Re: [UMN_MAPSERVER-USERS] Using JOIN for querying

[cris ricana]

Hi!

here is a last+1 try:
You mentioned the template file: On umn.mapserver.ch I found this example:

LAYER
NAME "Liegenschaften"
STATUS ON
DATA "data/Liegenschaften"
TYPE LINE
HEADER "NUMMER FLAECHE NAME PLZ ORT"
  JOIN
    NAME "Eigentümer"
    TABLE "data/eigentuemer.dbf"
    FROM NUMMER
    TO NR
    TYPE multiple
    TEMPLATE eigentuemer.htm
  END
  CLASS
    COLOR 0 0 0
    SYMBOL "continue"
    SIZE 2
    TEMPLATE liegenschaften.htm
  END
END

Here are two template files. The Layer has a template file in his own
CLASS Section. Is your mapfile analog to this file?

Manfred

Thank you very much Manfred! Because of your suggestion, I think I got
the answer to this problem.

When you emailed me the JOIN example above, it got me thinking if its
actually my mapfile that was wrong. My mapfile is actually correct as
I compare it with the example, as well as the template file, so
perhaps I made a mistake in the main html that gives the query option.
Sure enough, I made the mistake of using the name of the join as the
layer name. I realize now that even if we join two dataset, the
resulting join do not exist as a new layer (i.e.. as something we can
call and use) but more or less as an extension of the old layer. I
think this is true for other kind of datasets, dbfs and csvs. I also
want to try JOIN using direct mysql access, but got the following
error:

Content-type: text/html msMySQLJoinConnect(): Query error. MySQL
support not available (compile with --with-mygis)

Which is what I expect since I didn't compile mapserver with using
MySQL in mind. Anyway here's my JOIN statement if you're interested,
Greg:

(...)
JOIN
   CONNECTIONTYPE mysql
   CONNECTION "localhost:root::LGU"
# The syntax is
# CONNECTION "<hostname>:<username>:<password>:<database>"
   TABLE "lgu_DB_DHDPSEX"
   FROM "Place_ID"
   NAME "reg_join"
   TO "PLACE_ID"
   TYPE single
   TEMPLATE "reg_bndary2000_dbf.html"
END #join

In the template file where the results should show, use the format
"[<join_name>_<fieldname>]" for dbf and mysql joins and
"[<join_name>_<fieldname>]" for csv joins. For example, I'm using
[reg_join_2] to access data from the 2nd column of my csv file.

Please tell me what happens if you try it, Greg.

Maricris