|
|
 Apache UIMA News
|
ruta-3.3.0
– Apache UIMA Ruta 3.3.0 released
|
The Apache UIMA team is happy to announce the release of Ruta 3.3.0.
Changes include:
- Issue #100: Enable type system discovery via SPI in OSGI
- Issue #102: RutaResourceLoader should consider TCCL
- Issue #107: Enable loading uimaFIT annotators and other classes through the TCCL
- Issue #111: Support copy/paste clipboard for feature values in annotation browser view
- Issue #110: Usage of labels within macro action definition is broken
- Issue #118: Merge type system modules into core module
- Issue #122: Resolve split package between ruta-core and ruta-core-ext
- Issue #104: Upgrade dependencies
Please see the
the complete list of issues fixed for more details of the changes.
|
uimaj-3.4.1
– Apache UIMA Java SDK 3.4.1 released
|
The Apache UIMA team is happy to announce the release of the Apache UIMA Java SDK 3.4.1.
Changes include:
- Issue #304: Unable to deserialize CAS if last element on the heap is an empty array
- Issue #306: Cannot install Ruta 3.3.0 workbench against UIMA 3.4.0 Eclipse
- Issue #303: Build fails on Windows if namespace contains whitespace
- Issue #302: Update dependencies
Please see the
the complete list of issues fixed for more details of the changes.
|
08 Feb 2023
– Apache UIMA uimaFIT 3.4.0 released
|
The Apache UIMA team is happy to announce the release of uimaFIT 3.4.0.
Changes include:
- Issue #195: Detect descriptors via SPI
- Issue #196: Provide OSGI metadata in JARs
- Issue #205: Managed CASes in tests should consider validators
- Issue #209: Ability to override validator for a test
- Issue #211: Conveniently skip validation
- Issue #215: Improve descriptor scanning performance when there are many classloaders
- Issue #220: Add getType signature accepting a type system
- Issue #207: Cannot override CAS content set in a BeforeEach method
- Issue #213: Avoid broken typesystem when using ManagedCas
- Issue #198: Remove version overrides in Maven plugin modules
- Issue #218: Update dependencies
Please see the
the complete list of issues fixed for more details of the changes.
|
02 Feb 2023
– Apache UIMA Java SDK 3.4.0 released
|
The Apache UIMA team is happy to announce the release of the UIMA Java SDK 3.4.0.
Changes include:
- [UIMA-6474] Switch to getDataPathElements() in UIMA context
- Issue #226: Provide SPI interfaces to locate descriptors
- Issue #225: Improve deployment in OSGi environments
- Issue #245: Utility method for loading type systems in SPI providers
- Issue #247: RelativePathResolver should consider TCCL
- Issue #268: UIMA components log and then re-throw exceptions which usually leads to errors being logged twice
- Issue #252: Potential failure to look up FsGenerator3 in OSGI-like contexts
- Issue #265: Unable to release without auto-staging
- Issue #266: Unable to install UIMA 3.3.1 Eclipse Plugins in Eclipse 2022-09
- Issue #267: UIMA Log4jLogger_impl not compatible with log4j 2.18.0+
- Issue #272: select on FSArray seems broken
- Issue #275: Improved error message made it even more misleading
- Issue #285: NPE while deserializing an XMI in a PEAR context
- [UIMA-6440] Stage release artifacts as part of the build
- [UIMA-6443] Fresh Eclipse update site for every release
- [UIMA-6462] Avoid deploy broken checksum files for p2content.xml and artifacts.xml in feature modules
- [UIMA-6463] Use toolchains to ensure compatibility with Java 1.8
- [UIMA-6436] Move maintainer documentation from website into maintainer guide
- Issue #230: Remove version overrides in Maven plugin modules
- Issue #228: Move the UimaDecompiler class
- Issue #283: Update issue report generation
- Issue #292: Make CAS serialization tests more robust
- [UIMA-6459] Upgrade dependencies
- Issue #270: Update dependencies
Please see the the complete list of issues fixed in the
GitHub issue report and the
Jira issue report for more details of the changes.
|
21 Dec 2022
– Retirement of several Apache UIMA sub-projects
|
The Apache UIMA team announces the retirement of
- Apache UIMA Java SDK v2.x - please migrate to v3.x
- Apache uimaFIT v2.x - please migrate to v3.x
- Apache UIMA Ruta v2.x - please migrate to v3.x
- Apache UIMA DUCC - there is currently no replacement
- Apache UIMA AS v2.x - there is currently no replacement
|
31 Oct 2022
– Apache UIMA Java SDK 3.3.1 released
|
The Apache UIMA team is happy to announce the release of the UIMA Java SDK 3.3.1.
Changes include:
- Issue #255: File handle leak accessing performanceTuning.properties
- Issue #240: Helper annotation created by SelectFS should not survive
- Issue #238: Form 6 serializes non-reachable FSes but should not
- Issue #235: Misleading error message when JCas type is not registered
- [UIMA-6479] PearPackagingMavenPlugin has ancient JUnit dependency
- [UIMA-6473] CasToComparableText is broken
- Issue #222: Support comparing test files irrespective of line endings
- [UIMA-6480] Add tests with empty arrays to CAS de/ser-suite
- [UIMA-6454] Update dependencies
- [UIMA-6463] Use toolchains to ensure compatibility with Java 1.8
- [UIMA-6469] Cleaning up file handling code
Please see the the complete list of issues fixed in the
GitHub issue report and the
Jira issue report for more details of the changes.
|
06 Aug 2022
– Apache UIMA Java JSON CAS I/O
|
The JSON CAS I/O implementation for use with the UIMA Java SDK allows to serialize UIMA CAS
data to JSON and to de-serialize the data back from JSON again, loading it into a CAS object.
The aim of this library is to facilitate the data interoperability of UIMA data across
different platform and programming languages. For example, the implementation contains
functionality to deal with the different character offset counting strategies used by
different languages such as Java and Python. A Python-based implementation of the UIMA
JSON CAS format is available as part of the third-party
DKPro Cassis library.
This is the first public release based on the JSON serialization of the Apache UIMA CAS
draft specification version 0.4.0. Please note that the implementation and the specification
are not yet final. For this reason, is not yet recommended to use this library in scenarios
where data needs to be stored or archived over an extended period of time as future versions
of the implementation and specification may introduce incompatibilities with the current
version. Good usage scenarios are for example short-term data exchange between different
UIMA implementations such as for example in network communication.
