Fridolin Somers
d0d150a0c5
Add the sort-string-utf.chr for sorting Greek catalog (it can also contain latin records). Developped in collaboration with Giannis Kourmoulis <ikourmou@lib.auth.gr> Test plan : - Install using "gr" in "Primary language for Zebra indexing" - Index a greek catalog - Sort by title and check sorting is correct Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz> Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com> Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
319 lines
4.1 KiB
Text
319 lines
4.1 KiB
Text
# Generic character map for Latin and Greek characters.
|
||
|
||
# Define the basic value-set. *Beware* of changing this without re-indexing
|
||
# your databases.
|
||
encoding utf-8
|
||
|
||
# basic character set
|
||
lowercase {0-9}{a-z}αβγδεζηθικλμνξοπρστυφχψω
|
||
uppercase {0-9}{A-Z}ΑΒΓΔΕΖΗΘΙΚΛΜΝΞΟΠΡΣΤΥΦΧΨΩ
|
||
|
||
# Breaking characters
|
||
space {\001-\040}!"#$%&'\()*+,-./:;<=>?@\[\\]^_`\{|}~’{\x88-\x89}{\x98-\x9C}¡¿«»
|
||
|
||
# Characters to be considered equivalent
|
||
equivalent aáàãåâăąȧǎȁȃ
|
||
equivalent Æӕä(ae)
|
||
equivalent ā(aa)
|
||
equivalent iíìîịĩĭįǐȉȋ
|
||
equivalent ï(ie)
|
||
equivalent ī(ii)
|
||
equivalent uúùûũŭųűǔȕȗ
|
||
equivalent ü(ue)
|
||
equivalent ū(uu)
|
||
equivalent eéèêẽĕęėěȅȇ
|
||
equivalent ëē(ee)
|
||
equivalent oóòõôŏǫȯőǒȍȏ
|
||
equivalent Œœöø(oe)
|
||
equivalent ō(oo)
|
||
equivalent αάΑΆ(α΄)(Α΄)
|
||
equivalent εέΕΈ(ε΄)(Ε΄)
|
||
equivalent ιίϊΙΊΪ(ι΄)(Ι΄)
|
||
equivalent ηήΗΉ(η΄)(Η΄)
|
||
equivalent οόΟΌ(ο΄)(Ο΄)
|
||
equivalent υύϋΥΎΫ(υ΄)(Υ΄)
|
||
equivalent ωώΩΏ(ω΄)(Ω΄)
|
||
equivalent Σσς
|
||
|
||
# Supplemental mappings
|
||
# mapping diacritics to plain ASCII
|
||
# A
|
||
map ā a
|
||
map Ā a
|
||
map á a
|
||
map Á a
|
||
map à a
|
||
map À a
|
||
map ã a
|
||
map à a
|
||
map å a
|
||
map Å a
|
||
map â a
|
||
map  a
|
||
map ä a
|
||
map Ä a
|
||
map ă a
|
||
map Ă a
|
||
map ą a
|
||
map Ą a
|
||
map ȧ a
|
||
map Ȧ a
|
||
map ǎ a
|
||
map Ǎ a
|
||
map ȁ a
|
||
map Ȁ a
|
||
map ȃ a
|
||
map Ȃ a
|
||
map ⱥ a
|
||
map Ⱥ a
|
||
# AE
|
||
map ӕ (ae)
|
||
map Ӕ (ae)
|
||
# I
|
||
map ī i
|
||
map Ī i
|
||
map ï i
|
||
map Ï i
|
||
map î i
|
||
map Î i
|
||
map í i
|
||
map Í i
|
||
map ì i
|
||
map Ì i
|
||
map ị i
|
||
map Ị i
|
||
map ĩ i
|
||
map Ĩ i
|
||
map ĭ i
|
||
map Ĭ i
|
||
map į i
|
||
map Į i
|
||
map ǐ i
|
||
map Ǐ i
|
||
map ȉ i
|
||
map Ȉ i
|
||
map ȋ i
|
||
map Ȋ i
|
||
map ɨ i
|
||
map Ɨ i
|
||
# U
|
||
map ū u
|
||
map Ū u
|
||
map ü u
|
||
map Ü u
|
||
map ù u
|
||
map Ù u
|
||
map ú u
|
||
map Ú u
|
||
map û u
|
||
map Û u
|
||
map ũ u
|
||
map Ũ u
|
||
map ŭ u
|
||
map Ŭ u
|
||
map ų u
|
||
map Ų u
|
||
