You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@uima.apache.org by sc...@apache.org on 2015/06/25 17:35:01 UTC
svn commit: r1687557 - in /uima/uimaj/trunk: README RELEASE_NOTES.html
Author: schor
Date: Thu Jun 25 15:35:01 2015
New Revision: 1687557
URL: http://svn.apache.org/r1687557
Log:
no Jira - update the uimaj readme and release notes for 2.8.0 release
Modified:
uima/uimaj/trunk/README
uima/uimaj/trunk/RELEASE_NOTES.html
Modified: uima/uimaj/trunk/README
URL: http://svn.apache.org/viewvc/uima/uimaj/trunk/README?rev=1687557&r1=1687556&r2=1687557&view=diff
==============================================================================
--- uima/uimaj/trunk/README (original)
+++ uima/uimaj/trunk/README Thu Jun 25 15:35:01 2015
@@ -1,6 +1,8 @@
- Apache UIMA (Unstructured Information Management Architecture) v2.7.0 SDK
- -------------------------------------------------------------------------
+ Apache UIMA (Unstructured Information Management Architecture)
+ --------------------------------------------------------------
+
+This file applies to UIMA-SDK version 2.8.0 and subsequent versions.
Backwards Compatibility
-----------------------
@@ -10,17 +12,23 @@ expect that upgrading will not break exi
From time to time, non-backward compatible changes are made, such as increased
checking for error situations. This may cause existing pipelines to fail. In
-most cases, there are new JVM -D properties that can be used to disable new
-features, until the underlying issues are corrected or accommodated.
+most cases, there are JVM -D properties that can be used to disable these changes
+until the underlying issues are corrected or accommodated.
+Out of the box,
+ it will do additional checking and fixups for some hard-to-find user errors; see
+ the first section above coping with backwards compatibility issues.
+
The JVM -D property names for these, and their meaning, are:
-Duima.allow_duplicate_add_to_indexes
+
Adding the same exact FS instance to the indexes multiple times no longer
results in duplicate entries in the indexes. You may restore the previous
behavior using this flag.
-Duima.disable_auto_protect_indexes
+
UIMA now checks and automatically protects indexes if users (after adding
a feature structure to the indexes) updates a feature which is being used
as a key. For production runs, after insuring there are no such updates
@@ -29,6 +37,7 @@ The JVM -D property names for these, and
which may make things run slightly faster.
-Duima.report_fs_update_corrupts_index
+
See https://issues.apache.org/jira/browse/UIMA-4135.
Updating Features which are used in Set and Sorted indexes
as "keys" may corrupt the indexes, if the Feature Structure (FS) has been
@@ -43,13 +52,14 @@ The JVM -D property names for these, and
the string While FS was in the index, the feature
-Duima.disable_enhanced_check_wrong_add_to_index
+
See https://issues.apache.org/jira/browse/UIMA-4099.
Feature Structures which are subtypes of AnnotationBase may only be
added to the View corresponding to their Sofa reference. From version
2.7.0, there is additional checking of this which can be disabled
if needed for backward compatibility.
- See the UIMA References http://uima.apache.org/d/uimaj-2.7.0/references.html#ugr.ref.config
+ See the UIMA References http://uima.apache.org/d/uimaj-current/references.html#ugr.ref.config
for a table of these and other settings.
Building from the Source Distribution
@@ -75,23 +85,18 @@ For more details, please see http://uima
What's New
----------
- Please refer to the RELEASE_NOTES.html
-
- Backwards Compatibility
- -----------------------
- This version has changes which can affect backwards compatibility. Out of the box,
- it will do additional checking and fixups for some hard-to-find user errors; see
- the first section above coping with backwards compatibility issues.
-
+ Please refer to the RELEASE_NOTES.html
Supported Platforms
--------------------
Apache UIMA requires Java version 7 or later; it has been tested with Sun/Oracle Java SDK 7, and 8,
and IBM Java 7 and 8.
+
Running the Eclipse plugin tooling for UIMA requires you start Eclipse using a Java 7 or later, as well.
-The supported platforms are: Windows, Linux, and Mac OS X.
-Other platforms and Java (7+) implementations should work, but have not been significantly tested.
+
+The supported platforms are: Windows, Linux, and Mac OS X. Other platforms and Java (7+)
+implementations should work, but have not been significantly tested.
Many of the scripts in the /bin directory invoke Java. They use the value of the environment variable, JAVA_HOME,
to locate the Java to use; if it is not set, they invoke "java" expecting to find an appropriate Java in your PATH.
