Use a shared data type for FR citations.

regparser/commands/annual_version.py

@@ -6,6 +6,7 @@
 from regparser.history.annual import find_volume
 from regparser.history.versions import Version
 from regparser.index import dependency, entry
+from regparser.notice.citation import Citation
 from regparser.notice.fake import build as build_fake_notice
 from regparser.tree.gpo_cfr import builder
 
@@ -40,7 +41,7 @@ def create_version_entry_if_needed(volume, cfr_part):
     if version_id not in [c.path[-1] for c in version_dir.sub_entries()]:
         (version_dir / version_id).write(
             Version(version_id, effective=volume.publication_date,
-                    fr_volume=volume.vol_num, fr_page=1))
+                    fr_citation=Citation(volume.vol_num, 1)))
 
 
 @click.command()

regparser/commands/full_issuance.py

@@ -35,7 +35,7 @@ def process_version_if_needed(cfr_title, cfr_part, version_id):
     if deps.is_stale(version_entry):
         notice_xml = notice_entry.read()
         version = Version(version_id, notice_xml.effective,
-                          notice_xml.fr_volume, notice_xml.start_page)
+                          notice_xml.fr_citation)
         version_entry.write(version)
 
 

regparser/commands/proposal_versions.py

@@ -13,8 +13,8 @@ def proposal_versions(doc_number):
         raise dependency.Missing(str(notice), str(notice))
 
     notice = notice.read()
-    version = Version(doc_number, effective=None, fr_volume=notice.fr_volume,
-                      fr_page=notice.start_page)
+    version = Version(doc_number, effective=None,
+                      fr_citation=notice.fr_citation)
 
     for cfr_title, cfr_part in notice.cfr_ref_pairs:
         version_entry = entry.Version(cfr_title, cfr_part, doc_number)

regparser/commands/versions.py

@@ -61,8 +61,7 @@ def write_to_disk(xml, version_entry, delay=None):
     """Serialize a Version instance to disk"""
     effective = xml.effective if delay is None else delay.until
     if effective:
-        version = Version(xml.version_id, effective, xml.fr_volume,
-                          xml.start_page)
+        version = Version(xml.version_id, effective, xml.fr_citation)
         version_entry.write(version)
     else:
         logger.warning("No effective date for this rule: %s. Skipping",

regparser/grammar/delays.py

@@ -8,6 +8,7 @@
 from six.moves import reduce
 
 from regparser.grammar import utils
+from regparser.notice.citation import Citation as NoticeCitation
 
 
 @attr.attrs(slots=True, frozen=True)
@@ -20,12 +21,6 @@ class Delayed(object):
     """Placeholder token"""
 
 
-@attr.attrs(slots=True, frozen=True)
-class Notice(object):
-    volume = attr.attrib()
-    page = attr.attrib()
-
-
 effective_date = (
     utils.Marker("effective") + utils.Marker("date")
 ).setParseAction(EffectiveDate)
@@ -35,7 +30,7 @@ class Notice(object):
     Word(string.digits) +
     utils.Marker('FR') +
     Word(string.digits)
-).setParseAction(lambda m: Notice(int(m[0]), int(m[1])))
+).setParseAction(lambda m: NoticeCitation(m[0], m[1]))
 
 
 delayed = utils.Marker("delayed").setParseAction(Delayed)

regparser/history/delays.py

@@ -3,7 +3,8 @@
 from itertools import dropwhile, takewhile
 
 from regparser.grammar.delays import tokenizer as delay_tokenizer
-from regparser.grammar.delays import Delayed, EffectiveDate, Notice
+from regparser.grammar.delays import Delayed, EffectiveDate
+from regparser.notice.citation import Citation as NoticeCitation
 
 
 class FRDelay(namedtuple('FRDelay', ['volume', 'page', 'delayed_until'])):
@@ -29,7 +30,7 @@ def delays_in_sentence(sent):
 
     frs = list(takewhile(lambda t: not isinstance(t, Delayed), tokens))
     tokens = tokens[len(frs):]
-    frs = [t for t in frs if isinstance(t, Notice)]
+    frs = [t for t in frs if isinstance(t, NoticeCitation)]
 
     if not frs or not tokens:
         return []

regparser/history/versions.py

@@ -2,8 +2,11 @@
 from collections import namedtuple
 from datetime import datetime
 
+from regparser.notice.citation import Citation
+
+
 class Version(namedtuple('Version',
-                         ['identifier', 'effective', 'fr_volume', 'fr_page'])):
+                         ['identifier', 'effective', 'fr_citation'])):
     @property
     def is_final(self):
         return bool(self.effective)
@@ -13,11 +16,10 @@ def is_proposal(self):
         return not self.is_final
 
     def json(self):
-        result = self._asdict()
+        result = {'identifier': self.identifier,
+                  'fr_citation': self.fr_citation.asdict()}
         if self.is_final:
             result['effective'] = self.effective.isoformat()
-        else:
-            del result['effective']
 
         return json.dumps(result)
 
@@ -27,22 +29,20 @@ def from_json(json_str):
         effective = json_dict.get('effective')
         if effective:
             effective = datetime.strptime(effective, '%Y-%m-%d').date()
-        json_dict['effective'] = effective
-        return Version(**json_dict)
+        return Version(json_dict['identifier'], effective,
+                       Citation(**json_dict['fr_citation']))
 
     def __lt__(self, other):
         """Linearizing versions requires knowing not only relevant dates and
         identifiers, but also which versions are from final rules and which
         are just proposals"""
         if self.is_final and other.is_final:
-            left = (self.effective, self.fr_volume, self.fr_page,
-                    self.identifier)
-            right = (other.effective, other.fr_volume, other.fr_page,
-                     other.identifier)
+            left = (self.effective, self.fr_citation, self.identifier)
+            right = (other.effective, other.fr_citation, other.identifier)
             return left < right
         else:   # at least one of the two is a proposal
-            left = (self.fr_volume, self.fr_page, self.identifier)
-            right = (other.fr_volume, other.fr_page, other.identifier)
+            left = (self.fr_citation, self.identifier)
+            right = (other.fr_citation, other.identifier)
             return left < right
 
     @staticmethod

regparser/notice/citation.py

@@ -0,0 +1,13 @@
+import attr
+
+
+@attr.attrs(slots=True, frozen=True)
+class Citation(object):
+    volume = attr.attrib(convert=int)
+    page = attr.attrib(convert=int)
+
+    def formatted(self):
+        return '{0} FR {1}'.format(self.volume, self.page)
+
+    def asdict(self):
+        return attr.asdict(self)

regparser/notice/xml.py

@@ -16,6 +16,7 @@
 from regparser.history.delays import delays_in_sentence
 from regparser.index.http_cache import http_client
 from regparser.notice.amendments.fetch import fetch_amendments
+from regparser.notice.citation import Citation
 from regparser.notice.dates import fetch_dates
 from regparser.tree.xml_parser.xml_wrapper import XMLWrapper
 
@@ -339,7 +340,7 @@ def amendments(self):
 
     @property
     def fr_citation(self):
-        return '{0} FR {1}'.format(self.fr_volume, self.start_page)
+        return Citation(self.fr_volume, self.start_page)
 
     @property
     def title(self):
@@ -397,7 +398,7 @@ def as_dict(self):
                   'cfr_title': cfr_ref.title,
                   'dockets': self.docket_ids,
                   'document_number': self.version_id,
-                  'fr_citation': self.fr_citation,
+                  'fr_citation': self.fr_citation.formatted(),
                   'fr_url': self.fr_html_url,
                   'fr_volume': self.fr_volume,
                   # @todo - SxS depends on this; we should remove soon