def proc_statute_volume(path, options):
mods = etree.parse(path + "/mods.xml")
mods_ns = { "mods": "http://www.loc.gov/mods/v3" }
# Load the THOMAS committee names for this Congress, which is our best
# bet for normalizing committee names in the GPO data.
congress = mods.find( "/mods:extension[2]/mods:congress", mods_ns ).text
utils.fetch_committee_names(congress, options)
logging.warn("Processing %s (Congress %s)" % (path, congress))
package_id = mods.find( "/mods:extension[2]/mods:accessId", mods_ns ).text
for bill in mods.findall( "/mods:relatedItem", mods_ns ):
# MODS files also contain information about:
# ['BACKMATTER', 'FRONTMATTER', 'CONSTAMEND', 'PROCLAMATION', 'REORGPLAN']
if bill.find( "mods:extension/mods:granuleClass", mods_ns ).text not in [ "PUBLICLAW", "PRIVATELAW", "HCONRES", "SCONRES" ]:
continue
# Get the title and source URL (used in error messages).
title_text = bill.find( "mods:titleInfo/mods:title", mods_ns ).text.replace( '""', '"' )
source_url = bill.find( "mods:location/mods:url[@displayLabel='Content Detail']", mods_ns ).text
# Bill number
bill_elements = bill.findall( "mods:extension/mods:bill[@priority='primary']", mods_ns )
if len(bill_elements) == 0:
logging.error("No bill number identified for '%s' (%s)" % (title_text, source_url))
continue
elif len(bill_elements) > 1:
logging.error("Multiple bill numbers identified for '%s'" % title_text)
for be in bill_elements:
logging.error(" -- " + etree.tostring(be).strip())
logging.error(" @ " + source_url)
continue
else:
bill_congress = bill_elements[0].attrib["congress"]
bill_type = bill_elements[0].attrib["type"].lower()
bill_number = bill_elements[0].attrib["number"]
bill_id = "%s%s-%s" % (bill_type, bill_number, bill_congress)
# Title
titles = []
titles.append( {
"title": title_text,
"as": "enacted",
"type": "official",
"is_for_portion": False,
} )
# Subject
descriptor = bill.find( "mods:extension/mods:descriptor", mods_ns )
if descriptor is not None:
subject = descriptor.text
else:
subject = None
# Committees
committees = []
cong_committee = bill.find( "mods:extension/mods:congCommittee", mods_ns )
if cong_committee is not None:
chambers = { "H": "House", "S": "Senate", "J": "Joint" }
committee = chambers[cong_committee.attrib["chamber"]] + " " + cong_committee.find( "mods:name", mods_ns ).text
committee_info = {
"committee": committee,
"activity": [], # XXX
"committee_id": utils.committee_names[committee] if committee in utils.committee_names else None,
}
committees.append( committee_info )
# The 'granuleDate' is the enactment date?
granule_date = bill.find( "mods:extension/mods:granuleDate", mods_ns ).text
sources = [{
"source": "statutes",
"package_id": package_id,
"access_id": bill.find( "mods:extension/mods:accessId", mods_ns ).text,
"source_url": source_url,
"volume": bill.find( "mods:extension/mods:volume", mods_ns ).text,
"page": bill.find( "mods:part[@type='article']/mods:extent[@unit='pages']/mods:start", mods_ns ).text,
"position": bill.find( "mods:extension/mods:pagePosition", mods_ns ).text,
}]
law_elements = bill.findall( "mods:extension/mods:law", mods_ns )
# XXX: If <law> is missing, this assumes it is a concurrent resolution.
# This may be a problem if the code is updated to accept joint resolutions for constitutional amendments.
if ( law_elements is None ) or ( len( law_elements ) != 1 ):
other_chamber = { "HOUSE": "s", "SENATE": "h" }
actions = [{
"type": "vote",
"vote_type": "vote2",
"where": other_chamber[bill.find( "mods:extension/mods:originChamber", mods_ns ).text],
"result": "pass", # XXX
"how": "unknown", # XXX
# "text": "",
"acted_at": granule_date, # XXX
"status": "PASSED:CONCURRENTRES",
"references": [], # XXX
}]
else:
law_congress = law_elements[0].attrib["congress"]
law_number = law_elements[0].attrib["number"]
law_type = ( "private" if ( law_elements[0].attrib["isPrivate"] == "true" ) else "public" )
# Check for typos in the metadata.
if law_congress != bill_congress:
logging.error("Congress mismatch for %s%s: %s or %s? (%s)" % ( bill_type, bill_number, bill_congress, law_congress, source_url ) )
continue
actions = [{
"congress": law_congress,
"number": law_number,
"type": "enacted",
"law": law_type,
"text": "Became %s Law No: %s-%s." % ( law_type.capitalize(), law_congress, law_number ),
"acted_at": granule_date, # XXX
"status": "ENACTED:SIGNED", # XXX: Check for overridden vetoes!