Modified: uima/uimaj/trunk/RELEASE_NOTES.html
URL: http://svn.apache.org/viewvc/uima/uimaj/trunk/RELEASE_NOTES.html?rev=1687557&r1=1687556&r2=1687557&view=diff
==============================================================================
--- uima/uimaj/trunk/RELEASE_NOTES.html (original)
+++ uima/uimaj/trunk/RELEASE_NOTES.html Thu Jun 25 15:35:01 2015
@@ -20,10 +20,10 @@
-->
<html>
<head>
- <title>Apache UIMA v2.7.0 Release Notes</title>
+ <title>Apache UIMA v2.8.0 Release Notes</title>
</head>
<body>
-<h1>Apache UIMA (Unstructured Information Management Architecture) v2.7.0 Release Notes</h1>
+<h1>Apache UIMA (Unstructured Information Management Architecture) v2.8.0 Release Notes</h1>
<h2>Contents</h2>
<p>
@@ -81,123 +81,47 @@
<h2><a id="major.changes">Major Changes in this Release</a></h2>
-<h3>Java 7 minimum level</h3>
-<p>Java 7 is now the minimum level of Java required.</p>
-
-<h3>Several JVM properties support backwards compatibility</h3>
-<p>See the README and <a target="_blank" href="http://uima.apache.org/d/uimaj-2.7.0/references.html#ugr.ref.config">
-the Reference chapter</a> for a description of these.
-
-<h3>JSON serialization support</h3>
-<p>
- JSON serialization support is added for Type System Descriptions, and for CASs.
- Several formats for JSON CAS serialization are provided, please see the chapter in the
- UIMA reference documentation for details.
-</p>
-
-<h3>Sorted and Bag indexes no longer store multiple instances of identical FSs</h3>
-<p>
-The meaning of "bag" and "sorted" index has been made consistent with how these are handled
-when sending CASes to remote services. This means that adding the same identical FS to the
-indexes, multiple times, will no longer add duplicate index entries. And, removing a FS from an index
-is guaranteed that that particular FS will no longer be in any index. (Before, if you had
-added a particular FS to a sorted or bag index, multiple times, the remove behavior would
-remove just one of the instances). Deserialization of CASes sent to remote services has never
-added Feature Structures to the index multiple times, so this change makes that behavior
-consistent. For more details, see <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-3399">
-Jira issue UIMA-3399</a>.
-</p>
-
-<p>Because some users may need the previous behavior that permitted duplicates of identical Feature Structures in
-the sorted and bag indexes. this change can be disabled, by running the JVM with the defined property
-"uima.allow_duplicate_add_to_indexes".</p>
-
-
-<h3>Index corruption avoidance</h3>
-<p>To prevent potential index corruption, UIMA now recovers (unless disabled by
- <code>"-Duima.disable_auto_protect_indexes"</code>)
- from illegal modifications of features.</p>
-
- <p>These are modifications to features used as index keys, done while the Feature Structure
- being modified is currently in one or more indexes (see
- <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4135">Jira issue UIMA-4135</a>).</p>
-
- <p>Corruption is prevented by first removing the feature structure being updated,
- then doing the update, and then adding the feature structure back to the indexes.
- Because these actions can affect performance, it is recommended that you run with JVM property
- "-Duima.report_fs_update_corrupts_index" in order to see if any user code has this problem, and fix these via
- redesign, or by wrapping the affected areas with a form of <code>protectIndexes()</code>, which does the
- needed removes and add-backs under your control, so you can do several feature updates at once,
- before adding the feature structure back. <code>protectIndexes</code> is described in the
- <a target="_blank" href=
- "http://uima.apache.org/d/uimaj-2.7.0/references.html#ugr.ref.cas.updating_indexed_feature_structures">
- CAS Reference Chapter</a> and the CAS Javadocs; you can also use it with the JCas.</p>
-
- <p>Because this protection is automatic and hidden, if you are iterating over sorted or set indexes, the
- automatic recovery may cause unexpected ConcurrentModificationExceptions to be thrown by the iterator when
- advancing. To work around this, either stop modifying features which are used as keys in the index being iterated over,
- or use Snapshot iterators (see following).</p>
-
-<h3>New Snapshot iterators won't throw ConcurrentModificationException</h3>
-<p>A long-standing difficulty with Feature Structure iterators, namely, that adding to / removing from the underlying
- index being iterated over is not allowed while iterating (unless you use a moveToXXX kind of operation to "reset" the
- iterator), is addressed with a new class of "snapshot" iterators.
- These take a snapshot of the state of the index when the iterator is created; subsequent modifications to the index
- are then permitted, while the iterator continues to iterate over the snapshot it created; these iterators do not
- throw ConcurrentModificationException. The implementation of this feature is via a new method on FSIndex,
- <a target="_blank" href=
- "http://uima.apache.org/d/uimaj-2.7.0/apidocs/org/apache/uima/cas/FSIndex.html#withSnapshotIterators--">
- withSnapshotIterators()</a>, which
- creates a light-weight copy of the the FSIndex instance whose iterator method iterators gets the Snapshot kind.
- This approach allows using the new index in Java's "extended for" statement.</p>
-
- <p>The current implementation of the snapshot iterators makes a snapshot of the index being iterated over, at
- creation time, which has a cost in space and time.</p>
-
-<h3>Other changes</h3>
-<p>Some of the other major changes are listed here; for the complete list, see <a href="issuesFixed/jira-report.html">
-the Issues Fixed report</a>.</p>
-<ul><li> making the JCasGen Eclipse plugin work with more varieties of specifications for class paths.
