diff --git a/maps/bgnpcgn-kur-Arab-Latn-2007.yaml b/maps/bgnpcgn-kur-Arab-Latn-2007.yaml
new file mode 100644
index 00000000..b1348752
--- /dev/null
+++ b/maps/bgnpcgn-kur-Arab-Latn-2007.yaml
@@ -0,0 +1,172 @@
+---
+authority_id: bgnpcgn
+id: 2007
+language: kur
+source_script: Arab
+destination_script: Latn
+name: ROMANIZATION OF KURDISH -- BGN/PCGN 2007
+url: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/693727/ROMANIZATION_OF_KURDISH.pdf
+creation_date: 2007
+confirmation date: 2017-12
+description: |
+  The tabulation below is applicable to the Kurdish language as a
+  whole. It is based for the most part on the Hawar Roman alphabet used
+  in the Library of Congress Standard Kurdish Orthography Table, but it
+  also incorporates certain non-Hawar elements found in A Kurdish-English
+  Dictionary (Taufiq Wahby & C J Edmonds, OUP, 1966). The tabulation
+  covers both major varieties of the Kurdish language: Kurmanji and
+  Sorani. Kurmanji is spoken principally in Turkey and in Iraq north of
+  the Great Zab River (Dahūk/Dihok Governorate). It is generally written
+  in Roman script, and usually employs the Roman orthography. Sorani is
+  spoken principally in Iraq south of the Great Zab river (Arbīl/Hewlêr
+  and As Sulaymānīyah/Slêmanî governorates). It is generally written in
+  Perso-Arabic script, and usually employs the Perso-Arabic script
+  orthography.
+
+  Kurdish forms of geographical names in Turkey will usually be found
+  in Roman script, and so no romanization process will be required. The
+  digraph options for consonant letters '\u0686', '\u0634', and '\u063A'
+  will not be encountered for such names. In Iraq, Syria, and Iran,
+  Kurdish will usually be encountered in Perso-Arabic script, in which
+  case it should be romanized into the corresponding Roman script form.
+  Kurdish geographical names for places and features outside Turkey,
+  found in Roman script form, should, where necessary and if possible, be
+  tailored to fit the orthography of the Romanization shown below and
+  should employ the digraph options for consonant letters '\u0686',
+  '\u0634', and '\u063A'.
+
+notes:
+
+  - In pure Kurdish words hamza is borne by yā’ ( ئ ) and occurs only
+    before initial vowels; it is not romanized. Medial and final hamza in
+    Arabic borrowings are romanized by ’ (apostrophe – Unicode encoding
+    2019).
+
+  - The letters ث ذ ص ض ط ظ do not occur in pure Kurdish words. In Arabic
+    borrowings some writers retain these letters, others substitute س ز س ز
+    ت ز respectively. Only the letters ط ض and ص are catered for in the
+    Library of Congress tabulation, as reflected in lines 16-18 of the
+    above Consonant table. Words of obvious Arabic origin occurring in a
+    Kurdish toponymic environment will be treated as Kurdish rather than
+    Arabic, as will words of other non-Kurdish origins.
+
+  - The digraph options appearing in rows 6, 15 and 20 of the consonants
+    table should be used for Kurdish geographical names in Iraq, Iran, and
+    Syria. The single character options should be used for Kurdish
+    geographical names in Turkey.
+
+  - ڨ is used to represent v in foreign words. Some southern Kurdish
+    writers use it to represent the v in borrowings from northern Kurdish
+    dialects. و is pronounced as a v in the north and as a w elsewhere.
+
+  - Hā’ can be used as a vowel or a consonant. The initial (ه) and medial
+    (forms are used for the consonant h, Consonant table, row 31, while the
+    final (ه) and independent (forms are used to represent the vowel e,
+    Vowel table, row 1. Therefore, when used as a consonant, the final and
+    independent forms of hā’ will be seen as ‘ه’ instead of ‘and ‘ه’,
+    respectively. For example, مهه meh, (“month”). When used as ‘e’, the
+    hā’ behaves like the letters alif (ا) , wāw, dāl (د) , and rā (ر) , in
+    that it never joins to the following letter (i.e., it has no medial
+    form). Consequently, the following letter will display the initial
+    form, e.g. هەولێر Hewlêr (unless there is only one following letter, in
+    which case it will be written in the independent form, e.g. ماوەت
+    Mawet). As with other vowels (see special rules 2 and 3), initial e is
+    preceded by the kursî hamza, yielding initial ئه , e.g. ئهني enî
+    “forehead”.
