[2/4] tika git commit: Lower the priority of

Lower the priority of <html later in the file header

Keep the current higher magic priority for <html near the start of the
file header for detecting HTML files, but drop it for later in the file
header, to help avoid false positives, and aid with TIKA-1141


Project: http://git-wip-us.apache.org/repos/asf/tika/repo
Commit: http://git-wip-us.apache.org/repos/asf/tika/commit/d740f5d8
Tree: http://git-wip-us.apache.org/repos/asf/tika/tree/d740f5d8
Diff: http://git-wip-us.apache.org/repos/asf/tika/diff/d740f5d8

Branch: refs/heads/master
Commit: d740f5d8b2e42b1db42806ddd395e034cb416fd4
Parents: d8a2fc0
Author: Nick Burch <ni...@gagravarr.org>
Authored: Wed Feb 3 17:11:06 2016 +0000
Committer: Nick Burch <ni...@gagravarr.org>
Committed: Wed Feb 3 17:11:06 2016 +0000

----------------------------------------------------------------------
 .../org/apache/tika/mime/tika-mimetypes.xml        | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/tika/blob/d740f5d8/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml
----------------------------------------------------------------------
diff --git a/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml b/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml
index 1d7b42b..95f41e6 100644
--- a/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml
+++ b/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml
@@ -5432,12 +5432,6 @@
       <match value="&lt;head" type="string" offset="0:64"/>
       <match value="&lt;TITLE" type="string" offset="0:64"/>
       <match value="&lt;title" type="string" offset="0:64"/>
-      <!-- note on the offset value here: this can only be as big as
-           MimeTypes#getMinLength(). If you set the offset value to larger
-           than that size, the magic will only be compared to up to
-           MimeTypes#getMinLength() bytes.
-       -->
-      <match value="&lt;html" type="string" offset="0:8192"/>
       <match value="&lt;HTML" type="string" offset="0:64"/>
       <match value="&lt;BODY" type="string" offset="0"/>
       <match value="&lt;body" type="string" offset="0"/>
@@ -5449,6 +5443,17 @@
       <match value="&lt;H1" type="string" offset="0"/>
       <match value="&lt;!doctype HTML" type="string" offset="0"/>
       <match value="&lt;!DOCTYPE html" type="string" offset="0"/>
+      <match value="&lt;html" type="string" offset="0:128"/>
+    </magic>
+    <magic priority="20">
+      <!-- Lower priority match for <html anywhere near the top of the file -->
+      <!-- note on the offset value here: this can only be as big as
+           MimeTypes#getMinLength(). If you set the offset value to larger
+           than that size, the magic will only be compared to up to
+           MimeTypes#getMinLength() bytes. It should also only start after
+           the higher priority "start of file" one above
+       -->
+      <match value="&lt;html" type="string" offset="128:8192"/>
     </magic>
     <glob pattern="*.html"/>
     <glob pattern="*.htm"/>

---