A full list of issues addressed in this release can be found on the
issue tracker.
Please use the
mailing lists for feedback and the issue
tracker to report bugs.
|
28 Jun 2022
– Apache UIMA Ruta 3.2.0 released
|
The Apache UIMA team is happy to announce the release of Ruta 3.2.0.
Changes include:
- [UIMA-6411] - Avoid creation of `RutaBasics` for bad annotations
- [UIMA-6406] - Removing an annotation inside a `BLOCK` only takes effect outside the block
- [UIMA-6408] - No type check of features in `TRANSFER`
- [UIMA-6409] - Possible endless wildcard lookahead in combination with subtokens
- [UIMA-6414] - Missing match for optional after sidestep out of composed
- [UIMA-6404] - `@` with quantifier ignores matches
- [UIMA-6405] - Local variable not captured properly in a wildcard matching condition.
- [UIMA-6461] - Wrong argument to `contains()`
- [UIMA-6399] - `RutaPatternCache` prevents `CPEEngine` from terminating
- [UIMA-6383] - TRIE - Wordlist entry not annotated
- [UIMA-6394] - Label assignment in alternative match causes problems
Please see the
the complete list of issues fixed for more details of the changes.
|
03 Jun 2022
– Apache UIMA uimaFIT 3.3.0 released
|
The Apache UIMA team is happy to announce the release of uimaFIT 3.3.0.
Changes include:
- [UIMA-6431] - Use lambda functions as CAS processors
- [UIMA-6422] - `FSUtil.setFeature()` should offer signatures that accept a Feature
- [UIMA-6392] - Better delegate key generation in aggregate engine
- [UIMA-6424] - Upgrade uimaFIT to JUnit 5
- [UIMA-6426] - Upgrade to UIMA Java SDK 3.3.0
- [UIMA-6432] - Upgrade dependencies (uimaFIT 3.3.0)
- [UIMA-6384] - Parallelism argument in `CpePipeline` is ignored
- [UIMA-6385] - Potential resource key clash in environments with multiple classloaders
- [UIMA-6391] - `CpePipeline` should kill CPE if reader throws exception
- [UIMA-6396] - uimaFIT maven plugin mixes up test and compile scopes
- [UIMA-6417] - Problems setting numeric parameter values
- [UIMA-6446] - Complexities around enhancing classes with their resource name
Please see the
the complete list of issues fixed for more details of the changes.
|
09 May 2022
– Apache UIMA Java SDK 3.3.0 released
|
The Apache UIMA team is happy to announce the release of the UIMA Java SDK 3.3.0.
Changes include:
- [UIMA-6418] Added support for component parameters of type "long" and "double"
- [UIMA-6358] Added platform-independent methods for setting the datapath in a resource manager
- [UIMA-6374] Added an extensive CAS (de)serialization test suit
- [UIMA-6431] Added support for using lambda functions as CAS processors
- [UIMA-6412] Changed CPMEngine from using a thread group to using an executor service
- [UIMA-6389] Fixed exceptions being swallowed when using Log4J2 through its SLF4J API
- [UIMA-6386] Fixed wrong UIMA session being set on the ConfigurationManager in aggregates
- [UIMA-6390] Fixed NPE when trying to access config names of fresh context
- [UIMA-6378] Fixed build on Java 16
- [UIMA-6393] Fixed circular imports in descriptors breaking the resource manager cache
- [UIMA-6367] Fixed JCas cover annotation created in PEAR context being replaced by index operations
- [UIMA-6388] Fixed CAS.select(null) returning all annotations instead of throwing an exception
- [UIMA-6423] Fixed selecting a non-existing type returning all types instead of throwing an exception
- [UIMA-6421] Fixed range check when injecting a String value into StringArray slot to throw an exception
- [UIMA-6400] Fixed leaking ThreadLocal in UimaContextHolder
- [UIMA-6398] Fixed memory leak in UIMA loggers and loggers using the wrong classloader for i18n messages
- [UIMA-6413] Fixed memory leak in FSClassRegistry
- [UIMA-6377] Fixed spurious multipleReferencesAllowed warning when serializing empty arrays
- [UIMA-6372] Upgraded to JUnit 5
- [UIMA-6373] Format UIMA Core Java SDK codebase
Please see the
the complete list of issues fixed for more details of the changes.
|
02 July 2021
– Apache UIMA Java SDK 2.11.0 released
|
UIMA Java SDK 2.11.0 is in the 2.x.x series of core Java UIMA SDK; the latest series is 3.x.x;
see the news for 3.0.0 for important information about migrating from 2.x.x versions of UIMA.
Changes include:
- Added ability to serialize as XMIs as XML 1.1
- Added typed parameter support to PEARs
- Improved speed of constructing aggregate engines
- Fixed deep cloning of AnalysisEngineDescription
- Fixed bug causing Annotation to be returned when asking JCas for a specific type
- Fixed ability to install PEARs into directories containing XML special characters in the name
- Fixed index protection for cases that no FSes were indexed
- No longer ship Pack200-compressed versions of the Eclipse plugins
Please see the
the complete list of issues fixed for more details of the changes.
|
18 May 2020
– Apache UIMA Ruta 3.1.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special focus
lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, supports the
user and facilitates every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
This release provides compatibility to UIMA v3.
An upgrade from UIMA Ruta v2 to this version
requires also an update to UIMA v3 of the overall application and/or Eclipse installation.
Please refer to the UIMA v3 user's guide for detailed information:
https://uima.apache.org/d/uimaj-3.2.0/version_3_users_guide.html
UIMA Ruta Language and Analysis Engine:
- New config parameter for internal indexing strategy.
- New config parameter for indexing all debug information.
- New config parameter for type disambiguation.
- TextSeeder is the new default seeder (faster, no MARKUP annotations).
- Improved null check using label expression.
- Improved conjunctive and disjunctive rule elements.
- Improved conjunct rules.
- Improved implementation of SPLIT, FILL, ADD, REMOVE, AFTER.
- Improved documentation.
UIMA Ruta Workbench:
- Fixed problems launching script using newer Java versions (LayerInstantiationException).
- Fixed console logging when launching scripts.
- Improved compatibility with dark mode.