map ű u
|
||
map Ű u
|
||
map ǔ u
|
||
map Ǔ u
|
||
map ȕ u
|
||
map Ȕ u
|
||
map ȗ u
|
||
map Ȗ u
|
||
map ʉ u
|
||
map Ʉ u
|
||
map ů u
|
||
map Ů u
|
||
# E
|
||
map ê e
|
||
map Ê e
|
||
map ë e
|
||
map Ë e
|
||
map é e
|
||
map É e
|
||
map è e
|
||
map È e
|
||
map ē e
|
||
map Ē e
|
||
map ẽ e
|
||
map Ẽ e
|
||
map ĕ e
|
||
map Ĕ e
|
||
map ę e
|
||
map Ę e
|
||
map ė e
|
||
map Ė e
|
||
map ě e
|
||
map Ě e
|
||
map ȅ e
|
||
map Ȅ e
|
||
map ȇ e
|
||
map Ȇ e
|
||
map ɇ e
|
||
map Ɇ e
|
||
# O
|
||
map ô o
|
||
map Ô o
|
||
map ó o
|
||
map Ó o
|
||
map ò o
|
||
map Ò o
|
||
map ö o
|
||
map Ö o
|
||
map õ o
|
||
map Õ o
|
||
map ø o
|
||
map Ø o
|
||
map ō o
|
||
map Ō o
|
||
map ǒ o
|
||
map Ǒ o
|
||
map ŏ o
|
||
map Ŏ o
|
||
map ǫ o
|
||
map Ǫ o
|
||
map ȯ o
|
||
map Ȯ o
|
||
map ő o
|
||
map Ő o
|
||
map ȍ o
|
||
map Ȍ o
|
||
map ȏ o
|
||
map Ȏ o
|
||
map Œ (oe)
|
||
map œ (oe)
|
||
# R
|
||
map ṛ r
|
||
map Ṛ r
|
||
map r̥ r
|
||
map R̥ r
|
||
map ṝ r
|
||
map Ṝ r
|
||
map ř r
|
||
map Ř r
|
||
map ṟ r
|
||
map ɍ r
|
||
map Ɍ r
|
||
# L
|
||
map ḷ l
|
||
map Ḷ l
|
||
map ḹ l
|
||
map Ḹ l
|
||
map ḻ l
|
||
map Ḻ l
|
||
map ł l
|
||
map Ł l
|
||
# N
|
||
map ṅ n
|
||
map Ṅ n
|
||
map ñ n
|
||
map Ñ n
|
||
map ǹ n
|
||
map Ǹ n
|
||
map ń n
|
||
map Ń n
|
||
map ṉ n
|
||
map Ṉ n
|
||
map ṇ n
|
||
map Ṇ n
|
||
# T
|
||
map ṭ t
|
||
map Ṭ t
|
||
map ŧ t
|
||
map Ŧ t
|
||
# D
|
||
map ḍ d
|
||
map Ḍ d
|
||
map đ d
|
||
map Đ d
|
||
# S
|
||
map ś s
|
||
map Ś s
|
||
map ṣ s
|
||
map Ṣ s
|
||
map ş s
|
||
map Ş s
|
||
map š s
|
||
map Š s
|
||
map ṡ s
|
||
map Ṡ s
|
||
# M
|
||
map ṁ m
|
||
map Ṁ m
|
||
map ṃ m
|
||
map Ṃ m
|
||
# H
|
||
map ḥ h
|
||
map Ḥ h
|
||
map ẖ h
|
||
map ħ h
|
||
map Ħ h
|
||
# Z
|
||
map ź z
|
||
map Ź z
|
||
map ẓ z
|
||
map Ẓ z
|
||
map ż z
|
||
map Ż z
|
||
map ƶ z
|
||
map Ƶ z
|
||
#C
|
||
map ç c
|
||
map Ç c
|
||
map ć c
|
||
map Ć c
|
||
map č c
|
||
map Č c
|
||
map ȼ c
|
||
map Ȼ c
|
||
map Ċ c
|
||
map ċ c
|
||
map Ĉ c
|
||
map ĉ c
|
||
# G
|
||
map ǥ g
|
||
map Ǥ g
|
||
map ġ g
|
||
map Ġ g
|
||
map ǧ g
|
||
map Ǧ g
|
||
map ğ g
|
||
map Ğ g
|
||
# K
|
||
map ḵ k
|
||
map Ḵ k
|
||
map Ḱ k
|
||
map ḱ k
|
||
# Y
|
||
map ý y
|
||
map Ý y
|
||
map ɏ y
|
||
map Ɏ y
|
||
|
||
## mapping diacritics to plain on greek characters
|
||
map ά α
|
||
map Ά α
|
||
map έ ε
|
||
map Έ ε
|
||
map ί ι
|
||
map Ί ι
|
||
map ή η
|
||
map Ή η
|
||
map ό ο
|
||
map Ό ο
|
||
map ύ υ
|
||
map Ύ υ
|
||
map ώ ω
|
||
map Ώ ω
|
||
map Σ σ
|
||
map ς σ
|
||
|
||
# Remove prefixes
|
||
map (^ὁ\s) @
|
||
map (^ἡ\s) @
|
||
map (^τό\s) @
|
||
map (^τώ\s) @
|
||
map (^οἱ\s) @
|
||
map (^αἱ\s) @
|
||
map (^τά\s) @
|