status
stringclasses 1
value | repo_name
stringlengths 9
24
| repo_url
stringlengths 28
43
| issue_id
int64 1
104k
| updated_files
stringlengths 8
1.76k
| title
stringlengths 4
369
| body
stringlengths 0
254k
⌀ | issue_url
stringlengths 37
56
| pull_url
stringlengths 37
54
| before_fix_sha
stringlengths 40
40
| after_fix_sha
stringlengths 40
40
| report_datetime
timestamp[ns, tz=UTC] | language
stringclasses 5
values | commit_datetime
timestamp[us, tz=UTC] |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,158 |
["main/tests/cypress/cypress/e2e/create-project/preview_project.cy.js", "main/tests/cypress/cypress/support/commands.js"]
|
Flaky Cypress test in cypress/e2e/create-project/preview_project.cy.js "Tests disabling of automatic preview"
|
The test "Tests disabling of automatic preview" in `cypress/e2e/create-project/preview_project.cy.js` seems to be flaky as it fails in two PRs with unrelated changes (#6108 and #6156):
* https://github.com/OpenRefine/OpenRefine/actions/runs/6878198239/job/18707359425
* https://github.com/OpenRefine/OpenRefine/actions/runs/6878198239/job/18707359425
|
https://github.com/OpenRefine/OpenRefine/issues/6158
|
https://github.com/OpenRefine/OpenRefine/pull/6162
|
4583508a1650a8acbb8d70881d6a7acd59a1a312
|
0c1f98484e4f94e67b4b019645c150f69df73047
| 2023-11-15T14:25:37Z |
java
| 2023-11-16T10:09:17Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,154 |
["main/tests/cypress/cypress/e2e/project/grid/column/reconcile/actions/see-more-and-see-less.cy.js", "main/webapp/modules/core/langs/translation-en.json", "main/webapp/modules/core/scripts/views/data-table/cell-renderers/recon-renderer.js", "main/webapp/modules/core/styles/views/data-table-view.css"]
|
Only display up to three reconciliation candidates by default
|
<!-- Please provide a clear and concise description of your problem or unsatisfied needs. Ex. I'm always frustrated when [...] or, It would be easier if OpenRefine did [...]. This comment can be deleted if desired, but it will be hidden in your final submission. Please make sure that your new text is outside the enclosing angle brackets. -->
### Problem Description
During user interviews conducted in September on the redesign of the reconciliation dialog feature, a common pain point identified by reconciliation was the challenge of dealing with a large number of results displayed after selecting a reconciliation service. The feedback highlighted issues with both the accuracy and the overwhelming nature of the displayed results. Users expressed the need for improved visibility and control as they often had to sort through numerous irrelevant results and this also causes a distatsteful UI by expanding cells thereby creating too much whitespace.
### Proposed solution
To address the identified problem, I propose an enhancement to the reconciliation feature in OpenRefine. The suggested solution involves introducing a more user-friendly way for managing the display of reconciliation candidates.
- We start of displaying limited number of resuts, say limited to 2 by defaut.
- A "See More" button: Users can reveal additional candidates beyond the initially specified limit.
- A "See less" buton (link) to reverse the previous action and return back to the default display of 2 results.
This solution aims to provide users with greater control over the display of results, making the reconciliation process more efficient and user-friendly.
### Alternatives considered
An alternative is a straightforward "truncate results" option added to the dropdwon options of each reconciled cell. On click, a pop-up containing an input field appears where users can input the number of results they want displayed.
### Additional context
The proposed solution is a response to direct user feedback about the challenges faced when reconciling data, particularly in scenarios where a large number of candidates are presented. By allowing users to control the visibility of results, the proposed feature seeks to enhance the overall user experience, improving accuracy, and reducing the time and effort required for reconciliation tasks.
**Proposed defaukt display of results**
When user clicks on "See more"
**Alternative proposed**
|
https://github.com/OpenRefine/OpenRefine/issues/6154
|
https://github.com/OpenRefine/OpenRefine/pull/6244
|
5c4e76b548b3658e975414d848870ccaad5495f1
|
3bcc69df5e80916d927eeb682c42ac2eb9b38018
| 2023-11-13T12:12:50Z |
java
| 2024-01-13T09:43:58Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,143 |
["main/webapp/modules/core/scripts/dialogs/add-column-by-reconciliation.js"]
|
Missing space in "add column based on reconciled values" dialog
|
The dialog for "add column based on reconciled values" is missing a space between the "remove" and "configure" options.
### To Reproduce
Steps to reproduce the behavior:
1. First, reconcile a column
2. open the "add column based on reconciled values" dialog
3. Add a property
### Current Results
### Versions<!-- (please complete the following information)-->
- OpenRefine: 3.8-SNAPSHOT
|
https://github.com/OpenRefine/OpenRefine/issues/6143
|
https://github.com/OpenRefine/OpenRefine/pull/6151
|
0c1f98484e4f94e67b4b019645c150f69df73047
|
4d77409fe340fee73645c3084928c60ab6d90c39
| 2023-11-06T13:59:18Z |
java
| 2023-11-16T10:53:40Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,137 |
["main/src/com/google/refine/util/HttpClient.java", "main/tests/server/src/com/google/refine/operations/column/ColumnAdditionByFetchingURLsOperationTests.java"]
|
Add Column by Fetching URL silently fails on malformed URLs
|
If an invalid URL is used for the Add Column by Fetching URLs operation, the operation silently results in a null value instead of an error, making it much more difficult for the user to figure out what the problem is. This was recently on the forum as a cause of confusion for a user: https://forum.openrefine.org/t/fetching-urls-with-arabic-values/1036/1
This behavior was checked by a test, so I coded the new HTTP support to mimic it, but I think the test is wrong and we should fix both it and the code so that URL/URI errors get reported to the user in an intelligible fashion.
### To Reproduce
Steps to reproduce the behavior:
1. Create a project with a cell containing the text: https://transliterate.qcri.org/ar2en/احمد احمد
2. Select Add a Column by Fetching URLs
3. For "On error:" select "store error"
### Current Results
The operation silently fails with no error stored in the cell.
### Expected Behavior
The cell value should be set to an `EvalError` with the appropriate error code/message.
### Versions<!-- (please complete the following information)-->
- OpenRefine: 3.8-snapshot (ie current HEAD)
|
https://github.com/OpenRefine/OpenRefine/issues/6137
|
https://github.com/OpenRefine/OpenRefine/pull/6141
|
a64546cf2b6ba9780a3dadbbff4ea829d6db49a5
|
8dcf89da2eaa63a5c330057ca1484594c546c88a
| 2023-11-05T00:13:51Z |
java
| 2023-11-06T10:48:33Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,132 |
["main/tests/server/src/com/google/refine/commands/recon/GuessTypesOfColumnCommandTests.java", "main/tests/server/src/com/google/refine/model/recon/StandardReconConfigTests.java", "main/tests/server/src/com/google/refine/util/TestUtils.java"]
|
Fix Flaky Tests testGuessTypes and reconTest
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
## Description of the bug
The test cases [com.google.refine.commands.recon.GuessTypesOfColumnCommandTests.testGuessTypes](https://github.com/OpenRefine/OpenRefine/blob/master/main/tests/server/src/com/google/refine/commands/recon/GuessTypesOfColumnCommandTests.java#L64-L154) and [com.google.refine.model.recon.StandardReconConfigTests.reconTest](https://github.com/ThugJudy/OpenRefine/blob/3cb2bb3c33106a3908077cd9130d14d1765d94a1/main/tests/server/src/com/google/refine/model/recon/StandardReconConfigTests.java#L281-L389)
fail due to the non-deterministic behavior of the function ParsingUtilities.defaultWriter.writeValueAsString which does not guarantee the ordering of elements in the JSON created.
https://github.com/OpenRefine/OpenRefine/blob/master/main/src/com/google/refine/commands/recon/GuessTypesOfColumnCommand.java#L171
The JSON response is asserted as String and the order of the JSON elements is not maintained, thereby causing flakiness in the Tests that were mentioned above.
I found and confirmed the flaky behavior using an open-source research tool [NonDex](https://github.com/TestingResearchIllinois/NonDex), which shuffles implementations of nondeterminism operations.
### To Reproduce
The following command can be used to reproduce assertion failures and verify the fix:
```
mvn -pl main edu.illinois:nondex-maven-plugin:2.1.1:nondex -Dtest=com.google.refine.commands.recon.GuessTypesOfColumnCommandTests#testGuessTypes
mvn -pl main edu.illinois:nondex-maven-plugin:2.1.1:nondex -Dtest=com.google.refine.model.recon.StandardReconConfigTests#reconTest
```
### Current Results
The test fails upon running the nondex command due to the shuffling of the JSON elements inside the query.
### Expected Behavior
The test should pass consistently regardless of the shuffling of the elements.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: MacOS Venture Version 13.4.1 (22F82)cOs
- JRE or JDK Version: "11.0.20"
|
https://github.com/OpenRefine/OpenRefine/issues/6132
|
https://github.com/OpenRefine/OpenRefine/pull/6133
|
a195e4e99a41f77ffd449c8c8df21059b66a355e
|
ac8b9fb17ea40d67227f63d96d183c385e5f0821
| 2023-11-02T18:31:41Z |
java
| 2023-11-08T10:19:20Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,129 |
["main/src/com/google/refine/importers/TextFormatGuesser.java", "main/src/com/google/refine/importing/ImportingUtilities.java", "main/tests/data/persons.csv", "main/tests/data/persons.csv.bz2", "main/tests/server/src/com/google/refine/importing/ImportingUtilitiesTests.java"]
|
bzip2 import broken
|
It turns out that Apache Commons Compress was **not** a drop in replacement for Ant's implementation of bzip2 (see [COMPRESS-13](https://issues.apache.org/jira/browse/COMPRESS-13)) and we had no test coverage, so PR #2691 broke our bzip2 support on import.
_Originally posted by @tfmorris in https://github.com/OpenRefine/OpenRefine/issues/2691#issuecomment-1790022992_
|
https://github.com/OpenRefine/OpenRefine/issues/6129
|
https://github.com/OpenRefine/OpenRefine/pull/6130
|
1beeda249ceeb9cddcc4386efbdc156cb2b3e848
|
a195e4e99a41f77ffd449c8c8df21059b66a355e
| 2023-11-02T03:56:09Z |
java
| 2023-11-08T10:13:27Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,111 |
["main/src/com/google/refine/model/recon/StandardReconConfig.java", "main/src/com/google/refine/operations/recon/ReconOperation.java", "main/tests/server/src/com/google/refine/model/recon/StandardReconConfigTests.java"]
|
Improve error messages in reconciliation operation
|
The errors that are shown to the user when reconciliation fail are wordy and not very helpful to understand the source of the error.
Currently, the error message contains the JSON serialization of the entire batch. This was perhaps helpful when those error messages were written to the server logs, as they were necessary to figure out which part of the dataset was responsible for the error, but now that those errors are displayed in the grid, the context is already clear from the location of the error.
### Proposed solution
Review all the places where `recon.error` is set and make sure a helpful error message is set.
For instance:
* if we cannot connect to the reconciliation service, I would expect a message such as `Connecting to my-recon-service.com failed`
* if the reconciliation service returned an HTTP error, I would expect a messages such as `The reconciliation service returned HTTP error 403`
* if the reconciliation service returns an invalid JSON response, we could have something like `The reconciliation service returned an invalid response: expected ':' after field "queries"`
### Alternatives considered
When it is worth adding more information to the error, but that information can potentially take up a lot of space, then perhaps we could still use the server logs for that.
### Additional context
Follow up to #6006.
|
https://github.com/OpenRefine/OpenRefine/issues/6111
|
https://github.com/OpenRefine/OpenRefine/pull/6191
|
60d66e6b772bf757a7f5d8d5aaa0f4525a66465f
|
fa8829bdc10e034dcc119d337b573103ee10fd96
| 2023-10-23T10:55:32Z |
java
| 2023-12-15T06:59:29Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,110 |
["main/webapp/modules/core/scripts/views/data-table/cell-renderers/recon-renderer.js", "main/webapp/modules/core/styles/views/data-table-view.css"]
|
Cells with reconciliation errors should let the user match them manually
|
### To Reproduce
Steps to reproduce the behavior:
1. Reconcile a column via a recon service which returns errors (for instance, disable your internet connection during reconciliation)
### Current Results
The cells which store errors do not have a "Search for match" link like the ones which were successfully reconciled
### Expected Behavior
The "Search for match" link should be present and work as expected.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- OpenRefine: 3.8-SNAPSHOT
### Additional context
Follow-up to #6006
|
https://github.com/OpenRefine/OpenRefine/issues/6110
|
https://github.com/OpenRefine/OpenRefine/pull/6192
|
41b742851953d65462a990a75b875b37d7f95388
|
60d66e6b772bf757a7f5d8d5aaa0f4525a66465f
| 2023-10-23T10:46:14Z |
java
| 2023-12-15T06:57:48Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,099 |
["main/tests/server/src/com/google/refine/commands/recon/GuessTypesOfColumnCommandTests.java", "main/tests/server/src/com/google/refine/model/recon/StandardReconConfigTests.java", "main/tests/server/src/com/google/refine/operations/column/ColumnAdditionByFetchingURLsOperationTests.java"]
|
Windows CI job hangs for 6 hours
|
I suspect this was caused by a line ending issue which caused a test failure, but the job should exit on failure rather than hanging for 6 hours.
The failing job is https://github.com/OpenRefine/OpenRefine/actions/runs/6514940672/job/17696752789
and the errors in the log are below:
```
02:10:27.975 [main] ERROR FileProjectManager - Failed to load workspace from any attempted alternatives.
{ {
"columnGroups" : [ ], "columnGroups" : [ ],
"columns" : [ { "columns" : [ {
"cellIndex" : 0, "cellIndex" : 0,
"name" : "a", "name" : "a",
"originalName" : "a" "originalName" : "a"
}, { }, {
"cellIndex" : 1, "cellIndex" : 1,
"name" : "b", "name" : "b\ne",
"originalName" : "b" "originalName" : "b\ne"
} ], }, {
"keyCellIndex" : 0, "cellIndex" : 2,
"keyColumnName" : "a" "name" : "e",
} "originalName" : "e"
} ],
"keyCellIndex" : 0,
"keyColumnName" : "a"
}
java.lang.Exception: No column named director
at com.google.refine.operations.recon.ReconOperation$ReconProcess.populateEntries(ReconOperation.java:204)
at com.google.refine.operations.recon.ReconOperation$ReconProcess.run(ReconOperation.java:239)
at java.base/java.lang.Thread.run(Thread.java:1623)
Error: Exception in thread "Thread-2" java.lang.NullPointerException: Cannot invoke "java.util.List.iterator()" because "this._entries" is null
at com.google.refine.operations.recon.ReconOperation$ReconProcess.run(ReconOperation.java:247)
at java.base/java.lang.Thread.run(Thread.java:1623)
Terminate batch job (Y/N)?
Error: The operation was canceled.
```
The timestamps on the last few lines show the 6 hour hang
```
Sat, 14 Oct 2023 02:10:28 GMT Error: Exception in thread "Thread-2" java.lang.NullPointerException: Cannot invoke "java.util.List.iterator()" because "this._entries" is null
Sat, 14 Oct 2023 02:10:28 GMT at com.google.refine.operations.recon.ReconOperation$ReconProcess.run(ReconOperation.java:247)
Sat, 14 Oct 2023 02:10:28 GMT at java.base/java.lang.Thread.run(Thread.java:1623)
Sat, 14 Oct 2023 08:06:59 GMT Terminate batch job (Y/N)?
Sat, 14 Oct 2023 08:07:00 GMT Error: The operation was canceled.
```
There are also some earlier errors in the log, which may possibly be contributing, such as:
```
java.lang.IndexOutOfBoundsException: Index 1 out of bounds for length 0
at java.base/jdk.internal.util.Preconditions.outOfBounds(Preconditions.java:100)
at java.base/jdk.internal.util.Preconditions.outOfBoundsCheckIndex(Preconditions.java:106)
at java.base/jdk.internal.util.Preconditions.checkIndex(Preconditions.java:302)
at java.base/java.util.Objects.checkIndex(Objects.java:385)
at java.base/java.util.ArrayList.get(ArrayList.java:427)
at com.google.refine.commands.cell.EditOneCellCommand$EditOneCellProcess.createHistoryEntry(EditOneCellCommand.java:168)
at com.google.refine.process.QuickHistoryEntryProcess.performImmediate(QuickHistoryEntryProcess.java:71)
at com.google.refine.process.ProcessManager.queueProcess(ProcessManager.java:83)
at com.google.refine.commands.cell.EditOneCellCommand.doPost(EditOneCellCommand.java:127)
at com.google.refine.commands.cell.EditOneCellCommandTests.testEditOneCell(EditOneCellCommandTests.java:61)
```
|
https://github.com/OpenRefine/OpenRefine/issues/6099
|
https://github.com/OpenRefine/OpenRefine/pull/6100
|
f207431e7d9036b08b216d622344dad4e51cc550
|
e8ff5bd3a9b4681d4eb6be92073729e62cbf8945
| 2023-10-14T16:33:02Z |
java
| 2023-10-16T16:26:24Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,087 |
["refine", "refine.bat"]
|
Add support for Java 21
|
We currently only test up to Java 17 and issue a warning on startup if the Java version is higher. Since the Java 21 LTS release is out now, it would be good to support that.
### Proposed solution
Add CI testing for at least Java 21 (not sure we need any intermediary versions).
### Alternatives considered
Add CI testing support for all major versions in the supported range.
|
https://github.com/OpenRefine/OpenRefine/issues/6087
|
https://github.com/OpenRefine/OpenRefine/pull/6379
|
db368627ad6fd4919c1787bfbaec1947d00dd101
|
9a70da0c6743ff188be9e5c198c748a30bae4c89
| 2023-10-04T21:32:15Z |
java
| 2024-02-21T05:08:27Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,075 |
["extensions/gdata/pom.xml", "main/webapp/WEB-INF/jetty-web.xml", "main/webapp/WEB-INF/web.xml", "pom.xml", "server/src/com/google/refine/Refine.java", "server/src/com/google/util/threads/ThreadPoolExecutorAdapter.java"]
|
Upgrade to Jetty 10 / 11
|
Community support for Jetty 9 [ended](https://www.eclipse.org/lists/jetty-announce/msg00168.html) June 1, 2022. I t will continue to receive security updates, but it makes sense to move forward to at least Jetty 10. Jetty 11 has an incompatible API due to package name changes, so it requires a Butterfly update so is blocked by https://github.com/OpenRefine/simile-butterfly/issues/105
I had assumed that we'd eventually have to go to Jetty 11 and did the work to update, but Jetty 12 was just recently released and it provides support for Servlet 4, 5, and 6 in same implementation, potentially minimizing upgrade work, but it is a) brand new and b) requires Java 17 which is much more aggressive than OpenRefine has traditionally been with Java upgrades.