+
+  - In pure Kurdish words, the vowel ى is always long î, e.g. كانى ماسێ
+    Kanî Masê. When it represents îzafe, it is also romanized î and joined
+    by means of a hyphen to its preceding word e.g. پارێزگاى دهۆك Parêzga-î
+    Dihok.
+
+  - |
+    An inventory of letter-diacritic combinations, used in addition to
+    the unmodified letters of the basic Roman script in the Romanization of
+    Kurdish, with their Unicode encoding, is:
+
+      '‘': '\u2018' , '’': '2019'
+      'Ç': '00C7' , 'ç': '00E7'
+      'Ḍ': '1E0C' , 'ḍ': '1E0D'
+      'Ê': '00CA' , 'ê': '00EA'
+
+      # There is no single Unicode encoding for these letter-diacritic combinations.
+      'Ḧ': '0048+0308' , 'ḧ': '0068+0308'
+      'Î': '00CE' , 'î': '00EE'
+      'Ł': '0141' , 'ł': '0142'
+      'Ö': '00D6' , 'ö': '00F6'
+      'Ṟ': '1E5E' , 'ṟ': '1E5F'
+      'Ş': '015E' , 'ş': '015F'
+      'Ṣ': '1E62' , 'ṣ': '1E63'
+      'Ṭ': '1E6C' , 'ṭ': '1E6D'
+      'Û': '00DB' , 'û': '00FB'
+      'Ü': '00DC' , 'ü': '00FC'
+      'Ẍ': '1E8C' , 'ẍ': '1E8D'
+
+  - The Romanization column shows only lowercase forms but, when
+    romanizing, uppercase and lowercase Roman letters as appropriate should
+    be used.
+
+
+tests:
+  - source:
+    expected:
+
+map:
+  characters:
+    '\u0621': '’' # ء (see note 1 and 7)
+    '\u0628': 'b'  # ب
+    '\u067E': 'p'  # پ
+    '\u062A': 't'  # ت (see note 2)
+    '\u062C': 'c'  # ج
+    '\u0686': # چ (see notes 3 and 7)
+      - 'ch'
+      - 'ç'
+    '\u062D': 'ḧ'  # ح
+    '\u062E': 'x'  # خ
+    '\u062F': 'd'  # د
+    '\u0631': 'r'  # ر
+    '\u0695': 'ṟ'  # ڕ (Formerly written ڒ ڔ or رر according to typeface available; may vary on older sources. See note 7.)
+    '\u0632': 'z'  # ز (see note 2)
+    '\u0698': 'j'  # ژ
+    '\u0633': 's'  # س (see note 2)
+    '\u0634': # ش (see notes 3 and 7)
+      - 'sh'
+      - 'ş'
+    '\u0635': 'ṣ'  # ص (see notes 2 and 7)
+    '\u0636': 'ḍ'  # ض (see notes 2 and 7)
+    '\u0637': 'ṭ'  # ط  (see notes 2 and 7)
+    '\u0639': '‘'  # ع  (see note 7)
+    '\u063A': # غ (see notes 3 and 7)
+      - 'gh'
+      - 'ẍ'
+    '\u0341': 'f'  # ف
+    '\u06A8': 'v'  # ڨ  (see note 4)
+    '\u0642': 'q'  # ق
+    '\u06A9': 'k'  # ك
+    '\u06AF': 'g'  # گ
+    '\u0644': 'l'  # ل
+    '\u06B5': 'ł'  # ڵ  (Formerly written ڶ according to type available; may vary on older sources. See note 7)
+    '\u0645': 'm'  # م
+    '\u0646': 'n'  # ن
+    '\u0648': 'w'  # و  (see note 4)
+    '\u0647': 'h'  # ه  (see note 5)
+    '\u064A': 'y'  # ي
+
+    # VOWELS
+    'ه' : 'e' # See notes 1 and 5
+    'ئه' : 'e' # See notes 1 and 5
+    'ا' : 'a' # See note 1
+    'ئا' : 'a' # See note 1
+    'ي' : 'î' # See notes 1, 6 and 7
+    'ئي' : 'î' # See notes 1, 6 and 7
+    'ئ' : 'i'
+    'ێ' : 'ê' # See note 7
+    'ئێ' : 'ê' # See note 7
+    'و' : 'u'
+    'ئو' : 'u'
+    'وو' : 'û' # See note 7
+    'ئوو' : 'û' # See note 7
+    'ۆ' : 'o'
+    'ئۆ' : 'o'
+    'و' : 'ö' # Rare; previously written وي . See note 7
+    'ۊ' : 'ü' # Only appearing in some dialects and only in old sources. Often equated to /û/ (row 7 above). Sometimes written يو See note 7.