From ed8b1bcdad7c430af1eef5fbe78b6ec4eb3eb60e Mon Sep 17 00:00:00 2001
From: Přemysl Eric Janouch <p@janouch.name>
Date: Thu, 7 Oct 2021 03:37:12 +0200
Subject: Add sample dictionary downloaders/builders

---
 dicts/slovnik-cizich-slov.sh | 8 ++++++++
 1 file changed, 8 insertions(+)
 create mode 100755 dicts/slovnik-cizich-slov.sh

(limited to 'dicts/slovnik-cizich-slov.sh')

diff --git a/dicts/slovnik-cizich-slov.sh b/dicts/slovnik-cizich-slov.sh
new file mode 100755
index 0000000..cb99ea1
--- /dev/null
+++ b/dicts/slovnik-cizich-slov.sh
@@ -0,0 +1,8 @@
+#!/bin/sh -e
+# Slovník cizích slov, see https://slovnik-cizich-slov.abz.cz/web.php/o-slovniku
+# TODO: Skipping the optional pronunciation field, tabfile can't handle it yet,
+# but could be made to accept a lowercase sametypesequence
+curl -Lo- https://slovnik-cizich-slov.abz.cz/export.php | \
+iconv -f latin2 -t UTF-8 | perl -CSD -F\\\| -le '
+	print "$_\t" . $F[2] =~ s/\\/\\\\/gr =~ s/; /\\n/gr for split(", ", $F[0])
+' | sort -u | tabfile slovnik-cizich-slov
-- 
cgit v1.2.3-70-g09d2