[buddypress-trac] [BuddyPress] #3452: Long Inserted Image in Post doesn't Render in Activity Stream

buddypress-trac at lists.automattic.com buddypress-trac at lists.automattic.com
Sun Aug 14 16:22:35 UTC 2011

#3452: Long Inserted Image in Post doesn't Render in Activity Stream
 Reporter:  nahummadrid  |       Owner:
     Type:  defect       |      Status:  new
 Priority:  normal       |   Milestone:  1.5
Component:  Activity     |     Version:  1.5
 Severity:  normal       |  Resolution:
 Keywords:  has-patch    |

Comment (by boonebgorges):

 OK, good call. I hadn't seen that.

 In 3452.02.patch I implement your first two suggestions.

 As for the multibyte issue. In most places, I think that the non-multibyte
 equivalent will be fine.
 - Using strlen() instead of mb_strlen() means that sometimes the excerpt
 will end up being shorter than intended (because multibyte characters will
 have each of their bytes counted), but this is not the end of the world.
 (In languages where it makes a huge difference, because most characters
 are multibyte, a filter can be placed on 'bp_excerpt_length' to make the
 number larger by default. Imprecise, but better than nothing.)
 - Using strrpos() instead of mb_strrpos() - doesn't really make a
 difference, as long as we're consistent. mb_strrpos() is used in Cake to
 get the final position of a whitespace character, so that words aren't
 broken. strrpos( $truncate, ' ' ); and mb_strrpos( $truncate, ' ' ); will
 give different values, but as long as we're consistent in how they're used
 (and as long as we use either substr or mb_substr, as appropriate, when
 trimming) I don't think there will be an issue.

 The remaining problem with multibyte, aside from excerpt length, is that
 MB characters can sometimes get cut in the middle. When !$exact (ie when
 we only break on word breaks) it won't be an issue, since we'll be
 chopping everything after the last space anyway. For other cases, I've
 stolen some code from wp_html_excerpt() to remove partial entities at the
 end of the string.

 In 3452.02.patch, I have simply taken out the mb_ functions. What do you
 think instead of providing our own versions of these functions (maybe in
 bp-core-wpabstraction.php or something)? They would simply return the non-
 multibyte return value if the MB function doesn't exist. That way, the mb_
 functions could be used if they exist, which it seems like will be more
 likely to be true in parts of the world where it's more important.

Ticket URL: <https://buddypress.trac.wordpress.org/ticket/3452#comment:9>
BuddyPress <http://buddypress.org/>

More information about the buddypress-trac mailing list