Misclassifying text as text/html

[Plone Forge]

If a text file is uploaded without an extension in the filename,
text_type in lib/python/OFS/content_types.py is called to classify the text as text/html text/xml or text/plain. The very weak condition s.find('</') >= 0 is used for text/html, which means that if you are uploading a bunch of files via WebDAV to a ZWiki, without file extensions, anything with a </ in it gets classified as HTML, hence a Document not a WikiPage under Plone.

The code should be a bit more rigourous: a suggested patch is attached.

[Plone Forge (ploneforge)]

Uploaded: init__.py

Oops - sorry this code has moved in 2.9.8 to
Zope-2.9.8-final/lib/python/zope/app/content_types/__init__.py

The same logic applies as the code was duplicated. Attached is a slightly more elaborate patch that tries to be more efficient.

[Andreas Jung (ajung)]

Status: Pending => Resolved

Patch applied to 2.10 branch and trunk.

[Andreas Jung (ajung)]

Status: Resolved => Pending

ups...closed the wrong issue

[Andreas Jung (ajung)]

Status: Pending => Resolved

Patch applied to the zope.contenttype trunk:

Log message for revision 79688:
  Improved text_type(). Based on the patch from
  http://www.zope.org/Collectors/Zope/2355/

Changed:
  U zope.contenttype/trunk/CHANGES.txt
  U zope.contenttype/trunk/src/zope/contenttype/__init__.py
  U zope.contenttype/trunk/src/zope/contenttype/tests/testContentTypes.py