def ecfr_notice(title, cfr_part, notice, applies_to, act_title,
act_section, with_version=False, without_notice=False):
""" Generate RegML for a single notice from eCFR XML. """
# Get the notice the new one applies to
with open(find_file(os.path.join(cfr_part, applies_to)), 'r') as f:
reg_xml = f.read()
parser = etree.XMLParser(huge_tree=True)
xml_tree = etree.fromstring(reg_xml, parser)
doc_number = xml_tree.find('.//{eregs}documentNumber').text
# Validate the file relative to schema
validator = get_validator(xml_tree)
# Get the ecfr builder
builder = Builder(cfr_title=title,
cfr_part=cfr_part,
doc_number=doc_number,
checkpointer=None,
writer_type='XML')
# Fetch the notices from the FR API and find the notice we're
# looking for
builder.fetch_notices_json()
print([n['document_number'] for n in builder.notices_json])
notice_json = next((n for n in builder.notices_json
if n['document_number'] == notice))
# Build the notice
notice = builder.build_single_notice(notice_json)[0]
if 'changes' not in notice:
print('There are no changes in this notice to apply.')
return
# We've successfully fetched and parsed the new notice.
# Build a the reg tree and layers for the notice it applies to.
old_tree = build_reg_tree(xml_tree)
# Build the new reg tree from the old_tree + notice changes
last_version = doc_number
version = notice['document_number']
merged_changes = builder.merge_changes(version, notice['changes'])
reg_tree = compile_regulation(old_tree, merged_changes)
layer_cache = LayerCacheAggregator()
layers = builder.generate_layers(reg_tree,
[act_title, act_section],
layer_cache)
# Write the notice file
if not without_notice:
builder.write_notice(version,
old_tree=old_tree,
reg_tree=reg_tree,
layers=layers,
last_version=last_version)
# Write the regulation file for the new notice
if with_version:
builder.write_regulation(new_tree, layers=layers)
def parse_regulation(args):
""" Run the parser on the specified command-line arguments. Broken out into
separate function to assist in profiling.
"""
with codecs.open(args.filename, 'r', 'utf-8') as f:
reg = f.read()
doc_number = args.notice
act_title_and_section = [args.act_title, args.act_section]
# First, the regulation tree
reg_tree = Builder.reg_tree(reg)
builder = Builder(cfr_title=args.title,
cfr_part=reg_tree.label_id(),
doc_number=doc_number)
builder.write_notices()
# Always do at least the first reg
logger.info("Version %s", doc_number)
builder.write_regulation(reg_tree)
layer_cache = LayerCacheAggregator()
builder.gen_and_write_layers(reg_tree, act_title_and_section, layer_cache)
layer_cache.replace_using(reg_tree)
if args.generate_diffs:
generate_diffs(doc_number, reg_tree, act_title_and_section, builder,
layer_cache)
def parse_regulation(args):
""" Run the parser on the specified command-line arguments. Broken out into
separate function to assist in profiling.
"""
with codecs.open(args.filename, 'r', 'utf-8') as f:
reg = f.read()
doc_number = args.notice
act_title_and_section = [args.act_title, args.act_section]
# First, the regulation tree
reg_tree = Builder.reg_tree(reg)
builder = Builder(cfr_title=args.title,
cfr_part=reg_tree.label_id(),
doc_number=doc_number)
builder.write_notices()
# Always do at least the first reg
logger.info("Version %s", doc_number)
builder.write_regulation(reg_tree)
layer_cache = LayerCacheAggregator()
builder.gen_and_write_layers(reg_tree, act_title_and_section, layer_cache)
layer_cache.replace_using(reg_tree)
if args.generate_diffs:
generate_diffs(doc_number, reg_tree, act_title_and_section, builder, layer_cache)
def build_by_notice(filename,
title,
act_title,
act_section,
notice_doc_numbers,
doc_number=None,
checkpoint=None):
with codecs.open(filename, 'r', 'utf-8') as f:
reg = f.read()
file_digest = hashlib.sha256(reg.encode('utf-8')).hexdigest()
if checkpoint:
checkpointer = Checkpointer(checkpoint)
else:
checkpointer = NullCheckpointer()
# build the initial tree
reg_tree = checkpointer.checkpoint("init-tree-" + file_digest,
lambda: Builder.reg_tree(reg))
title_part = reg_tree.label_id()
if doc_number is None:
doc_number = Builder.determine_doc_number(reg, title, title_part)
checkpointer.suffix = ":".join(
["", title_part, str(args.title), doc_number])
# create the builder
builder = Builder(cfr_title=title,
cfr_part=title_part,
doc_number=doc_number,
checkpointer=checkpointer)
builder.fetch_notices_json()
for notice in notice_doc_numbers:
builder.build_notice_from_doc_number(notice)
builder.write_regulation(reg_tree)
layer_cache = LayerCacheAggregator()
act_title_and_section = [act_title, act_section]
builder.gen_and_write_layers(reg_tree, act_title_and_section, layer_cache)
layer_cache.replace_using(reg_tree)
if args.generate_diffs:
generate_diffs(reg_tree, act_title_and_section, builder, layer_cache)
def build_by_notice(filename, title, act_title, act_section,
notice_doc_numbers, doc_number=None, checkpoint=None):
with codecs.open(filename, 'r', 'utf-8') as f:
reg = f.read()
file_digest = hashlib.sha256(reg.encode('utf-8')).hexdigest()
if checkpoint:
checkpointer = Checkpointer(checkpoint)
else:
checkpointer = NullCheckpointer()
# build the initial tree
reg_tree = checkpointer.checkpoint(
"init-tree-" + file_digest,
lambda: Builder.reg_tree(reg))
title_part = reg_tree.label_id()
if doc_number is None:
doc_number = Builder.determine_doc_number(reg, title, title_part)
checkpointer.suffix = ":".join(
["", title_part, str(args.title), doc_number])
# create the builder
builder = Builder(cfr_title=title,
cfr_part=title_part,
doc_number=doc_number,
checkpointer=checkpointer)
builder.fetch_notices_json()
for notice in notice_doc_numbers:
builder.build_notice_from_doc_number(notice)
builder.write_regulation(reg_tree)
layer_cache = LayerCacheAggregator()
act_title_and_section = [act_title, act_section]
builder.gen_and_write_layers(reg_tree, act_title_and_section, layer_cache)
layer_cache.replace_using(reg_tree)
if args.generate_diffs:
generate_diffs(reg_tree, act_title_and_section, builder, layer_cache)
def parse_regulation(args):
""" Run the parser on the specified command-line arguments. Broken out into
separate function to assist in profiling.