UIMA Ruta Maven Plugin:
- Fail on error by default.
For a complete list of bugs and improvements included in this release please see
issue report.
|
06 May 2021
– Apache uimaFIT 3.2.0 released
|
uimaFIT 3.2.0 is a feature and bugfix release. On supported platforms, it serves mostly as
a drop-in replacement for previous uimaFIT 3.x versions. However, the behavior of the various
select methods was slightly adapted in edge cases to align with the update behavior of the
UIMA Java SDK SelectFS API. For details, please refer to the migration section in the
documentation in the Apache UIMA Java SDK 3.2.0.
Changes include:
- Added pluggable CAS validation support
- Added selectOverlapping to (J)CasUtil
- Added ability to Maven plugin to add generated resource folder as Maven resource folder
- Added typed PEAR parameter support
- Improved overhead of createTypeSystemDescription() and friends
- Improved compatibiltiy of the uimaFIT select-methods with the UIMA Java SDK SelectFS API
- Improved performance of selectCovering
- Fixed Maven plugin failing to import type systems from dependencies
- Fixed detection and usage of proper classloader e.g. for resolving imports
- Several additional bug fixes
Please see the
the complete list of issues fixed for more details of the changes.
|
03 May 2021
– Apache UIMA Java SDK 3.2.0 released
|
UIMA Java SDK 3.2.0 is in the 3.x.x series of core Java UIMA SDK; see the news for 3.0.0 for important
information about migrating from 2.x.x versions of UIMA.
Changes include:
- Added AnnotationPredicates utility class providing various predicates testing how annotations
relate to each other (e.g. covering, being covered by, following, preceding, etc.)
- Added single-int arg version of select.startAt()
- Added trim method to AnnotationFS
- Added ability to serialize as XMIs as XML 1.1
- Added ability to serialize as XMIs pretty-printed using CasIOUtils
- Added typed parameter support to PEARs
- Improve performance of setting up JCas classes by reducing sync lock contention
- Improved speed of constructing aggregate engines
- Many additional bug fixes
Please see the
the complete list of issues fixed for more details of the changes.
|
12 May 2020
– Apache uimaFIT 3.1.0 released
|
uimaFIT 3.1.0 is a minor feature and bugfix release. On supported platforms, it serves
mostly as a drop-in replacement for previous uimaFIT 3.x versions. However, due to a clash in
method signatures, you may notice incompatible changes in the methods provided
ExternalResourceFactory. For details, please refer to the migration section in the
documentation.
The complete list of fixes is here.
Notable changes in this release include:
- Added support for PEARs in AnalysisEngineFactory
- Added ExternalResourceFactory.createResource(...) methods for instantiating resources
- Added support for Charset-typed parameters in components
- Added ability to set number of threads in CpePipeline.runPipeline(...)
- Fixed issue with non-XML 1.0 characters in parameter values when running CPEs
- Fixed JCasIterable.iterator() destroying the ResourceManager before it is even used
- Fixed issue causing component parameters to be used when initializing resources
- Fixed CollectionReaderFactory.createReaderDescription(...) not discovering type prios and indexes
- Fixed clashing method signatures of ExternalResourceFactory.bindResource(...)
- Upgraded to UIMA 3.1.1
- Upgraded to Spring 4.3.26
This version of uimaFIT requires Java version 8 or later.
|
26 Apr 2020
– Apache uimaFIT 2.5.0 released
|
uimaFIT 2.5.0 is a minor feature and bugfix release to uimaFIT 2.4.0. It should should
serve as a drop-in replacement for previous uimaFIT 2.x versions.
The complete list of fixes is here.
Notable changes in this release include:
- Added support for PEARs in AnalysisEngineFactory;
- Added support for destroying auto-created resource managers if feasible;
- Added CasFactory;
- Added CasUtil.exists(...) method;
- Added support for Charset-typed parameters in components;
- Added ExternalResourceFactory.createResource(...) methods for instantiating resources;
- Added ability to set number of threads in CpePipeline.runPipeline(...);
- Fixed FSUtil.getFeature(...) failing when called with Object.class on multi-valued features;
- Fixed issue with non-XML 1.0 characters in parameter values when running CPEs;
- Fixed JCasIterable.iterator() destroying the ResourceManager before it is even used;
- Fixed CollectionReaderFactory.createReaderDescription(...) not discovering type prios and indexes;
- Fixed issue causing component parameters to be used when initializing resources;
- Upgraded to UIMA 2.10.4;
- Upgraded to Spring 4.3.26.
This version of uimaFIT requires Java version 8 or later.
|
17 Apr 2020
– Apache UIMA Ruta 3.0.1 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special focus
lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, supports the
user and facilitates every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
This release provides compatibility to UIMA v3 with the functionality of the version 2.8.1.
An upgrade from UIMA Ruta v2 to this version
requires also an update to UIMA v3 of the overall application and/or Eclipse installation.
Please refer to the UIMA v3 user's guide for detailed information:
https://uima.apache.org/d/uimaj-3.1.1/version_3_users_guide.html
Major Changes in this Release
- Fixed broken literal string matching.
- Fixed variable initialization with other variables.
- Fixed assignment of lists to variables.
- Fixed assignment of annotation comparison expression to variables.
- PARSE condition supports additional argument that represents the string to be parsed.
The UIMA Ruta Workbench and in particular the launching of UIMA Ruta
script in simple UIMA Ruta projects faces some incompatibilities with
newer Java versions. Please refer to the UIMA mailing lists for known
workarounds.
For a complete list of bugs and improvements included in this release please see
issue report.
|
11 Apr 2020
– Apache UIMA Ruta 2.8.1 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
- Fixed broken literal string matching.
- Fixed variable initialization with other variables.
- Fixed assignment of lists to variables.
- Fixed assignment of annotation comparison expression to variables.
- PARSE condition supports additional argument that represents the string to be parsed.
The UIMA Ruta Workbench and in particular the launching of UIMA Ruta
script in simple UIMA Ruta projects faces some incompatibilities with
newer Java versions. Please refer to the UIMA mailing lists for known
workarounds.
For a complete list of bugs and improvements included in this release please see
issue report.
|
09 Dec 2019
– Apache UIMA Ruta 3.0.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special focus
lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, supports the
user and facilitates every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
This release provides compatibility to UIMA v3 with the functionality of the version 2.8.0.
