summaryrefslogtreecommitdiff
path: root/src/leap/bitmask
diff options
context:
space:
mode:
authorNavaL <ayoyo@thoughtworks.com>2016-10-19 15:23:17 +0200
committerKali Kaneko (leap communications) <kali@leap.se>2016-10-19 16:48:57 -0400
commit0329b242c851ea1f05e795102638b01bbac7c231 (patch)
tree2f79783c08f8ef7d2b983f7548f5e450a66c120a /src/leap/bitmask
parent446280b452218569978da2a6d9ba5f8ef3e501b6 (diff)
[feat] leap.mail: expose content charset as a separated attribute
Before we used to have content_doc.content_type = 'text/plain; charset="utf-8"'. Currently, we have content_doc.content_type = 'text/plain'. This change will give us, on top of that: content_doc.charset = 'utf-8'. One can reconstruct the original one using .content_type and .charset.
Diffstat (limited to 'src/leap/bitmask')
-rw-r--r--src/leap/bitmask/mail/adaptors/soledad.py1
-rw-r--r--src/leap/bitmask/mail/walk.py1
2 files changed, 2 insertions, 0 deletions
diff --git a/src/leap/bitmask/mail/adaptors/soledad.py b/src/leap/bitmask/mail/adaptors/soledad.py
index 4a8008d6..665cef23 100644
--- a/src/leap/bitmask/mail/adaptors/soledad.py
+++ b/src/leap/bitmask/mail/adaptors/soledad.py
@@ -432,6 +432,7 @@ class ContentDocWrapper(SoledadDocumentWrapper):
content_disposition = ""
content_transfer_encoding = ""
content_type = ""
+ charset = ""
class __meta__(object):
index = "phash"
diff --git a/src/leap/bitmask/mail/walk.py b/src/leap/bitmask/mail/walk.py
index 61bbe0b7..6c9fc382 100644
--- a/src/leap/bitmask/mail/walk.py
+++ b/src/leap/bitmask/mail/walk.py
@@ -79,6 +79,7 @@ def get_raw_docs(msg):
'raw': part.get_payload(),
'phash': get_hash(part.get_payload()),
'content-type': part.get_content_type(),
+ 'charset': part.get_content_charset(),
'content-disposition': first(part.get(
'content-disposition', '').split(';')),
'content-transfer-encoding': part.get(