def get_record_num(SICNo, s_d, s_m, s_y, e_d, e_m, e_y):
sic_str = 'sic={sic}&'
s_d_str = 'startday={s_d}&'
s_m_str = 'startmonth={s_m}&'
s_y_str = 'startyear={s_y}&'
e_d_str = 'endday={e_d}&'
e_m_str = 'endmonth={e_m}&'
e_y_str = 'endyear={e_y}&'
search_url = 'https://www.osha.gov/pls/imis/AccidentSearch.search?'
search_url = search_url + 'p_logger=1&acc_description=fall&acc_Abstract=&acc_keyword=&'
search_url = search_url + sic_str.format(sic=SICNo)
search_url = search_url + 'naics=&Office=All&officetype=All&'
search_url = search_url + e_m_str.format(e_m=e_m)
search_url = search_url + e_d_str.format(e_d=e_d)
search_url = search_url + e_y_str.format(e_y=e_y)
search_url = search_url + s_m_str.format(s_m=s_m)
search_url = search_url + s_d_str.format(s_d=s_d)
search_url = search_url + s_y_str.format(s_y=s_y)
search_url = search_url + 'InspNr='
brw = BrowserShadow()
res = brw.open_url(search_url)
page_content = res.read()
page_soup = BeautifulSoup(page_content, "html.parser")
count_str_cell = page_soup.select('#maincontain')[0].select('.row-fluid')[0].select('.row-fluid')[0].select('.span3')[0].select('.text-right')
count_str = count_str_cell[0].get_text();
record_num = int(count_str[count_str.find('of') + 2 : len(count_str)])
#print(record_num)
#print(search_url)
return record_num
def parse_page_obtian_event(page_soup):
#event_content = page_soup.select('#maintain')[0].select('.row-fluid')[0].select('.table-responsive')[0].select('tbody')[0]
event_content = page_soup.select('#maincontain')[0].select('.row-fluid')[0].select('.table-responsive')[1].select('table')[0].find_all('tr')
for index in range(1 , len(event_content)):
item_list = event_content[index].find_all('td')
SummaryNr = item_list[2].find('a').get_text()
EventDate = item_list[3].get_text()
ReportID = item_list[4].get_text()
Fat = item_list[5].get_text()
if Fat.find('X') > -1 :
Fat = 1
else:
Fat = 0
SIC = item_list[6].find('a').get_text()
EventDesc = item_list[7].get_text()
#print(SummaryNr + EventData + ReportID + str(Fat) + SIC)
accident_url_str = item_list[2].find('a')['href']
accident_url_str = 'https://www.osha.gov/pls/imis/' + accident_url_str
#print(SummaryNr)
brw_d = BrowserShadow()
res_d = brw_d.open_url(accident_url_str)
page_content_d = res_d.read()
page_soup_d = BeautifulSoup(page_content_d, "html.parser")
# parse the detail information about the event
abstract_info = {'SummaryNr':SummaryNr , 'EventDate':EventDate, 'ReportID':ReportID, 'Fat':str(Fat), 'SIC':SIC, 'EventDesc':EventDesc}
parse_accident_details(page_soup_d , abstract_info)
def get_similar_matrix(data_dic):
output_f = open('ocup_comp_result.txt', 'w+')
comp_element = list(data_dic.keys())
with open('ocup_comp_records.csv', 'w', newline='') as comp_r_csv:
spamwriter = csv.writer(comp_r_csv,dialect='excel')
spamwriter.writerow(comp_element)
url = "https://www.metamind.io/language/relatedness/test?{test_p_str}"
brw = BrowserShadow()
similar_matrix_dic = {}
for each_item_r in comp_element:
r_list = []
for each_item_c in comp_element:
test_data = {"text_2":each_item_r, "text_1":each_item_c}
test_data_str = urlencode(test_data)
comp_url = url.format(test_p_str = test_data_str)
res = brw.open_url(comp_url)
page_content = res.read()
compar_result = page_content.decode()
result = eval(compar_result)
r_list.append(result["score"])
r_str = "row, col, comp_r, %s, %s, %s" % (each_item_r, each_item_c, result["score"])
output_f.write(r_str + "\n")
print(r_str)
similar_matrix_dic[each_item_r] = r_list
spamwriter.writerow(r_list)
output_f.close()
return similar_matrix_dic
def get_record_list(SICNo, s_d, s_m, s_y, e_d, e_m, e_y):
record_num = get_record_num(SICNo, s_d, s_m, s_y, e_d, e_m, e_y)
if record_num <= 0:
print("No Eligible Record has been retrieved!")
return
p_finish = 0
# if SICNo == 16:
# p_finish = 825
# else:
# p_finish = 0
p_show = 100
if record_num < p_show:
p_show = record_num
checked_num = 0
while 1:
if (p_finish + p_show) > record_num:
p_show = record_num - p_finish
if p_show == 0:
break
sic_str = 'sic={sic}&'
p_finish_str = 'p_finish={p_finish}&'
p_show_str = 'p_show={p_show}'
search_url = 'https://www.osha.gov/pls/imis/accidentsearch.search?'
