From 8fb38c386245c836bd2f08d7b3f9d62f8326bd2d Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Fri, 29 Oct 2010 12:17:54 -0600
Subject: [PATCH] FB2 Input: Make parsing of malformed FB2 files a little more
 robust

---
 resources/templates/fb2.xsl        |  6 +++---
 src/calibre/ebooks/fb2/input.py    | 12 ++++++++----
 src/calibre/ebooks/metadata/fb2.py |  3 ++-
 3 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/resources/templates/fb2.xsl b/resources/templates/fb2.xsl
index ed2d8359db..753bd97373 100644
--- a/resources/templates/fb2.xsl
+++ b/resources/templates/fb2.xsl
@@ -178,13 +178,13 @@
                     </xsl:if>
                     <xsl:if test="$section_toc_id != 'None'">
                         <xsl:element name="a">
-                            <xsl:attribute name="name">TOC_<xsl:value-of select="$section_toc_id"/></xsl:attribute>
+                            <xsl:attribute name="id">TOC_<xsl:value-of select="$section_toc_id"/></xsl:attribute>
                         </xsl:element>
                     </xsl:if>
 					<a name="TOC_{generate-id()}"></a>
 					<xsl:if test="@id">
 						<xsl:element name="a">
-							<xsl:attribute name="name"><xsl:value-of select="@id"/></xsl:attribute>
+							<xsl:attribute name="id"><xsl:value-of select="@id"/></xsl:attribute>
 						</xsl:element>
 					</xsl:if>
 					<xsl:apply-templates/>
@@ -194,7 +194,7 @@
 				<xsl:element name="h6">
 					<xsl:if test="@id">
 						<xsl:element name="a">
-							<xsl:attribute name="name"><xsl:value-of select="@id"/></xsl:attribute>
+							<xsl:attribute name="id"><xsl:value-of select="@id"/></xsl:attribute>
 						</xsl:element>
 					</xsl:if>
 					<xsl:apply-templates/>
diff --git a/src/calibre/ebooks/fb2/input.py b/src/calibre/ebooks/fb2/input.py
index e541de46f7..2b08a716cc 100644
--- a/src/calibre/ebooks/fb2/input.py
+++ b/src/calibre/ebooks/fb2/input.py
@@ -40,14 +40,18 @@ class FB2Input(InputFormatPlugin):
                 accelerators):
         from calibre.ebooks.metadata.opf2 import OPFCreator
         from calibre.ebooks.metadata.meta import get_metadata
-        from calibre.ebooks.oeb.base import XLINK_NS, XHTML_NS
+        from calibre.ebooks.oeb.base import XLINK_NS, XHTML_NS, RECOVER_PARSER
         NAMESPACES = {'f':FB2NS, 'l':XLINK_NS}
         log.debug('Parsing XML...')
-        raw = stream.read()
+        raw = stream.read().replace('\0', '')
         try:
-            doc = etree.fromstring(raw.replace('\0', ''))
+            doc = etree.fromstring(raw)
         except etree.XMLSyntaxError:
-            doc = etree.fromstring(raw.replace('& ', '&amp;'))
+            try:
+                doc = etree.fromstring(raw, parser=RECOVER_PARSER)
+            except:
+                doc = etree.fromstring(raw.replace('& ', '&amp;'),
+                        parser=RECOVER_PARSER)
         stylesheets = doc.xpath('//*[local-name() = "stylesheet" and @type="text/css"]')
         css = ''
         for s in stylesheets:
diff --git a/src/calibre/ebooks/metadata/fb2.py b/src/calibre/ebooks/metadata/fb2.py
index 6e0d56dfa0..3636b89df4 100644
--- a/src/calibre/ebooks/metadata/fb2.py
+++ b/src/calibre/ebooks/metadata/fb2.py
@@ -22,7 +22,8 @@ def get_metadata(stream):
                 'xlink':XLINK_NS})
     tostring = lambda x : etree.tostring(x, method='text',
             encoding=unicode).strip()
-    root = etree.fromstring(stream.read())
+    parser = etree.XMLParser(recover=True, no_network=True)
+    root = etree.fromstring(stream.read(), parser=parser)
     authors, author_sort = [], None
     for au in XPath('//fb2:author')(root):
         fname = lname = author = None