An upgrade from UIMA Ruta v2 to this version
requires also an update to UIMA v3 of the overall application and/or Eclipse installation.
Please refer to the UIMA v3 user's guide for detailed information:
https://uima.apache.org/d/uimaj-3.0.0/version_3_users_guide.html
This release contains changed in the analysis engine descriptor and the type system descriptors
and therefore requires an update of the Ruta projects. This can be accomplished for example by
right-click on the Ruta project, selecting "UIMA Ruta" and then "Update Project".
UIMA Ruta Language and Analysis Engine:
- Depends on UIMA 3.1.1 and uimaFIT 3.0.0
- Provides compatibility to UIMA v3
UIMA Ruta Workbench:
- Depends on UIMA v3 and provides compatibility to UIMA v3
For a complete list of bugs and improvements included in this release please see
issue report.
|
28 Nov 2019
– Apache UIMA Ruta 2.8.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
This release contains changed in the analysis engine descriptor and the type system descriptors
and therefore requires an update of the Ruta projects. This can be accomplished for example by
right-click on the Ruta project, selecting "UIMA Ruta" and then "Update Project".
UIMA Ruta Language and Analysis Engine:
- The analysis engine is able to generate debug information about inlined rules which includes also an extension of the ruta type system.
- Type expressions in dot notation for annotation expressions a new supported: a1:ANY a2:ANY{a1.type==a2.type -> Type};
- Matching on string literals is no more restricted to single RutaBasic annotations, e.g., it is now possible to write: "This is a test"{-> Test};
- Dictionary matching is now more robust concerning white spaces in the word list. The parameter dictRemoveWS is now also set to true by default.
- Fixed anchors at composed rule elements.
- Labels at macro actions are supported now.
- Fixed several bugs.
UIMA Ruta Workbench:
- New view for visualizing the explanation of inlined rules.
- Fixed problem with blocked build processes in Ruta projects with many scripts.
- Fixed bugs.
For a complete list of bugs and improvements included in this release please see
issue report.
|
08 Nov 2019
– Apache UIMA Java SDK 3.1.1 released
|
UIMA Java SDK 3.1.1 is in the 3.x.x series of core Java UIMA SDK; see the news for 3.0.0 for important
information about migrating from 2.x.x versions of UIMA.
Changes include:
- Build updates to build from the GIT repositories, after the uimaj project's SCM was moved
from SVN to GitHub.
- Fixed a performance issue when running with 100's of CASes using the same type system.
Please see the
the complete list of issues fixed for more details of the changes.
|
15 Aug 2019
– Apache UIMA Java SDK 3.1.0 released
|
UIMA Java SDK 3.1.0 is in the 3.x.x series of core Java UIMA SDK; see the news for 3.0.0 for important
information about migrating from 2.x.x versions of UIMA.
Changes include:
- JCasGen now generates FSArray values with generic types, if the component type is specified.
This change was responsible for bumping up the middle number of the version spec.
- Fixes to allow the Eclipse plugin for the Component Descriptor Editor to run with the latest Eclipse version.
- Improved support of 0-length typed FSArrays
Please see the
the complete list of issues fixed for more details of the changes.
|
09 Aug 2019
– Apache UIMA Java SDK 2.10.4 released
|
This is a minor bug fix release.
The major changes in this release include:
- The Eclipse plugin for the Component Descriptor editor was fixed to work with
the latest version of Eclipse.
- Internationalized Exceptions more reliably find their message bundles.
- An additional use case for running pipeline within pipelines is now supported.
- An override was added to enable using External DTDs in XML descriptors.
- The destroy cleanups around PEARs was improved.
Please see the
the complete list of issues fixed for more details of the changes.
|
26 Apr 2019
– Apache UIMA DUCC 3.0.0 released
|
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system providing tooling,
management, and scheduling facilities to automate the scale-out of applications written to the UIMA framework.
Core UIMA provides a generalized framework for applications that process unstructured information such as human
language, but does not provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster management of the resources.
DUCC defines a formal job model that closely maps to a standard UIMA pipeline. Around this job model DUCC
provides cluster management services to automate the scale-out of UIMA pipelines over computing clusters.
The complete list of fixes is here.
The major changes in this release are:
- Support for UIMA v2 and v3
- Created new pull service that can be run with or without DUCC
- Enable DUCC to run without shared file system
- Add new DUCC stop options, including quiesce
- Add "CASes processed" to annotator performance metrics
- Upgraded to Cassandra Server v.3.11.3, Cassandra Driver v.3.6.0, Jetty
v.9.4.14.v20181114, guava v.18.0, joda v.2.4, commons.lang v.3.1,
commons.math v.3.2, netty v.4.0.44, snappy v.1.1.1.7
- Other bug fixes
This version of UIMA-DUCC requires Java version 8.
|
16 Apr 2019
– Apache uimaFIT 3.0.0 released
|
uimaFIT 3.0.0 is a major release compatible with UIMA v3. It includes a few changes in the
API which do not make it a drop-in replacement - minor changes to your code may be required
when you upgrade.
The complete list of fixes is here.
The major changes in this release are:
- Upgrade to UIMA 3.0.2.
- Requires Java 8
- New CasFactory (equivalent to JCasFactory but for the CAS API).
- Added CasUtil.exists() method.
- Added CasUtil.selectSingleFS() method.
- Use List instead of Collection in return types where possible.
- Fixed problem that different signatures of createExternalResourceDescription shadow
each other.
- Destroy auto-created resource managers if feasible
- Upgraded dependencies
This version of uimaFIT requires Java version 8 or later.
|
10 Apr 2019
– Apache UIMA Java SDK 3.0.2 released
|
UIMA Java SDK 3.0.2 is in the 3.x.x series of core Java UIMA SDK; see the news for 3.0.0 for important
information about migrating from 2.x.x versions of UIMA.
Changes include:
- Better backwards binary compatibility with UIMA version 2, to allow more migrations
to happen without recompiling user code (except for JCas classes).
- Some bug fixes, especially around the subiterator, for edge cases.