### Proposed solution
Update to Jetty 10 and defer updating to Jetty 11 until other breaking changes are needed in Butterfly.
### Alternatives considered
We could potentially just live with Jetty 9.4 as long as security updates keep coming, hoping to hang on long enough make the transition to Jetty 12 in one jump, but it seems like we could get the rug pulled out on short notice.
### Additional context
<!-- Add any other context or screenshots about the feature request here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/6075
|
https://github.com/OpenRefine/OpenRefine/pull/6076
|
d11994c859351d62a8a039522db21d4b4a0744fa
|
b4154aab26b205889da7eaa8dd44b982e5143b19
| 2023-09-29T04:18:10Z |
java
| 2023-10-04T20:06:20Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,071 |
["main/webapp/modules/core/scripts/facets/list-facet.js"]
|
Missing space when using facets
|
Missing space between "name" and "count" when using facets:
https://github.com/OpenRefine/OpenRefine/blob/adb0e5f1d2a615e6cb310fc9a84a3c36ce1d03cc/main/webapp/modules/core/scripts/facets/list-facet.js#L181-L182
|
https://github.com/OpenRefine/OpenRefine/issues/6071
|
https://github.com/OpenRefine/OpenRefine/pull/6072
|
f76d08f9d73999fa20e9367c1b17d6f16ea931a0
|
de4a98d4faf80012f288a0b85947b433b75a3cdf
| 2023-09-27T11:47:00Z |
java
| 2023-09-28T18:10:14Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,064 |
["extensions/wikibase/module/scripts/dialogs/statement-configuration-dialog.js", "extensions/wikibase/src/org/openrefine/wikibase/schema/strategies/QualifiersStatementMerger.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/schema/strategies/QualifiersStatementMergerTests.java"]
|
Unable to add additional values to statements on existing Qids when uploading via OR schema
|
[LSEPPR BHL schema.txt](https://github.com/OpenRefine/OpenRefine/files/12712961/LSEPPR.BHL.schema.txt)
Using an schema in OpenRefine I am able to successfully add new statements to a set of existing Qids, but where a statement already exists and I want to add a second value - eg instance of, published in the second value is not added, though my reference for that value is added.
For example I experimented with [After Merkel: Germany from peace to war - Wikidata](https://www.wikidata.org/wiki/Q122664150). It already contains an instance of scholarly article statement. And I also want to add a value of academic journal article.
When my general schema didn't work I tried to isolate the issue by creating a schema simply to add the academic journal article value. But when I upload those changes to Wikidata, although OR indicates the update has happened there is no evidence of it on the history tab of the item - and the second value is still not added.
I understand from the OR forum that I am using the correct config settings
A full description of the issue and what has been suggested so far is here https://forum.openrefine.org/t/adding-additional-statements-to-wikidata-property/306/12 but I am still unable to resolve the issue.
I'm working in Windows 10, on Chrome, and OR 3.7.4
[vol 3 issue 1 LSEPPR for BHL upload.xls](https://github.com/OpenRefine/OpenRefine/files/12712921/vol.3.issue.1.LSEPPR.for.BHL.upload.xls)
I've attached spreadsheet of the data, and the OR schema.
Thanks
|
https://github.com/OpenRefine/OpenRefine/issues/6064
|
https://github.com/OpenRefine/OpenRefine/pull/6066
|
38e117d8dfacc9e93cc26aa8404b1134c3acb636
|
21aab5441051182e249644c33f66767bd0ffdee8
| 2023-09-25T08:33:26Z |
java
| 2023-10-09T14:12:31Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,063 |
["main/src/com/google/refine/commands/Command.java"]
|
Mangled Unicode characters in yellow message after matching using "search for match" dialog
|
After matching a value, there's a yellow message shown at the top of the page. If the match was done using the "Search for match" dialog, many Unicode characters are turned into "?".
### To Reproduce
Steps to reproduce the behavior:
1. Create a new project with the following lines:
* Māori
* Omaha–Ponca
* Võro
2. Select "Start reconciling" from the menu for "Column 1"
3. Select the Wikidata reconciliation service
4. Unselect "Auto-match candidates with high confidence"
5. Click "Start reconciling"
6. Click "Search for match" for one of the entries
7. Select the matching item from the dropdown
### Current Results
The yellow message shown after matching shows Māori as "M?ori" and "Omaha–Ponca" as "Omaha?Ponca", but displays "Võro" correctly.
All of the names are displayed correctly if you click on the tick to accept the match instead of using "Search for match".
### Expected Behavior
The Unicode characters should not turn into question marks.
### Screenshots
After using "Search for match":
After clicking on the tick:
(taken from a longer list of names, so the row numbers don't match)
### Versions<!-- (please complete the following information)-->
- Operating System: Ubuntu
- Browser Version: <!-- e.g. Chrome 19, Firefox 61, Safari, NOTE: OpenRefine does not support IE but works OK in most cases -->
- JRE or JDK Version: openjdk 17.0.8.1
- OpenRefine: Version 3.7.5 [a04fb5f]
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, you can share it selectively by email to developers who work on the issue -->
### Additional context
The non-ASCII characters in these three names are:
* U+0101 LATIN SMALL LETTER A WITH MACRON
* U+2013 EN DASH
* U+00F5 LATIN SMALL LETTER O WITH TILDE
It seems to only affect characters beyond U+00FF, so something is probably trying to use ISO 8859-1 (Latin-1).
Looking at the network requests seems to confirm that:
When using "Search for match", the browser sends the data to `/command/core/recon-judge-similar-cells`, and when clicking on the tick, it sends it to `/command/core/recon-judge-one-cell`.
In both cases it uses the header `Content-Type: application/x-www-form-urlencoded; charset=UTF-8`.
`/command/core/recon-judge-similar-cells` gives a response with the header `Content-Type: application/json;charset=iso-8859-1` while `/command/core/recon-judge-one-cell` gives a response with `Content-Type: application/json;charset=utf-8`.
|
https://github.com/OpenRefine/OpenRefine/issues/6063
|
https://github.com/OpenRefine/OpenRefine/pull/6242
|
714504e96ae472b9296a10cf1315b8513959a8d6
|
41b742851953d65462a990a75b875b37d7f95388
| 2023-09-22T23:38:25Z |
java
| 2023-12-14T03:52:53Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,059 |
["server/src/com/google/util/logging/IndentingLayout.java"]
|
Logger config not logging exception information
|
We have lots of examples of error logging of the form
```
logger.error("Error generating parser UI initialization data for XML file", e);
```
https://github.com/OpenRefine/OpenRefine/blob/63fe765d0f1c63b6ffcf017ad862a58aa3a845df/main/src/com/google/refine/importers/XmlImporter.java#L188
and none of them appear to be logging exception information, as is obviously intended.
We need to figure out what's broken in the logging configuration so that this information gets logged again. This could be a side effect of Jetty changing from implementing their own logging.
|
https://github.com/OpenRefine/OpenRefine/issues/6059
|
https://github.com/OpenRefine/OpenRefine/pull/6376
|
10cd91c13299b4ee7a61014ddc59a8e1bc1680ad
|
b9028ed7808f0d2b493930e2109f76671b9ca697
| 2023-09-14T18:31:50Z |
java
| 2024-02-21T06:06:22Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,058 |
["main/src/com/google/refine/commands/Command.java", "main/src/com/google/refine/commands/HttpUtilities.java", "main/src/com/google/refine/importers/ExcelImporter.java", "main/src/com/google/refine/importers/XmlImporter.java", "main/src/com/google/refine/importing/DefaultImportingController.java", "main/tests/cypress/cypress/e2e/create-project/create_project.cy.js", "main/tests/cypress/cypress/fixtures/AAT-with-xsi.xml", "main/tests/cypress/cypress/fixtures/AAT.xml", "main/webapp/modules/core/scripts/index/default-importing-controller/controller.js"]
|
XML import preview should report errors to the user
|
When attempting to import the Getty AAT vocabulary [XML file](https://web.archive.org/web/20220418022604/http://aatdownloads.getty.edu/VocabData/aat_xml_1121.zip) the XML parser throws an error about the `xsi` namespace prefix being undefined, but this error is swallowed and never shown to the user (or even logged on the console)
### To Reproduce
Steps to reproduce the behavior:
1. Create a project using the URL above
### Current Results
Create Project preview page shows an empty display with the message "Click on the first XML element corresponding to the first record to load." (even though there are no XML elements displayed) and the console logs
```
13:49:52.532 [..e.importers.XmlImporter] Error generating parser UI initialization data for XML file (474ms)
```
### Expected Behavior
Console displays the exception which is currently being swallowed
```
ParseError at [row,col]:[1,205]
Message: http://www.w3.org/TR/1999/REC-xml-names-19990114#AttributePrefixUnbound?Vocabulary&xsi:noNamespaceSchemaLocation&xsi
```
User is shown an interpretable (hopefully) error message, along with the line and column where the error occurred, and, ideally, the text of the offending line.
### Versions
- OpenRefine: 3.7.5
### Datasets
https://web.archive.org/web/20220418022604/http://aatdownloads.getty.edu/VocabData/aat_xml_1121.zip
|
https://github.com/OpenRefine/OpenRefine/issues/6058
|
https://github.com/OpenRefine/OpenRefine/pull/6370
|
9e82b2a790e943b47127e81698fb3bb48847648e
|
21aaf8a367f1b2ef42ccbdf43f96d72af2737153
| 2023-09-14T18:05:05Z |
java
| 2024-02-28T10:13:55Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,047 |
["main/src/com/google/refine/sorting/StringCriterion.java", "main/tests/server/src/com/google/refine/operations/row/RowReorderOperationTests.java"]
|
case-sensitive sort does not produce expected result
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
### To Reproduce
Steps to reproduce the behavior:
1. Create a project with the supplied "temp.csv"
2. Configure a sort on the single column as follows
3. Apply the sort by "Reorder rows permanently"
### Current Results
<!-- What results occurred or were shown. -->
Sort only moves the character "c"
### Expected Behavior
<!-- A clear and concise description of what you expected to happen or to show. -->
Here an example within Excel on what the sort should be showing
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 10
- Browser Version: Chromium 116.0.5845.180
- JRE or JDK Version: using the "openrefine-win-with-java-3.7.4.zip" release
- OpenRefine: openrefine-win-with-java-3.7.4
### Datasets
[temp.csv](https://github.com/OpenRefine/OpenRefine/files/12572271/temp.csv)
### Additional context
<!-- Add any other context about the problem here. -->
Looking at the oldest version of JDK that OpenRefine supports: https://docs.oracle.com/en/java/javase/11/docs/api/java.base/java/text/Collator.html#TERTIARY
https://github.com/OpenRefine/OpenRefine/blob/2b07cfa5ced9bcfdc6d485fe1849af1c79ac320a/main/src/com/google/refine/sorting/StringCriterion.java#L57C50-L57C50
Shouldn't the Collator over here be set to "TERTIARY" or even "IDENTICAL"?
Checking the blame history shows the Collator initially set to the "IDENTICAL" strength.
|
https://github.com/OpenRefine/OpenRefine/issues/6047
|
https://github.com/OpenRefine/OpenRefine/pull/6050
|
5184fdf013b225cb019cca24ee115bfb677da377
|
e6f87fb9c2f74af0e2caf7c8597288c10fa57b5d
| 2023-09-11T04:50:53Z |
java
| 2023-09-13T12:59:28Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 6,031 |
["main/webapp/modules/core/scripts/reconciliation/recon-dialog.js", "main/webapp/modules/core/scripts/reconciliation/standard-service-panel.js", "main/webapp/modules/core/styles/reconciliation/standard-service-panel.css"]
|
Property selection in reconciliation dialog is unreliable
|
In the reconciliation dialog, there is a panel on the right-hand side to configure which columns should be matched with which properties offered by the service:
There are a couple of problems with this UI.
### Redundant checkbox
The checkbox before each column name is redundant with the property field. It only makes sense to check this box if the associated property field is filled, and if the property field is filled then the box should be checked.
In other words, none of the following states are useful to the user:
* 
* 
Therefore, the value of the checkbox could be inferred directly from the state of the associated input field.
### Unclear auto-completion status in property field
The property field does not indicate clearly if the property name typed by the user has been correctly matched to a property offered by the reconciliation service.
When typing a property name in that field, the user is presented with a few options in a drop-down menu (if the service supports it):
If the user picks an option in the list, the property will be correctly used for reconciliation. The state of the input field after doing so looks like this:
If instead the user ignores the suggestions and clicks anywhere else, without picking an option, then the input field will look exactly the same, but the property will not be used for reconciliation because it is not associated with any valid property id:
This can be quite frustrating because once reconciliation has run, it is not necessarily easy to determine if a property was indeed used for reconciliation or not (depending on how the service uses additional properties for scoring). Moreover, one has to wait for the entire reconciliation process to finish before being able to spot this problem (in 3.x).
### Proposed solution
* Remove the checkbox
* Mark the property fields as having an error (similarly to what we do in the Wikibase schema) if they are non-empty but have not been validated with auto-completion. Refuse to start reconciliation as long as any property field is an error state.
### Alternatives considered
Find a new design for this UI, which could potentially also solve #2075 in the same go.
### Additional context
To be considered for NFDI-funded improvements to reconciliation. Cc @lozanaross @ayushrai206 and @OpenRefine/designers
|
https://github.com/OpenRefine/OpenRefine/issues/6031
|
https://github.com/OpenRefine/OpenRefine/pull/6060
|
87ce8324c0de2e5de497b938eda9b5d101d380dd
|
91f2dd04bb07bcea62ef546b572901319f132181
| 2023-08-31T09:37:57Z |
java
| 2023-10-21T08:59:44Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,998 |
["main/tests/cypress/cypress/e2e/project/grid/column/reconcile/add_entity_identifiers.cy.js", "main/webapp/modules/core/langs/translation-en.json", "main/webapp/modules/core/scripts/views/data-table/add-q-column-dialog.html", "main/webapp/modules/core/scripts/views/data-table/menu-reconcile.js"]
|
UI improvements for the "Add entity identifiers column" action
|
When using the "Add entity identifiers column action", the UI currently looks like this:
### Proposed improvements
* the input field for the column name should be focused when the dialog opens
* the input field could be pre-filled with an automatically generated column name which does not conflict with any existing column name (for instance, if the base column is called "Name", the default column name could be "matched id for Name", perhaps?)
* the "Submit Query" label for the validation button does not make sense, as we are not submitting a query. It should be "OK" or "Create column", I would say. It should be marked as primary (default) button, which generally appears in bold I think
* the title of the dialog does not clearly separate the user-provided data (column name) from the rest. Instead of being `Add column containing entity identifiers on Name`, it could be `Add column of entity identifiers from column "Name"` (note the double quotes)
### Additional context
Discovered while reviewing #5960.
Maybe @OpenRefine/designers have more thoughts about this?
|
https://github.com/OpenRefine/OpenRefine/issues/5998
|
https://github.com/OpenRefine/OpenRefine/pull/6021
|
b80e9836441ad16b983abeb432fa9d18a650fe42
|
257a603b49257d4b8d35052a810aeb6a090ac69f
| 2023-07-30T10:22:24Z |
java
| 2023-08-31T21:51:10Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,991 |
["main/src/com/google/refine/expr/functions/Length.java", "main/tests/server/src/com/google/refine/expr/functions/LengthTests.java"]
|
length(dict) should return number of elements, not serialized length
|
`'{"a":"b","c":"d"}'.parseJson().length()` should evaluate to 2, not 17, as it currently does.
By contrast, `'["abc","def"]'.parseJson().length()` does correctly evaluate to 2
### Current Results
`'{"a":"b","c":"d"}'.parseJson().length()` evaluates to 17
### Expected Behavior
`'{"a":"b","c":"d"}'.parseJson().length()` evaluates to 2
|
https://github.com/OpenRefine/OpenRefine/issues/5991
|
https://github.com/OpenRefine/OpenRefine/pull/5994
|
d3b9e7ae7c5817dc88e1439705c685c627c4991a
|
463313f06863470a0489f10e5368a1a189501cc6
| 2023-07-27T19:26:29Z |
java
| 2023-07-29T02:21:29Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,983 |
[".gitignore"]
|
Add internal yarn files to .gitignore
|
In #5960, @ayushrai206 added files such as `main/tests/cypress/.yarn/cache/@babel-helper-string-parser-npm-7.22.5-448ff0e489-836851ca5e.zip` to some commits by mistake.
On my machine, yarn does not seem to create those `.yarn` directories but it feels like they should be added to our `.gitignore` file(s) to prevent that. The same goes for `.yarnrc.yml`.
|
https://github.com/OpenRefine/OpenRefine/issues/5983
|
https://github.com/OpenRefine/OpenRefine/pull/5984
|
16c81699572bb604182404b3675e85ee3c2f11b6
|
6dffe684ab0075c0569f040418d4bb145f94a825
| 2023-07-25T15:11:10Z |
java
| 2023-07-26T07:00:07Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,980 |
["main/webapp/modules/core/scripts/index/open-project-ui.js"]
|
OpenRefine opens the "Open project" tab by default
|
Until recently, when starting OpenRefine, one would be presented with the first panel ("Create project").
In the current development version (master branch), the "Open project" tab opens instead.
### To Reproduce
Steps to reproduce the behavior:
1. First, do `./refine`
2. Then, observe the panel that appears in the browser.
### Current Results
The "Open project" panel is shown.
### Expected Behavior
The "Create project" panel is shown.
