[go: nahoru, domu]

Open Bug 129407 Opened 23 years ago Updated 2 years ago

Japanese ascii characters are not converted to standard ascii even though they look the same

Categories

(MailNews Core :: Internationalization, defect)

defect

Tracking

(Not tracked)

Future

People

(Reporter: u32858, Unassigned)

Details

(Keywords: intl)

Attachments

(2 files)

From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.8+) Gecko/20020227
BuildID:    2002022700

Hello
Today I found out that as well as Japanese duplications of !! ?? .。 ,、 There
is one that affects more than all the others.

The japanese '@' does not work if I add it as I acidentally did today. Some one
emailed me with emailaddress@hotmail.com so i added it the addres book but it
did not find it when i typed in the nick name.

Could a check be made when importing and adding new email addresses that if the
Japanese '@' is used it is replaced with the standard '@'

This also applys to all the input fields on mozilla, and composer

Reproducible: Always
Steps to Reproduce:
1.try to email emailaddress@hotmail.com
2.try to add it to your address book


Actual Results:  3.see that the emailaddress@hotmail.com requires '@' instead

Expected Results:  should convert japanese AT to standard '@' AT
set browser to utf-8 as headers were missing for this email and site
Since with the current build, email address area only supports ascii, so typing
Japanese @ is not considered a right action. But we need to convert it to
English @ for address collecting. JG, could you attach a testing mail to this
report?
Thanks.
Status: UNCONFIRMED → NEW
Ever confirmed: true
Keywords: intl
make sure you use ISO-2022-JP to view it on the charset menu in mozilla
Status: NEW → ASSIGNED
Target Milestone: --- → Future
Hello

I have changed the topic to "Japanese ascii characters are not converted to
standard ascii even though they look the same"

I hope this is ok for me to change,it but it affects all letters not just @

japanese ascii: 123456789@12345.com
standard ascii: 123456789@12345.com

Also the alphabet has the same problems, not just numbers. the address book
allows them to be input, but does not convert them

messenger can not find the name if i add it then
Summary: Japanese @ is not supported even though it looks identical → Japanese ascii characters are not converted to standard ascii even though they look the same
This is to do with "double width characters" which CJK typically use on
computers for some stuff since CJK characters are usually twice as wide as
most Roman characters.  Often things look nicer when both types of characters
line up perfectly.  So they came up with these double width characters.
Unicode begrudgingly supports these as well as the normal width characters
but they generally lead to problems or at least misunderstandings such as here.
Double width characters are *not* ASCII characters.
OK, I have seen them called J-ASCII, but how they can be Japanese and American
in the same acronim I don't quite follow.

The problem is more too do with the lack of checking for valid email addresses
in the address book though.

Another test case,

1. enter an name in your address book, test: test@blue.com,  <- make sure the
comma is there
2. type test in compose
3. see test <"test"@blue.com,>  pops up, another problem

Perhaps I should rename this bug "Improve validation of email address characters
and convert double space Japanese latin characters to the permitted ASCII for
email" ?

JG
I don't think "Japanese ascii" should be converted to "standard ascii".
Many people use "Japanese ascii" to avoid auto-collecting.
Hello Kazhik,

I think the email protocol only works with US-ASCII for email addresses, i dont
think Japense ascii should be converted for email body, only the email address
fields.

Perpahs this is covered by an RFC

JG
jg, what's your point? do you mean "Japanese ascii" characters should be
converted to ascii when editing the address field in the mail compose window?
Can you quote RFC?
Hello kazhik,
Currently, i can enter "tom" "04@423.com" in the address book and the
address book will accept it. Then when i try and email "tom" it does not work.

I think it would be useful if when i enter "04@423.com" in the address
book it was conveted to "04@423.com" <- US-ASCII.

JG

ps i use utf-8 for this message.
text/plain; charset=utf-8
utf8 does not seem to be forced, i created an attachement but no diffence, does
anyone know how to set it?

This is an example J-ASCII email address in utf8
rental@open.net

This might look normal, but all characters are wrong, they will not work

in standard ASCII it is:
rental@open.net


but unfortunalty some people have this as there email address, or email me like
this, it is easy to make the mistake if the font used to display is very similar

There is NO checking for valid emails in the adress book, this is a major
problem, i could even enter $"£$"$£ as the email adress and it would be accepted.

My point is that these rental@open.net email addresses can be
detected and converted to standard email addresses.

The adddress book allows me to add them, but the message composition window
never finds them if you search by email adress.

Ji: can this be upgraded and set a milestone?

Regards
JG



bug 150946 is for e-mail address validation
Product: MailNews → Core
Product: Core → MailNews Core
QA Contact: ji → i18n
On Thunderbird 3.1 with full-width test case, It causes error "xxx is not a valid e-mail address" when sending mail.  It is feature request issue.
Assignee: nhottanscp → nobody
Status: ASSIGNED → NEW
OS: Linux → All
Hardware: x86 → All
Severity: minor → S4
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: