From ab0ee308818f8d04b8a2bb0d472334598265a004 Mon Sep 17 00:00:00 2001
From: Magnus Hagander <magnus@hagander.net>
Date: Wed, 16 Nov 2016 13:00:50 +0100
Subject: [PATCH] Support loading docs in utf-8 from pg version 10

---
 tools/docs/docload.py | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/tools/docs/docload.py b/tools/docs/docload.py
index faf1892e..12b26957 100755
--- a/tools/docs/docload.py
+++ b/tools/docs/docload.py
@@ -29,7 +29,17 @@ def load_doc_file(filename, f):
 				indent='auto',
 			)
 
-	contents = unicode(f.read(),'latin1')
+	# Postgres 10 started using xml toolchain and now produces docmentation in utf8. So we need
+	# to figure out which version it is.
+	rawcontents = f.read()
+	if rawcontents.startswith('<?xml version="1.0" encoding="UTF-8"'):
+		# Version 10, use utf8
+		encoding = 'utf-8'
+	else:
+		encoding = 'latin1'
+
+	contents = unicode(rawcontents, encoding)
+
 	tm = re_titlematch.search(contents)
 	if tm:
 		title = tm.group(1)
-- 
2.39.5