Add emoji-updater script

* Exports emoji list from unicode.org in LatinIME compatible xml format

Change-Id: I6b86ac3253ee775c0c1c206ac62bf34a8f317093

1 files changed, 30 insertions, 0 deletions

diff --git a/emoji-updater/emoji-updater.py b/emoji-updater/emoji-updater.py
new file mode 100755
index 0000000..a0293c7
--- /dev/null
+++ b/emoji-updater/emoji-updater.py
@@ -0,0 +1,30 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+import sys
+
+import requests
+from lxml import etree
+
+if __name__ == '__main__':
+    if len(sys.argv) < 2:
+        sys.exit(f'usage: {sys.argv[0]} [url|https://unicode.org/emoji/charts-12.0/full-emoji-list.html]')
+
+    url = sys.argv[1]
+    req = requests.get(url=url)
+
+    parser = etree.HTMLParser(recover=True, encoding='utf-8')
+    doc = etree.fromstring(text=req.content, parser=parser)
+
+    for tr in doc.xpath('.//tr'):
+        mediumhead = tr.xpath('.//th[@class="mediumhead"]/a')
+
+        if len(mediumhead) > 0:
+            print(f'        <!-- {mediumhead[0].text} -->')
+            continue
+
+        code = tr.xpath('.//td[@class="code"]/a')
+
+        if len(code) > 0:
+            codes = ','.join([x[2:] for x in code[0].text.split()])
+            print(f'        <item>{codes}</item>')
+            continue