### Versions<!-- (please complete the following information)-->
- Operating System: Debian
- Browser Version: both firefox and chromium
- OpenRefine: 3.8-SNAPSHOT [9042ff7]
|
https://github.com/OpenRefine/OpenRefine/issues/5980
|
https://github.com/OpenRefine/OpenRefine/pull/6102
|
e8ff5bd3a9b4681d4eb6be92073729e62cbf8945
|
135050ee35363625daac165961b4fa5f589c76dd
| 2023-07-19T12:26:24Z |
java
| 2023-10-17T07:37:00Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,977 |
["main/webapp/modules/core/styles/views/data-table-view.css"]
|
Cursor suggests action-on-click when over a column name even though there is no action
|
I often need to copy the name of a column and although this is simple enough by selecting the text of the column name with a double-click or click-drag, I notice that the cursor turns to the gloved finger (the same type of cursor you get when over a url link) when in this vicinity, suggesting that there might be some action on a single click. As far as I can tell, however, there is only an action when you click on the triangle-in-square next to the column name (i.e. to bring up the column actions menu). So although this is a very minor thing and there is no loss of functionality, I think it would make the interface look more consistent and clean if the cursor were the I-bar type (for text selection) rather than the finger when it is just over the column name (and only switch to the finger when over the triangle for the action menu).
|
https://github.com/OpenRefine/OpenRefine/issues/5977
|
https://github.com/OpenRefine/OpenRefine/pull/5869
|
8e3612c102d7feb5dda44dda0c12cf15bf9c837d
|
9074c2df2b7a527360040e4b166207b78dc99d8e
| 2023-07-17T14:57:10Z |
java
| 2023-05-25T08:56:35Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,947 |
["README.md", "packaging/pom.xml"]
|
SNAPSHOT release link for MacOs is broken
|
https://oss.sonatype.org/service/local/artifact/maven/content?r=snapshots&g=org.openrefine&a=openrefine&v=3.8-SNAPSHOT&c=mac&p=dmg
in the readme file
Also here
https://openrefine.org/docs/manual/installing
the link is broken
https://github.com/OpenRefine/OpenRefine-snapshot-releases/releases
|
https://github.com/OpenRefine/OpenRefine/issues/5947
|
https://github.com/OpenRefine/OpenRefine/pull/5972
|
721351c6f8815e0fe3b73087cdd8bc7d3e87322d
|
c12852cf1db0f3c6538f750d7aeded8886a0f8dc
| 2023-06-22T13:15:37Z |
java
| 2023-07-14T06:56:53Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,926 |
["main/webapp/modules/core/langs/translation-en.json", "main/webapp/modules/core/langs/translation-es.json", "main/webapp/modules/core/scripts/reconciliation/recon-manager.js"]
|
Reconciliation services can be added multiple times
|
When adding a new reconciliation service in the reconciliation dialog, we do not check if the service being added is already in the list.
### To Reproduce
Steps to reproduce the behavior:
1. First, add a reconciliation service
2. Then, add it again
### Current Results
The service is added twice.
### Expected Behavior
The service list should not have any duplicates.
### Additional context
Reported by @fsteeg. Maybe an interesting issue for @ayushrai206?
|
https://github.com/OpenRefine/OpenRefine/issues/5926
|
https://github.com/OpenRefine/OpenRefine/pull/5964
|
21d9be29a7abfc3beda9fa9ab6d697901c76fbfa
|
951173ec23fe6d7d265e38dc01a355ecff9a5ed3
| 2023-06-15T14:08:06Z |
java
| 2023-07-10T20:34:26Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,908 |
["main/src/com/google/refine/operations/recon/ReconOperation.java", "main/tests/server/src/com/google/refine/operations/recon/ReconOperationTests.java", "main/webapp/modules/core/scripts/views/data-table/menu-reconcile.js"]
|
Make reconciliation judgment facets resizable like other text facets
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
The facets for reconciliation judgment seem to be intentionally not resizable and it is not clear to me it is a good thing.
The number of choices display in the facet can grow after the facet was created, which will cause the choices to overflow and make the facet unusable.
Also, the visual inconsistency between this facet and others is confusing to me as a user.
### To Reproduce
Steps to reproduce the behavior:
1. First, reconcile a column, which creates a judgment facet - this one is resizable.
2. Second, delete that facet and re-create it from the Reconcile -> Facets menu. It is not resizable.
3. Then, mark some cells as being matched to new items
4. Finally, observe the glitch in facet rendering
### Current Results
Initially (note the absence of resizing slider in the judgment facet):
After marking items as new (note the overflowing "Facet by choice counts"):
### Expected Behavior
Make the facet resizable like all others and do not overflow when choices change.
### Versions<!-- (please complete the following information)-->
- Operating System: Debian
- Browser Version: Firefox
- JRE or JDK Version: 17
- OpenRefine: 3.8-SNAPSHOT (master)
|
https://github.com/OpenRefine/OpenRefine/issues/5908
|
https://github.com/OpenRefine/OpenRefine/pull/5918
|
b33eb122256c26c8da58408336b5bf1a5f6fe739
|
3fe107310206e11a1ee57ecbb0c28c34d16b99b6
| 2023-06-08T09:56:12Z |
java
| 2023-06-14T02:56:12Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,907 |
["main/webapp/modules/core/scripts/reconciliation/standard-service-panel.js"]
|
Default reconciliation types are rendered without name (only id)
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
The default types proposed by a reconciliation service are proposed to the user by showing only their id, not their human-readable label.
### To Reproduce
Steps to reproduce the behavior:
1. First, import this dataset: https://opendata.paris.fr/explore/dataset/lieux-de-tournage-a-paris/ (CSV with `;` delimiter: https://opendata.paris.fr/api/explore/v2.1/catalog/datasets/lieux-de-tournage-a-paris/exports/csv?lang=fr&timezone=Europe%2FBerlin&use_labels=true&delimiter=%3B)
2. Then, open the reconciliation dialog to reconcile the "Titre" column to Wikidata
### Current Results
<!-- What results occurred or were shown. -->
Observe that the last type in the list appears as an identifier (without a name).
### Expected Behavior
Like other type options in the list, the type should be provided as a pair of (name, id).
<!-- A clear and concise description of what you expected to happen or to show. -->
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: Linux
- Browser Version: Firefox
- JRE or JDK Version: 17
- OpenRefine: 3.8-SNAPSHOT
### Additional context
Follow-up to #4224.
|
https://github.com/OpenRefine/OpenRefine/issues/5907
|
https://github.com/OpenRefine/OpenRefine/pull/5917
|
7a9c45e6d84f5a8310bda6fa1b07bee9de8c3802
|
b33eb122256c26c8da58408336b5bf1a5f6fe739
| 2023-06-08T09:35:50Z |
java
| 2023-06-13T10:02:30Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,893 |
["extensions/wikibase/module/scripts/dialogs/schema-list-item.html", "extensions/wikibase/module/styles/dialogs/schema-management-dialog.css"]
|
Manage Wikibase schemas should use a list (auto numbered) to improve display
|
Currently, the Manage Wikibase schemas dialog shows slight grey rows for each schema to be managed.
If one row only appears, then this looks awkward because the grey row with a schema name looks like some kind of header for the dialog, instead of a row element. It would be better to tweak the display so that rows get numbered automatically.
### Current Results
### Expected Behavior
Rows are numbered as in this example (number just before the schema row for `test`).
If a 2nd schema was defined then I would expect numbered rows 1. and 2.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: <!-- e.g. iOS, Windows 10, Linux, Ubuntu 18.04 -->
- Browser Version: <!-- e.g. Chrome 19, Firefox 61, Safari, NOTE: OpenRefine does not support IE but works OK in most cases -->
- JRE or JDK Version: <!-- output of "java -version" e.g. JRE 1.8.0_181 -->
- OpenRefine: <!-- e.g. OpenRefine 3.0 Beta] -->
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, you can share it selectively by email to developers who work on the issue -->
### Additional context
From Forum support topic https://forum.openrefine.org/t/where-is-export-import-schema-option-in-or-wikidata-extension-now/784
|
https://github.com/OpenRefine/OpenRefine/issues/5893
|
https://github.com/OpenRefine/OpenRefine/pull/5976
|
58ed7a8a6bd6f82be7ea21036c044d1e5a04d5db
|
a465eeaba489e3838bde6f8f8c144cc672bbde87
| 2023-06-04T01:17:38Z |
java
| 2023-07-30T12:09:29Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,879 |
["main/tests/cypress/cypress/e2e/project/grid/all-column/collapse_all.cy.js", "main/tests/cypress/cypress/e2e/project/grid/column/view/collapse.cy.js", "main/tests/cypress/cypress/e2e/project/grid/column/view/collapse_left.cy.js", "main/tests/cypress/cypress/e2e/project/grid/column/view/collapse_right.cy.js", "main/tests/cypress/cypress/e2e/project/grid/column/view/expand-left.cy.js", "main/tests/cypress/cypress/e2e/project/grid/column/view/expand-right.cy.js", "main/webapp/modules/core/langs/translation-en.json", "main/webapp/modules/core/scripts/views/data-table/data-table-view.js", "main/webapp/modules/core/styles/views/data-table-view.css"]
|
Add a visual indicator/button for expanding collapsed columns
|
### Proposed solution
@wetneb [pointed out](https://github.com/OpenRefine/OpenRefine/pull/5869#issuecomment-1567269708) that there is no longer an indicator that one can click the header of a collapsed column to expand it. We should bring this back but rather than just have the mouse cursor change lets add a proper button(semantically not necessarily visually) so it works from the keyboard this time around.
### Alternatives considered
The cheap solution of making sure the cursor changes one again on hover when a column is collapsed.
### Additional context
I removed the cursor effect in #5869 and I think that's still a good change given the odd behavior of changing the cursor on non-clickable expanded columns as well.
|
https://github.com/OpenRefine/OpenRefine/issues/5879
|
https://github.com/OpenRefine/OpenRefine/pull/5981
|
cef6a82778f04fe67cc939d220e0f31f64db2cfc
|
857d9caa93a0f849f6a6203345f42be3d5938849
| 2023-05-29T17:00:38Z |
java
| 2024-02-14T12:16:40Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,878 |
["extensions/wikibase/module/styles/dialogs/wikibase-dialog.css", "main/webapp/modules/core/MOD-INF/controller.js", "main/webapp/modules/core/langs/translation-en.json", "main/webapp/modules/core/scripts/reconciliation/recon-dialog-2.html", "main/webapp/modules/core/scripts/reconciliation/recon-dialog-2.js", "main/webapp/modules/core/scripts/reconciliation/recon-dialog.html", "main/webapp/modules/core/scripts/reconciliation/recon-dialog.js", "main/webapp/modules/core/scripts/views/data-table/column-header-ui.js", "main/webapp/modules/core/scripts/views/data-table/menu-reconcile.js", "main/webapp/modules/core/styles/reconciliation/recon-dialog.css"]
|
Show endpoint URL in reconciliation service list
|
When selecting a reconciliation service (in the reconciliation dialog), the user only sees the reconciliation services listed by name.
It would be useful to also show the reconciliation endpoint, for instance to:
* distinguish more easily between services provided by the official data source or third-party ones
* spot any local services, which might require starting up the local reconciliation service before using it
* inspect any other aspects of the URL which are not reflected in the name
### Proposed solution
Add the reconciliation endpoint URL underneath the name (for each service).
### Alternatives considered
There are surely other ways to present this information.
|
https://github.com/OpenRefine/OpenRefine/issues/5878
|
https://github.com/OpenRefine/OpenRefine/pull/6118
|
78f939d60c12bab930e57f8ceeec8946fba128c2
|
f856713fd81af6629178e38d93fe55016529204b
| 2023-05-29T12:39:06Z |
java
| 2023-12-07T10:33:20Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,808 |
[".github/workflows/pull_request.yml", ".github/workflows/snapshot_release.yml", "main/tests/cypress/build-test-matrix.js", "refine"]
|
Improve/Shorten pull_request.yml e2e test names
|
The PR ui_tests names are cut off in the 'Checks' section. Because of this, the runtime is not visible:
Also, ui_test is actually an e2e test and not just a UI test. So, it should be changed to e2e_test to be more accurate.
Proposed Changes:
* Update the scripts so that the e2e tests look something like this:
CI / e2e_test Project
CI / e2e_test Column
CI / e2e_test Facet
CI / e2e_test Reconcile
CI / e2e_test Misc
CI / e2e_test Undo
* Change "refine ui_tests" to "refine e2e_tests"
|
https://github.com/OpenRefine/OpenRefine/issues/5808
|
https://github.com/OpenRefine/OpenRefine/pull/5817
|
83f0407953ea1c3845e68487355d0c460489c2dd
|
df484ddd4ddad47731a039ca0af48137746b7c55
| 2023-04-22T01:36:35Z |
java
| 2023-04-25T23:33:35Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,800 |
["main/tests/cypress/cypress/support/openrefine_api.js"]
|
Project deletion after Cypress tests is inefficient
|
After a Cypress test is run, we attempt to delete all projects created in the test.
When multiple tests are run in the same runner, the list that keeps track of which projects were created grows longer and so we attempt to delete many projects which were already deleted earlier. This is inefficient and can be cleared.
### To Reproduce
Steps to reproduce the behavior:
1. Run all Cypress tests locally (using the "Run all 80 specs" option)
### Current Results
After running a couple of tests, you can observe that the clean-up phase is making many requests to the backend to delete projects which already no longer exist.
### Expected Behavior
Only the projects created in the said test are deleted.
### Screenshots
|
https://github.com/OpenRefine/OpenRefine/issues/5800
|
https://github.com/OpenRefine/OpenRefine/pull/5801
|
a009a98079277668ffbc990257fb474628d4e777
|
921d122b5f61e685a6336db3ec0b21e65971370a
| 2023-04-21T10:45:23Z |
java
| 2023-04-22T06:01:59Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,799 |
["main/src/com/google/refine/model/changes/MassRowChange.java", "main/src/com/google/refine/model/changes/RowRemovalChange.java", "main/tests/server/src/com/google/refine/model/CacheTests.java", "main/tests/server/src/com/google/refine/operations/row/RowRemovalOperationTests.java"]
|
Duplicates facet computing incorrectly after undoing row removal
|
I often use the 'duplicates' built-in facet and recently I noticed that it was reporting incorrectly after I had undone a row removal operation. I find this is reproducible - see the attached images and the dataset I used for this demo:
1. Load a file with a column that has duplicates
2. Apply the **Facet > Customized facets > Duplicates facet** to that column and note that it shows the duplicates as a 'true' set
3. Remove rows (whichever way you like) so that there are no longer duplicates and note that the facet is now showing only 'false'. (In my example here I starred the second line, faceted by star, and removed matching, but I could have done this a little easier by faceting on Col1, selecting "B", and then removing)
4. Undo the row removal and observe that the facet is still showing only 'false' even though all the duplicates are back. You can even add a new duplicates facet on the column and it will only show 'false'
<img width="499" alt="Original load" src="https://user-images.githubusercontent.com/36742596/233584964-994b9fc8-3c4f-470a-afc1-ad7c09b9487b.png">
<img width="496" alt="Undo row removal" src="https://user-images.githubusercontent.com/36742596/233584973-1cc44900-d50d-4eeb-b971-08cb149bcd16.png">
<img width="495" alt="Facet not returning to original state" src="https://user-images.githubusercontent.com/36742596/233584968-d6130e53-8afa-47eb-aedf-5635ae5434ce.png">
Note that in my use case I often modify the facet to show the duplication count (i.e. `facetCount(value, 'value', 'Col2')`) but the problem is the same.
### Versions<!-- (please complete the following information)-->
- Operating System: MacOS 12.6.5
- Browser Version: Safari 16.4.1
- OpenRefine: 3.7.2 [f7ad526]
[Simple.txt](https://github.com/OpenRefine/OpenRefine/files/11293737/Simple.txt)
|
https://github.com/OpenRefine/OpenRefine/issues/5799
|
https://github.com/OpenRefine/OpenRefine/pull/5804
|
468116e6d4110a1f6cb000ca090c2afa3fb3712d
|
83f0407953ea1c3845e68487355d0c460489c2dd
| 2023-04-21T08:25:58Z |
java
| 2023-04-25T21:09:05Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,795 |
["refine.bat"]
|
The free memory info is inconsistent between refine and refine.bat
|
I noticed that refine.bat prints out the Java version and "Getting Free Ram...", and then it outputs info in a support.log.
The Linux refine script outputs a banner "You Have xxx M of free memory" but does not write a support.log.
I think it is simpler to just output everything on the command-line rather than output extra info to a support.log.
I would modify refine.bat to output the same banner message that refine script has for the free memory.
Windows refine.bat
```
>refine.bat run
Java 17 (17.0.2)
Getting Free Ram...
21:23:11.511 [ refine_server] Starting Server bound to '[127.0.0.1:3333](http://127.0.0.1:3333/)' (0ms)
21:23:11.517 [ refine_server] refine.memory size: 16384M JVM Max heap: 17179869184 bytes (6ms)
```
Windows support.log
```
-----------------------
PROCESSOR_ARCHITECTURE = AMD64
JAVA_HOME = C:\Program Files\Java\jdk-17.0.2
java release = 17
java -version = 17.0.2
freeRam = 31875 MB
REFINE_MEMORY = 16384M
-----------------------
```
Linux refine run
```
$ ./refine run
Using refine.ini for configuration
-------------------------------------------------------------------------------------------------
You have 7948M of free memory.
Your current configuration is set to use 1400M of memory.
OpenRefine can run better when given more memory. Read our FAQ on how to allocate more memory here:
https://openrefine.org/docs/manual/installing#increasing-memory-allocation
-------------------------------------------------------------------------------------------------
21:36:21.505 [ refine_server] Starting Server bound to '[127.0.0.1:3333](http://127.0.0.1:3333/)' (0ms)
21:36:21.510 [ refine_server] refine.memory size: 1400M JVM Max heap: 1468006400 bytes (5ms)
21:36:21.534 [ refine_server] Initializing context: '/' from '/home/github/OpenRefine/main/webapp' (24ms)
```
|
https://github.com/OpenRefine/OpenRefine/issues/5795
|
https://github.com/OpenRefine/OpenRefine/pull/5798
|
0c79d90ae941778d319cc99d676cea9b963a9514
|
e09153ca000d238e5dbeb343d176ae783c1a49cd
| 2023-04-19T22:31:52Z |
java
| 2023-04-21T10:49:54Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,778 |
["refine"]
|
Refine script support for the Google extension
|
The refine.ini declares three variables for the Google extension: GDATA_CLIENT_ID, GDATA_CLIENT_SECRET, and GDATA_API_KEY
https://github.com/OpenRefine/OpenRefine/blob/53466223ef11aad83422d5989db995cda8ff69ec/refine.ini#L29
refine checks for these, and if all three are set it passes corresponding definitions to the OpenRefine Java process
```
-Dext.gdata.clientid=%GDATA_CLIENT_ID% -Dext.gdata.clientsecret=%GDATA_CLIENT_SECRET% -Dext.gdata.apikey=%GDATA_API_KEY%
```
The Linux refine script parses for `-k <google_api_key>`, but it doesn't do anything. This option is mentioned in the docs, but not in the `refine /h` usage statement.
The scripts should be made consistent.
Options (One of the following):
* Remove the -k option from the Linux refine script.
