test diacritic composition with latvian pangram

Compare the same sentence with precomposed and decomposed characters,
which should look alike with correct support for diacritics composition.
The (alternate) Latvian pangram features accents both above and below
letters, and does not match automated Unicode decomposition because
typographically preferred commas accents are used instead of cedillas.

diff --git a/unicode.txt b/unicode.txt
index 8679eb612f079f31d01fbf5962077e8ca965ac89..0f0688f9f94e8f0c4e26ac375a62923da34c8969 100644 (file)
--- a/unicode.txt
+++ b/unicode.txt
@@ -72,9 +72,12 @@ Proper typography:
   • the euro symbol: │ 14.95 € │
                      ╰─────────╯
 
-Combining characters:
+Precomposed and combining diacritics:
 
-  STARGΛ̊TE SG-1, a = v̇ = r̈, a⃑ ⊥ b⃑
+  Četri psihi faķīri vēlu vakarā zāģēja guļbūvei durvis, fonā šņācot mežam.
+  Četri psihi fak̦īri vēlu vakarā zāg̓ēja gul̦būvei durvis, fonā šn̦ācot mežam.
+
+  STARGΛ̊TE, a = v̇ = r̈, a⃑ ⊥ b⃑
 
 Pangrams: