Forgot password? | Forgot username? | Register
  • Index
  • » Users
  • » keithmaguire
  • » Profile

Posts

Posts

We're interested in how people record the location information for items collected in a location known at the time by a different name to its current name.

In Collection Events how do you record the collection location when people may want to search by either the historical or current place name? Is there a place name history/synonym field we're not seeing?

Thanks
Keith

Hi everyone

In the hope that it is useful to other institutions here are instructions for adding a small google map to the catalogue Page View. We here at the South Australian Museum have found it very useful.

I gave a talk at the Australian Users conference the other week on the variety of uses of XSL reports. We use them for most of our reporting now as the same techniques work for mapping and Excel report and Admin tasks and all sorts.  Here's the slides from the talk.

In the hope that others will find the information useful I've put the example files for most of the reports on Github. It's in the nature of XSLT that the files are quite obtuse and tricky to wrap your head around at first so I've tried to add comments into the XSLT files and add a little bit of supporting documentation.

Regards
Keith

14-Oct-15 16:16:02
Looking to improve how we use EMu to manage location stocktakes

We manage this through the Condition tab in the Catalogue module. When an item is checked just for stocktake purposes it gets its condition recorded as 'sighted' with a date. If it is checked for a condition check or conservation work the appropriate condition is recorded along with a date.

A report of everything with a condition check after the last stocktake returns everything sighted since that date.

In modules with a large number of tabs it can be difficult to spot that a piece of data has been entered in a rarely used field.

This can result in an attachment being made in error. It also makes it time-consuming to check if an existing record meets your needs as as you have to check all the tabs to make sure there isn't a value recorded that doesn't apply to your new record.

It would be great if, in record view, some kind of indicator appeared in the tabs names at the bottom of the window when the tab is empty. Then its absence would indicate that at least one field in that tab does have a value. We're not thinking anything elaborate, an asterisk or suchlike would do the trick.

16-Apr-15 09:35:49
Any examples of how this has been done?

Hi Ellen

I'm possibly a few months late noticing this, but if you are still looking for details of peoples' Collection Events incorporating Sites modules if you email me at keith.maguire@samuseum.sa.gov.au I can send through images and descriptions of our module and practices

Regards
Keith

Hi Melissa

If I've correctly understood what you're asking you can do it like this:

The thing to do is check that the field in the second row of the table has a value. 

Enter all the other bits of your search, say prefix or department or whatever to narrow it down.

Get the back-end name for the field in the table that you are interested in (Using the question-mark arrow or F1).

Choose "Show Search" from the File menu

From the search that appears there
1. change "where true and" to "where"
2. add "and exists FIELDBACKENDNAME{2}" to the end

FIELDBACKENDNAME{2} means FIELDBACKENDNAME on the second row of the table.

To show an example: if I want to search for all our Ichthyology records that have two "other numbers" recorded I fill in "Ichthyology" in the collection field and select "Show Search". That gives me:

select all
from ecatalogue
where true and
(
    exists
    (
        CatCollectionName_tab
        where
        (
            CatCollectionName contains '"Ichthyology"'
        )
    )
)

I then remove "true and" and add on the appropriate field name with {2} after it to the end, giving

select all
from ecatalogue
where
(
    exists
    (
        CatCollectionName_tab
        where
        (
            CatCollectionName contains '"Ichthyology"'
        )
    )
)
and exists CatOtherNumbersSource_tab{2}

Which gets up the results I was after

Emu really doesn't like these types of searches so after your results are returned you need to update the amount of records to an accurate result. You do that by, in list view, going to the last entry in the results and then back to the first. This will allow EMu to get an accurate count of the records.

I hope that does what you need

Regards
Keith

22-Aug-14 10:23:40
Category: Using EMu

Hi

I've found this very useful for cleaning up the cruft that has accumulated in our database over many years, migrations and imports. I'm sharing it in the hope that others may find it useful too.

This comes from a support request where Helena in the Melbourne office gave me the TexQL command for a particular search and suggested it could also be run as a shell script. I've just generalised the search to save typing out the column names repeatedly.

The actual search used is:

Code:


select count(record), $query
from
(
        nest
        (
                select $query from $TABLE
        )
        on $query
        forming record
)
where count(record) > $MAXNOTRETURNED;

Everything else is error-checking and building the variables

It allows you to, say, look in the Parties module and return all the duplicates occurrences of SummaryData, NamFirst and NamLast.

It doesn't tell you what records are duplicates, it shows you what values are duplicated and you can then do a search using those values to get them up and have a check-see if you've got good candidates for merging.