Please see the
the complete list of issues fixed for more details of the changes.
|
24 Feb 2019
– Apache UIMA Ruta 2.7.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
UIMA Ruta Language and Analysis Engine:
- Requires Java 8
- New language feature: label expressions at actions for directly assigning/reusing newly created annotations. Example: Document{-> a:T1, CREATE(T2, "ref" = a)};
- New language feature: new type of rule element for completely optional match which does not require an existing annotation and therefore also works at the boundary of a window/document. Example: NUM _{-PARTOF(CW)};
- Type lists can be used as matching condition.
- Initial default value of string and annotations variables is now null.
- Comparison of annotation and annotation list are now supported.
- New configuration parameter 'inferenceVisitors'.
- New configuration parameter 'maxRuleMatches'.
- New configuration parameter 'maxRuleElementMatches'.
- New configuration parameter 'rulesScriptName'.
- Inlined rules as condition are only evaluated if the rule element match was successful.
- Multiple inlined rule blocks are allowed at one rule element.
- String features with allowed values are supported.
- PlainTextAnnotator supports vertical tabs.
- Various improvements for WORDTABLE.
- Thrown exceptions include script name.
- Fixed values of label for failed matches.
- Fixed inlined rules as condition at wildcards.
- Fixed resetting of annotation-based variables.
- Fixed various bugs of wildcards.
- Fixed CONTAINS condition for annotations overlapping the window.
- Fixed COUNT condition.
- Fixed setting variables by configuration parameter.
UIMA Ruta Workbench:
- Query View support more CAS formats.
- Fixed order of scripts in Applied Rules view.
- Fixed reporting of non-existing problems in editor.
For a complete list of bugs and improvements included in this release please see
issue report.
|
29 Nov 2018a
– Apache UIMA Java SDK 3.0.1 released
|
UIMA Java SDK 3.0.1 is in the 3.x.x series of core Java UIMA SDK; see the news for 3.0.0 for important
information about migrating from 2.x.x versions of UIMA.
Changes include:
- The code and the build was adjusted to compile using both Java 8 and Java 11.
- The class loading when using the UIMA Class Loader now
will first try loading from that class loader's URL list,
and if not found will try using the Context Class Loader (if set), and if not found,
then will default to how it previously searched for classes.
Previously the Context Class Loader was never consulted.
- The PEAR installation no longer lock files.
- Pear Specifier parameters now passed to the wrapped Analysis Engine
- A destroy() method was added to the ResourceManager which closes its UIMAClassLoader
Specific changes in this release unique to version 3 include:
- Multiple small adjustments for edge cases for the select framework, to make it match what uimaFIT does.
- Bug fixes for several of the select methods, with associated documentation and javadoc updates/fixes
Please see the
the complete list of issues fixed for more details of the changes.
|
29 Nov 2018b
– Apache UIMA Java SDK 2.10.3 released
|
This is a minor bug fix release.
The major changes in this release include:
- The minimum Java version was changed from 7 to 8. The code and the build
was adjusted to compile using both Java 8 and Java 11.
- The class loading when using the UIMA Class Loader now will first try loading
from that class loader's URL list, and if not found will try using the
Context Class Loader (if set), and if not found, then will delegate to the parent.
Previously the Context Class Loader was never consulted.
- Fixed some edge case performance issues.
- Fixed PEAR installation to no longer lock files.
- Pear Specifier parameters now passed to the wrapped Analysis Engine
- Implemented a destroy() method for ResourceManager which closes its UIMAClassLoader
Please see the
the complete list of issues fixed for more details of the changes.
|
27 Apr 2018
– Apache UIMA ConceptMapper addon 2.10.2 released
|
UIMA ConceptMapper 2.10.2 updates the Concept Mapper, one of the annotators in the Addons, to use the
current UIMA APIs, allows configuration inputs to come from Jars on the classpath, and adds buffering to the I/O
for compiled dictionaries.
The complete list of fixes is here.
|
17 Apr 2018
– Apache UIMA Asynchronous Scaleout (UIMA-AS) 2.10.3 released
|
UIMA AS 2.10.3 is a maintenance release that contains fixes and improvements over UIMA-AS 2.10.2.
The complete list of fixes is here.
The major changes in this release are:
- Modified client code to assign unique ClientID to broker connection
- Fixed ClassCastException when async aggregate initializes delegate with JMS Service Descriptor
- Fixed broken classpath and logging for UIMA-AS run configurations
This version of UIMA-AS requires Java version 8 or later.
|
19 Mar 2018
– Apache UIMA DUCC 2.2.2 released
|
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system providing tooling,
management, and scheduling facilities to automate the scale-out of applications written to the UIMA framework.
Core UIMA provides a generalized framework for applications that process unstructured information such as human
language, but does not provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster management of the resources.
DUCC defines a formal job model that closely maps to a standard UIMA pipeline. Around this job model DUCC
provides cluster management services to automate the scale-out of UIMA pipelines over computing clusters.
The complete list of fixes is here.
The major changes in this release are:
- Upgraded UIMA (2.10.2), UIMA-AS (2.10.2), ActiveMQ (5.15.2), Spring Framework (4.3.9), Camel (2.19.3), XStream (1.4.10)
- Requires java 8 for builds and runtime. User jobs and services must use java 8.
- Fixed rogue process detector to prevent truncation of user names
- Reduced size of publications between the PM and Agents
- Improved Agent's detection of defunct processes
- Improved logging of system events
- ducc_watcher improvements
- Job and managed reservation specs now saved in DB
- DUCC Web Server (WS) queries DB too often
- DUCC Web Server (WS) should use Resource Manager (RM) calculated value for machine's memory available
- Other bug fixes
This version of UIMA-DUCC requires Java version 8.
|
05 Mar 2018
– Apache UIMA Java SDK 3.0.0 released
|
UIMA Java SDK 3.0.0 is a new version of the core Java UIMA SDK.
This version requires a migration for pre V3 versions which use JCas classes, which will take a little
planning. For those users who do not wish to migrate, please continue to use 2.10.2, which will
continue to be maintained.
Version 3 re-implements Feature Structures as Java Objects; this allows them to be garbage collected.
Many of the internal data structures have been improved and several new capabilities have been added.
Here is a brief overview of the new features, capabilities, and migration tooling.
Requirements: Java 8 and, for any user-defined JCas cover classes, regeneration or migration of those
(a migration tool is provided).
Except for migration of user-defined JCas classes, this new version is designed to backwards compatible with existing
version 2 UIMA pipelines. The
Version 3 User's Guide has more details.
