do not duplicate the true/false indicators for the metadata vocab/value matrix
use recordNum instead of studentId for transposed data
use "recordNum" rather than "studentId"
omit studentId values from normal resultsetstrip the html tags from the qti elements for both return formats
disable output escaping for CDATA html elements
split vocab/value into two columns (can easily switch back to single column)
question choices and metadata selection controlled by a counter - can easily increase the number returned...note: probably need to make this a parameter for user or be intelligent about figuring what the widest question is (ie: does it have 5,10,20! metadata entries?)
include short answer prompts in the resultset view
include prompt for both multiple choice and short answer questions when downloading
omit the demographic data - this is not well-defined for the prototype
omit duplicate values in the for the standard metadata header section (columns for each question)
omit the metadata vocab/values from the normal header section since they will be used in the "matrix-style" header section
add vocab name/value matrix above data rows.currently there is only a single column for the concatenated value of the vocab name + "/" + vocab value. it's so much easier!
reduce the extra whitespace on the transpose
hide the "transpose" checkbox except when on the data cart page
use name attributes for selection entity-attributes.note: this is still a static version of which question-level metadata is being included in the resultset
only "correct" non-null values
include qPrompt and qChoiceneed to quote the csv so that commas in the question text do not throw it off
Now the metadata attributes show as rows above the data.very close!todo: special handling for thesaurus (ie NBII/Bloom) metadata fields
pivot constrained to a List of specific attributes: studentId (record number), score, response.This is the "lower half" of the return set (data).now: on to the metadata header rows
Now doing the dataquery/download in the FIRST skin's jsp rather than embedded in the Metacat servlet.TODO: implement the very-custom data return format agreed upon by the FIRST team
include pivot options for dataquery
Use only one eml common stylesheet directory for all post 2.0.0 beta versions of eml. Point all eml versions to that directory in skin xml files.
include NSF grant number
use "prompt" xpath that matches current QTI serialization
handle case when there is a single question on the assessment
correct spelling of "Assessment"
change "blah blah" to something more respectable
add skin properties files so that first can be enabled as the default when configuring metacat in the admin interface.note: not checking in the modifcation to metacat.properties file - that will remain a deployment-specific change and not included in the standard distribution
handle mixture of datapackages: some with demographic data, some without.takes advantage of the [new] StaticSelectionItem feature in the datamanager
include demographic data if present in the datapackageTODO: handle a mixture of datapackages when performing the union (i.e. cases when one package as demographic data and the other does not)http://bugzilla.ecoinformatics.org/show_bug.cgi?id=3519
assessment title and type are now elements rather than attributes in edmlhttp://bugzilla.ecoinformatics.org/show_bug.cgi?id=3345
added "generic" search term picklisthttp://bugzilla.ecoinformatics.org/show_bug.cgi?id=3520
declare questionIds array for each documenthttp://bugzilla.ecoinformatics.org/show_bug.cgi?id=3620
add "Assessment Database" to the header bannerremove "II" from the banner
call it FIRST without the numerals
call it the "Data Cart" instead of just the "Cart"
add institution logos to footer
use solid green for chalkboard borders
pull out the IE/FF differences into common method that returns the document object for an iFrame (works with IE now!)
try IE work around for iframe access
update the status of search items WRT the cart when adding/removing from within the search results view (when you add an assessment, the icon changes to the "remove" icon)
swap out new add/remove cart icons, ensure left and right chalkboard borders continue all the way down the page with the content
make metadata field selections only from within the cart
add page header
logout redirects to the login page (just needed to use _top)
redesign the look and feel - save point.-cart editing is done by clicking individual assessments rather than doing them in batches.-cart status is shown in the search results (but still need a refresh for after adding cart items from search...)
Redirects to the index.jsp page.
improve clarity of the available cart actions by:-hiding/showing the logical options for a given screen (only allowed to remove when viewing the cart)-refresh when cart changes - more positive feedback that pushing buttons actually did something.
make "add to cart" more obvious (with icon, too!)
add login feature for first. integrates with existing metacat-skin-based login approach.only show cart for authenticated users.
add logout link.do not search for cart documents if the cart is empty.
add spinner for long searches
highlight the cart contents on reload
use regular hyperlink to viewing question (item) details - nested form elements were causing grief after being included with ajax calls.
refresh the entire cart view when modifications are made.add icons to many of the control features (download, remove...)
show selected metadata fields in the cart view and allow them to be removed from said view
added Scriptaculous effects to many of the ajax calls.neat-o!
can now select and save metadata fields from a document and use them when merging across documents in the "cart"
use full width of the screen
add "cart" information (won't work for others until i commit the backing classes in meetacat servlet...but need to track the changes for this now).
[optionally] view assessment details from within results set
parameterize the div used for ajax results
correct targets for links in iframe
rearrange pages, search features, etc...
search within specific fields - with option for matching any vs. all
search method for retrieving specific documents (matching docid)
added place holders for structured search fields
make as-you-type search the default.remove assessment vs. question search checkboxes...confusing.
refactor search methods into common search.js file.start on making ajax-based search from the index page.state of things: now merging across assessments and including question-level metadata
use the generalized (and now built-in) data query mechanism for downloading "merged" data.
refactor into reusable methods such that queries can be constructed from different form interfaces.
fix the optional "include questions" checkbox so that it actually toggles the option.
added support for merging response data with question-level metadata
only include checked metadata
detokenize
add dataquery construction via js.used in the first XSL-housed style sheets
switch to jsp
de-tokenize (finally)
render the pages using the common settings
switch to jsp (first of many steps...)
include changes to the first/metacat integration projectmerging/union across multiple datapackages now supported
save point before extending to allow merging across assessments
use the first (0) column as the index for questionId when selecting only a subset of the assessment data
add handling for qti metadata in resultset view
adjust xpaths to match changes to edml/qti output by FirstMorpho
Changes to remove ANT tokens for the turnkey install project
add support for selecting records for only one question from a set of one or more assessment response data files
no padding on the table header lefthand side
render multiple mattext elements if they exist...
return immediately if there is no resultset to displayuse getString() to display resultset for each column
add CSV download as an option for retrieving datasets.NOTE: depends on opencsv jar in the /first project...
include both assessments and questions in the results of a search (driven by option in search form)
add ability to retrieve response data using datamanager lib (note that this requires both the datamanager.jar and firstMetacatIntegration.jar from /eml and /first modules respectively...)yes, this is more ajax...and still mostly Proof Of Concept-ish
organize css references
less emphasis on emphasis class (reduced font size)
included assessmentId in search results.TODO: search for both assessment and question doctypes in same query and have the result page display them intelligently (segregate by doc type).
organize the bulk of the stylesheets in their own xsl directory.note that these xsl files are being maintained in the "first" cvs module and should be copied from there.