The script is used like this (in a shell on the EMu server):
duplicates.sh -t [the name of the module] -n [the minimum number of duplications that should be reported(optional)] [field1] [field2] [...] [fieldn]

To illustrate, this is an excerpt checking to see the occurrences where the Name of an organisation, its address and the summary data occur more than three times. The results are returned with the number of duplications first and then the column values in the order they were entered.

Code:


$duplicates.sh -t eparties -n 3 NamOrganisation AddPhysStreet SummaryData
columns searched: NamOrganisation AddPhysStreet SummaryData (3 columns)

(4,'CSIRO',NULL,'CSIRO')
(4,'Elder Expedition',NULL,'Elder Expedition')
(4,'Fisheries Officer',NULL,'Fisheries Officer')

I added the -n option to prevent getting disheartened at the amount of duplicates. It allows you to work on the worst offenders first by putting in a high number and then when you've sorted them out put in a lower number and so on.

I use the script in combination with a very full list view and lots of clicking through tabs in records.

I can't overemphasise that it is up to you to confirm that records which appear to be duplicates really are!

I assume no responsibility and make no guarantees, use at your own risk, read through the script and verify it does what I've said it does etc. but I do hope it's of use to some of you. It's attached as duplicates.sh, you'll need to remove the first line -added so the forum uploader wouldn't reject it- and change the permissions as appropriate

Regards
Keith

12-May-14 10:12:34
Working on color-coded Page Views but having xslt trouble
Category: Using EMu

No worries Kate, glad to be able to help and delighted that it worked!
Keith

26-Mar-14 15:41:14
Working on color-coded Page Views but having xslt trouble
Category: Using EMu

Hi Kate

Ah, looking at your example HTML showed how the example I was using wouldn't work - I've been enclosing everything in DIVs when working on my own page view so I was overly attached to that and it messes up the tables. I was also changing big lumps of the page instead of just the particular fields

The strategy you're working with totally works, but may end up being a bit laborious as you add tests for all the options.

What appears to work is again keeping the colours in the css named by department or event type. But using the XSL to change the class of the td to include the value exported from EMu.

So whereas normally the values tds in the table have the class "atomvalue", they get changed to instead read for example "Technology atomvalue"

I added a test to make sure that this is only done for the two fields of interest, this stops any values that appear in the other fields and coincidentally match the css classes from turning the table technicolor. There would be a reasonable chance of that happening, if the value was a sentence such as in the description each word would be added as a class to the td.

The attached file works with the xmldataFIX.xml file. You could get it to work with xmldataOld.xml if you needed to I reckon by tweaking line 324 - the xsl:when bit

(I say "it works" - I mean it worked for me when I tested it in Firefox, hopefully it also works for you in EMu)

As before, do let me know if it doesn't work or if it's nearly-but-not-quite-there

Regards
Keith

xsl-for-page-view.xsl

19-Mar-14 13:21:09
Working on color-coded Page Views but having xslt trouble
Category: Using EMu

Hi Kate


Hmm... our SecDepartment_tab is a table too, which then contains a tuple which contains an atom called SecDepartment. So in theory it should work... Maybe I forgot to include something in the original typing out of the answer.

Would you be able to attach a copy of the xml produced for the page view? Using that I (hopefully) should be able to get it working for you

Assuming you're using Windows, run the page view XSL report as a report for a single entry. You can find the xml for the report in the folder
C:\Users\YOUR_LOGIN_NAME\AppData\Local\Temp
You may have to set your computer to show hidden files - when searching choose Tools > Folder Options, then in the View tab select "Show hidden files, folders and drives"

The file you're after will be called xmldata.xml and will be one of the most recent ones created so sort by date modified and look for that file.

Regards
Keith

07-Mar-14 14:10:54
Working on color-coded Page Views but having xslt trouble
Category: Using EMu

Hi Kate

I think the thing to do is associate the color with some new classes in the CSS which are named after your collections. Then use the xsl to create a DIV with the class that matches the colour you want.

The forum wouldn't accept multiple xml code blocks to explain and demonstrate this so I've attached a text file below with contains a much longer and more detailed explanation.

Please do let me know if any of the explanation doesn't make sense, or if it does make sense but doesn't work as expected

Regards
Keith

(South Australian Museum)
conditional-page-colours.txt

  • Index
  • » Users
  • » keithmaguire
  • » Profile

Board Info

Board Stats
 
Total Topics:
601
Total Polls:
0
Total Posts:
1362
User Info
 
Total Users:
827
Newest User:
Marcus Swann (Axiell Melbourne)
Members Online:
2
Guests Online:
152