The Release Notes
describe what's changed since the last 3.0.0-beta release.
The Eclipse update site is available, but located in a
different spot,
in order to prevent accidentally installing v3; this version requires Java 8 and the JCasGen generates
v3 style classes. If you want to use this and also need to keep using v2 components, please install these
Eclipse plugins into a separate, newly obtained eclipse download.
|
07 Feb 2018
– Apache UIMA Asynchronous Scaleout (UIMA-AS) 2.10.2 released
|
UIMA AS 2.10.2 is a maintenance release that contains fixes and improvements over UIMA-AS 2.9.0.
The complete list of fixes is here.
The major changes in this release are:
- Updated to use ActiveMQ 5.15.2 (requires Java 1.8+)
- Added dependency on UIMA SDK 2.10.2
- Fixed merging of type priorities
- Added support for targeting specific service instance
- Fixed per CAS Performance Metrics breakdown
- Fixed eclipse examples import
- Fixed dd2spring issues
- Fixed JMS listener recoverey after service restart
- Other bug fixes
This version of UIMA-AS requires Java version 8 or later.
|
14 Nov 2017
– Apache uimaFIT 2.4.0 released
|
uimaFIT 2.4.0 is a minor feature and bugfix release to uimaFIT 2.3.0. It should should
serve as a drop-in replacement for previous uimaFIT 2.x versions.
The complete list of fixes is here.
The major changes in this release are:
- Fixed bug: indexCovered should not return reference annotation;
- Fixed bug: JCasUtil.selectAt has different meaning than expected;
- Upgrade to UIMA 2.10.2;
- Added FSUtil.isMultiValuedFeature signature that works without a feature structure;
- Maven plugin "generate" goal should include type system imports.
This version of uimaFIT requires Java version 7 or later.
|
30 Aug 2017
– Apache UIMA DUCC 2.2.1 released
|
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system
providing tooling, management, and scheduling facilities to automate the scale-out of
applications written to the UIMA framework. Core UIMA provides a generalized framework for
applications that process unstructured information such as human language, but does not
provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster
management of the resources. DUCC defines a formal job model that closely maps to a standard
UIMA pipeline. Around this job model DUCC provides cluster management services to automate
the scale-out of UIMA pipelines over computing clusters.
The complete list of fixes is here.
The major changes in this release are:
- The userid of a privileged DUCC installation does not have to be "ducc"
- ducc-mon login can be used on systems where users do not have password login
- The DUCC head-node daemons may be moved to another host without breaking working applications
- The deployment descriptor for a UIMA-AS service can be loaded from the classpath
- Interactive applications run correctly with viaducc (fixed lost inputs)
- Files created by DUCC jobs inherit the permissions of the launching shell's umask.
- DUCC performance breakdown for scaled synchronous pipelines is now correct
- Fixed javadoc method headers to enable ducc build with java 8
- Fixed JP communication threads wait logic when JD returns no work
- GC stats not available sometimes from remote JP
- Other bug fixes
- Many DUCC webpage improvements
This version of UIMA-DUCC requires Java version 7 or version 8.
|
29 Aug 2017
– Apache UIMA Java SDK 2.10.1 released
|
This is a minor bug fix release. The major changes in this release include:
- https://issues.apache.org/jira/browse/UIMA-5411 - for external override values,
make the class/data path search for external settings explicit by requiring a path: prefix.
This is a syntax change for this feature introduced in 2.10.0.
- https://issues.apache.org/jira/browse/UIMA-5532 fixed some edge case issues around
xmi and JSON serialization of Feature values marked as multiple-references-allowed.
Please see the
the complete list of issues fixed for more details of the changes.
|
24 Jul 2017
– Apache UIMA Ruta 2.6.1 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
UIMA Ruta Language and Analysis Engine:
- Fixed problem with empty annotation list expressions
- Fixed stacked feature expressions
- Fixed false positive match in conjunct rules
- Fixed mixing FOREACH with BLOCK blocks
- Fixed min/max quantifier
- Fixed float variables
- Fixed dynamic internal reindexing
- Requires UIMA 2.10.0 and uimaFIT 2.3.0
- Other bugfixes
UIMA Ruta Workbench:
- Fixed Query View
- Fixed Annotation Testing View
- Fixed error reporting in editor
For a complete list of bugs and improvements included in this release please see
issue report.
|
4 Apr 2017
– Apache UIMA Java SDK 2.10.0 released
|
The UIMA core Java SDK adds some new capabilities and fixes some bugs.
- External Override values are supported within UIMA Descriptor import statements
and within the fileUrl value of fileResourceSpecifiers.
- External Overrides: support was added for searching the classpath and UIMA Datapath to find the
external overrides.
If a relative file name is specified but the file is not found,
the classpath and datapath will be searched.
This approach will be changed in the next release
to make a clearer distinction between file and classpath resources
in order to avoid the ambiguity of the current design; note
that this change will not be backwards-compatible.
- to improve integration with Java, the JCas cover classes
for the built-in UIMA types FSArray, FSList,
StringArray, and StringList now support the Iterable interface.
- a new report on API changes, api-change-report.html,
is included as part of the release.
- several bugs were fixed; please see
issuesFixed for a list of changes.
Please see the RELEASE_NOTES
for a summary of the changes.
|
29 Mar 2017
– Apache uimaFIT 2.3.0 released
|
uimaFIT 2.3.0 is a minor feature and bugfix release to uimaFIT 2.2.0. It should should
serve as a drop-in replacement for previous uimaFIT 2.x versions.
The complete list of fixes is here.
The major changes in this release are:
- Support for language and mime-type capabilities;
- Faster indexCovered implementation;
- Document auto-detection of index and type priority definitions;
- Fixed selectPreceding not skipping over undesired annotations;
- Fixed calling typeSystemInit on readers and other life-cycle methods;
This version of uimaFIT requires Java version 7 or later.
|
10 Mar 2017
– Apache UIMA Ruta 2.6.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
UIMA Ruta Language and Analysis Engine:
- Annotation expressions can be restricted using feature matches and conditions
- Several new configuration parameters for RutaEngine
- Experimental features to optimize internal indexing (for experienced users)
- Minimal support of feature structures in feature match expressions
- API change report for ruta-core
- Typesystem descriptors with JCasGen classes are located in separate artifact
- Implementation of RutaBasic is located in separate artifact
- Many bug fixes and improvements, especially for label expressions
UIMA Ruta Workbench:
- Direct debugging of launched scripts in Java is supported
- Improved error messages in launcher
- Removed restriction of classpath size causing problems in launcher
- Deactivated noVM preference
- Changed UI to set annotation mode in views
- Launcher uses project encoding
- Bug fixes
This release requires an update of script projects and its descriptors in the UIMA Ruta Workbench.
