Ticket #1028: batchimport-2.txt

File batchimport-2.txt, 12.5 KB (added by Jari Häkkinen, 14 years ago)
2Generic item importer
5This is a description of an import plug-in that can be used to import almost any
6kind of item from a tab-separated text file (or compatible). The plug-in should
7be able to create new items or update existing items. In both cases it should be
8able to set values for:
10 * Simple properties. Eg. string values, numeric values, dates, etc.
11 * Single-item references: Eg. protocol, label, software, owner, etc.
12 * Multi-item references: Eg. the labeled extracts of a hybridization,
13   pooled samples, etc. In some cases a multi-item reference is bundled
14   with simple values. Eg. used quantity of a source biomaterial, the array
15   index a labeled extract is used on, etc. Multi-item references are never
16   removed by the importer, only added or updated. Removing an item from a
17   multi-item reference is a manual procedure to be done using the web
18   interface.
20The importer should not be able to set values for annotations since this is handled
21by the already existing annotation importer plug-in. The annotation importer
22and item importer should have similar behaviour and functionality to minimize
23the learning cost for users.
25Key features
28The item importer is only expected to work on a single type of item at each use
29and should read data from a single file.
31The importer should be able to work in 'dry-run' mode. Eg. everything is performed
32as if a real import is taking place, but the work (transaction) is not committed
33to the database.
35The importer should implement user-controlled error handling. Summary results,
36eg. number items importer, number of failed items, etc. should be kept track of
37and reported as the job status message. More details about the error handling
38options can be found later in this document.
40The plug-in should support logging more detailed error message. Eg. reasons for
41failed items, line numbers, etc. to a separate file (in the BASE file system).
42[IMPLEMENTATION NOTE] This file needs to be handled in a separate transaction,
43otherwise a complete failure will erase the file as well when the transaction
44is rolled back.
46The plug-in should be configurable and be able to store file parsing settings,
47including column mappings, etc. as plug-in configurations.
49File format
52The input file should be organised into columns separated by a specified character.
53Eg. tab, comma, etc. Fixed-width columns are not supported. Eg. the file is a
54file that can be parsed with the FlatFileParser class.
56The first data line contains the column headers which defines the contents of each
57column. The column headers can be mapped to item properties at each use of the
58plug-in or by saving predefined settings as a plug-in configuration. This also
59includes separator character and other information that is needed to parse the
60file. Saved configurations should implement auto-detection functionality.
62Data for a single item may be split onto multiple lines. The first line contains
63simple properties and single-item references, and the first multi-item reference.
64If there are more multi-item references they should be on the following lines with
65empty values in all other columns, except for the column holding the item
66identifier, which must have the same value on all lines. If the following
67lines contains other data, this should be ignored, or it may be considered an
68error condition. It may be caused by giving two items the same name by accident.
70When reading data for an item the plug-in needs to know if it should create a new
71item or update an existing item. First, we need to know the method for identifying
72items. Depending on the item type there are two or three options:
74 * Using the internal 'id'. This is always unique.
75 * Using the 'name'. This may or may not be unique.
76 * Some items have an 'externalId'. This may or may not be unique.
78The plug-in should ask the user which method to use. The user must also tell the
79plug-in among which items it should look for an item with a given 'name' or
80'externalID'. There are four options, and the user may select one or several
81of them:
83 * Owned by the logged in user
84 * Shared to the logged in user
85 * In the current project
86 * Owned by other users (only available if the logged in user has enough
87   permissions, eg. generic read permission for the item type)
89If the 'id' method is used, the above options are not used. When the plug-in
90is looking for an item there are three possible outcomes.
92 * No item is found. This can be handled in different ways:
93   - An error condition which aborts the plug-in
94   - The line is ignored
95   - A new item is created
96 * One item is found. This is the item that is going to be updated.
97 * More than one item is found. This can be handled in different ways:
98   - An error condition which aborts the plug-in
99   - The line is ignored
102Parsing the data.
105Simple properties
107We need to know the data type of each property as a Type object.
108The string values can then be parsed with Type.parseString().
110Single-item references
113This is either the 'id', 'name' or 'externalId' of the item. The plug-in should
114support those cases by a single column mapping and an option to select which
115method to use.
118This creates two input parameters for each columns which may be too many...
119Alternative options are:
121 * A global option for all item references. This doesn't give the user
122   any chance to use different method for different items.
123 * A global option with an 'auto' alternative that uses the 'id' method
124   for numeric values, otherwise first the 'name' and if no item is
125   found the 'externalId'.
127[JH: I think one of the alternatives would be better than a plethoria
128of parameters ... I like the auto idea but will it create though
129conditions on names used on items?]
133When looking for item references the plug-in doesn't have to use the same
134setting for 'owned by', 'shared to', etc as when looking for the main items.
135In fact, this is not desired since many of those items are owned by the root
136user or a system administrator. Eg. labels, software, hardware, etc. I don't
137think it is practical to have another option for selecting this for each type
138of item reference, or....???
139well... maybe the code can be prepared for it, but the default should be to
140look among all items the user has access to...