* Remove all references to google extension from the refine scripts. The user would have to add variables to the JAVA_OPTIONS in refine.ini to enable the google extension.
```
JAVA_OPTION= -Dext.gdata.clientid=%GDATA_CLIENT_ID% -Dext.gdata.clientsecret=%GDATA_CLIENT_SECRET% -Dext.gdata.apikey=%GDATA_API_KEY%
```
|
https://github.com/OpenRefine/OpenRefine/issues/5778
|
https://github.com/OpenRefine/OpenRefine/pull/5823
|
f6966a0d479e9415382c5b98c1a10affaf99f5ff
|
086bb76f03251074e3d31075cc514f16c28e4118
| 2023-04-08T04:23:42Z |
java
| 2023-05-01T20:38:25Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,776 |
["refine", "refine.bat"]
|
Can't set REFINE_MEMORY using refine
|
Same as #5622 but for the Linux refine script.
Command line options should be set last, right now refine.ini takes precedence.
|
https://github.com/OpenRefine/OpenRefine/issues/5776
|
https://github.com/OpenRefine/OpenRefine/pull/5777
|
089616ca2bb40351bd19cc4831a89fdd13c6938c
|
a7b66bcd338b9091590d263ac87527a326a6b959
| 2023-04-06T06:32:09Z |
java
| 2023-04-10T09:24:27Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,774 |
["main/webapp/modules/core/scripts/index/edit-metadata-dialog.js"]
|
Dragging project metadata dialog causes it to jump to the top left of the screen
|
Dragging the project metadata dialog causes it to jump to the top left of the screen:
https://user-images.githubusercontent.com/42903164/230260917-6ca9b741-4d03-4251-8b5a-41e864e0d1f8.mp4
This is due to a regression introduced by: #4527
|
https://github.com/OpenRefine/OpenRefine/issues/5774
|
https://github.com/OpenRefine/OpenRefine/pull/5775
|
4476435a9f8a52e8f554cf076ec1f920a8b892d2
|
8b564dd4ab4e37f35231f9cec55e240b5d952ec4
| 2023-04-06T03:43:41Z |
java
| 2023-04-14T07:53:29Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,773 |
["refine", "refine.bat"]
|
Refine.ini is read too late in refine bash script too
|
Just like in #5384, refine.ini is read after the block checking `JAVA_HOME` in the [refine bash script](https://github.com/OpenRefine/OpenRefine/blob/3.7/refine#L834).
### Versions
- Operating System: macOS Ventura 13.0.1
- OpenRefine: OpenRefine 3.7
|
https://github.com/OpenRefine/OpenRefine/issues/5773
|
https://github.com/OpenRefine/OpenRefine/pull/5777
|
089616ca2bb40351bd19cc4831a89fdd13c6938c
|
a7b66bcd338b9091590d263ac87527a326a6b959
| 2023-04-05T22:14:30Z |
java
| 2023-04-10T09:24:27Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,772 |
["main/src/com/google/refine/ProjectMetadata.java", "main/tests/server/src/com/google/refine/commands/project/SetProjectMetadataCommandTests.java"]
|
Error when editing CustomMetadata field
|
When trying to edit the customMetadata field in the about section of a project to add new values, the OpenRefine server crashes and throws a ``` java.lang.IllegalArgumentException: Can not set java.util.Map field com.google.refine.ProjectMetadata._customMetadata to java.lang.String```. The issue is the same as well when interacting with the OpenRefine API through ```set-project-metadata``` endpoint as well.
### To Reproduce
Steps to reproduce the behavior:
1. First, open the Open Project section to the left of the screen
2. Then, choose any project (with empty customMetadata) and press on about.
3. Finally, next to the customMetada field, press the edit button and add any JSON-like format.
### Current Results
The newly added values in the customMetadata field is not saved when the page is reloaded and the server crashes (error shown below).
### Expected Behavior
The newly added values in the customMetadata field should be saved when the page is reloaded.
### Screenshots
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 10
- Browser Version: Chrome Version 111.0.5563.147
- JRE or JDK Version: openjdk 1.8.0_345 <!-- output of "java -version" e.g. JRE 1.8.0_181 -->
- OpenRefine: 3.7
|
https://github.com/OpenRefine/OpenRefine/issues/5772
|
https://github.com/OpenRefine/OpenRefine/pull/5973
|
56b872dd6d8d303dd1e57f75d76fceaa56508353
|
15a029576b03eb3f1adde80bd1e3b76800befbaa
| 2023-04-05T08:50:06Z |
java
| 2023-07-16T17:37:43Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,750 |
["main/webapp/modules/core/scripts/util/i18n.js"]
|
OpenRefine displaying IDs instead of German translation
|
OpenRefine doesn't display the German translation, instead the IDs are shown.
### To Reproduce
Steps to reproduce the behavior:
1. Install `OpenRefine 3.7.1`
2. Open `Language Settings`
3. Choose `Deutsch`
4. Look at the top, where `core-index/slogan` is written instead of the slogan
5. Compare with weblate (https://hosted.weblate.org/translate/openrefine/translations/de/?offset=1&q=+core-index%2Fslogan&sort_by=-priority%2Cposition&checksum=) the slogan is translated since two years
### Current Results
I'm seeing translation ids
### Expected Behavior
I see the translation
### Screenshots
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 10
- Browser Version: Firefox 111.0.1 (64-Bit)
- JRE or JDK Version: 17
- OpenRefine: OpenRefine 3.7.1
|
https://github.com/OpenRefine/OpenRefine/issues/5750
|
https://github.com/OpenRefine/OpenRefine/pull/5758
|
d440908f7454f3ea9789b50db8d952ba097d8e3f
|
e78a12d1db7a9ba8c7b3b9da118ee8742abae876
| 2023-03-29T13:51:13Z |
java
| 2023-03-30T19:13:52Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,749 |
["main/webapp/modules/core/scripts/facets/list-facet.js", "main/webapp/modules/core/scripts/project/browsing-engine.js"]
|
Facet sizing broken if the facet panel is not open when the facet is created
|
Creating a facet while the Undo / Redo tab is active makes the facet initially collapsed, breaking the automatic sizing introduced in #5514.
### Current results
### Expected Behavior
The facet should be properly sized automatically
### Versions<!-- (please complete the following information)-->
- Browser Version: Firefox 102.8.0esr <!-- e.g. Chrome 19, Firefox 61, Safari, NOTE: OpenRefine does not support IE but works OK in most cases -->
- OpenRefine: 3.8-SNAPSHOT
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, you can share it selectively by email to developers who work on the issue -->
### Additional context
Ideally, the facet panel should also be switched to automatically, because when I create a facet I want to see it directly (either to look at its statistics or to use it to filter rows / records). This means that the fix should be rather simple and should not worry about resizing an element that is not currently rendered: just switch to the facet panel whenever new facets are created. That being said, I am not completely sure about what to do when facets are automatically created without an action from the user, for instance after a reconciliation operation is run.
|
https://github.com/OpenRefine/OpenRefine/issues/5749
|
https://github.com/OpenRefine/OpenRefine/pull/5767
|
53466223ef11aad83422d5989db995cda8ff69ec
|
089616ca2bb40351bd19cc4831a89fdd13c6938c
| 2023-03-29T09:26:12Z |
java
| 2023-04-10T09:04:51Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,747 |
["main/webapp/modules/core/scripts/index/open-project-ui.js"]
|
Seemingly dead frontend code refering to create-project-from-upload command
|
The following reference to the `create-project-from-upload` command seems to be dead code.
https://github.com/OpenRefine/OpenRefine/blob/76152500a5af8c38b88d34e15b8b01fcf6d0da6e/main/webapp/modules/core/scripts/index/open-project-ui.js#L316-L344
This request is bound to an event listener on ` #upload-file-button`, but I cannot find any such element in our code base.
https://github.com/OpenRefine/OpenRefine/blob/76152500a5af8c38b88d34e15b8b01fcf6d0da6e/main/webapp/modules/core/scripts/index/open-project-ui.js#L42-L58
The other elements whose values are encoded in the request also do not seem to exist anymore.
One should double check that this is not needed and remove this code (and any other parts that are also never executed anymore).
|
https://github.com/OpenRefine/OpenRefine/issues/5747
|
https://github.com/OpenRefine/OpenRefine/pull/5770
|
43229916af0340d9f87757de2d51fb5ad4608aab
|
1de933ac36645d60abeeda70d29ad0f321d93077
| 2023-03-28T19:06:29Z |
java
| 2023-04-04T07:55:30Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,743 |
["main/webapp/modules/core/MOD-INF/controller.js", "main/webapp/modules/core/styles/common.css", "main/webapp/modules/core/styles/pure.css"]
|
Choose File Button - MAC
|
Today, I downloaded OpenRefine on my Mac, having previously used Windows. When attempting to upload a file, I received an error message stating, "You must specify a data file to import." I was confused and unsure of what I had done wrong. After reviewing the page again, I realized that I had clicked on the wrong button - the "Next" button caught my eye instead of the "Choose File" button. The problem was that the "Choose File" button was not as noticeable, as it was a white button with regular text and only a small shadow around it, which is different from what is used on Windows.
### Proposed solution
A possible solution would be to either implement the same design as on Windows or to make the "Choose File" button a blue text icon that is consistent with the OpenRefine design.
### Additional context
shot on mac
<img width="1300" alt="Screenshot 2023-03-28 at 06 06 36" src="https://user-images.githubusercontent.com/23156958/228260389-fbcbf837-96f4-485f-972c-a7c174283276.png">
shot on windows
|
https://github.com/OpenRefine/OpenRefine/issues/5743
|
https://github.com/OpenRefine/OpenRefine/pull/5771
|
e113e847b5cd0096f3c082a88b527d64fc28a70f
|
696a93ff9e0adb4d9e77096dbbaaf7e3cbf30a61
| 2023-03-28T14:00:35Z |
java
| 2023-04-05T07:55:20Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,741 |
["main/webapp/modules/core/styles/index/create-project-ui.css"]
|
Unexpected style change of progress bar during project creation
|
The appearance of the progress bar during project creation has recently changed, likely unintentionally
### Current Results
### Expected Behavior
<!-- A clear and concise description of what you expected to happen or to show. -->
### Versions<!-- (please complete the following information)-->
master branch
### Additional context
Introduced by the migration from LESS to CSS:
9de62f67f3 (#5525)
|
https://github.com/OpenRefine/OpenRefine/issues/5741
|
https://github.com/OpenRefine/OpenRefine/pull/5742
|
2b30ecd344cf9694ff2124e6973c8193bc6a8073
|
9d120c8a8e27d54753295f59664488aefac4202f
| 2023-03-28T09:55:56Z |
java
| 2023-03-28T10:55:53Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,738 |
["main/webapp/modules/core/scripts/views/data-table/data-table-view.js"]
|
(Un)star/flag action reverts unrelated cell reconciliation judgments
|
Starring or flagging an individual row reverts any manual reconciliation action on a cell in that row.
### To Reproduce
Steps to reproduce the behavior:
1. First, reconcile a column
2. Then, make a reconciliation judgment in one of the cells of that column. The action should only impact that single cell (not similar cells).
3. Finally, change the star or flag status of the same row
### Current Results
The reconciliation action of step 2 is visibly reverted (but only client-side, not server side: if the grid is refreshed, we see that it was not actually reverted).
### Expected Behavior
The action in step 2 should not be visibly reverted.
### Screenshots
### Versions<!-- (please complete the following information)-->
- Operating System: Linux
- Browser Version: Firefox
- JRE or JDK Version: 17
- OpenRefine: 3.8-SNAPSHOT
### Additional context
Reported on the forum: https://forum.openrefine.org/t/unstarring-reverts-un-reconciliation/672
|
https://github.com/OpenRefine/OpenRefine/issues/5738
|
https://github.com/OpenRefine/OpenRefine/pull/5842
|
5e6581837d146b68f499baac6d7cec8f88b02ade
|
7298361736875f09fc494497c805bf01f937f0a3
| 2023-03-27T19:34:36Z |
java
| 2023-05-13T11:00:58Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,716 |
["extensions/wikibase/module/scripts/wikibase-suggest.js"]
|
Localization of auto-complete fields in Wikibase schema broken since API update
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
The behaviour of the Wikibase API's endpoint for auto-completion of entities (`wbsearchentities`) has changed and since that, the drop-down menu suggesting entities in the Wikibase schema editor displays entities in English regardless of the user's language.
### To Reproduce
Steps to reproduce the behavior:
1. First, set OpenRefine's language to something else than English, for instance French
2. Then, open a project or create a new one
3. In the Wikibase extension menu, make sure Wikidata is selected as the target Wikibase instance
4. Create a Wikibase schema and type a few letters (such as "mac") in the entity subject field (the only text field present in the schema editor by default)
### Current Results
Auto-completion results are shown in English.
<!-- What results occurred or were shown. -->
### Expected Behavior
The results should be shown in the language selected by the user.
<!-- A clear and concise description of what you expected to happen or to show. -->
### Screenshots
### Versions<!-- (please complete the following information)-->
- Operating System: irrelevant
- Browser Version: Firefox
- JRE or JDK Version: 17
- OpenRefine: 3.7.1
### Additional context
Our code makes requests of the form:
https://www.wikidata.org/w/api.php?action=wbsearchentities&language=fr&search=mac&type=item&format=json
Those are missing the `uselang=fr` parameter, for them to return the expected results:
https://www.wikidata.org/w/api.php?action=wbsearchentities&language=fr&uselang=fr&search=mac&type=item&format=json
It should just be a matter of adding this extra parameter.
|
https://github.com/OpenRefine/OpenRefine/issues/5716
|
https://github.com/OpenRefine/OpenRefine/pull/5809
|
549f5ee343970ecb3204ede878ded0046e11bb8a
|
702728cd45bd0aa3ee174daf675a4e1bfdf87c73
| 2023-03-21T18:13:28Z |
java
| 2023-04-24T18:12:40Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,714 |
["main/webapp/modules/core/scripts/util/dialog.js"]
|
Dialog mover
|
I don't know if this problem is only on the 3.7.1 version or on other versions too but on opening the reconciliation dialogue and moving it up to view the list down below, I couldn't move it back down (mover doesn't display on the footer section) which means I can't move it back down as screen scroller is also not showing anymore since it is a dialogue
**Proposal**
Add mover when hovering on footers section in a dialogue and probably to other similar dialogue that maybe affected by this
|
https://github.com/OpenRefine/OpenRefine/issues/5714
|
https://github.com/OpenRefine/OpenRefine/pull/5724
|
5273a3d3553755a31add210d3dd69de1923b7d55
|
b558ff64ca2d026f3412c23a32ed8c7138b5d349
| 2023-03-21T16:37:53Z |
java
| 2023-03-26T19:10:39Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,698 |
["refine"]
|
refine script ignores compilation errors and attempts to run the app anyway
|
The `refine` script on Linux does not do a good job at stopping if it needs to compile OpenRefine and compilation errors occur.
### To Reproduce
Steps to reproduce the behavior (from a git clone of the repository):
1. First, clean with `./refine clean`
2. Then, introduce some compilation error somewhere, such as by doing `rm ./extensions/wikibase/src/org/openrefine/wikibase/qa/QAWarning.java`
3. Finally, run `./refine`
### Current Results
The compilation fails in the Wikibase extension but the `refine` script still attempts to start OpenRefine:
```
...
[ERROR] extensions/wikibase/src/org/openrefine/wikibase/qa/scrutinizers/NewEntityScrutinizer.java:[104,17] cannot find symbol
[ERROR] symbol: class QAWarning
[ERROR] location: class org.openrefine.wikibase.qa.scrutinizers.NewEntityScrutinizer
[ERROR] extensions/wikibase/src/org/openrefine/wikibase/qa/scrutinizers/NewEntityScrutinizer.java:[104,39] cannot find symbol
[ERROR] symbol: class QAWarning
[ERROR] location: class org.openrefine.wikibase.qa.scrutinizers.NewEntityScrutinizer
[ERROR] -> [Help 1]
[ERROR]
[ERROR] To see the full stack trace of the errors, re-run Maven with the -e switch.
[ERROR] Re-run Maven using the -X switch to enable full debug logging.
[ERROR]
[ERROR] For more information about the errors and possible solutions, please read the following articles:
[ERROR] [Help 1] http://cwiki.apache.org/confluence/display/MAVEN/MojoFailureException
[ERROR]
[ERROR] After correcting the problems, you can resume the build with the command
[ERROR] mvn <args> -rf :wikibase
15:54:05.840 [ refine_server] Starting Server bound to '127.0.0.1:3333' (0ms)
15:54:05.975 [ refine_server] refine.memory size: 1400M JVM Max heap: 1468006400 bytes (135ms)
15:54:06.359 [ refine_server] Initializing context: '/' from '/home/user/OpenRefine/main/webapp' (384ms)
15:54:06.948 [ refine_server] Starting autoreloading scanner... (589ms)
15:54:13.478 [ refine] Starting OpenRefine 3.8-SNAPSHOT [acca2dd]... (6530ms)
15:54:13.495 [ refine] initializing FileProjectManager with dir (17ms)
...
```
### Expected Behavior
The `refine` script should stop at the end of the failing Maven build.
