Faced with the perplexing use of a space before a comma in
oils_i18n_gettext() keys, db-seed-i18n.py's regex failed. The solution
for a sloppy regex? More powerful regex!
Signed-off-by: Dan Scott <email address hidden>
Signed-off-by: Thomas Berezansky <email address hidden>
Support the "standard" i18n build process for internationalization by
integrating the TPAC string extraction (via xgettext.pl) and update /
install bits. Also, add the base POT file to support translation in
Launchpad.
Signed-off-by: Dan Scott <email address hidden>
Signed-off-by: Thomas Berezansky <email address hidden>
TPAC: Protect against requests for non-existent records
If a record request is received for a record that does not exist (say,
if 100 records have been loaded and someone submits a request for
/eg/opac/record/399), rather than dying with a server error, return
immediately.
Eventually we'll want to provide an explicit "record was not
found error" in the TPAC record details; for now the TPAC just displays
zero bib detail.
Signed-off-by: Dan Scott <email address hidden>
Signed-off-by: Thomas Berezansky <email address hidden>
Protect against empty default_CD_modifiers in opensrf.xml
If the app settings for the open-ils.search service in opensrf.xml had
an empty (but not entirely missing) <default_CD_modifiers> element, then
the open-ils.storage log would contain entries like the following:
Use of uninitialized value $class in hash element at
/usr/local/share/perl5/OpenILS/Application/Storage/QueryParser.pm
Protect against that problem with more defensive code in QueryParser.pm.
Signed-off-by: Mike Rylander <email address hidden>
Signed-off-by: Dan Scott <email address hidden>
7fae54a...
by
Lebbeous Fogle-Weekley <email address hidden>
Avoid too-strict controller checks when fleshing for cstore, reporter-store
Move controller-verificiation from the beginning of recursive
doFieldmapperSearch() to nearer the end, right before we recurse, so we
only check the child class's controller if we're pcrud. If we're not
pcrud, code that calls doFieldmapperSearch() from the outside will have
already done appropriate testing of the controller.
Org Unit Custom Tree supports add directly after delete
When deleting a node from the custom tree, it's necessary to save the
tree store after the final call to store.deleteItem, or the delete
operation will not complete. Without this, it's not possible to add a
remove org unit back into the custom tree without reloading the UI.
Signed-off-by: Bill Erickson <email address hidden>
Signed-off-by: Mike Rylander <email address hidden>
Repairs a problem with detecting and storing the sort order of sibling
nodes within custom trees. The original problem was the result of
comparing parent nodes for non-siblings because of a depth-first tree
traversal. Now we process siblings in level-order before processing
child nodes.
Signed-off-by: Bill Erickson <email address hidden>
Signed-off-by: Mike Rylander <email address hidden>
The use of an explicit ID in creating a new
action_trigger.event_definition row was followed by the use of the
CURRVAL() function to grab the most recently issued value for
the action_trigger.event_defintion_id_seq sequence - which happened to
be 100, due to a previously issued SETVAL() on the sequence.
This resulted in references to non-existent event definitions and thus
an error in creating the stock database schema.
By using an explicit ID for the dependent row, we avoid this error.
Signed-off-by: Dan Scott <email address hidden>
Signed-off-by: Thomas Berezansky <email address hidden>