cco/cybertools
3
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix boundary (add -- prefix) for MHT documents

Browse source
This commit is contained in:
Helmut Merz 2013-01-18 16:35:34 +01:00
parent 9b09d91f16
commit da946ff560
1 changed files with 3 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
docgen/mht.py
View file

@ -37,7 +37,7 @@ class MHTFile(object):
encoding = 'Windows-1252' encoding = 'Windows-1252'
bodyMarker = 'lxdoc_body' bodyMarker = 'lxdoc_body'
foldernameSuffix = 'Dateien' foldernameSuffix = 'Dateien'
indexes = dict(body=2, filelist=-2) indexes = dict(body=1, filelist=-2)
path = documentName = None path = documentName = None
@ -52,7 +52,7 @@ class MHTFile(object):
def __init__(self, data, body): def __init__(self, data, body):
self.data = data self.data = data
self.msg = email.message_from_string(data) self.msg = email.message_from_string(data)
self.boundary = self.msg.get_boundary() self.boundary = '--' + self.msg.get_boundary()
self.parts = data.split(self.boundary) self.parts = data.split(self.boundary)
self.body = body self.body = body
self.htmlDoc = HTMLDoc(body) self.htmlDoc = HTMLDoc(body)
@ -61,8 +61,7 @@ class MHTFile(object):
for idx, part in enumerate(self.msg.walk()): for idx, part in enumerate(self.msg.walk()):
docPath = part['Content-Location'] docPath = part['Content-Location']
contentType = part.get_content_type() contentType = part.get_content_type()
#print '***', idx, docPath, contentType if idx == self.indexes['body']:
if idx == self.indexes['body'] - 1:
self.path, docname = os.path.split(docPath) self.path, docname = os.path.split(docPath)
self.documentName, ext = os.path.splitext(docname) self.documentName, ext = os.path.splitext(docname)
if contentType.startswith('image/'): if contentType.startswith('image/'):