### Versions<!-- (please complete the following information)-->
- Operating System: Linux
- JRE or JDK Version: 17
- OpenRefine: master branch
## Additional context
This could be solved with some bash flags, see https://github.com/OpenRefine/OpenRefine/issues/5241
|
https://github.com/OpenRefine/OpenRefine/issues/5698
|
https://github.com/OpenRefine/OpenRefine/pull/5720
|
a9ba31313ed74b84e58b4c5af514bace0488448e
|
8ce76dda813f3deb868d755dd8903cf4856b2432
| 2023-03-16T14:56:44Z |
java
| 2023-03-22T20:39:11Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,696 |
["main/tests/server/src/com/google/refine/grel/controls/ForRangeTests.java", "main/tests/server/src/com/google/refine/importers/ExcelImporterTests.java"]
|
Locale-dependent tests failing with a French locale
|
We have certain tests which fail when run on a machine with a French locale set:
```
[ERROR] Failures:
[ERROR] ForRangeTests.testForRangeWithFloatingPointValues:82->RefineTest.parseEval:352 Wrong result for expression: forRange(0,0.6,0.1,v,v.toString('%.3f')).join(',') expected [0.000,0.100,0.200,0.300,0.400,0.500] but found [0,000,0,100,0,200,0,300,0,400,0,500]
[ERROR] ExcelImporterTests.readXlsxAsText:284 expected [100.00%] but found [100,00%]
```
This should not be the case.
|
https://github.com/OpenRefine/OpenRefine/issues/5696
|
https://github.com/OpenRefine/OpenRefine/pull/5697
|
68256734bbc212239bf366e42cdf9ae1c517962a
|
7a26fe23bbd1d5acbd28ed78f2008bff35214656
| 2023-03-16T07:50:09Z |
java
| 2023-03-16T15:44:45Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,695 |
["main/src/com/google/refine/expr/functions/math/FactN.java", "main/src/com/google/refine/expr/functions/math/Multinomial.java", "main/tests/server/src/com/google/refine/grel/GrelTests.java"]
|
Fix code scanning alert in Multinomial - Implicit narrowing conversion in compound assignment
|
<!-- Warning: The suggested title contains the alert rule name. This can expose security information. -->
Tracking issue for:
- [x] https://github.com/OpenRefine/OpenRefine/security/code-scanning/43
Use `long`s for calculations instead of `int`s.
https://github.com/OpenRefine/OpenRefine/blob/90efb305f8ede391d9f1e59c20bb80ef4a10b466/main/src/com/google/refine/expr/functions/math/Multinomial.java#L51-L60
|
https://github.com/OpenRefine/OpenRefine/issues/5695
|
https://github.com/OpenRefine/OpenRefine/pull/5699
|
9ecf6e9a383f04c6968be3d7d20c4f72f226170e
|
99b46045dd9aa3036ebe80d5319fb5611eca70e3
| 2023-03-15T23:45:39Z |
java
| 2023-03-16T19:01:01Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,680 |
["refine"]
|
The "Something is already running on <URL>.." error is incorrect
|
### To Reproduce
Steps to reproduce the behavior:
1. First, try running OpenRefine on an invalid interface: `./refine -i gjksjd -p 333sd4`
### Current Results
The "Something is already running on <URL>..." will be displayed even although noting is running there
### Expected Behavior
The check running script is only checking if OpenRefine is running on the given interface, not if "something" is running, the error message should therefore, not say that something is already running there but rather give a more generic error. "OpenRefine isn't running on $URL."?
https://github.com/OpenRefine/OpenRefine/blob/ea017a7ff614aea7d0220e6230c6db18b6322622/refine#L143
|
https://github.com/OpenRefine/OpenRefine/issues/5680
|
https://github.com/OpenRefine/OpenRefine/pull/5689
|
ea1764ac08493a418f036d08f34bde55dc098dfd
|
80ce9cb704aaf55d2a349b397e9c1063601249ea
| 2023-03-13T13:23:57Z |
java
| 2023-03-15T07:51:26Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,665 |
["main/src/com/google/refine/importing/ImportingJob.java"]
|
On project creation memory usage reported incorrectly
|
### To Reproduce
Steps to reproduce the behavior:
1. Find a large data file that will take some time to create a project
2. Then, create a project from the data file
3. While the project creation is happening observe the memory usage reporting
### Current Results
Memory usage is shown as consuming millions of percent of the available memory
### Expected Behavior
A relative accurate (or at least possible!) indication of memory usage is expected
### Screenshots
Screenshot from a recent forum posting
Source: https://forum.openrefine.org/t/large-csv-loading/478
### Additional context
I think this commit is the problem https://github.com/OpenRefine/OpenRefine/commit/6a7046323181dd0094058cc64d1171c166488cf4
I suspect the order of calculation for the memory means that only the Runtime.getRuntime().freeMemory() is divided by 1048576 to get the value in MB, leaving the Runtime.getRuntime().totalMemory() in bytes - which then leaves the percentage calculation as wildly inaccurate
|
https://github.com/OpenRefine/OpenRefine/issues/5665
|
https://github.com/OpenRefine/OpenRefine/pull/5666
|
acc3a5596754dce0653f682cffb5b0bc3e184347
|
6aaf84d7623077ba413f20fb3dfb2ec041c7457e
| 2023-03-03T14:10:41Z |
java
| 2023-03-06T08:10:39Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,656 |
["extensions/wikibase/src/org/openrefine/wikibase/qa/scrutinizers/FileNameScrutinizer.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/qa/scrutinizers/FileNameScrutinizerTest.java"]
|
Wikimedia Commons Wikibase not detecting non-Latin script characters
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
I've been trying to upload a batch of files that have names containing both Latin and [Odia](https://en.wikipedia.org/wiki/Odia_script) characters. For instance, Wikimedia Commons allows filenames with Odia characters (for instance, this file](https://commons.wikimedia.org/wiki/Category:Audio_files_created_using_Kathabhidhana), but OpenRefine does not. See the screenshot of the error I see below. cc @wetneb
### To Reproduce
Steps to reproduce the behavior:
1. Create a file with a file name such as "ଫାଇଲ.wav" (that's a WAVE file with a filename with in the Odia script)
2. Try to upload to Wikimedia Commons using OpenRefine. It fails at the Wikibase (Upload edits to Wikibase)
### Current Results
<!-- What results occurred or were shown. -->
It fails (see below)
<img width="752" alt="OpenRefine-issue-with-Odia" src="https://user-images.githubusercontent.com/1258090/222007841-b59c7218-db7a-44aa-b5cf-45b88375e354.png">
### Expected Behavior
<!-- A clear and concise description of what you expected to happen or to show. -->
The upload should normally happen, just like any other file.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: MacOS<!-- e.g. iOS, Windows 10, Linux, Ubuntu 18.04 -->
- Browser Version: Brave 1.46.171<!-- e.g. Chrome 19, Firefox 61, Safari, NOTE: OpenRefine does not support IE but works OK in most cases -->
- JRE or JDK Version: <!-- output of "java -version" e.g. JRE 1.8.0_181 -->
- OpenRefine: 3.7<!-- e.g. OpenRefine 3.0 Beta] -->
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, you can share it selectively by email to developers who work on the issue -->
1. [Sample dataset containing the same files I was trying to upload](https://drive.google.com/drive/u/0/folders/1lyg_63oZCd7fIwWFvqU4v0AwKOIHK8FQ)
### Additional context
<!-- Add any other context about the problem here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5656
|
https://github.com/OpenRefine/OpenRefine/pull/6227
|
fdc28849d8a8daa9eb77d78e03d07079f01ca5f0
|
6976b18af2ddda86d1087612159cfbfdb2cba06c
| 2023-03-01T01:33:47Z |
java
| 2023-12-09T09:05:44Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,654 |
["main/tests/cypress/build-test-matrix.js"]
|
Fix set-output deprecation warning in prepare_ui_test_matrix
|
[prepare_ui_test_matrix](https://github.com/OpenRefine/OpenRefine/actions/runs/4276187262/jobs/7453816454#step:4:23)
The `set-output` command is deprecated and will be disabled soon. Please upgrade to using Environment Files. For more information see: https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/
|
https://github.com/OpenRefine/OpenRefine/issues/5654
|
https://github.com/OpenRefine/OpenRefine/pull/5660
|
0ffc69fed164d0916188c6ff82037d6bce883361
|
606ad716f1bf91e24f04b31e79a18fc2fc42c1e2
| 2023-02-28T19:47:19Z |
java
| 2023-03-02T10:06:45Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,637 |
["main/pom.xml", "packaging/THIRD-PARTY.txt", "packaging/manual_licenses.xml", "packaging/pom.xml"]
|
Fix licensing issues during maven test and packaging
|
There are a few warnings regarding license information for some dependencies (transitive as well).
We can handle this noise and be a good citizen of open source license information for our dependencies by using and providing more context through `license-maven-plugin:download-licenses`
Below are the current warnings sometimes exhibited depending on connectivity issues:
### To Reproduce
Steps to reproduce the behavior:
1. First, do `mvn test`
2. Then, notice WARNINGs which happens because of the `generate-sources` lifecycle event in maven which happens during test/packaging.
### Current Results
```
[INFO] --- license-maven-plugin:2.0.0:download-licenses (default) @ packaging ---
[WARNING] No license information available for: oro:oro:2.0.8
[WARNING] Unable to retrieve license from URL 'https://oss.oracle.com/licenses/CDDL+GPL-1.1' for dependency 'javax.servlet:javax.servlet-api:4.0.1': Connect to oss.oracle.com:443 [oss.oracle.com/138.1.117.31] failed: Read timed out
[WARNING] Unable to retrieve license from URL 'https://raw.githubusercontent.com/OpenRefine/OpenRefine/master/LICENSE.txt' for dependency 'org.openrefine:server:3.8-SNAPSHOT': The requested name is valid, but no data of the requested type was found (raw.githubusercontent.com)
[WARNING] There were 3 download errors - check the warnings above
```
### Expected Behavior
The maven plugin: `download-licenses` has several configuration options and we can fallout with ignoring the errors and hiding them through a config flag, but to be a good citizen of open source we should provide manual license information through a `licensesConfigFile` which we can simply store as `manual_licenses.xml` in the `/packaging` folder.
This will eliminate warnings during CI as well as locally for folks that experience connectivity issues occasionally (like me!).
It will also provide the right context and license information for the dependencies that are dual or triple licensed or that no longer are reachable via a URL.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 11
- Browser Version: n/a
- JRE or JDK Version: JDK 17
- OpenRefine: master branch
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, you can share it selectively by email to developers who work on the issue -->
### Additional context
I'm going to tackle this since I've done this before for another OSS project and I'm almost done with the basic configuration for a PR soon.
|
https://github.com/OpenRefine/OpenRefine/issues/5637
|
https://github.com/OpenRefine/OpenRefine/pull/5638
|
a2199db761898d1b68b7dd73eaec1315998076f6
|
d5f8c40d5bf3ef77171bf8185c63334f0b3141b3
| 2023-02-17T09:48:48Z |
java
| 2024-02-15T07:58:20Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,633 |
["main/src/com/google/refine/expr/WrappedRow.java", "main/src/com/google/refine/expr/functions/Cross.java", "main/tests/server/src/com/google/refine/expr/functions/CrossTests.java"]
|
Field lookup of .record incorrect for lists of WrappedRows
|
Evaluation of the `record` field accessor on lists of wrapped rows, such as returned by the `cross()` function incorrectly uses the global context from the bindings when looking up the row index, rather than using the `rowIndex` field in the associated `WrappedRow`s.
### To Reproduce
Steps to reproduce the behavior:
1. Use the following data to create a new project from the clipboard
```
001,035$a,cross,245$a,020$a
53773,(CaOOAMICUS)000034856822,,Why I love my mummy /,7270208
53806,(CaOOAMICUS)000034856822,,Why I love my daddy /,7877617
```
3. Choose add column based on this column, using the "035$a" column with the expression: `cell.cross("","035$a").record.index`
4. Inspect the results in the preview display
### Current Results
row | value | cell.cross("","035$a").record.
-- | -- | --
1 | (CaOOAMICUS)000034856822 | [ 0, 0]
2 | (CaOOAMICUS)000034856822 | [ 1, 1 ]
### Expected Behavior
row | value | cell.cross("","035$a").record.
-- | -- | --
1 | (CaOOAMICUS)000034856822 | [ 0, 1 ]
2 | (CaOOAMICUS)000034856822 | [ 0, 1 ]
### Versions
- OpenRefine: 3.6.2, 3.7
|
https://github.com/OpenRefine/OpenRefine/issues/5633
|
https://github.com/OpenRefine/OpenRefine/pull/5634
|
c89e3c2e4dc70cea85a84432894f5c2e3363dfd7
|
2f29f502bea20979d703df6b9d8a3e937e7e45b8
| 2023-02-16T01:02:28Z |
java
| 2023-02-28T19:38:08Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,628 |
["main/tests/cypress/cypress/e2e/project/grid/column/reconcile/actions/create_new_item_for_each.cy.js"]
|
Failing Cypress test in the CI
|
Our CI is red since 37a443d735811be58892d9dcbd5c36c835cddc6f, although the cause is likely external. We should fix this.
|
https://github.com/OpenRefine/OpenRefine/issues/5628
|
https://github.com/OpenRefine/OpenRefine/pull/5632
|
20381488b6c020c415a9e77b66401019b34bda74
|
dde64fd6a6efbd58ddd3eba0ad8d34468654a744
| 2023-02-15T12:49:54Z |
java
| 2023-02-16T07:03:10Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,620 |
["server/src/com/google/refine/Refine.java"]
|
disambiguate JVM Max heap used in startup logging
|
The current logger string does not tell the user the unit of measurement for `JVM Max heap:`
```log
09:57:55.910 [ refine_server] refine.memory size: 1400M JVM Max heap: 1468006400 (2ms)
```
### Proposed solution
add `bytes` to the end of the logger string
```log
09:57:55.910 [ refine_server] refine.memory size: 1400M JVM Max heap: 1468006400 bytes (2ms)
```
### Alternatives considered
<!-- If there alternative solutions that you have considered or think should be considered, please list them here -->
### Additional context
<!-- Add any other context or screenshots about the feature request here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5620
|
https://github.com/OpenRefine/OpenRefine/pull/5621
|
b2553c916c18d0ae26971b19ea07b03c41f91159
|
d7ba91dc39cb6bc7d33a0d2ed3cf8e39f6e9a6de
| 2023-02-12T02:05:31Z |
java
| 2023-02-13T07:28:19Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,610 |
["main/webapp/modules/core/scripts/project/browsing-engine.js", "main/webapp/modules/core/styles/project/sidebar.css"]
|
Facet sidebar shifts when it refreshes
|
The issue is as @thadguidry says:
> Unneccesary jumping up/down of Facet panels while users click on selections. The div browsing-panel-indicator will quickly show a hidden spinner “Refreshing facets…” which is above the browsing-panel-controls and hence causes all child elements underneath to suddenly move down and then back up. The effect to a user is disconcerting with jerky motion.
|
https://github.com/OpenRefine/OpenRefine/issues/5610
|
https://github.com/OpenRefine/OpenRefine/pull/5611
|
979b9dcd665338954d662164f34f51bf5bd9f718
|
d5c325652cc2cea732211c6345d2a43f0521490d
| 2023-02-08T01:06:05Z |
java
| 2023-02-08T22:55:12Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,609 |
[".github/workflows/unassign_issues.yml"]
|
Unassign contributors automatically after a delay
|
[Our issues labeled "good first issue"](https://github.com/OpenRefine/OpenRefine/issues?q=is%3Aopen+is%3Aissue+label%3A%22good+first+issue%22) tend to attract new contributors, which is great.
However, contributors often abandon the issue without unassigning themselves.
This means that the pool of good first issues available for prospective contributors shrinks artificially.
As a preparation for our participation in Outreachy/GSoC I have cleared assignees of all good first issues today (after checking that they had been assigned for a long time). But I think we should not have to do this manually.
There is a GitHub Action which seems to do just that:
https://github.com/marketplace/actions/unassign-contributor-after-days-of-inactivity
Any resistance to trying this out?
As an experiment I would first restrict this to the "good first issue" tag, and set a fairly generous delay - perhaps 3 months?
|
https://github.com/OpenRefine/OpenRefine/issues/5609
|
https://github.com/OpenRefine/OpenRefine/pull/5612
|
d7ba91dc39cb6bc7d33a0d2ed3cf8e39f6e9a6de
|
37a443d735811be58892d9dcbd5c36c835cddc6f
| 2023-02-07T18:13:30Z |
java
| 2023-02-14T07:40:40Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,606 |
["main/webapp/modules/core/scripts/facets/list-facet.js"]
|
Facets do not keep their size positions any longer in 3.8 snapshot
|
REGRESSION: Text Facet (and maybe others?) is not keeping it's dimensions after I click to include a selection from the list.
See Video below (soon).
### To Reproduce
Steps to reproduce the behavior:
1. Create Text Facet
2. Then, click on include links for a few selections
### Current Results
Notice that upon every click on a *include* link, the facet window shrinks back to default size!
### Expected Behavior
Facet should remain with the size the user has changed it to (especially for long lists for easier browsing/analysis)
### Screenshots
https://user-images.githubusercontent.com/986438/216807735-ce179a85-b5d6-4f27-ae40-e1e36fa92ea5.mp4
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 11
- Browser Version: Edge latest
- JRE or JDK Version: JDK 17
- OpenRefine: Version 3.8-SNAPSHOT [4d7571d]
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, you can share it selectively by email to developers who work on the issue -->
### Additional context
<!-- Add any other context about the problem here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5606
|
https://github.com/OpenRefine/OpenRefine/pull/5608
|
59bbbadb74c2d87990eb97f3d07bc528fe2fdb82
|
c722190f823eed84aa638c4cef2df891022ea48a
| 2023-02-05T07:48:59Z |
java
| 2023-02-07T23:42:29Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,603 |
["docs/package.json", "docs/yarn.lock"]
|
Respect "Batch size" if specified by reconciliation manifest
|
Reconciliation manifests can sometimes have a [ `batchSize` property set in their manifest](https://reconciliation-api.github.io/specs/0.2/#service-manifest) indicating the maximum number of reconciliation queries in a reconciliation query batch.
OpenRefine currently sends 10 reconciliation queries per batch rather than the value specified by the `batchSize` property. By respecting the `batchSize` property many users would likely see less HTTP-requests and faster reconciliation.
Subset of #2549
|
https://github.com/OpenRefine/OpenRefine/issues/5603
|
https://github.com/OpenRefine/OpenRefine/pull/4215
|
6960b7db66f771b0f0762de53f3155eb26177711
|
aca71f6e3d2b477af1936385bb4783af9cfa2184
| 2023-02-04T10:53:37Z |
java
| 2021-10-16T08:11:32Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,587 |
["main/src/com/google/refine/expr/functions/strings/Split.java", "main/tests/server/src/com/google/refine/expr/functions/strings/SplitTests.java"]
|
Fix split() function to match its documentation
|
The documentation for `split()` says:
"Returns the array of strings obtained by splitting s by sep. The separator can be either a string or a regex pattern. If preserveTokens is true, empty segments are preserved."
This implies, but doesn't explicitly say, that (all) empty tokens are elided, but the actual implementation naively follows the default behavior of the external functions used by the implementation, which differ from this (and from each other).
- `org.apache.commons.lang3.StringUtils.splitByWholeSeparator()` - returns empty tokens if there are trailing separator characters, which I suspect is unintentional from looking at the implementation, but it's a long standing behavior that I suspect won't change.
- `java.util.regex.Pattern.split()` - documented to return an empty token for leading pattern patch.
Neither of these behaviors matches the OpenRefine documentation and, worse, they differ from each other.
