msg filter not working on certain messages
I'm using Thunderbird 45.7.1 under Win 7.
I want to tag some junk mail using "Body Contains" + a word cut & pasted from the message. The message looks the same displayed as Plain Text or Original HTML.
The filter is not working on these messages but does work on a copy I forwarded to myself. So, I checked the message source which says "Content-Type: text/plain; charset="ibm852"; Content-Transfer-Encoding: 8bit". The forwarded copy is "Content-Type: text/plain; charset=UTF-8; format=flowed; Content-Transfer-Encoding: 7bit".
I'm guessing the problem is the filter is not working on the ibm852 charset. Am I on the right track? How do I fix it?
Ñemoĩporã poravopyre
My husband found a work-around; it isn't perfect but will probably suffice. I added a new category for the filter to work from via the customize option in the drop down menu of items to filter. I added "Content-Type" and contains "ibm852". This filter worked on the spam currently annoying me. The article Matt linked to was helpful in deciding that this method will rarely have false hits. Thanks, Matt!
Emoñe’ẽ ko mbohavái ejeregua reheve 👍 0Opaite Mbohovái (5)
The body test available to a filter is only the visible text when you read the message, not the actual body text you see when you view the message source. Are you perhaps using source to try and test on a non visible part of the body?
I think I am using the visible part. I copied a word from the junk mail & pasted to create the filter. I ran the filter on all inbox messages. It worked on another message which had the same word but did not work on the junk mail I copied from. The only difference I could find was the coding info I got from looking at the message source.
ok, I will bite. What is the word
The encoding difference "should" not be an issue as email is generally UNICODE and the charset="ibm852" is a sub set of that in being ANSI. However it will bring up different content/letters in some instances and if your spammer is using it to hide the real content for the apparent it might be enough to fool the particularly literal message filters.
I do not claim any real expertise in character encoding, but the Wikipedia article on the the topic would indicate that it is basically obsolete https://en.wikipedia.org/wiki/Code_page_852 and as ANSI is a subset of unicode I really don't think it should be an issue, but it might
Well, the intended words were "my perverted friend" but when that failed I went with "activity" which I had in another email. (And which I put into a test email to myself.) No funny characters or extra accents are visible, like I see in "Russian Girls" type of spam.
The filter works on a forwarded copy of the email so I may try copying it from the profiles structure to try it on another (newer) system.
Thank you for your responses. I will check out that article.
Ñemoĩporã poravopyre
My husband found a work-around; it isn't perfect but will probably suffice. I added a new category for the filter to work from via the customize option in the drop down menu of items to filter. I added "Content-Type" and contains "ibm852". This filter worked on the spam currently annoying me. The article Matt linked to was helpful in deciding that this method will rarely have false hits. Thanks, Matt!