"references": [], # XXX
}]
status, status_date = bill_info.latest_status( actions )
bill_data = {
'bill_id': bill_id,
'bill_type': bill_type,
'number': bill_number,
'congress': bill_congress,
'introduced_at': None, # XXX
'sponsor': None, # XXX
'cosponsors': [], # XXX
'actions': actions, # XXX
'history': bill_info.history_from_actions( actions ),
'status': status,
'status_at': status_date,
'enacted_as': bill_info.slip_law_from( actions ),
'titles': titles,
'official_title': bill_info.current_title_for( titles, "official" ),
'short_title': bill_info.current_title_for( titles, "short" ), # XXX
'popular_title': bill_info.current_title_for( titles, "popular" ), # XXX
'subjects_top_term': subject,
'subjects': [],
'related_bills': [], # XXX: <associatedBills> usually only lists the current bill.
'committees': committees,
'amendments': [], # XXX
'sources': sources,
'updated_at': datetime.datetime.fromtimestamp(time.time()),
}
if not options.get('textversions', False):
bill_info.output_bill( bill_data, options )
# XXX: Can't use bill_versions.fetch_version() because it depends on fdsys.
version_code = "enr"
bill_version_id = "%s%s-%s-%s" % ( bill_type, bill_number, bill_congress, version_code )
bill_version = {
'bill_version_id': bill_version_id,
'version_code': version_code,
'issued_on': status_date,
'urls': { "pdf": bill.find( "mods:location/mods:url[@displayLabel='PDF rendition']", mods_ns ).text },
'sources': sources,
}
utils.write(
json.dumps(bill_version, sort_keys=True, indent=2, default=utils.format_datetime),
bill_versions.output_for_bill_version(bill_version_id)
)
# Process the granule PDF.
# - Hard-link it into the right place to be seen as bill text.
# - Run "pdftotext -layout" to convert it to plain text and save it in the bill text location.
pdf_file = path + "/" + sources[0]["access_id"] + "/document.pdf"
if os.path.exists(pdf_file):
dst_path = fdsys.output_for_bill(bill_data["bill_id"], "text-versions/" + version_code, is_data_dot=False)
if options.get("linkpdf", False):
os.link(pdf_file, dst_path + "/document.pdf") # a good idea
if options.get("extracttext", False):
logging.error("Running pdftotext on %s..." % pdf_file)
if subprocess.call(["pdftotext", "-layout", pdf_file, dst_path + "/document.txt"]) != 0:
raise Exception("pdftotext failed on %s" % pdf_file)
return {'ok': True, 'saved': True}
def proc_statute(path, options):
mods = etree.parse(path + "/mods.xml")