### Current Results
From https://github.com/OpenRefine/OpenRefine/pull/5551#issuecomment-1373703149 grel examples by @ostephens
```
" abc ".split(" ") -> [ "abc", "" ]
" abc ".split(/(?U)[\W]+/) -> [ "", "abc" ]
```
### Expected Behavior
Both cases above return a single token "abc"
### Versions
- OpenRefine: 3.6
### Additional context
While this is technically a breaking change because it changes the behavior, it would bring the behavior in line with the documentation and user expectations.
|
https://github.com/OpenRefine/OpenRefine/issues/5587
|
https://github.com/OpenRefine/OpenRefine/pull/5588
|
1f21cfa32b58de1d8aaa5c6e393ad80f2f3d6339
|
0ffc69fed164d0916188c6ff82037d6bce883361
| 2023-01-27T17:41:44Z |
java
| 2023-03-02T03:10:15Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,586 |
["refine"]
|
Clean up the refine script
|
The `refine` script has plenty of broken/dead code that we should clean up. Here is a probably non-exhaustive list:
* A lot of the packaging logic (`./refine dist`) is now handled directly in Maven so that command should basically just call `mvn package -DskipTests=true`
* A lot of the linters/checkers are broken or not used
* The `./refine ui_tests` needlessly downloads reconcile-csv, which is no longer used for Cypress tests. Also, it seems to fail to run the Cypress tests on my end
* The `./refine test` and `./refine server_tests` do the same thing because `ui_tests` are not included there.
* The automatic download of Maven is broken (#5484)
|
https://github.com/OpenRefine/OpenRefine/issues/5586
|
https://github.com/OpenRefine/OpenRefine/pull/5644
|
26ed466278421691f5dbfdd62ac012a449694bd9
|
7cacc9117f0f4bcc7792392f2afe0e1f2e712fa2
| 2023-01-27T16:21:15Z |
java
| 2023-02-28T00:15:07Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,583 |
["packaging/pom.xml", "pom.xml"]
|
OpenRefine 3.7 beta 2 for Windows not executing
|
OpenRefine 3.7 beta 2 for Windows OS is not installing when attempting to download and install from the .exe file with latest version of java already installed using latest version of Chrome
### To Reproduce
Steps to reproduce the behavior:
1. Download zip file for Windows OS, OpenRefine 3.7 beta 2 with java
2. Unzip file
3. Update and install latest version of java on local machine and latest version of Chrome
4. Open openrefine.exe (application file)
### Current Results
Command line loaded as expected, but application file didn't execute to open the actual application, with latest Chrome browser installed execution of the file opens a new tab which displays the OpenRefine website in the Java section of the page (https://openrefine.org/docs/manual/installing#java)
### Expected Behavior
Expecting to open the application and execute the file
### Screenshots
NA
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 10 Enterprise
- Browser Version: Chrome Version 109.0.5414.120 (Official build)
- JRE or JDK Version: Version 8, update 361 (build 1.8.0_361-b09)
- OpenRefine: OpenRefine 3.7 beta 2 with java
### Datasets
NA
### Additional context
NA
|
https://github.com/OpenRefine/OpenRefine/issues/5583
|
https://github.com/OpenRefine/OpenRefine/pull/5931
|
221a6196a36c50b3f1e289ab12214cb6cf0d0769
|
2782b9495b084990d32b906ce1552d0531bc6133
| 2023-01-25T15:04:33Z |
java
| 2023-06-19T06:26:39Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,581 |
["main/webapp/modules/core/scripts/dialogs/clustering-dialog.js"]
|
Cluster and edit changes space character in values to nbsp
|
When using cluster and edit, when you select an existing value by clicking on the hyperlinked option, it replaces any spaces in the original values with non-breaking spaces. See [here](https://forum.openrefine.org/t/odd-space-issue-causing-blank-down-issues/330).
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
### To Reproduce
Steps to reproduce the behavior:
1. Use cluster and edit on a column of data
2. In the interface, select one of the clustering option by clicking on the hyperlinked options. Select merge selected and close.
3. New values will all have replaced standard spaces with non-breaking spaces
### Current Results
Space replaced with non-breaking space which can cause inconsistencies with rest of data
### Expected Behavior
Should not alter original data in this way
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: Mac <!-- e.g. iOS, Windows 10, Linux, Ubuntu 18.04 -->
- Browser Version: Chrome <!-- e.g. Chrome 19, Firefox 61, Safari, NOTE: OpenRefine does not support IE but works OK in most cases -->
- JRE or JDK Version: <!-- output of "java -version" e.g. JRE 1.8.0_181 -->
- OpenRefine: 3.6.2
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, you can share it selectively by email to developers who work on the issue -->
### Additional context
<!-- Add any other context about the problem here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5581
|
https://github.com/OpenRefine/OpenRefine/pull/5584
|
2fd133e638c677f0904a3bde168637fa02cc68d2
|
4d7571dad1301acaa3b3295c8dc70ff02bf87b96
| 2023-01-24T20:37:13Z |
java
| 2023-01-26T09:41:04Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,569 |
[".github/workflows/pull_request.yml"]
|
Simplify the CI/CD workflows to use the builtin Maven cache feature in setup-java action that now supports it
|
We can now remove the actions/cache steps and replace them with a config setting under `setup-java` action. It automatically uses actions/cache already under the hood and will create the fileHash for the Maven `**/pom.xml` automatically.
Documention: https://github.com/actions/setup-java/blob/main/README.md#caching-packages-dependencies
Usage such as:
```
steps:
- uses: actions/checkout@v3
- uses: actions/setup-java@v3
with:
distribution: 'temurin'
java-version: '17'
cache: 'maven'
- name: Build with Maven
run: mvn -B package --file pom.xml
```
### To Reproduce
Steps to reproduce the behavior:
1. Run a PR workflow.
### Current Results
Extra steps not need for `actions/cache` separately.
### Expected Behavior
cleaner and less verbose build/test output in logs.
|
https://github.com/OpenRefine/OpenRefine/issues/5569
|
https://github.com/OpenRefine/OpenRefine/pull/5570
|
dc01d58aff29c3e0aca58366acb22aea3b2b777b
|
3df47341164177f6c39891c7b7ab49e5f1261cea
| 2023-01-17T07:41:52Z |
java
| 2023-01-17T13:49:36Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,558 |
["main/webapp/modules/core/langs/translation-en.json", "main/webapp/modules/core/scripts/reconciliation/standard-service-panel.html", "main/webapp/modules/core/scripts/reconciliation/standard-service-panel.js"]
|
Link to reconciliation service's documentation
|
Some reconciliation services has a link to documentation specified in their service manifest. Currently an OpenRefine user would need to do the following to go to the documentation from OpenRefine:
1. click the "Access service API" link in the main reconciliation dialog
2. browse the JSON manifest to see if there is a documentation link
3. click it
It should be easier.
### Proposed solution
The main reconciliation dialog should have a "Documentation" link next to the "Access service API" link.
|
https://github.com/OpenRefine/OpenRefine/issues/5558
|
https://github.com/OpenRefine/OpenRefine/pull/5784
|
3402a31267f94418e549850ce041cebe22d5adda
|
e27e0e979484b4d55738fa3adc208928033766f9
| 2023-01-11T17:22:50Z |
java
| 2023-06-07T13:07:18Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,548 |
["extensions/database/module/MOD-INF/controller.js", "extensions/database/module/scripts/database-extension.js", "extensions/database/module/scripts/index/database-import-form.html", "extensions/database/module/scripts/index/database-source-ui.js", "extensions/database/module/styles/bootstrap.css", "extensions/database/module/styles/database-import.css", "extensions/database/module/styles/pure.css"]
|
The database extension overrides CSS styles from core
|
The database extensions loads Yahoo's prue.css library which styles elements regardless of classes and as a result one must always ensure that core got styles which take priority to avoid side effects.
### To Reproduce
Inspect any element on either the index or project page and check for styles coming from prue.css.
|
https://github.com/OpenRefine/OpenRefine/issues/5548
|
https://github.com/OpenRefine/OpenRefine/pull/5550
|
7a26fe23bbd1d5acbd28ed78f2008bff35214656
|
9ecf6e9a383f04c6968be3d7d20c4f72f226170e
| 2023-01-05T14:32:53Z |
java
| 2023-03-16T17:23:20Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,531 |
["main/src/com/google/refine/LookupCacheManager.java", "main/tests/server/src/com/google/refine/expr/functions/CrossTests.java"]
|
cross function produces null when there is no match
|
I notice that the result of the `cross()` function is not an array when there is no match found - instead, the result is `null`. This seems to contradict the reference documentation which states "Returns an array of **zero** or more rows..." (that is, we should expect an an empty array, `[]`, when there are no matching rows). So if, for instance, you are wanting a count of the number of matches using `cross(...).length()` (which is something I do) then you currently get "Error: length expects an array or a string" where you expect to get 0.
Incidentally, if the function gets fixed to produce an empty array, as per the documentation, then the result of a common operation where you just want to retrieve the first matching row (i.e. `cross(...)[0]`) will, when there is no match, change to "Error: java.lang.ArrayIndexOutOfBoundsException: Index 0 out of bounds for length 0" - currently the result is just null (since `null[0]` is also null). So I'm wondering if the null return was implemented as a specific exception to avoid this ungainly instance. Should we consider making the result of indexing an empty array (e.g. `[][0]`) null instead of producing the error?
I'm looking at OpenRefine 3.6.2 [579a6f7].
|
https://github.com/OpenRefine/OpenRefine/issues/5531
|
https://github.com/OpenRefine/OpenRefine/pull/5713
|
6b2d889e82a8f218952706cdaa568709e58cb2c8
|
c9bdba47d7554ceb00f15420f7111142ca2bef2b
| 2022-12-23T12:00:08Z |
java
| 2023-03-24T08:34:46Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,523 |
["main/webapp/modules/core/scripts/reconciliation/standard-service-panel.html", "main/webapp/modules/core/scripts/reconciliation/standard-service-panel.js"]
|
Suggest property query includes a type when none is selected
|
When reconciling a column, using a service that supports a `/suggest/property` endpoint, if I select `Reconcile against no particular type`, OpenRefine includes the most recently selected type in the query.
For example, in this screenshot, I selected `Reconcile against no particular type` and then started typing `test` in the `As property` box.
<img width="926" alt="Screen Shot 2022-12-14 at 4 40 06 PM" src="https://user-images.githubusercontent.com/16618618/207953210-0b2413b5-d5c7-436a-9ae6-e381bbc3ca2f.png">
### Current Behavior
In that example, OpenRefine sent the following request to the reconciliation service:
`https://wikidata.reconci.link/en/suggest/property?prefix=test&spell=always&exact=false&scoring=schema&prefixed=true&type=Q223393`
### Expected Behavior
I would expect the type to be sent in the suggest/property query if one is selected so that a reconciliation service has the option to filter the suggested properties by the included type.
If `Reconcile against no particular type` is selected, then I expect no type to be included in the query so that the reconciliation service does no filtering on the suggested properties.
### Versions<!-- (please complete the following information)-->
- Operating System: MacOS Monterey
- Browser Version: Chrome
- JRE or JDK Version: JDK 13.0.2
- OpenRefine: Version 3.6.2
### Additional context
I'm creating a reconciliation service and this affects how we need to implement the internal logic of the suggest property endpoint.
|
https://github.com/OpenRefine/OpenRefine/issues/5523
|
https://github.com/OpenRefine/OpenRefine/pull/5543
|
512bd55aab0ff5d77d853c4e1ef167e427552553
|
f72e8e699a0359df7379a0e405b863966179e51c
| 2022-12-15T19:56:30Z |
java
| 2023-01-04T08:46:13Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,520 |
["main/src/com/google/refine/grel/controls/ForRange.java", "main/tests/server/src/com/google/refine/grel/controls/ForRangeTests.java"]
|
Enhance `forRange` to accept negative increments
|
`forRange` allows iteration based on a defined range of numbers and an increment, but currently only works if the increment is positive. Current examples:
```
forRange(0,10,1,i,i) -> [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 ]
forRange(10,0,-1,i,i) -> [ ]
```
This is different to the `range()` function where a negative increment is accepted:
```
range(0,10,1) -> [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 ]
range(10,0,-1) -> [ 10, 9, 8, 7, 6, 5, 4, 3, 2, 1 ]
```
### Proposed solution
forRange could be enhanced to work with a negative increment
It might also be a good idea to reject invalid arguments - for example currently:
```
forRange(10,0,1,i,i)
```
As there is no way of incrementing by 1 to get from 10 to zero, this should be rejected as invalid parameters, whereas instead it currently returns an empty array
### Alternatives considered
It is possible to use `range` combined with `forEach`:
```
forEach(range(10,0,-1),v,v) -> [ 10, 9, 8, 7, 6, 5, 4, 3, 2, 1 ]
```
This isn't hugely more complicated but it seems that `range` and `forRange` should be consistent with each other
|
https://github.com/OpenRefine/OpenRefine/issues/5520
|
https://github.com/OpenRefine/OpenRefine/pull/5629
|
9f8937460869a34918ec7b40f84ca835a6c41fc3
|
558771eeff8d756428297c0333efc16e18d81c34
| 2022-12-14T14:49:10Z |
java
| 2023-02-27T07:05:50Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,513 |
["main/webapp/modules/core/scripts/facets/list-facet.js", "main/webapp/modules/core/styles/project/facets.css"]
|
Fit height of text facet to its contents when few options are present
|
When creating a text facet in a column where few different values are present, I get a facet with a lot of whitespace:
### Proposed solution
When few options are present, the facet should shrink to a smaller size, without introducing a scroll bar:
When more options are available, it could potentially expand more than its default size, but that should not happen indefinitely: we should introduce a scroll bar past a certain length.
The facets should remain resizable by the user in any case.
### Alternatives considered
None.
### Additional context
Prompted by @ostephens's video on [doing the Advent of Code 2022 with OpenRefine](http://www.meanboyfriend.com/overdue_ideas/2022/12/advent-of-code-with-openrefine/)
|
https://github.com/OpenRefine/OpenRefine/issues/5513
|
https://github.com/OpenRefine/OpenRefine/pull/5514
|
56839d84f7269f9040d7221d5592478f4261e5cd
|
2fd133e638c677f0904a3bde168637fa02cc68d2
| 2022-12-10T16:56:10Z |
java
| 2023-01-26T08:01:48Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,497 |
["main/tests/server/src/com/google/refine/importers/ExcelImporterTests.java"]
|
XLSX test failure
|
A test instability seems to have been uncovered by the CI: https://github.com/OpenRefine/OpenRefine/actions/runs/3627223760/jobs/6116914771
```
Error: Tests run: 704, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 38.319 s <<< FAILURE! - in TestSuite
Error: readXlsxAsText(com.google.refine.importers.ExcelImporterTests) Time elapsed: 0.04 s <<< FAILURE!
java.lang.AssertionError: expected [2022-12-06 06:55] but found [2022-12-06 06:56]
at com.google.refine.importers.ExcelImporterTests.readXlsxAsText(ExcelImporterTests.java:274)
```
The same problem appears in a different job of the same run:
https://github.com/OpenRefine/OpenRefine/actions/runs/3627223760/jobs/6116914844
|
https://github.com/OpenRefine/OpenRefine/issues/5497
|
https://github.com/OpenRefine/OpenRefine/pull/5504
|
2ec45092c3990d632ef2fa29b1dfe746194e8ef4
|
70776e13592bbcae0adf0f5b15f58208f5aa0e1f
| 2022-12-06T07:09:22Z |
java
| 2022-12-07T07:36:29Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,492 |
[".github/workflows/snapshot_release.yml"]
|
Host OpenRefine releases on GitHub releases again
|
The downloads for the 3.6 release series are hosted on Maven Central (Sonatype repository). Unfortunately, the download statistics are not freely accessible and only count `.jar` downloads, which are irrelevant for us at this stage.
We should switch back to hosting packaged artifacts on GitHub releases as this comes with openly available download statistics.
|
https://github.com/OpenRefine/OpenRefine/issues/5492
|
https://github.com/OpenRefine/OpenRefine/pull/5493
|
24ef30580fe7f83830bd30770430acc7b88ae079
|
1e9184c7dd2ef4eaa7099704cfa3e4322b523ea8
| 2022-12-05T06:22:28Z |
java
| 2022-12-06T06:54:06Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,488 |
["main/src/com/google/refine/expr/functions/TimeSinceUnixEpochToDate.java", "main/tests/server/src/com/google/refine/expr/functions/TimeSinceUnixEpochToDateTest.java"]
|
TimeSinceUnixEpochToDate loses precision for millis & microseconds
|
Because it uses the 1-argument version of `Instant.ofEpochSecond()` instead `Instant.ofMillis()` or the 2-arg version `ofEpochSecond`, it's maximum precision is 1 second. The tests use test data without any fractional seconds, so this isn't detected.
### To Reproduce
Look at the updated tests in the associated PR.
### Current Results
Precision is truncated to seconds.
### Expected Behavior
Full precision for the provided number of milliseconds or microseconds.
|
https://github.com/OpenRefine/OpenRefine/issues/5488
|
https://github.com/OpenRefine/OpenRefine/pull/5489
|
c72ca44840a1a9b0774f044e6cf66b3ffe1fe4e5
|
70f833e4150197c30d817b2889ac6804dda2716b
| 2022-12-02T20:44:04Z |
java
| 2022-12-05T05:57:48Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,485 |
["main/src/com/google/refine/exporters/OdsExporter.java", "main/src/com/google/refine/importers/OdsImporter.java"]
|
Remove deprecation warning suppression from ODF Importer/Exporter
|
Since the update to 0.9.0 in #4326 with the resurrection of the historical API, there are no longer deprecation warnings, so this annotation can be removed.
|
https://github.com/OpenRefine/OpenRefine/issues/5485
|
https://github.com/OpenRefine/OpenRefine/pull/5486
|
9725cea0693ca7c3061208baf6217b546b849a21
|
c5debd930067010fe8c5ae9d0ad202d35a75b157
| 2022-12-02T19:16:11Z |
java
| 2022-12-03T09:32:48Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,484 |
["refine", "refine.bat"]
|
refine script fails with weird message if maven download fails
|
Apparently the current Maven version is 3.8.6 and when it fails to download 3.8.5 it gives the error:
```
tar: Error opening archive: Unrecognized archive format
Error: Error while expanding apache-maven-3.8.5-bin.tar.gz
```
when it feeds the 404 HTML page to `tar`
### To Reproduce
1. Uninstall `mvn`
2. Run `refine` script
### Current Results
```
$ ./refine test
Using refine.ini for configuration
-------------------------------------------------------------------------------------------------
You have 1352M of free memory.
Your current configuration is set to use 1024M of memory.
