Filter tags on the fly - workaround for #168 #329
Conversation
octonato
force-pushed
the
quick-fix-events-by-tag
branch
from
22b5cea to
ed0a38c
Compare
octonato
force-pushed
the
quick-fix-events-by-tag
branch
from
ed0a38c to
5fccc55
Compare
| private[dao] def perfectlyMatchTag( | ||
| dbPublisher: DatabasePublisher[JournalRow], | ||
| tag: String): Source[JournalRow, NotUsed] = { | ||
| Source.fromPublisher(dbPublisher).filter(_.tags.exists(tags => tags.split(",").contains(tag))) |
There was a problem hiding this comment.
I just realised that it's possible to change the tag separator and therefore the split must take that into consideration. I will change this.
| */ | ||
| private[dao] def perfectlyMatchTag( | ||
| dbPublisher: DatabasePublisher[JournalRow], | ||
| tag: String): Source[JournalRow, NotUsed] = { |
There was a problem hiding this comment.
doesn't matter, but this is conceptually a Flow, and could be defined as such
There was a problem hiding this comment.
I can refactor it. I don't have the reflex to think in terms of Flow. Will work on that. ;-)
|
|
||
| # The tag separator to use when tagging events with more than one tag. | ||
| # This property affects jdbc-journal.tagSeparator and jdbc-read-journal.tagSeparator. | ||
| tagSeparator = "," |
There was a problem hiding this comment.
This property was defined in two different places, but we should have only one. We should use the same on the write and read side.
jdbc-journal.tagSeparator and jdbc-read-journal.tagSeparator are now referencing this value.
| */ | ||
| private[dao] def perfectlyMatchTag(tag: String, separator: String) = | ||
| Flow[JournalRow].filter(_.tags.exists(tags => tags.split(separator).contains(tag))) | ||
| } |
There was a problem hiding this comment.
@patriknw, refactor to a Flow
There was a problem hiding this comment.
It's now taking into consideration the separator that is defined in config file.
|
Looks good |
As discussed here: #318
We plan to cut a release for users that can't immediately migrate to the upcoming 4.0.0 (because of schema migrations).
This is a quick fix for the most annoying bug we have. When using sharded tags (like in Lagom), the
eventsByTagcall may return events with unrelated tags. For instance, if you search for the tag "foo1", you also get events tag "foo10".The workaround is to filter out any partial match, in memory. Once merged and release, this can immediately benefit existing Lagom users.