Journal entries retain their original wording and hashes. These display-only

from pathlib import Path
import json,re
root=Path(__file__).resolve().parents[1]
def read(n):return json.loads((root/'model'/f'{n}.json').read_text())
# Journal entries retain their original wording and hashes. These display-only
# substitutions repair compressed drafting in the readable continuation.
PROSE_WORDS = {
 'cumulativehead': 'cumulative head', 'separatecross': 'separate cross',
 'roundedlifespans': 'rounded lifespans', 'theaccepted': 'the accepted',
 'anyfixed': 'any fixed', 'theirprefixes': 'their prefixes',
 'inverseprefix': 'inverse prefix', 'lowerbranch': 'lower branch',
 'fullrefinement': 'full refinement', 'primaryheldhead': 'primary held head',
 'theirhead': 'their head', 'scalarforms': 'scalar forms',
 'primaryspines': 'primary spines', 'connectedplacement': 'connected placement',
 'biographycorrespondence': 'biography correspondence',
 'localpartition': 'local partition', 'translatededges': 'translated edges',
 'alternatecuts': 'alternate cuts', 'affineconversion': 'affine conversion',
 'historicalredaction': 'historical redaction',
 'statisticalrarity': 'statistical rarity', 'canonicalsource': 'canonical source',
 'mainexplanation': 'main explanation', 'onversion': 'on version',
 'existingidentity': 'existing identity', 'Sourcefiles': 'Source files',
 'sourcefiles': 'source files', 'sourcehash': 'source hash',
 'sequentialjournal': 'sequential journal', 'softwarecheck': 'software check',
 'thethree': 'the three', 'upperprefix': 'upper prefix',
 'sourceprefix': 'source prefix', 'theprefix': 'the prefix',
 'andknown': 'and known', 'theknown': 'the known', 'secondarybranch': 'secondary branch',
 'thefixed': 'the fixed', 'theidentical': 'the identical',
 'thedeclarednative': 'the declared native', 'fornative': 'for native',
 'itshead': 'its head', 'allgive': 'all give',
 'notneedanynew': 'not need any new', 'theall': 'the all',
 'itslocal': 'its local', 'lowerlegs': 'lower legs',
 'requiredreversal': 'required reversal', 'localdefects': 'local defects',
 'thedigit': 'the digit', 'particularadditive': 'particular additive',
 'thelower': 'the lower', 'Everyplace': 'Every place',
 'withoutcarry': 'without carry', 'andthethree': 'and the three',
 'theoriginal': 'the original', 'theexplicit': 'the explicit',
 'declarednative': 'declared native', 'andunchangedlifespans': 'and unchanged lifespans',
 'declaredYear': 'declared Year', 'fullcontrollers': 'full controllers',
 'mainreader': 'main reader', 'thecurrentreader': 'the current reader',
 'sourcebindings': 'source bindings', 'independentnumerical': 'independent numerical',
 'headpairs': 'head pairs', 'finalresearch': 'final research',
 'theconnected': 'the connected', 'sourcecontrolled': 'source controlled',
 'unchangedlifespans': 'unchanged lifespans', 'andunchanged': 'and unchanged',
 'lowerleg': 'lower leg', 'upperleg': 'upper leg',
 'headclass': 'head class', 'fullword': 'full word',
 'inversehead': 'inverse head', 'Executeaccepted': 'Execute accepted',
 'Fallhead': 'Fall head', 'itslocalsegment': 'its local segment',
 'thefull': 'the full', 'Noahrefinement': 'Noah refinement',
 'Thenewtables': 'The new tables', 'andformulas': 'and formulas',
 'nextsubheading': 'next subheading', 'thefoot': 'the foot',
 'layoutfix': 'layout fix', 'finalvisualreview': 'final visual review',
 'onpage': 'on page', 'bothfollowing': 'both following',
 'pagebreak': 'page break', 'establishedoperations': 'established operations', 'wholepaths': 'whole paths', 'locatedsourceinputs': 'located source inputs', 'common-grammar': 'common grammar', 'globalminimality': 'global minimality', 'post-Flood910': 'post-Flood 910',
}

def _plain_prose(s):
 # Split ordinary compressed CamelCase, including acronym-to-word boundaries.
 s=re.sub(r'(?<=[a-z])(?=[A-Z])|(?<=[A-Z])(?=[A-Z][a-z])',' ',s)
 for old,new in sorted(PROSE_WORDS.items(), key=lambda item: -len(item[0])):
  s=re.sub(r'(?<![A-Za-z])'+re.escape(old)+r'(?![A-Za-z])',new,s)
 s=re.sub(r'\bpre Noah\b','pre-Noah',s)
 s=re.sub(r'\b(MT|LXX|SP)(?=F(?:[/\s]|$))',r'\1 ',s)
 s=re.sub(r'(?<=\S)(?=§)',' ',s)

 # Protect identifiers, section numbers, dates and literal file paths before
 # separating prose from adjacent numbers. In particular, C1483, File52a,
 # SHA256, python3 and §4A.3 must retain their identities.
 protected=[]
 def hold(match):
  protected.append(match.group())
  return chr(0xE000+len(protected)-1)
 token_pattern=(r'§\d+[A-Za-z]?(?:\.\d+[A-Za-z]?)*'
                r'|\bC\d+\b|\bFile_?\d+[a-z]?\b'
                r'|\bSHA(?:1|256|512)\b|\bpython\d+\b'
                r'|\b\d{4}-\d{2}-\d{2}\b'
                r'|\b(?:[A-Za-z0-9_.-]+/)*[A-Za-z0-9_.-]+\.(?:json|md|py|pdf|zip)\b')
 s=re.sub(token_pattern,hold,s)
 s=re.sub(r'(?<=[A-Za-z])(?=\d)|(?<=\d)(?=[A-Za-z])',' ',s)
 s=re.sub(r'([,;:])(?=[A-Za-z0-9+−])',r'\1 ',s)
 s=re.sub(r'(?<=[A-Za-z])(?=[+−]\d)',' ',s)
 s=re.sub(r'(?<=[A-Za-z0-9])(?=[\ue000-\uf8ff])',' ',s)
 s=re.sub(r'(?<=[\ue000-\uf8ff])(?=[A-Za-z])',' ',s)
 for i,token in enumerate(protected):
  s=s.replace(chr(0xE000+i),token)
 return s

def prose(s):
 # Inline code and inline/display math are opaque: never rewrite commands,
 # identifiers or equations while polishing neighboring human prose.
 chunks=re.split(r'(`+[^`]*`+|\\\([^\n]*?\\\)|\\\[[^\n]*?\\\])',s)
 return ''.join(chunk if i%2 else _plain_prose(chunk)
                for i,chunk in enumerate(chunks))