Module: GovCodes::Dafecd::RiSdi::Title

Defined in:
lib/gov_codes/dafecd/ri_sdi/title.rb

Overview

Conservative title normalization for RI/SDI records.

These helpers mirror the AFSC SpecialtyParser's title handling exactly (sanitize -> title-case verbatim, flag probable pdf glue) but live in their own module so the RI/SDI extraction shares NONE of the AFSC ladder pipeline's mutable state — the already-shipped enlisted/officer artifacts cannot be perturbed by anything done here.

As in the AFSC pipeline, no de-gluing is attempted: a run-together title like "ENLISTEDAIDE" becomes "Enlistedaide" and is flagged via #glued? for the deferred, governed de-glue pass. Clean titles are authored out of band.

Constant Summary collapse

DECORATIVE = 
Patterns::DECORATIVE
UNICODE_DASHES = 
Patterns::UNICODE_DASHES

Class Method Summary collapse

Class Method Details

.glued?(name) ⇒ Boolean

Heuristic proxy for pdf-reader glue: an unusually long single token (excluding preserved acronyms). Reported, not relied upon.

Returns:

  • (Boolean) 


66
67
68
69
70
71
72
 
# File 'lib/gov_codes/dafecd/ri_sdi/title.rb', line 66

def glued?(name)
  return false if name.nil?
  name.split(/\s+/).any? do |token|
    bare = token.delete("()/-")
    bare.length >= 12 && bare.match?(/\A[A-Za-z]+\z/)
  end
end

.real?(title) ⇒ Boolean

A real title starts with a capital letter, a digit, or an opening paren — never the lowercase word that begins a wrapped prose sentence (the "SDI 8P000, completion of ..." false positive).

Returns:

  • (Boolean) 


59
60
61
62
 
# File 'lib/gov_codes/dafecd/ri_sdi/title.rb', line 59

def real?(title)
  t = title.to_s.strip
  !t.empty? && t.match?(/\A[A-Z0-9(]/)
end

.sanitize(line) ⇒ Object

Remove decorative symbol glyphs, normalize unicode dashes, and collapse whitespace runs so a title line reduces to its plain-text content.



27
28
29
 
# File 'lib/gov_codes/dafecd/ri_sdi/title.rb', line 27

def sanitize(line)
  line.gsub(DECORATIVE, "").gsub(UNICODE_DASHES, "-").gsub(/\s+/, " ").strip
end

.title_line?(stripped) ⇒ Boolean

A title line is short, starts with a capital / paren / digit, and is not a prose sentence. Accepts ALL-CAPS and Title-Case forms.

Returns:

  • (Boolean) 


33
34
35
36
37
38
39
 
# File 'lib/gov_codes/dafecd/ri_sdi/title.rb', line 33

def title_line?(stripped)
  return false if stripped.length > 90
  return false unless stripped.match?(/\A[A-Z0-9(]/)
  return false if stripped.split(/\s+/).size > 12
  return false if stripped.match?(/[a-z]\.\s+[A-Z]/) # mid-line sentence break = prose
  stripped.match?(/[A-Za-z]/)
end

.titleize(raw) ⇒ Object

Conservative title-case: capitalize each alphabetic word, preserving parenthesized acronyms and all-caps-with-digit tokens.



43
44
45
 
# File 'lib/gov_codes/dafecd/ri_sdi/title.rb', line 43

def titleize(raw)
  raw.split(/\s+/).map { |word| titleize_word(word) }.join(" ")
end

.titleize_word(word) ⇒ Object 



47
48
49
50
51
52
53
54
 
# File 'lib/gov_codes/dafecd/ri_sdi/title.rb', line 47

def titleize_word(word)
  return word if word.match?(/\A\([A-Z0-9&\/.-]{2,}\)\z/)
  return word if word.match?(/\A[A-Z]*\d[A-Z0-9]*\z/) && word.match?(/[A-Z0-9]/)

  word.split(/([\/-])/).map { |seg|
    seg.match?(/[A-Za-z]/) ? seg.capitalize : seg
  }.join
end