search_url = search_url + sic_str.format(sic=SICNo)
search_url = search_url + 'sicgroup=&naics=&acc_description=fall&acc_abstract=&acc_keyword=&inspnr=&fatal=&officetype=All&office=All&'
search_url = search_url + 'startmonth=07&startday=24&startyear=2015&endmonth=07&endday=23&endyear=1984&keyword_list=&p_start=&'
search_url = search_url + p_finish_str.format(p_finish=p_finish)
search_url = search_url + 'p_sort=&p_desc=DESC&p_direction=Next&'
search_url = search_url + p_show_str.format(p_show=p_show)
#print(search_url)
brw = BrowserShadow()
res = brw.open_url(search_url)
page_content = res.read()
page_soup = BeautifulSoup(page_content, "html.parser")
# prase the specified record
"""collect the records into mysql database"""
parse_page_obtian_event(page_soup)
checked_num = checked_num + p_show
#print(checked_num)
if checked_num == record_num:
break
else:
p_finish = p_finish + p_show
def compare_2_sentences(text_1, text_2):
url = "https://www.metamind.io/language/relatedness/test?{text}"
data_dic = {"text_1":text_1, "text_2":text_2}
text_str = urlencode(data_dic)
url = url.format(text = text_str)
brw = BrowserShadow()
res = brw.open_url(url)
page_content = res.read()
compar_result = page_content.decode()
result = eval(compar_result)
return result["score"]
def update_employee_detail_info_from_OSHA():
user = '******'
pwd = '123456'
host = '127.0.0.1'
db = 'reported_fall_event'
cnx = mysql.connector.connect(user=user, password=pwd, host=host, database=db)
cursor = cnx.cursor()
query_sql = "select SummaryNr from case_employees order by ID"
cursor.execute(query_sql)
employee_list = cursor.fetchall()
brw = BrowserShadow()
need_update = 0
for each_employee in employee_list:
SummaryNr_value = each_employee[0].strip()
if SummaryNr_value != "202315776" and need_update == 0:
continue
else:
need_update = 1
print("SummaryNr :" , SummaryNr_value)
employee_detail_info_url = "https://www.osha.gov/pls/imis/accidentsearch.accident_detail?id={SummaryNr}"
employee_detail_info_url = employee_detail_info_url.format(SummaryNr = SummaryNr_value)
res = brw.open_url(employee_detail_info_url)
page_content = res.read()
page_soup = BeautifulSoup(page_content, "html.parser")
event_content_details = page_soup.find_all('tr')
keyword_position = -1
for index in range(0, len(event_content_details)):
if event_content_details[index].get_text().find('Keywords') > -1 :
keyword_position = index
# do not process if there is no keyword
if keyword_position == -1:
continue
proj_type_has = 1
if event_content_details[keyword_position + 1].find_all('th')[0].get_text().find('End Use') < 0:
proj_type_has = 0
if proj_type_has == 0:
index_employee_start = keyword_position + 2
else:
index_employee_start = keyword_position + 4
for index in range(index_employee_start, len(event_content_details)):
current_employee_eid = event_content_details[index].find_all('td')[0].get_text().strip()
current_employee_href = event_content_details[index].find_all('td')[0].find('a')['href']
current_employee_url_str = 'https://www.osha.gov/pls/imis/' + current_employee_href
print(current_employee_url_str)
res_employee_info = brw.open_url(current_employee_url_str)
page_content_employee = res_employee_info.read()
page_soup_employee = BeautifulSoup(page_content_employee, "html.parser")
employee_content_details = page_soup_employee.find_all('tr')
inspection_Nr_position = -1
for index in range(0, len(employee_content_details)):
if employee_content_details[index].get_text().find('Inspection') > -1 :
inspection_Nr_position = index
break
if inspection_Nr_position == -1:
continue
TaskAssigned = employee_content_details[inspection_Nr_position + 13].find_all('td')[1].get_text()
HumanFactor = employee_content_details[inspection_Nr_position + 10].find_all('td')[1].get_text()
EnvironmentFactor = employee_content_details[inspection_Nr_position + 9].find_all('td')[1].get_text()
EventType = employee_content_details[inspection_Nr_position + 8].find_all('td')[1].get_text()
SourceInjury = employee_content_details[inspection_Nr_position + 7].find_all('td')[1].get_text()
PartBody = employee_content_details[inspection_Nr_position + 6].find_all('td')[1].get_text()
update_employee_injury_info_sql = ("update case_employees set "
"TaskAssigned = %s, HumanFactor = %s, EnvironmentFactor = %s, EventType = %s, SourceInjury = %s, PartBody = %s"
"where EID = %s and SummaryNr = %s")
update_values = (TaskAssigned, HumanFactor, EnvironmentFactor, EventType, SourceInjury, PartBody, current_employee_eid, SummaryNr_value)
cursor.execute(update_employee_injury_info_sql, update_values)
cnx.commit()
print(TaskAssigned, '--', HumanFactor, '--', EnvironmentFactor, '--', EventType, '--', SourceInjury, '--', PartBody)
print("Update has done!")
return 1
