View Issue Details

IDProjectCategoryView StatusLast Update
0008263mantisbtscriptingpublic2007-11-23 19:00
Reportergiallu Assigned Tojreese  
PrioritynormalSeverityminorReproducibilityhave not tried
Status closedResolutionfixed 
Product Version1.1.0rc2 
Target Version1.1.0rc3Fixed in Version1.1.0rc3 
Summary0008263: smaller issue in email address recognition
Description

it seems that putting a email address at the end of a line (so it has an additional full stop after it) creates a weird markup on output.

Test follows: test@example.com.

first seen on 0008252

TagsNo tags attached.

Relationships

has duplicate 0008573 closedgiallu Invalid formatting for email addresses 

Activities

vboctor

vboctor

2007-11-13 10:42

manager   ~0016187

Reminder sent to: paulr

IIRC, this is your favourite area :)

jreese

jreese

2007-11-15 16:27

reporter   ~0016219

This happens whenever there is a period and THEN a line feed. This triggers another atom in the email regex, which with our regex, only checks for the first char of an atom to be 'not @', which matches the < of the
nl2br() replacement for the \n. In turn, because the
has a space in it, that forces the end of the email atom regex, and the <br portion is interpretted as part of the address, while the /> portion gets thrown onto the next line.

I've tried tweaking our existing email regex, but it seems very brittle, and does not properly handle an email address ending in a period even without a newline afterwards. We could technically pull Perl's RFC-compliant regex, but it's about 2000 character's long, so I will look for a good, robust, compact alternative to use in place of our existing regex.

jreese

jreese

2007-11-20 09:26

reporter   ~0016252

This has been fixed in 1.1.0rc3 and 1.2.0-SVN.