142[JH: I think we should settle for all items the user has access to and
143ignore the setting of 'owned by', 'shared to', etc. As I understand it
144then is that the filtering is only used to find items to be updated
145(or created).]
147In any case, there are three outcomes:
149 * No item is found.
150   - This can be an error condition that aborts the plug-in. If it is a
151     required property this will always happen.
152   - The link is ignored. No call is made to setABC() method. Note that this
153     case is different from having an empty column in which case
154     setABC(null) would be called.
155 * One item is found. This is the item we link to.
156 * Multiple items are found.
157   - This can be an error condition that aborts the plug-in.
158   - The link is ignored.
161Multi-item references
164This should work in the same way as single item references.
167Using the plug-in
170Configuring the plug-in is done with the usual wizard. There will be plenty of
171parameters so it is probably a good idea to use a multi-step wizard. This may have
172to be tried out by actual users before we make any final decisions.
174Step 1
177The user selects a file and enter values for the regular expressions and other
178options for parsing the file. Column mappings are also specifiec in this
179step. The "Test with file" function should be supported. Parameters that are
182 * A file to parse
183 * Data header: Regular expression for finding the start of data
184 * Data splitter: Regular expression that splits data lines into columns
185 * Remove quotes: boolean option that removes "quotes" around values
186 * Ignore: Regular expression that matches lines to be ignored
187 * Data footer: Regular expression for finding the end of data
188 * Min/max data columns: The number of columns a data line must have, otherwise
189   it is ignored
190 * Character set: The characeter set (eg. iso-8859-1, utf-8, etc.) used in the file
191 * Decimal separator: if dot or comma is used as a decimal separator for numeric values
193The above parameter are the same as those found in many of the existing import
196Since each type of item has different properties, colum mapping parameters vary
197from case to to case. Column mapping parameters may need to be divided into
198subsections for clarity.
200For the ID property we need one column mapping parameter, one enum parameter
201to select which identification method to use and one enum parameter for selecting
202which items to search (multi-choice).
204For simple properties we need a single column mapping parameter.
206For single-item references we need one column mapping parameter and one enum
207parameter to select which identification method to use.
209All options (except the file to parse) in this step should also be available
210to store as a plug-in configuration. But in this case we need a 'step 0' which
211asks us about which type of item the configuration is to be used with. Otherwise
212we don't know which properties, etc. to provide column mappings for.
214[JH: Isn't is possible to deduce the item type from the context in the
215GUI? I mean, the user selects to use the batchimporter from some
216context and the maybe this information could be passed on to the core.]
218Step 2
221This step is mainly about error handling options. Default values are marked with
224 * Default error handling: *fail*, skip line
225 * Item not found: fail, *create*, skip line
226 * Multiple items found: *fail*, skip line
227 * Referenced item not found: *fail*, ignore, skip line
228 * Multiple referenced items found: *fail*, ignore, skip line
229 * Missing a required property: *fail*, ignore if updating, skip line
230 * String too long: *fail*, crop, ignore
231 * Invalid numeric value: *fail*, ignore
232 * Numeric value out of range: *fail*, ignore
233 * A log file for detailed error messages
235If there are multi-item references the 'skip line' option above means that we
236should skip all lines that are related to the same item.
238Implementation details
241We need some kind of basic, generic functionality that is handling the file parsing,
242property mapping, item lookup, error handling, logging, etc.
244We also need functionality that is specific for each type of item the plug-in
245should support. We need to know which properties that exists on the items. For each
246property we need to know the data type or if the property is a single-item or '
247multi-item reference. We need factory methods for creating new items etc.
249If possible, it should also be relatively easy to extend the item importer with
250support for other item types in the future.
252One possible approach is to use an abstract base class for the common functionality.
253This class defines some abstract methods that must be implemented by subclasses
254where each subclass handles a single type of item. This approach makes it relatively
255easy to add support for other item types just by creating a new subclass. This
256approach creates a separate plug-in for each item type. Eg. SampleImporter,
257ExtractImporter, etc.
260Item specific functionality that is needed
263 * List of properties that can be imported. For each property we must know:
264   - if it is a simple value, a single-item reference or a multi-item reference
265   - the data type, eg. string, float, int, SAMPLE, PROTOCOL, etc.
266   - if the property is required or not
267   - name and description and other details for better user experience
268   This may be implemented as a 'Property' interface, with concrete implementations
269   for each type. The implementation should also know how to set a value on items.
270   Eg. NameProperty.setValue(item, theName) --> item.setName(theName). Note! some
271   properties may require multiple parameters. Eg. setting the source item and
272   used quantity: BioMaterialEvent.addSource(source, usedQuantity).
274 * A factory method for creating new items. Some items can be created without
275   any parameters (eg. Sample.getNew()), some requires one or more parameters
276   (eg. LabeledExtract.getNew(Label).
278 * Find an item. This functionality is alredy implemented by the annotation importer,
279   but it is not very flexible since it uses reflection to find the getQuery() method.
280   The annotation importer only works with items were the getQuery() method doesn't
281   require any parameters.
283Supported item types