OpenRefine can run better when given more memory. Read our FAQ on how to allocate more memory here:
https://docs.openrefine.org/manual/installing#increasing-memory-allocation
-------------------------------------------------------------------------------------------------
tar: Error opening archive: Unrecognized archive format
Error: Error while expanding apache-maven-3.8.5-bin.tar.gz
```
### Expected Behavior
Ideally, it figures out the current version. At a minimum, say that Maven is a requirement and tell the user to install it.
### Versions<!-- (please complete the following information)-->
- Operating System: Mac OS
- OpenRefine: HEAD
|
https://github.com/OpenRefine/OpenRefine/issues/5484
|
https://github.com/OpenRefine/OpenRefine/pull/5646
|
298495f4ad0b7c4fa258c687d00d279989ecc273
|
c89e3c2e4dc70cea85a84432894f5c2e3363dfd7
| 2022-12-01T19:00:22Z |
java
| 2023-02-28T08:10:32Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,476 |
["conf/openrefine.l4j.ini", "main/src/com/google/refine/util/HttpClient.java", "main/tests/server/src/com/google/refine/util/HttpClientTests.java"]
|
Proxy configuration still not working
|
In OpenRefine 3.6.2, the configuration of proxy is half working after the implementation of #3919.
Indeed, the nonProxyHosts parameter is no taking into account which prevents the use case where some services are outside (access through the proxy) and others are inside the institution (access without the proxy).
The following configuration in openrefine.l4j.ini:
```
-Dhttp.proxyHost=myproxy.mydomain.org
-Dhttp.proxyPort=8080
-Dhttp.nonProxyHosts="*.mydomain.org|localhost|127.0.0.1"
```
should use the proxy ``myproxy.mydomain.org`` for all the hosts except the ones in ``*.mydomain.org``
currently the proxy isused for all the hosts
|
https://github.com/OpenRefine/OpenRefine/issues/5476
|
https://github.com/OpenRefine/OpenRefine/pull/5477
|
76152500a5af8c38b88d34e15b8b01fcf6d0da6e
|
89e6dbe2fe8390a55065e997ff657b89ef065d5f
| 2022-11-27T10:12:26Z |
java
| 2023-02-09T10:39:59Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,465 |
["main/src/com/google/refine/grel/Parser.java", "main/tests/server/src/com/google/refine/grel/GrelTests.java"]
|
Negative integer literals (w/unary minus sign) incorrectly interpreted as floats
|
If you have an integer-only expression you expect the result to remain integer, but here are a couple of instances where using a negative integer can result in a floating point result (note that the only way I know how to test for the resultant type is to use toString):
1. If the first (or only) value of the expression is a negative integer, as in
`toString(-1+2,"%d")`
2. If a variable in a with() or forEach() function is a negative integer, as in
`with(-1,a,toString(2+a,"%d"))`
In both of these cases the result is
_Error: java.util.IllegalFormatConversionException: d != java.lang.Double_
but if you use "%f" in the string conversion you get 1.000000 .
I am using OpenRefine Version 3.6.2 in Safari Version 16.1 on MacOS 12.6.1 .
|
https://github.com/OpenRefine/OpenRefine/issues/5465
|
https://github.com/OpenRefine/OpenRefine/pull/5467
|
c0f7c4d5b5abb4d9fbb93e825e2b6c7faf29e8cf
|
481586c6260b69d9f45ebbcc75fca61c9ec71ed1
| 2022-11-23T11:56:35Z |
java
| 2022-11-24T06:13:08Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,464 |
["main/tests/server/src/com/google/refine/importers/ExcelImporterTests.java"]
|
XLS tests are locale-dependent
|
The tests for XLS importer can fail depending on the locale they are run with.
### To Reproduce
Steps to reproduce the behavior: run `mvn test` in the `fr.FR` locale (unfortunately there does not seem to be a CLI argument to set that).
### Current Results
```
[ERROR] readXlsxAsText(com.google.refine.importers.ExcelImporterTests) Time elapsed: 0.1 s <<< FAILURE!
java.lang.AssertionError: expected [1.1] but found [1,1]
at com.google.refine.importers.ExcelImporterTests.readXlsxAsText(ExcelImporterTests.java:224)
```
### Expected Behavior
The tests should pass.
### Versions<!-- (please complete the following information)-->
- OpenRefine: master branch
### Additional context
Intuitively the locale dependency of OpenRefine's behaviour is expected here, since we are just preserving the string representation of values in the XLS file, which are locale dependent, so I would say we should just ensure the tests accept that.
It could be even better to give the user the option to select the locale in which the values should be represented, but I am not even sure POI lets us configure that.
|
https://github.com/OpenRefine/OpenRefine/issues/5464
|
https://github.com/OpenRefine/OpenRefine/pull/5470
|
104f03e781eb0a3a488cd2b88ac1c9d43ab1442f
|
f510b3362587766975bfffc14e9c90f45a5706bc
| 2022-11-23T08:39:03Z |
java
| 2022-12-01T03:11:23Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,463 |
["main/src/com/google/refine/expr/functions/strings/ReplaceEach.java", "main/tests/server/src/com/google/refine/expr/functions/strings/ReplaceEachTest.java"]
|
replaceEach function not working and mis-documented
|
At https://docs.openrefine.org/manual/grelfunctions#replaceeach-s-a-find-a-replace in the GREL manual, the documentation describes a function replaceEach with syntax
replaceEach(s, a find, a replace)
But the example in the subsequent description uses a function called replaceEachFromArrays, and if you try this example you do indeed get an Unknown function error. However, using replaceEach with the syntax described also always produces an error
_Error: replaceEach expects 3 arguments: string, array of strings to replace, array of replacement strings_
even when using the correct argument types. Try, for instance,
`"The cow jumps over the moon and moos".replaceEach(["th", "moo"], ["ex", "mee"])`
I am using OpenRefine Version 3.6.2 [579a6f7] in Safari Version 16.1 on MacOS 12.6.1 .
|
https://github.com/OpenRefine/OpenRefine/issues/5463
|
https://github.com/OpenRefine/OpenRefine/pull/5650
|
6e2edf04022e96cdc4680f128e13000c211fdb78
|
a04cf0a7f708be1b983c22c8a00c54bb3e449400
| 2022-11-22T21:07:09Z |
java
| 2023-03-03T18:15:03Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,452 |
["extensions/wikibase/src/org/openrefine/wikibase/qa/scrutinizers/QuantityScrutinizer.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/qa/scrutinizers/QuantityScrutinizerTest.java"]
|
Warning for missing unit when making schema to Wikidata
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
### To Reproduce
Steps to reproduce the behavior:
1. Have a dataset ready to upload to Wikidata
2. Make a schema for a quantity property, but where Wikidata does not expect units, for example students count (P2196)
This video shows this error (link is time stamped to when I discover it): https://youtu.be/oSXxraHgbLs?t=1594
### Current Results
<!-- What results occurred or were shown. -->
Under issues a warning like this is issued:
> **Units missing for students count (P2196)**
> Values such as `whatever number you had` on [Wikidata item] are expected to have units.
### Expected Behavior
<!-- A clear and concise description of what you expected to happen or to show. -->
No warning should be given.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
This was in Chrome Version 107.0.5304.87 on Ubuntu 22.04 using OpenRefine in Wikimedia PAWS, 28 October, 2022.
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, ping us on our [mailing list](https://groups.google.com/forum/#!forum/openrefine) -->
[Studenter-universitet.csv](https://github.com/OpenRefine/OpenRefine/files/10043481/Studenter-universitet.csv)
|
https://github.com/OpenRefine/OpenRefine/issues/5452
|
https://github.com/OpenRefine/OpenRefine/pull/6359
|
05ff957f27aae8548bd241b78c0b0da7818df977
|
a2199db761898d1b68b7dd73eaec1315998076f6
| 2022-11-18T17:26:27Z |
java
| 2024-02-14T20:59:19Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,449 |
["extensions/wikibase/module/scripts/schema-alignment-tab.html", "extensions/wikibase/module/styles/schema-alignment.css"]
|
UI fix: improve position of add media / add item button in Schema tab
|
This issue is transferred from a comment in #5050, so it's easier to track separately.
The UI fix in question is to move the "add item" / "add media" button to the left of the schema building window and thus make it more prominent. See screenshots below (context-specific to Wikibase chosen):
Complete wireframes (including development parameters are here): https://xd.adobe.com/view/3ae13acb-7afb-4ea5-8252-351a1622abba-6fb2/
_Originally posted by @lozanaross in https://github.com/OpenRefine/OpenRefine/issues/5050#issuecomment-1193187217_
|
https://github.com/OpenRefine/OpenRefine/issues/5449
|
https://github.com/OpenRefine/OpenRefine/pull/5676
|
696a93ff9e0adb4d9e77096dbbaaf7e3cbf30a61
|
da839f62323317dfc207e76ab69f0ed00df31637
| 2022-11-17T18:31:44Z |
java
| 2023-04-05T11:19:03Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,433 |
["extensions/wikibase/module/langs/translation-en.json", "extensions/wikibase/src/org/openrefine/wikibase/editing/MediaFileUtils.java", "extensions/wikibase/src/org/openrefine/wikibase/qa/scrutinizers/FileNameScrutinizer.java", "extensions/wikibase/src/org/openrefine/wikibase/schema/WbMediaInfoEditExpr.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/schema/WbMediaInfoEditExprTest.java"]
|
Improve handling of file name normalization for file upload in Wikibase extension
|
Currently, uploading files with the Wikibase extension will fail if the proposed filenames contain characters that should be normalized, such as `:` or `/`.
### Proposed solution
Instead, normalize those characters ahead of time (at schema evaluation time) and generate a QA warning indicating that this normalization has taken place.
### Alternatives considered
We could also not do the normalization and just block the upload with a critical QA warning, forcing the user to do the normalization themselves (so that they control how it is done), but that is needlessly rude I would say.
### Additional context
Bug found by @trnstlntk and @ookgezellig.
|
https://github.com/OpenRefine/OpenRefine/issues/5433
|
https://github.com/OpenRefine/OpenRefine/pull/5435
|
243c60e71e4c8a85d671fc93488935125ac3052c
|
d65b6da5d5287897bdd0872da4bf8e8af618f919
| 2022-11-12T16:10:04Z |
java
| 2022-11-19T21:18:00Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,431 |
["main/src/com/google/refine/commands/project/ImportProjectCommand.java", "main/tests/cypress/cypress/e2e/create-project/preview_project.cy.js", "main/tests/cypress/cypress/e2e/import-project/import_project.cy.js", "main/webapp/modules/core/langs/translation-en.json", "main/webapp/modules/core/scripts/index/import-project-ui.html", "main/webapp/modules/core/scripts/index/import-project-ui.js"]
|
Import project archive from URL
|
From the OR start screen you choose "Import project", which enables the user to select a local project tar.gz file for upload into OR
It would be handy to also have the possiblilty the import a project archive directly from URL, such as (for instance) https://github.com/KBNLwikimedia/OpenRefine-WikimediaCommons-Workshop/blob/main/Dataset-WorkshopOpenRefine-WikiconNL19112022-xlsx.openrefine.tar.gz?raw=true
|
https://github.com/OpenRefine/OpenRefine/issues/5431
|
https://github.com/OpenRefine/OpenRefine/pull/5892
|
ce94ad79a7e60ee23776098afd5ea7fb3a490522
|
81daa2a6b80e098cdec1411afa5c2266f4f524d9
| 2022-11-12T12:22:36Z |
java
| 2023-07-01T04:15:21Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,430 |
["extensions/wikibase/module/langs/translation-en.json", "extensions/wikibase/src/org/openrefine/wikibase/editing/MediaFileUtils.java", "extensions/wikibase/src/org/openrefine/wikibase/qa/scrutinizers/FileNameScrutinizer.java", "extensions/wikibase/src/org/openrefine/wikibase/schema/WbMediaInfoEditExpr.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/schema/WbMediaInfoEditExprTest.java"]
|
Add QA warning for file upload with filename starting with a namespace prefix
|
When uploading a file to a Wikibase, if the filename supplied by the user starts with `File:` (or another namespace in use on the wiki) then the upload will fail.
### To Reproduce
Steps to reproduce the behavior:
1. First, create a wikimedia commons schema for file upload
2. Then, make sure the filenames start with `File:`
3. Finally, attempt the upload
### Current Results
No file is uploaded
### Expected Behavior
The upload dialog should warn that the wiki will reject the uploads based on the filenames.
### Versions<!-- (please complete the following information)-->
- OpenRefine: 3.7-snapshot
### Additional context
The list of namespace prefixes and their aliases can be extracted from the API, such as https://fr.wikipedia.org/w/api.php?action=query&meta=siteinfo&siprop=namespaces&formatversion=2.
Identified by @trnstlntk and @ookgezellig at the Wikimedia NL hackathon today
|
https://github.com/OpenRefine/OpenRefine/issues/5430
|
https://github.com/OpenRefine/OpenRefine/pull/5435
|
243c60e71e4c8a85d671fc93488935125ac3052c
|
d65b6da5d5287897bdd0872da4bf8e8af618f919
| 2022-11-12T12:14:09Z |
java
| 2022-11-19T21:18:00Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,420 |
["extensions/wikibase/src/org/openrefine/wikibase/editing/MediaFileUtils.java", "extensions/wikibase/src/org/openrefine/wikibase/updates/MediaInfoEdit.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/editing/MediaFileUtilsTest.java"]
|
[OR 3.7 Snapshot 20221108] Upload local media files to WM Commonms not working/stalls
|
Continuing from https://github.com/OpenRefine/OpenRefine/issues/5418, I've adapted my OR 3.7 snapshot project to include uploads from local files (instead of URLs):
I build my OR schema
, and when I try to upload to Commons, all seems OK, without error messages
So I press the "Upload edits" button, and the upload to Commons seems to begin...
but nothing happens, no files actually uploaded... seems to stall for ever...
I check the console: error messages
================================
13:42:01.709 [..mWikibaseEditsOperation] Performing edits (31ms)
13:42:01.709 [..ting.EditBatchProcessor] Requesting documents (0ms)
Exception in thread "Thread-3" java.lang.NullPointerException
at org.wikidata.wdtk.wikibaseapi.MediaInfoIdQueryAction.getMediaInfoIds(MediaInfoIdQueryAction.java:68)
at org.wikidata.wdtk.wikibaseapi.WikibaseDataFetcher.getMediaInfoIdsByFileName(WikibaseDataFetcher.java:377)
at org.wikidata.wdtk.wikibaseapi.WikibaseDataFetcher.getMediaInfoIdsByFileName(WikibaseDataFetcher.java:337)
at org.wikidata.wdtk.wikibaseapi.WikibaseDataFetcher.getMediaInfoIdByFileName(WikibaseDataFetcher.java:309)
at org.openrefine.wikibase.editing.MediaFileUtils$MediaUploadResponse.getMid(MediaFileUtils.java:283)
at org.openrefine.wikibase.updates.MediaInfoEdit.uploadNewFile(MediaInfoEdit.java:251)
at org.openrefine.wikibase.editing.EditBatchProcessor.performEdit(EditBatchProcessor.java:167)
at org.openrefine.wikibase.operations.PerformWikibaseEditsOperation$PerformEditsProcess.run(PerformWikibaseEditsOperation.java:262)
at java.base/java.lang.Thread.run(Unknown Source)
|
https://github.com/OpenRefine/OpenRefine/issues/5420
|
https://github.com/OpenRefine/OpenRefine/pull/5421
|
4c0d5318422c6f0020fe63aaa05bb8b14c5b15ba
|
7d4220027197de9b28b9160de54c09b001d7b405
| 2022-11-09T12:46:47Z |
java
| 2022-11-12T06:37:49Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,418 |
["extensions/wikibase/src/org/openrefine/wikibase/qa/scrutinizers/NewEntityScrutinizer.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/qa/scrutinizers/NewEntityScrutinizerTest.java"]
|
[OR 3.7 Snapshot 20221108] Upload URLs of media files not accepted (as they used to be)
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
I'm using the latest (20221108) snapshot version of OR. I want to upload images from the KB (national library of the Netherlands) to Wikimedia Commons.
The URLs to the full images are typically formated as follows: http://resolver.kb.nl/resolve?urn=urn:gvn:KONB16:533939704&role=page&count=1&size=large. When I open this URL in a browser, it displays the image to be uploaded to Commons.
I've made an OR project where I try to make a new Commons file to be called "
File:De Nieuwe Haven van Texel met het dorp 't Schilt in 't verschiet - Nederlandsche havengezichten enz. - KONB16-533939704 - Prent 1 van 18.jpg", with the source the mentioned URL
(See the attached Excel "Dataset_WorkshopOpenRefine_WikiconNL19112022.xlsx" for the exact data I want to upload -
[Dataset_WorkshopOpenRefine_WikiconNL19112022.xlsx](https://github.com/OpenRefine/OpenRefine/files/9970167/Dataset_WorkshopOpenRefine_WikiconNL19112022.xlsx))
When I try to upload to Commons, OR gives an error, that the URL-format is invalid and can't be accepted.
I would expect such URLs to be accepted (again), because with an earlier version of OR snapshot 3.7 , I uploaded all files in https://commons.wikimedia.org/wiki/Category:400_years_The_Netherlands_%E2%80%93_Japan_(Memory_of_The_Netherlands), using the same URL-syntax, see column H in the Excel "Nederland-Japan.xlsx". So in the past these URLs were accepted by OR.
[Nederland-Japan.xlsx](https://github.com/OpenRefine/OpenRefine/files/9970154/Nederland-Japan.xlsx)
### Versions<!-- (please complete the following information)-->
- Operating System: Window 10 Pro
- Browser Version: Firefox 106.0.4 (32-bits)
- JRE or JDK Version: <!-- output of "java -version" e.g. JRE 1.8.0_181 -->
- OpenRefine: 3.7 Snapshot (20221108)
|
https://github.com/OpenRefine/OpenRefine/issues/5418
|
https://github.com/OpenRefine/OpenRefine/pull/5419
|
db079bfc2f1e1e6ce91dc1d9ae7d8e8610c12f0d
|
ab7e871eceff052e98778f9ddc96ee0955f9ace7
| 2022-11-09T11:01:40Z |
java
| 2022-11-10T11:01:26Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,394 |
["main/src/com/google/refine/expr/functions/xml/SelectXml.java"]
|
Help for select() has parameter order backwards
|
The help text for the GREL `select()` function has the parameter order backwards. I haven't checked, but I wouldn't be surprised if the error propagated to the user manual.