There are several ways to achieve this. The recommended way is to right-click on the UIMA Ruta project and select "UIMA Ruta -> Convert to UIMA Ruta project", which will update all provided descriptors.
Then, select the project and press "Project -> Clean..." in the menu, which will regenerate all descriptors of your scripts based on the new templates.
Projects built with the UIMA Ruta Maven Plugin require no manual effort.
For a complete list of bugs and improvements included in this release please see
issue report.
|
23 Feb 2017
– Apache UIMA DUCC 2.2.0 released
|
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system
providing tooling, management, and scheduling facilities to automate the scale-out of
applications written to the UIMA framework. Core UIMA provides a generalized framework for
applications that process unstructured information such as human language, but does not
provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster
management of the resources. DUCC defines a formal job model that closely maps to a standard
UIMA pipeline. Around this job model DUCC provides cluster management services to automate
the scale-out of UIMA pipelines over computing clusters.
The complete list of fixes is here.
The major changes in this release are:
- Ships with the latest UIMA-AS v2.9.0 and UIMA SDK 2.9.0
- Ships with ActiveMQ v5.14.0
- Added support for static failover and capability to move the head node
- Fixed DUCC OR "warm" restart issues
- Fixed DUCC startup script to fail if the DB doesn't start
- Fixed DUCC shutdown sequence bug which prevented agents from stopping if Broker was shutdown first
- Fixed Rogue process detector to detect and cleanup orphan services
- Deprecated ducc.agent.node.metrics.sys.gid.max and replaced with ducc.agent.rogue.process.sys.uid.max
- Enhanced DUCC Job Driver (JD) to provide individual work item performance breakdowns
- Modified DUCC to restrict broker use to ducc user only
- On process launch failure, agent supplies a reason for failure for display in ducc mon
- Added duplicate daemon detector to prevent starting duplicate DUCC daemon
- Many DUCC Database improvements
- Many DUCC webpage improvements
This version of UIMA-DUCC requires Java version 7 or version 8.
|
15 Dec 2016
– Apache UIMA Asynchronous Scaleout (UIMA-AS) 2.9.0 released
|
UIMA AS 2.9.0 is a maintenance release that contains fixes and improvements over UIMA-AS 2.8.1.
The complete list of fixes is here.
The major changes in this release are:
- Updated to use Activemq 5.14.0
- Added dependency on UIMA SDK 2.9.0
- Fixed http based service connectivity
- Added support for automatic recovery of temp queues after broker restart.
- Fixed per CAS Performance Metrics breakdown
- Fixed per process CPU and RSS reporting
- Fixed example runtime configurations
- Fixed error recovery on exception while deserializing a CAS
- "Pinned" JMX MBeans to a specific deployment to enable orderly cleanup
- Fixed support of AMQ white listing of packages
- Added support to disable JMX via a new argument -Duima.as.enable.jmx=false
- Fixed dd2spring issues
- Updated version checker to test compatibility with UIMA SDK.
This version of UIMA-AS requires Java version 7 or later.
|
28 Sep 2016
– Apache UIMA Ruta 2.5.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
UIMA Ruta Language and Analysis Engine:
- Depends on UIMA 2.9.0 and uimaFIT 2.2.0
- New block FOREACH: iterate over pattern and reuse the matched annotation in contained rules
- Label expressions across inlined rules are supported, behave like local variables now
- Feature expressions on label expressions are supported
- Setting values of variables by configuration parameter
- Macros support variables
- New type function: typeFromString
- Wordlist initialization with string expressions: simplified specification of language-dependent dictionary usage
- New maven archetype for typical UIMA Ruta projects
- Improved and faster maven plugins
- Parameters additionalScripts and additionalEngines are now optional
- Removed engine loader concept and parameter completely
- Removed redundant Document type
- Improved datapath support
- Improved resource manager support
- Many bug fixes and improvements
UIMA Ruta Workbench:
- Improved support of classpath-lookup in maven-based projects
- Launch configuration supports various serialization formats (with preference)
- Bug fixes
This release requires an update of script projects and its descriptors in the UIMA Ruta Workbench.
There are several ways to achieve this. The recommended way is to right-click on the UIMA Ruta project and select "UIMA Ruta -> Convert to UIMA Ruta project", which will update all provided descriptors.
Then, select the project and press "Project -> Clean..." in the menu, which will regenerate all descriptors of your scripts based on the new templates.
Projects built with the UIMA Ruta Maven Plugin require no manual effort.
For a complete list of bugs and improvements included in this release please see
issue report.
|
30 Aug 2016
– Apache UIMA Java SDK 2.9.0 released
|
The UIMA core Java SDK adds some new capabilities and fixes some bugs.
- Two new features were added.
- CasIOUtls - a collection of general-purpose CAS load/save utilities, with support for
including the type system and index definition information along with the Feature Structures.
- A new set of APIs allowing use of the external shared settings files to be used directly, rather than only by
overriding UIMA configuration parameters. These APIs give direct access by the external settings file's key names to the
key values.
- Many bugs were fixed, including:
- A regression (since version 2.8.0) which broke the unambiguous AnnotationIndex iterator in some cases.
- fixes for the Eclipse tooling plugins to work with the latest Eclipse (4.6, Neon) version.
- Sofa mapping support for flow controllers.
- Fixing a stack overflow error in AnnotationViewer when switching Sofas.
- Fixing 2 JSON serialization issues: mishandling of arrays of FeatureStructures having null entries,
and mishandling of namespaces for uninstantiated types found as supertypes.
Please see the RELEASE_NOTES
for a summary of the changes.
|
08 Aug 2016
– Apache UIMA DUCC 2.1.0 released
|
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system providing tooling,
management, and scheduling facilities to automate the scale-out of applications written to the UIMA framework.