"""
with codecs.open(args.filename, 'r', 'utf-8') as f:
reg = f.read()
file_digest = hashlib.sha256(reg.encode('utf-8')).hexdigest()
act_title_and_section = [args.act_title, args.act_section]
if args.checkpoint:
checkpointer = Checkpointer(args.checkpoint)
else:
checkpointer = NullCheckpointer()
# First, the regulation tree
reg_tree = checkpointer.checkpoint(
"init-tree-" + file_digest,
lambda: Builder.reg_tree(reg))
title_part = reg_tree.label_id()
doc_number = checkpointer.checkpoint(
"doc-number-" + file_digest,
lambda: Builder.determine_doc_number(reg, args.title, title_part))
if not doc_number:
raise ValueError("Could not determine document number")
checkpointer.suffix = ":".join(
["", title_part, str(args.title), doc_number])
# Run Builder
builder = Builder(cfr_title=args.title,
cfr_part=title_part,
doc_number=doc_number,
checkpointer=checkpointer)
builder.write_notices()
# Always do at least the first reg
logger.info("Version %s", doc_number)
builder.write_regulation(reg_tree)
layer_cache = LayerCacheAggregator()
builder.gen_and_write_layers(reg_tree, act_title_and_section, layer_cache)
layer_cache.replace_using(reg_tree)
if args.generate_diffs:
generate_diffs(doc_number, reg_tree, act_title_and_section,
builder, layer_cache, checkpointer)
reg_tree = Builder.reg_tree(reg)
builder = Builder(cfr_title=int(sys.argv[2]),
cfr_part=reg_tree.label_id(),
doc_number=doc_number)
# Didn't include the provided version
if not any(n['document_number'] == doc_number for n in builder.notices):
print "Could not find notice_doc_#, %s" % doc_number
exit()
builder.write_notices()
# Always do at least the first reg
logger.info("Version %s", doc_number)
builder.write_regulation(reg_tree)
layer_cache = LayerCacheAggregator()
builder.gen_and_write_layers(reg_tree, sys.argv[4:6], layer_cache)
layer_cache.replace_using(reg_tree)
if len(sys.argv) < 7 or sys.argv[6].lower() == 'true':
all_versions = {doc_number: reg_tree}
for last_notice, old, new_tree, notices in builder.revision_generator(
reg_tree):
version = last_notice['document_number']
logger.info("Version %s", version)
all_versions[version] = new_tree
builder.doc_number = version
builder.write_regulation(new_tree)
layer_cache.invalidate_by_notice(last_notice)
builder.gen_and_write_layers(new_tree, sys.argv[4:6], layer_cache,
notices)
lambda: Builder.determine_doc_number(reg, args.title, title_part))
if not doc_number:
raise ValueError("Could not determine document number")
checkpointer.suffix = ":".join(
["", title_part, str(args.title), doc_number])
# Run Builder
builder = Builder(cfr_title=args.title,
cfr_part=title_part,
doc_number=doc_number,
checkpointer=checkpointer)
builder.write_notices()
# Always do at least the first reg
logger.info("Version %s", doc_number)
builder.write_regulation(reg_tree)
layer_cache = LayerCacheAggregator()
builder.gen_and_write_layers(reg_tree, act_title_and_section, layer_cache)
layer_cache.replace_using(reg_tree)
# this used to assume implicitly that if gen-diffs was not specified it was
# True; changed it to explicit check
if args.generate_diffs:
all_versions = {doc_number: reg_tree}
for last_notice, old, new_tree, notices in builder.revision_generator(
reg_tree):
version = last_notice['document_number']
logger.info("Version %s", version)
all_versions[version] = new_tree