# copyright 2007 ETH Zurich, DMATL
# author lorenz textor
from Products.Silva.Folder import Folder as SilvaFolder
from Products.Silva.Publication import Publication
from OFS.Folder import Folder
from Products.PageTemplates.ZopePageTemplate import ZopePageTemplate
from Products.PythonScripts.PythonScript import PythonScript
from Products.SilvaExternalSources.CodeSource import CodeSource
# script which walks through all folders below object obj and fixes non-ascii page templates and python scripts inside code sources
# P R O B L E M
# -> management_page_charset sets the encoding header for the management views
# -> was utf-8 until SilvaExternalSources 0.13, now iso-8859-1(why?)
# -> objects in code sources acquire this setting (e.g. page templates or python scripts)
# -> to work with silva this objects should be utf-8, not iso-8859-1
# solution:
# - edit /Products/SilvaExternalSources/ExternalSource.py (otherwise CodeSources have wrong encoding)
# -management_page_charset = 'iso-8859-1'
# +management_page_charset = 'UTF-8'
# S O L U T I O N
# -> now we leave SilvaExternalSources source code as it is and do the following:
# - decode all python scripts and page templates inside code sources with non-ascii characters with utf-8 charset and reencode with iso-8859-1
# - change "Data encoding" for this code sources to iso-8859-1
def fixTemplatesAndScripts(obj):
"""walks through all folders below object obj and fixes non-ascii page templates and python scripts inside code sources"""
# obj = context.restrictedTraverse('/silva')
# get files
def walk(objects):
for id,val in objects:
if isinstance(val,SilvaFolder) or isinstance(val,Publication) or isinstance(val,Folder):
# step into subfolder
walk(val.objectItems())
elif isinstance(val,ZopePageTemplate):
zpts.append(val)
elif isinstance(val,PythonScript):
pyts.append(val)
# scan files
def test(object):
content = object.read()
for char in content:
if ord(char)>127:
return True
return False
# fix file
def fix(zpt):
# check if in codesource
inCodeSource = False
pathParts = zpt.absolute_url_path().split('/')
for i in range(len(pathParts)-2):
container = obj.restrictedTraverse('/'.join(zpt.absolute_url_path().split('/')[0:-1-i]))
if isinstance(container,CodeSource):
inCodeSource = True
break
if inCodeSource:
try:
zpt.write(zpt.read().decode('utf-8').encode('iso-8859-1'))
output.append('%s (fixed)
' % zpt.absolute_url_path())
container.set_data_encoding('iso-8859-1')
output.append('%s (CodeSource changed to iso-8859-1)
' % (container.absolute_url_path()))
except:
output.append('%s (error while trying to fix)
' % zpt.absolute_url_path())
else:
output.append('%s (not fixed because not in code source)
' % zpt.absolute_url_path())
zpts = []
pyts = []
output = ['']
walk(obj.objectItems())
# scan page tamplates
output.append('Fixing PageTemplates with non-ascii characters
------------------------------------------------
')
for zpt in zpts:
if test(zpt):
fix(zpt)
output.append("%d PageTemplates scanned.
" % len(zpts))
# scan python scripts
output.append('Fixing PythonScripts with non-ascii characters
------------------------------------------------
')
for pyt in pyts:
if test(pyt):
fix(pyt)
output.append("%d PythonScripts scanned.
" % len(pyts))
output.append('
Be sure to test all fixed pages afterwards!
')
output.append('')
return ''.join(output)