From d0513b69248953feae89dfa63a286aededeb22a1 Mon Sep 17 00:00:00 2001
From: Andrew Sidwell <andy@entai.co.uk>
Date: Thu, 15 May 2008 16:58:20 +0000
Subject: Add a script that fetches the current HTML5 entity list and parses it
 into a C struct format, suitable for "#include"ing.

svn path=/trunk/hubbub/; revision=4161
---
 build/get-entities/entities.xslt | 11 +++++++++++
 build/get-entities/get_entities  | 13 +++++++++++++
 2 files changed, 24 insertions(+)
 create mode 100644 build/get-entities/entities.xslt
 create mode 100755 build/get-entities/get_entities

(limited to 'build')

diff --git a/build/get-entities/entities.xslt b/build/get-entities/entities.xslt
new file mode 100644
index 0000000..cd371ec
--- /dev/null
+++ b/build/get-entities/entities.xslt
@@ -0,0 +1,11 @@
+<?xml version="1.0"?>
+<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
+<xsl:output method='text' version='1.0' encoding='UTF-8' indent='yes' />
+<xsl:preserve-space elements="t" />
+
+<xsl:template match="/table">
+  <xsl:for-each select="tr">{ "<xsl:value-of select="td[1]/code" />", 0x<xsl:value-of select="substring(td[2],4)" />},
+</xsl:for-each>
+</xsl:template>
+
+</xsl:stylesheet>
diff --git a/build/get-entities/get_entities b/build/get-entities/get_entities
new file mode 100755
index 0000000..abc67b3
--- /dev/null
+++ b/build/get-entities/get_entities
@@ -0,0 +1,13 @@
+#!/bin/bash
+# Get entities from the whatwg.org SVN repository.
+
+(rm -f *.inc entities.xml) || exit 1
+(wget http://svn.whatwg.org/webapps/entities-legacy.inc &&
+	wget http://svn.whatwg.org/webapps/entities-unicode.inc) ||
+	(echo "Failed to get entities" && exit 1)
+(echo "<table>" &&
+	cat entities-legacy.inc &&
+	cat entities-unicode.inc &&
+	echo "</table>") > entities.xml
+(xsltproc entities.xslt entities.xml |
+	sort -d) > entities.inc
-- 
cgit v1.2.3