[EP-tech] BUG? First creators disrespected until another run of import --update
Florian Heß
hess at ub.uni-heidelberg.de
Thu Apr 25 15:38:41 BST 2013
Hi,
We're experiencing a weird phenomenon in E-Prints version 3.3.10: The
first creators are not imported on the first run of
# The record isn't existing yet
$ ~/bin/import $repo --user $userid --enable-import-fields
--enable-file-imports --update eprint XML /tmp/savifa1867.xml
# Now, record is there but without creator
$ ~/bin/import $repo --user $userid --enable-import-fields
--enable-file-imports --update eprint XML /tmp/savifa1867.xml
# Now, everything is fine
Between the two runs no changes to /tmp/savifa1867.xml have been made.
Now, please look at the relevant data of another record with multiple
creators ...
After 1st run of bin/import command:
mysql> select oa.source_opus, oa.reihenfolge rfo, e.pos rf,
oa.creator_name, e.creators_name_family from savifadok.opus_autor oa
left outer join eprint_creators_name e on oa.source_opus=e.eprintid AND
oa.reihenfolge=e.pos+1;
+-------------+------+------+-------------------------------------+
| source_opus | rfo | rf | creator_name | creators_name_family
+-------------+------+------+--------------------+----------------+
[...]
| 181 | 1 | 0 | Yule, Paul | Brandtner |
| 181 | 2 | 1 | Brandtner, Martin | Misra |
| 181 | 3 | 2 | Misra, B. | Mohanty |
| 181 | 4 | NULL | Mohanty, P. | NULL |
[...]
After the 2nd:
[...]
| 181 | 1 | 0 | Yule, Paul | Yule |
| 181 | 2 | 1 | Brandtner, Martin | Brandtner |
| 181 | 3 | 2 | Misra, B. | Misra |
| 181 | 4 | 3 | Mohanty, P. | Mohanty |
My creators definition in cfg.d/eprint_fields.pl:
{
'name' => 'creators',
'type' => 'compound',
'multiple' => 1,
'fields' => [
{
'sub_name' => 'name',
'type' => 'name',
'hide_honourific' => 1,
'hide_lineage' => 1,
'family_first' => 0,
},
#{
# 'sub_name' => 'id',
# 'type' => 'text',
# 'input_cols' => 20,
# 'allow_null' => 1,
#}
],
'input_boxes' => 1,
},
Dear Developers, I'd love to provide more information if needed. :-)
This issue is of no gravity for us as multiple migration runs have been
done on a regular basis for each of our four instances (thanks to having
done the metadata preparation in the legacy repository).
Nevertheless I like to hear if you can reproduce the problem in the
first place.
Kind regards,
Florian
--
UB Heidelberg (Altstadt)
Plöck 107-109, 69117 HD
Abt. Informationstechnik
Tel. 06221 / 54 3550
http://www.ub.uni-heidelberg.de/
More information about the Eprints-tech
mailing list