None # This portal has DOB as N/A for every defendent
Race = driver.find_element_by_xpath(
'//*[@id="fd-table-2"]/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[7]/td[2]').text.strip()
Sex = driver.find_element_by_xpath(
'//*[@id="mainTableContent"]/tbody/tr/td/table/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[6]/td[2]').text.strip()
FirstName = None
MiddleName = None
LastName = None
PartyID = None
# Only collect PII if configured
if settings['collect-pii']:
# Navigate to party profile
full_name = driver.find_element_by_xpath(
'//*[@id="mainTableContent"]/tbody/tr/td/table/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[1]/td[2]').text.strip()
MiddleName = None
LastName = None
if ',' in full_name:
name_split = full_name.split(',')[1].lstrip().split()
FirstName = name_split[0]
MiddleName = " ".join(name_split[1:])
LastName = full_name.split(',')[0]
else:
# If there's no comma, it's a corporation name.
FirstName = full_name
PartyID = driver.find_element_by_xpath(
'//*[@id="mainTableContent"]/tbody/tr/td/table/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[8]/td[2]').text.strip() # PartyID is a field within the portal system to uniquely identify defendants
record = Record(_id, _state, _county, case_number, CaseNum, AgencyReportNum, PartyID, FirstName, MiddleName,
LastName, Suffix, DOB, Race, Sex, ArrestDate, FilingDate, OffenseDate, DivisionName, CaseStatus,
DefenseAttorney, PublicDefender, Judge, list(Charges.values()), ArrestingOfficer,
ArrestingOfficerBadgeNumber)
ScraperUtils.write_csv(output_file, record, settings['verbose'])
def test_parse_plea_case_numbers_blank(self):
assert ScraperUtils.parse_plea_case_numbers("", [1, 2, 3]) == []
def test_parse_plea_case_numbers__no_charge_mentioned(self):
plea = "PLEA OF NOT GUILTY"
assert ScraperUtils.parse_plea_case_numbers(plea, [1]) == []
def test_parse_plea_case_numbers_messy(self):
# Test a really ugly plea docket I found in one case
plea = "PLEA OF NOT GUILTY/DENIAL, WAIVER OF ARRAIGNMENT, DEMAND FOR NOTICE OF EXPERT TESTIMONY, DEMAND FOR DISCOVERY, DEMAND FOR STATEMENT OF PARTICULARS, DEMAND FOR JURY TRIAL, DESIGNATION OF E-MAIL ADDRESSES PURSUANT TO RULE 2.516 1/28/2020"
assert ScraperUtils.parse_plea_case_numbers(plea, [1, 2, 3]) == []
def test_parse_plea_case_numbers_multiple_case_numbers(self):
plea = "DEFENDANT ENTERED PLEA OF NOLO CONTENDERE SEQ: 1,2,3,4,5"
assert ScraperUtils.parse_plea_case_numbers(
plea, [1, 2, 3, 4, 5, 6]) == [1, 2, 3, 4, 5]
def test_parse_plea_case_numbers_no_charge_numbers(self):
plea = "DEFENDANT ENTERED PLEA OF NOLO CONTENDERE SEQ: 1,2,3,4,5"
assert ScraperUtils.parse_plea_case_numbers(plea, []) == []
def test_parse_plea_case_numbers_one_charge_mentioned(self):
plea = "DEFENDANT ENTERED PLEA OF : NOLO-CONTENDERE SEQ 2"
assert ScraperUtils.parse_plea_case_numbers(plea, [1, 2]) == [2]
def scrape_record(case_number):
"""
Scrapes a record once the case has been opened.
:param case_number: The current case's case number.
"""
# Wait for court summary to load
for i in range(FLAGS.connect_thresh):
try:
WebDriverWait(driver, 5).until(
EC.presence_of_element_located((By.ID, 'summaryAccordion')))
except TimeoutException:
if i == FLAGS.connect_thresh - 1:
raise RuntimeError(
'Summary details did not load for case {}.'.format(
case_number))
else:
driver.refresh()
# Get relevant page content
summary_table_col1 = driver.find_elements_by_xpath(
'//*[@id="summaryAccordionCollapse"]/table/tbody/tr/td[1]/dl/dd')
summary_table_col2 = driver.find_elements_by_xpath(
'//*[@id="summaryAccordionCollapse"]/table/tbody/tr/td[2]/dl/dd')
summary_table_col3 = driver.find_elements_by_xpath(
'//*[@id="summaryAccordionCollapse"]/table/tbody/tr/td[3]/dl/dd')
# Wait for court dockets to load
for i in range(FLAGS.connect_thresh):
try:
WebDriverWait(driver, 5).until(
EC.presence_of_element_located((By.ID, 'gridDocketsView')))
except TimeoutException:
if i == FLAGS.connect_thresh - 1:
raise RuntimeError(
'Dockets did not load for case {}.'.format(case_number))
else:
driver.refresh()
charges_table = driver.find_elements_by_xpath(
'//*[@id="gridCharges"]/tbody/tr')
docket_public_defender = driver.find_elements_by_xpath(
"//*[contains(text(), 'COURT APPOINTED ATTORNEY') and contains(text(), 'ASSIGNED')]"
)
docket_attorney = driver.find_elements_by_xpath(
"//*[contains(text(), 'DEFENSE') and contains(text(), 'ASSIGNED')]")
docket_pleas = driver.find_elements_by_xpath(
"//*[contains(text(), 'PLEA OF')]")
docket_attachments = driver.find_elements_by_class_name('casedocketimage')
r = BenchmarkRecordBuilder()
r.id = str(uuid.uuid4())
r.state = FLAGS.state
r.county = FLAGS.county
r.portal_id = case_number
r.case_num = Pii.String(summary_table_col2[1].text.strip())
r.agency_report_num = summary_table_col1[4].text.strip()
r.arrest_date = None # Can't be found on this portal
r.filing_date = summary_table_col1[2].text.strip()
r.offense_date = None # Can't be found on this portal
r.division_name = summary_table_col3[3].text.strip()
r.case_status = summary_table_col3[1].text.strip()
# Create list of assigned defense attorney(s)
defense_attorney_text = list(map(lambda x: x.text, docket_attorney))
r.defense_attorney = ScraperUtils.parse_attorneys(defense_attorney_text)
# Create list of assigned public defenders / appointed attorneys
public_defender_text = list(map(lambda x: x.text, docket_public_defender))
r.public_defender = ScraperUtils.parse_attorneys(public_defender_text)
# Get Judge
r.judge = Pii.String(summary_table_col1[0].text.strip())