mods_ns = { "mods": "http://www.loc.gov/mods/v3" }
# Load the THOMAS committee names for this Congress, which is our best
# bet for normalizing committee names in the GPO data.
congress = mods.find( "/mods:extension[2]/mods:congress", mods_ns ).text
utils.fetch_committee_names(congress, options)
logging.warn("Processing %s (Congress %s)" % (path, congress))
for bill in mods.findall( "/mods:relatedItem", mods_ns ):
titles = []
titles.append( {
"title": bill.find( "mods:titleInfo/mods:title", mods_ns ).text.replace( '""', '"' ),
"as": "enacted",
"type": "official",
} )
descriptor = bill.find( "mods:extension/mods:descriptor", mods_ns )
if descriptor is not None:
subject = descriptor.text
else:
subject = None
# MODS files also contain information about:
# ['BACKMATTER', 'FRONTMATTER', 'CONSTAMEND', 'PROCLAMATION', 'REORGPLAN']
if bill.find( "mods:extension/mods:granuleClass", mods_ns ).text not in [ "PUBLICLAW", "PRIVATELAW", "HCONRES", "SCONRES" ]:
continue
committees = []
cong_committee = bill.find( "mods:extension/mods:congCommittee", mods_ns )
if cong_committee is not None:
chambers = { "H": "House", "S": "Senate", "J": "Joint" }
committee = chambers[cong_committee.attrib["chamber"]] + " " + cong_committee.find( "mods:name", mods_ns ).text
committee_info = {
"committee": committee,
"activity": [], # XXX
"committee_id": utils.committee_names[committee] if committee in utils.committee_names else None,
}
committees.append( committee_info )
bill_elements = bill.findall( "mods:extension/mods:bill", mods_ns )
if ( bill_elements is None ) or ( len( bill_elements ) != 1 ):
logging.error("Could not get bill data for %s" % repr(titles) )
continue
else:
bill_congress = bill_elements[0].attrib["congress"]
bill_type = bill_elements[0].attrib["type"].lower()
bill_number = bill_elements[0].attrib["number"]
bill_id = "%s%s-%s" % (bill_type, bill_number, bill_congress)
actions = []
law_elements = bill.findall( "mods:extension/mods:law", mods_ns )
# XXX: If <law> is missing, this assumes it is a concurrent resolution.
# This may be a problem if the code is updated to accept joint resolutions for constitutional amendments.
if ( law_elements is None ) or ( len( law_elements ) != 1 ):
other_chamber = { "HOUSE": "s", "SENATE": "h" }
action = {
"type": "vote",
"vote_type": "vote2",
"where": other_chamber[bill.find( "mods:extension/mods:originChamber", mods_ns ).text],
"result": "pass", # XXX
"how": "unknown", # XXX
# "text": "",
"acted_at": bill.find( "mods:extension/mods:granuleDate", mods_ns ).text, # XXX
"status": "PASSED:CONCURRENTRES",
"references": [], # XXX
}
else:
law_congress = law_elements[0].attrib["congress"]
law_number = law_elements[0].attrib["number"]
law_type = ( "private" if ( law_elements[0].attrib["isPrivate"] == "true" ) else "public" )
action = {
"congress": law_congress,
"number": law_number,
"type": "enacted",
"law": law_type,
"text": "Became %s Law No: %s-%s." % ( law_type.capitalize(), law_congress, law_number ),
"acted_at": bill.find( "mods:extension/mods:granuleDate", mods_ns ).text, # XXX
"status": "ENACTED:SIGNED", # XXX: Check for overridden vetoes!
"references": [], # XXX
}
actions.append( action )
# Check for typos in the metadata.
if law_congress != bill_congress:
logging.error("Congress mismatch for %s%s: %s or %s?" % ( bill_type, bill_number, bill_congress, law_congress ) )
continue
status, status_date = bill_info.latest_status( actions )
bill_data = {
'bill_id': bill_id,
'bill_type': bill_type,
'number': bill_number,
'congress': bill_congress,
'introduced_at': None, # XXX
'sponsor': None, # XXX
'cosponsors': [], # XXX
'actions': actions, # XXX
'history': bill_info.history_from_actions( actions ),
'status': status,
'status_at': status_date,
'enacted_as': bill_info.slip_law_from( actions ),
'titles': titles,
'official_title': bill_info.current_title_for( titles, "official" ),
'short_title': bill_info.current_title_for( titles, "short" ), # XXX
'popular_title': bill_info.current_title_for( titles, "popular" ), # XXX
# 'summary': summary,
'subjects_top_term': subject,
'subjects': [],
'related_bills': [], # XXX: <associatedBills> usually only lists the current bill.
'committees': committees,
'amendments': [], # XXX
'updated_at': datetime.datetime.fromtimestamp(time.time()),
}
bill_info.output_bill( bill_data, options )
# XXX: Can't use bill_versions.fetch_version() because it depends on fdsys.
version_code = "enr"
bill_version_id = "%s%s-%s-%s" % ( bill_type, bill_number, bill_congress, version_code )
bill_version = {
'bill_version_id': bill_version_id,
'version_code': version_code,
'issued_on': status_date,
'urls': { "pdf": bill.find( "mods:location/mods:url[@displayLabel='PDF rendition']", mods_ns ).text },
}
import json, bill_versions
utils.write(
json.dumps(bill_version, sort_keys=True, indent=2, default=utils.format_datetime),
bill_versions.output_for_bill_version(bill_version_id)
)
return {'ok': True, 'saved': True}