- Jira issues: UIMA-<a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4080">4080</a>/
- <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4081">4081</a></li>
-
- <li>moveTo(a_Feature_Structure) or creating a new iterator to start at a feature sometimes went to the wrong place.
- Jira issues: <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4094">UIMA-4094</a>
- and <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4105">UIMA-4105</a>.</li>
-
- <li>deserialization of deltaCAS when modifying existing indexed Feature Structures could corrupt the indexes.
- Jira issue: <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4100">UIMA-4100</a>.</li>
-
- <li>default bag indexes will now be created if there are only Set indexes.
- Jira issue: <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4111">UIMA-4111</a>.</li>
+<h3>Iterating over sorted indexes performance improvement</h3>
+<p>Sorted index usage by iterators is monitored, and when it is determined that
+a particular index is no longer being updated and is experiencing poor performance due to
+merging of subindexes for the subtypes being iterated over, a one-time conversion
+of this index to a "flattened" version is done; the flattened form incurs the cost
+of merging subindexes once, and for subsequent uses, iteration speed is improved.
+Any update that affects this flattened index will cause it to be discarded and
+not used for subsequent iterations. Performance improvement should be seen for the
+perhaps common use case where one annotator creates and indexes instances of a type and its subtypes,
+and subsequent annotators read and perhaps update these,
+but don't add or remove entries of this type/subtypes from the index.
+Jira issue <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4357">UIMA-4357</a></p>
+
+<h3>Other major changes</h3>
+<ul><li>The Document Analyzer's Java-based Cas Viewer was extended to support being
+able to select which annotations are highlighted, using a selection approach where you can
+specify a type plus a feature of that type plus one or more values of that feature.
+ Jira issue <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-3374">UIMA-3374</a></li>
- <li><p>Xmi CAS Serialization now checks to see that list and array feature values marked as multipleReferencesAllowed=false
- (or not marked at all, which defaults to multipleReferencesAllowed=false) are not multiply-referenced. If they are,
- they continue to be serialized as if they are independent objects (as was previously done), but now a new
- warning message is issued. Because there can be a huge number of these messages, they are automatically
- throttled down, to prevent running out of room in the error logs.</p>
- <p>The message strings for these look like "Feature [some-feature-name] is marked multipleReferencesAllowed=false,
- but it has multiple references. These will be serialized in duplicate."</p></li>
-
- <li>The CasCopier now checks to insure that the range type of the target feature has the same name as the
- range type of the source feature, which catches errors when two different type systems are used for the source and
- target. For example, this now prevents a feature with range type "uima.cas.Float" from being copied
- into one with range type "uima.cas.String".
- </li>
+ <li>Performance for iterators over UIMA Set Indexes was improved.
+ Jira issue: <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4345">UIMA-4345</a>.</li>
</ul>
- <h3>Performance change highlights</h3>
+ <h3>API changes</h3>
+ <p>Note that we have moved to a more formal semantic versioning, following
+ the ideas in <a target="_blank" href="http://semver.org/">http://semver.org/</a>.</p>
+ <p>
+ There were no API changes that affect backward compatibility. Some new APIs were added or extended.</p>
<ul>
- <li>Iterators obtained for Bag indexes and from the method getAllIndexedFS( ... type ... ) are
- by definition, unordered. The implementation for these iterators is now
- much faster, taking advantage of the unordered aspect of these things and returning items in a more
- efficient sequence.
- Jira issue: <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4166">UIMA-4166</a>.</li>
- <li>CasCopier has been reimplemented and is approximately 5-10 times faster.</li>
- </ul>
+ <li>Add back the API <code>alreadyCopied</code> that was accidentally removed in the CasCopier class in version 2.7.0.
+ Jira issue: <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4428">UIMA-4428</a>.</li>
- <p>The complete list of fixes is <a href="issuesFixed/jira-report.html">here</a>.
-
+ <li>add new variants of <code>getAllIndexedFS, getIndex, and getAnnotationIndex</code> which take an additional argument
+ that specifies the JCas class, and permits these to return the correct generically typed indexes.
+ Jira issue: <a target="_blank" href="https://issues.apache.org/jira/browse/UIMA-4299">UIMA-4299</a>.</li>
+ </ul>
+
+<h2><a id="list.issues">Full list of JIRA Issues Fixed in this Release</a></h2>
+Click <a href="issuesFixed/jira-report.html">issuesFixed/jira-report.hmtl</a> for the list of
+issues fixed in this release.
+
<h2><a id="get.involved">How to Get Involved</a></h2>
<p>
The Apache UIMA project really needs and appreciates any contributions,
@@ -214,9 +138,6 @@ issues you find at
<a href="http://issues.apache.org/jira/browse/uima">http://issues.apache.org/jira/browse/uima</a>
</p>
-<h2><a id="list.issues">List of JIRA Issues Fixed in this Release</a></h2>
-Click <a href="issuesFixed/jira-report.html">issuesFixed/jira-report.hmtl</a> for the list of
-issues fixed in this release.
</body>
</html>
\ No newline at end of file