# Download docket attachments.
# Todo(OscarVanL): This could be parallelized to speed up scraping if save-attachments is set to 'all'.
if FLAGS.save_attachments:
for attachment_link in docket_attachments:
attachment_text = attachment_link.find_element_by_xpath(
'./../../td[3]').text.strip()
if FLAGS.save_attachments == 'filing':
if not ('CITATION FILED' in attachment_text
or 'CASE FILED' in attachment_text):
# Attachment is not a filing, don't download it.
continue
ScraperUtils.save_attached_pdf(
driver, output_attachments,
'{}-{}'.format(case_number, attachment_text),
FLAGS.portal_base, attachment_link, 20, FLAGS.verbose)
Charges = {}
for charge in charges_table:
charge_builder = ChargeBuilder()
charge_cols = charge.find_elements_by_tag_name('td')
count = int(charge_cols[0].text.strip())
charge_builder.count = count
charge_desc = charge_cols[1].text
charge_builder.description, charge_builder.statute = (
ScraperUtils.parse_charge_statute(charge_desc))
charge_builder.level = charge_cols[2].text.strip()
charge_builder.degree = charge_cols[3].text.strip()
# plea = charge_cols[4].text.strip() # Plea is not filled out on this portal.
charge_builder.disposition = charge_cols[5].text.strip()
charge_builder.disposition_date = charge_cols[6].text.strip()
Charges[count] = charge_builder.build()
r.charges = list(Charges.values())
# Pleas are not in the 'plea' field, but instead in the dockets.
for plea_element in docket_pleas:
plea_text = plea_element.text.strip()
plea = ScraperUtils.parse_plea_type(plea_text)
plea_date = plea_element.find_element_by_xpath(
'./../td[2]').text.strip()
plea_number = ScraperUtils.parse_plea_case_numbers(
plea_text, list(Charges.keys()))
# If no case number is specified in the plea, then we assume it applies to all charges in the trial.
if len(plea_number) == 0:
for charge in Charges.values():
charge.plea = plea
charge.plea_date = plea_date
else:
# Apply plea to relevant charge count(s).
for count in plea_number:
Charges[count].plea = plea
Charges[count].plea_date = plea_date
r.arresting_officer = None # Can't be found on this portal
r.arresting_officer_badge_number = None # Can't be found on this portal
profile_link = driver.find_element_by_xpath(
"//table[@id='gridParties']/tbody/tr/*[contains(text(), 'DEFENDANT')]/../td[2]/div/a"
).get_attribute('href')
# profile_link = driver.find_element_by_xpath('//*[@id="gridParties"]/tbody/tr[1]/td[2]/div[1]/a').get_attribute(
# 'href')
load_page(profile_link, 'Party Details:', FLAGS.verbose)
r.suffix = None
r.dob = None # This portal has DOB as N/A for every defendent
r.race = driver.find_element_by_xpath(
'//*[@id="fd-table-2"]/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[7]/td[2]'
).text.strip()
r.sex = driver.find_element_by_xpath(
'//*[@id="mainTableContent"]/tbody/tr/td/table/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[6]/td[2]'
).text.strip()
# Navigate to party profile
full_name = driver.find_element_by_xpath(
'//*[@id="mainTableContent"]/tbody/tr/td/table/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[1]/td[2]'
).text.strip()
r.middle_name = None
r.last_name = None
if ',' in full_name:
r.first_name, r.middle_name, r.last_name = ScraperUtils.parse_name(
full_name)
else:
# If there's no comma, it's a corporation name.
r.first_name = Pii.String(full_name)
r.party_id = driver.find_element_by_xpath(
'//*[@id="mainTableContent"]/tbody/tr/td/table/tbody/tr[2]/td[2]/table[2]/tbody/tr/td[2]/table/tbody/tr[8]/td[2]'
).text.strip(
) # PartyID is a field within the portal system to uniquely identify defendants
record = r.build()
ScraperUtils.write_csv(FLAGS.output, record, FLAGS.verbose)