### Current Results
> select | (string s, element e)
### Expected Behavior
select | (element e, string s)
### Versions<!-- (please complete the following information)-->
- OpenRefine: 3.6.2
|
https://github.com/OpenRefine/OpenRefine/issues/5394
|
https://github.com/OpenRefine/OpenRefine/pull/5502
|
2f38095d04dd5bdc690df6fcda082e13f348c0a7
|
a6f9095c7149ea1ba2ae4d39803f3f6cc7e2ec25
| 2022-10-30T23:44:55Z |
java
| 2022-12-08T08:03:39Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,390 |
["main/src/com/google/refine/expr/functions/ToDate.java", "main/src/com/google/refine/importers/ExcelImporter.java", "main/tests/server/src/com/google/refine/importers/ExcelImporterTests.java"]
|
Unsupported datatypes should be imported as strings
|
There are a large number of Excel date, time, duration, post code, etc formats which are not supported by OpenRefine, but are currently imported as either numbers or datetimes. Anything unsupported should be imported as a string instead of a number or datetime.
### To Reproduce
Steps to reproduce the behavior:
1. Import an Excel worksheet containing dates such as `09:46:00 AM` in a column
2. Open the project
### Current Results
The cell is converted to `1899-12-31T09:46:00Z`.
### Expected Behavior
The column should be kept as a string.
### Additional context
I have not found any way to disable cast on import.
|
https://github.com/OpenRefine/OpenRefine/issues/5390
|
https://github.com/OpenRefine/OpenRefine/pull/5468
|
fc3a4765d67a1b6e923f34dba627164582adac14
|
ff84c80c5ae64a619e123d69f637f6bf1afcb130
| 2022-10-27T13:15:47Z |
java
| 2022-12-02T17:15:26Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,389 |
["main/src/com/google/refine/expr/functions/ToDate.java", "main/src/com/google/refine/importers/ExcelImporter.java", "main/tests/server/src/com/google/refine/importers/ExcelImporterTests.java"]
|
All numbers are imported as floats by XLS(X) importer
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
### To Reproduce
Steps to reproduce the behavior:
1. Export a table containing an integer column
### Current Results
A cell showing the integer `123` is exported as `123.0`. This makes SQL exports fail with the following error: `MESSAGE: 123.0 is not compatible with column type :INT`.
### Expected Behavior
What you see is what you get: the software should not export cells in a different format.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: Linux
- Browser Version: Firefox
- JRE or JDK Version: OpenJDK Runtime Environment 17.0.4+8-nixos
- OpenRefine: 3.6.2
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, ping us on our [mailing list](https://groups.google.com/forum/#!forum/openrefine) -->
### Additional context
<!-- Add any other context about the problem here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5389
|
https://github.com/OpenRefine/OpenRefine/pull/5468
|
fc3a4765d67a1b6e923f34dba627164582adac14
|
ff84c80c5ae64a619e123d69f637f6bf1afcb130
| 2022-10-27T13:09:03Z |
java
| 2022-12-02T17:15:26Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,388 |
["main/src/com/google/refine/exporters/sql/SqlCreateBuilder.java", "main/src/com/google/refine/exporters/sql/SqlInsertBuilder.java"]
|
SQL export does not put quote around column names
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
### To Reproduce
Steps to reproduce the behavior:
1. Export project as SQL
### Current Results
Column names are not quoted. If a column is named `Plot`, `Plot VARCHAR(255) NULL,` will be printed. Creating such table in Postgres would result in a lowercase `plot` column, if the column name is not quoted.
### Expected Behavior
Column names should be quoted always or when needed.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: Linux
- Browser Version: Firefox
- JRE or JDK Version: OpenJDK Runtime Environment 17.0.4+8-nixos
- OpenRefine: 3.6.2
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, ping us on our [mailing list](https://groups.google.com/forum/#!forum/openrefine) -->
### Additional context
<!-- Add any other context about the problem here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5388
|
https://github.com/OpenRefine/OpenRefine/pull/5674
|
e5677efb0c818552060f27d6ee922b333ba9eb48
|
c4f4a11aa9cad3797d19a2dd1b06d89fc690a827
| 2022-10-27T13:00:57Z |
java
| 2023-04-18T09:18:41Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,387 |
["main/src/com/google/refine/ProjectManager.java", "main/src/com/google/refine/ProjectMetadata.java", "main/src/com/google/refine/commands/project/SetProjectTagsCommand.java", "main/src/com/google/refine/commands/row/GetRowsCommand.java", "main/src/com/google/refine/history/History.java", "main/src/com/google/refine/model/Project.java", "main/tests/server/src/com/google/refine/ProjectManagerTests.java", "main/tests/server/src/com/google/refine/RefineTest.java", "main/tests/server/src/com/google/refine/history/HistoryTests.java", "main/tests/server/src/com/google/refine/model/ProjectStub.java"]
|
Rows are counted only after the project is opened
|
<!-- Describe the bug - Please add a clear and concise description of the bug above this line. You can delete this line if you want. It will be hidden in the final bug report -->
### To Reproduce
Steps to reproduce the behavior:
1. First, upload a file via API
2. Then, open the projects' list
### Current Results
`Rows` column shows 0 rows.
### Expected Behavior
Rows should be displayed.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: Linux
- Browser Version: Firefox
- JRE or JDK Version: OpenJDK Runtime Environment 17.0.4+8-nixos
- OpenRefine: 3.6.2 and 3.6.1
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, ping us on our [mailing list](https://groups.google.com/forum/#!forum/openrefine) -->
### Additional context
Opening the project link updates the `rows` column.
Workaround: GET request toward `/command/core/get-rows?project=PROJECT&start=0&limit=0`.
|
https://github.com/OpenRefine/OpenRefine/issues/5387
|
https://github.com/OpenRefine/OpenRefine/pull/5655
|
002310972764fbf34105764f18cfec44b520df96
|
4d2519f38db8b4f5d1ec42598a4fae24b193e830
| 2022-10-27T08:57:53Z |
java
| 2023-03-17T20:47:05Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,385 |
["extensions/wikibase/src/org/openrefine/wikibase/editing/EditBatchProcessor.java", "extensions/wikibase/tests/src/org/openrefine/wikibase/editing/EditBatchProcessorTest.java"]
|
OpenRefine upload to Wikidata stalls on deleted items
|
When you have a list of items and reconcile them "as identifiers", it often happens that some of the items had already been deleted from Wikidata. But they still reconcile.
However, when the Wikidata upload gets to this item, it gives the following error:
Exception in thread "Thread-12" java.lang.NullPointerException
And then it basically just stops the upload...
Steps to reproduce:
Reconcile Q56640589 as Wikidata identifier
try to import something
Expected behavior:
Skip the deleted item. Or do not reconcile it at all.
This is OpenRefine 3.5.2.
Haven't tried it yet in a newer version.
|
https://github.com/OpenRefine/OpenRefine/issues/5385
|
https://github.com/OpenRefine/OpenRefine/pull/5895
|
dd768bd3a0bae11637cdca3b10f9527d6fe43c08
|
27a66a5ee67eca02d208be8fc74cbc22be11ae7f
| 2022-10-25T12:28:13Z |
java
| 2023-06-15T09:37:01Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,384 |
["refine.bat"]
|
refine.ini is read too late in refine.bat
|
When using the Windows kit with embedded Java with refine.bat starter, the refine.ini file is read after the "Check JAVA_HOME".
So a manually set JAVA_HOME variable in refine.ini is never read before the JAVA_HOME check.
The whole "Check JAVA_HOME" block should be moved below the "Read ini file" block
### To Reproduce
Steps to reproduce the behavior:
1. Use Windows kit with embedded Java
2. Open Windows Command Line in openrefine directory
3. Run refine.bat or refine.bat /c <path_to_refine.ini>
### Current Results
**JAVA_HOME check fails**
H:\openrefine>refine.bat
H:\openrefine>rem Previous line hides the remarks from being displayed with the prompt for each line
H:\openrefine>rem Change current working directory to directory of the batch script
...
H:\openrefine>rem --- First two utilities for exiting --------------------------------------------
H:\openrefine>goto endUtils
H:\openrefine>set OPTS=
H:\openrefine>rem --- Check JAVA_HOME ---------------------------------------------
H:\openrefine>if not "" == "" goto gotJavaHome
H:\openrefine>echo You must set JAVA_HOME to point at your Java Development Kit installation
You must set JAVA_HOME to point at your Java Development Kit installation
H:\openrefine>echo.
H:\openrefine>echo If you don't know how to do this, follow the instructions at
If you don't know how to do this, follow the instructions at
H:\openrefine>echo.
H:\openrefine>echo http://bit.ly/1c2gkR
http://bit.ly/1c2gkR
H:\openrefine>echo.
H:\openrefine>goto fail
H:\openrefine>echo Type 'refine /h' for usage.
Type 'refine /h' for usage.
H:\openrefine>goto end
### Expected Behavior
**Here, I've moved the "Check JAVA_HOME" block after the "Read ini file" block and openrefine starts as expected:**
H:\openrefine>refine.bat
H:\openrefine>rem Previous line hides the remarks from being displayed with the prompt for each line
H:\openrefine>rem Change current working directory to directory of the batch script
....
H:\openrefine>rem --- First two utilities for exiting --------------------------------------------
H:\openrefine>goto endUtils
H:\openrefine>set OPTS=
H:\openrefine>rem --- Argument parsing --------------------------------------------
H:\openrefine>if """" == """" goto readIniFile
H:\openrefine>rem --- Read ini file -----------------------------------------------
H:\openrefine>if "" == "" set REFINE_INI_PATH=refine.ini
H:\openrefine>if not exist refine.ini (
echo The system cannot find the file refine.ini
exit /B 1
)
H:\openrefine>echo Using refine.ini for configuration
Using refine.ini for configuration
....
H:\openrefine>(set JAVA_HOME=H:\openrefine\server\target\jre )
....
H:\openrefine>(
echo -----------------------
echo PROCESSOR_ARCHITECTURE = AMD64
echo JAVA_HOME = H:\openrefine\server\target\jre
echo java release = 11
echo java -version = 11.0.16.1
echo freeRam =
echo REFINE_MEMORY = 1400M
echo -----------------------
....
### Versions<!-- (please complete the following information)-->
- Operating System: All Windows versions
- Browser Version: n/a
- JRE or JDK Version: n/a
- OpenRefine: OpenRefine 3.6.2, but also OpenRefine 3.5.2
|
https://github.com/OpenRefine/OpenRefine/issues/5384
|
https://github.com/OpenRefine/OpenRefine/pull/5404
|
475cf76d4147604a0df5e890f730bd4c8241c495
|
a41f9b30b92d85a8d3679454d3d03bb9a3474926
| 2022-10-25T12:19:23Z |
java
| 2022-11-04T07:04:39Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,371 |
["docs/docs/manual/grelfunctions.md", "docs/versioned_docs/version-3.4/manual/grelfunctions.md", "docs/versioned_docs/version-3.5/manual/grelfunctions.md", "docs/versioned_docs/version-3.6/manual/grelfunctions.md"]
|
Error in the GREL partition documentation
|
There is an error in the GREL partition documentation on docs.openrefine.org
Reported on the mailing list https://groups.google.com/g/openrefine/c/KwUyRM7VBB4/m/FVkX6T2bAAAJ
> This is a piece of text of partition() documentation:
>
> partition(s, s or p fragment, b omitFragment (optional))[](https://docs.openrefine.org/manual/grelfunctions#partitions-s-or-p-fragment-b-omitfragment-optional)
> Returns an array of strings [ a, fragment, z ] where a is the substring within s before the first ..............................The expresion "abcdefgh".partition(/c.e/) will output [“abc”, "cde", defgh” ].
>
> I think that:
>
> "abcdefgh".partition(/c.e/)
>
> [ "ab", "cde", "fgh" ]
>
> and not:
>
> [“abc”, "cde", defgh” ]
|
https://github.com/OpenRefine/OpenRefine/issues/5371
|
https://github.com/OpenRefine/OpenRefine/pull/5377
|
271459984d37519259fb5b8ca70885765c22b8dd
|
3233f6ed4bb8e1a75b969eed34800cd1034fd1f3
| 2022-10-24T11:29:07Z |
java
| 2022-10-25T00:52:02Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,359 |
["refine.bat"]
|
Windows refine.bat needs to be updated for new arguments matching Linux refine script
|
Windows refine script arguments `/d` and `/x` no longer match the Linux refine script and causes confusion for those that maintain across both OS's.
We introduced a few new arguments in the Linux refine script over the last couple of years
- `-d <path> path to the data directory`
- `-x <name=value> additional configuration parameters to pass to OpenRefine`
- `-v <level> verbosity level [from low to high: error,warn,info,debug,trace]`
and replaced two with different arguments (which should no longer are `/d` and `/x` as above but instead mimic the Linux refine script as
- `--debug enable JVM debugging (on port 8000)`
- `--jmx enable JMX monitoring (for jconsole and jvisualvm)`
### To Reproduce
Steps to reproduce the behavior:
1. Start OpenRefine from Cmd Prompt with `refine /d "D:\myOpenRefineWorkspace\"`
2. Notice that log output does not show the data directory being set but instead JVM debugging is enabled
### Current Results
`refine.bat /?` is not consistent with the output from Linux with `./refine -h`
### Expected Behavior
both startup scripts should provide identical arguments, except for differences in shell syntax.
Additionally, `/debug` and `/jmx` should be the corresponding argument syntax in `refine.bat`.
We should also update our user documentation pages where necessary for Windows to reflect the updated arguments.
### Screenshots
<!-- If applicable, add screenshots to help explain your problem. -->
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 11
- Browser Version: Edge latest
- JRE or JDK Version: JDK 11
- OpenRefine: Version 3.7-SNAPSHOT [ea70840]
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, ping us on our [mailing list](https://groups.google.com/forum/#!forum/openrefine) -->
### Additional context
<!-- Add any other context about the problem here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5359
|
https://github.com/OpenRefine/OpenRefine/pull/5673
|
ac33e2e8f8980cc1648686d8e2bb9889032e0233
|
ea017a7ff614aea7d0220e6230c6db18b6322622
| 2022-10-21T05:38:09Z |
java
| 2023-03-11T01:40:20Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,358 |
["extensions/wikibase/module/scripts/schema-alignment.js", "main/tests/cypress/cypress/e2e/extensions/wikibase/wikibase_schema.cy.js", "main/webapp/modules/core/scripts/project/history-panel.js"]
|
No issues notification circle for Entity problems such as missing description or subject item id
|
We don't show a notification circle number next to the issues tab in the Schema view for Entity problems that result in the schema being incomplete.
### To Reproduce
Steps to reproduce the behavior:
1. Use Clipboard importer in OpenRefine and copy and paste the following simple 2 row example dataset
| A | B | C |
| - | - | - |
| 1 | 2 | 3 |
| 4 | 5 | 6 |
2. Click **Create Project**
3. Click on **Wikidata Extension** -> **Edit Wikibase Schema**
4. Click **Add item** button
5. On Terms click **add term** link
6. From dropdown change Alias to Description but do not type anything into boxes.
7. On top header, click on **Issues** tab to ensure there are 2 issues reported for the empty entity.
### Current Results
No notification circle is shown on the Issues tab for the count of 2 issues.
### Expected Behavior
The issues tab should show the notification circle with number 2 signifying 2 issues are reported under the tab.
### Screenshots
No number is shown on the issues tab
However, when the issues tab is clicked, you can see 2 entity issues do exist
I expected to see a number 2 next to the issues tab as in this example:
### Versions<!-- (please complete the following information)-->
- Operating System: Windows 11
- Browser Version: Edge latest
- JRE or JDK Version: embedded in OpenRefine
- OpenRefine: Version 3.7-SNAPSHOT [ea70840]
### Datasets
<!-- If you are allowed and are OK with making your data public, it would be awesome if you can include or attach the data causing the issue or a URL pointing to where the data is.
If you are concerned about keeping your data private, ping us on our [mailing list](https://groups.google.com/forum/#!forum/openrefine) -->
### Additional context
<!-- Add any other context about the problem here. -->
|
https://github.com/OpenRefine/OpenRefine/issues/5358
|
https://github.com/OpenRefine/OpenRefine/pull/5859
|
1b5815f73771c83408ad27d3951453be33654210
|
0bb8c5d45b039a9366d4146627fa608d80c29b6a
| 2022-10-20T00:46:32Z |
java
| 2023-05-27T19:53:33Z |
closed
|
OpenRefine/OpenRefine
|
https://github.com/OpenRefine/OpenRefine
| 5,353 |
["main/pom.xml"]
|
JaCoCo fails to instrument marc4j CodeTableGenerated class due to size
|
The test suite gets an error attempting to use JaCoCo to instrument `org/marc4j/converter/impl/CodeTableGenerated` due to its size. This is non-fatal, but it should be explicitly excluded from instrumentation to reduce the console clutter.
```
java.lang.instrument.IllegalClassFormatException: Error while instrumenting org/marc4j/converter/impl/CodeTableGenerated with JaCoCo 0.8.8.202204050719/5dcf34a.
at org.jacoco.agent.rt.internal_b6258fc.CoverageTransformer.transform(CoverageTransformer.java:94)
at java.instrument/java.lang.instrument.ClassFileTransformer.transform(ClassFileTransformer.java:246)
at java.instrument/sun.instrument.TransformerManager.transform(TransformerManager.java:188)
at java.instrument/sun.instrument.InstrumentationImpl.transform(InstrumentationImpl.java:563)
at java.base/java.lang.ClassLoader.defineClass1(Native Method)
at java.base/java.lang.ClassLoader.defineClass(ClassLoader.java:1016)
at java.base/java.security.SecureClassLoader.defineClass(SecureClassLoader.java:151)
at java.base/jdk.internal.loader.BuiltinClassLoader.defineClass(BuiltinClassLoader.java:825)
at java.base/jdk.internal.loader.BuiltinClassLoader.findClassOnClassPathOrNull(BuiltinClassLoader.java:723)
at java.base/jdk.internal.loader.BuiltinClassLoader.loadClassOrNull(BuiltinClassLoader.java:646)
at java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:604)
at java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:168)
at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:522)
at java.base/java.lang.Class.forName0(Native Method)
at java.base/java.lang.Class.forName(Class.java:377)
at org.marc4j.converter.impl.AnselToUnicode.loadGeneratedTable(AnselToUnicode.java:210)
```
|
https://github.com/OpenRefine/OpenRefine/issues/5353
|
https://github.com/OpenRefine/OpenRefine/pull/5364
|
98e01e8df1b1db8082cf16950e6127af81709715
|
a247881135741be7bd5afadba5298f349737c1f1
| 2022-10-19T00:35:51Z |
java
| 2022-10-22T02:39:02Z |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.