Core UIMA provides a generalized framework for applications that process unstructured information such as human
language, but does not provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster management of the resources.
DUCC defines a formal job model that closely maps to a standard UIMA pipeline. Around this job model DUCC
provides cluster management services to automate the scale-out of UIMA pipelines over computing clusters.
The complete list of fixes is here.
The major changes in this release are:
- Ubuntu and RHEL 7 support
- cgroup enhancements
- uses standard cgroups organization
- supports cgroup swappiness setting, restricting any swapping if desired
- DUCC state and history storage moved from flat files to Cassandra DB, reducing storage size 5x
- Ships with the latest UIMA-AS v2.8.1
- Ships with recent ActiveMQ v5.13.2
- DUCC's UIMA-AS services support failover and ssl connectors
- Many DUCC webpage improvements
- Clear user display of DUCC classes and relation to machines
- Robust handling of dynamic changes to DUCC class and nodepool definitions
- Full support of nodepools with different quantum
- DUCC broker access restricted to user ducc
- Eliminate need for user home directories located on a shared filesystem
- Built-in Job error handler programmable per job
- Migration utility for DUCC updates
- Change to vary-off behavior to facilitate cluster management
- Horizontal stacking of services instance allocations
- java-viaducc improvements including separation of stdout from stderr respoonses
- An alert banner is displayed on ducc-mon pages if daemons are down
- Promoted DUCC from sandbox to the regular Apache project in the SVN
This version of UIMA-DUCC requires Java version 7 or version 8.
|
20 May 2016
– Apache UIMA Asynchronous Scaleout (UIMA-AS) 2.8.1 released
|
UIMA AS 2.8.1 is a maintenance release that contains fixes and improvements over UIMA-AS 2.6.0.
The complete list of fixes is here.
The major changes in this release are:
- Upgraded ActiveMQ, Spring Framework and Jetty
- Upgraded dependency on UIMA-SDK to v.2.8.1
- Fixed per CAS Performance Metrics breakdown for async deployments
- Added new feature to allow warm up of a JVM service instance before real processing begins,
by feeding it a specified set of CASes before the instance connects to the service input queue.
This version of UIMA-AS requires Java version 7 or later.
|
06 Apr 2016
– Apache uimaFIT 2.2.0 released
|
uimaFIT 2.2.0 is a minor feature and bugfix release to uimaFIT 2.1.0. It should should
serve as a drop-in replacement for previous uimaFIT 2.x versions.
The complete list of fixes is here.
The major changes in this release are:
- New FSUtil class with methods to get/set feature values
- New selectAt method
- Improved compatibility with thread context classloaders
- Upgrades to dependencies including UIMA SDK and Spring Framework
- Use of iteratorWithSnapshot in select methods
- Limited support of UIMA arrays (prototypical)
- Many, many bug fixes and improvements
This version of uimaFIT requires Java version 7 or later.
|
15 Feb 2016
– Apache UIMA Ruta 2.4.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
UIMA Ruta Language and Analysis Engine:
- Explicit referencing of annotations with variables, labels and addresses
- Helper methods for applying rules directly in Java code
- New action for splitting annotations
- New block for resetting match context
- Import of uimaFIT analysis engines with mandatory parameters
- Macros for conditions and actions (prototypical)
- Limited support of UIMA arrays (prototypical)
- Many, many bug fixes and improvements
UIMA Ruta Workbench:
- More support of external files
- Bug fixes
This release requires an update of script projects and its descriptors in the UIMA Ruta Workbench,
e.g, by deleting all descriptors and updating the script project.
Right-click on a project and select "UIMA Ruta -> Convert to UIMA Ruta project.
Projects built with the UIMA Ruta Maven Plugin require no manual effort.
For a complete list of bugs and improvements included in this release please see
issue report.
|
11 Aug 2015
– Apache UIMA DUCC 2.0.0 released
|
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system providing tooling,
management, and scheduling facilities to automate the scale-out of applications written to the UIMA framework.
Core UIMA provides a generalized framework for applications that process unstructured information such as human
language, but does not provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster management of the resources.
DUCC defines a formal job model that closely maps to a standard UIMA pipeline. Around this job model DUCC
provides cluster management services to automate the scale-out of UIMA pipelines over computing clusters.
This is a major release containing new features and numerous bug fixes. What's new:
Added support for classpath isolation - User's code runs with the classpath it supplies
Added support for non-preemptive (NP) workloads
Modified DUCC error handler interface API
Extended CGROUPS support by adding CPU control
Changed DUCC flow controller typesystem location
Added support for queueing resource requests that were previously unfulfillable
Please see the RELEASE_NOTES
for a summary of the changes.
|
22 July 2015
– Apache UIMA Java SDK 2.8.0 released
|
Note: this version has been replaced by 2.8.1 (above), and has been removed
from the download page (but is available on the archives, if needed).
This release is a bug fix release, plus one enhancement. The version number incremented because
we restored some APIs that were accidentally dropped in 2.7.0.
An improvement was made to the performance of iterators over sorted indexes, such as the
built-in Annotation index. This improvement may be significant when iterating over types where the type has
many subtypes.
Index usage by iterators in this situation is now 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 overhead is reduced. Any update that affects this flattened index will cause it to be discarded
and not used for subsequent iterations.
Performance improvement may be seen for use cases
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.
For more details, see Jira issue
UIMA-4357.
For more details, please see the RELEASE_NOTES.
|
08 June 2015
– Apache UIMA Ruta 2.3.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
Major Changes in this Release
This release requires an update of script projects and its descriptors in the UIMA Ruta Workbench,
e.g, by deleting all descriptors and updating the script project.
Right-click on a project and select "UIMA Ruta -> Convert to UIMA Ruta project
UIMA Ruta Language and Analysis Engine:
- Requires Java 7
- Maven Plugin for UIMA Ruta
- Extensions and simplifications of the language (external resources, null values, more syntactic sugar)
- Improvements to Analysis Engines (RutaEngine, HtmlConverter)
- More exemplary script projects
- Many, many bug fixes
UIMA Ruta Workbench:
- Requires Java 7
- Works now also in other project structures (Maven projects)
- Bug fixes
For a complete list of bugs and improvements included in this release please see
issue report.
|
06 March 2015
– Apache UIMA Java SDK 2.7.0 released
|
The UIMA Java SDK now requires Java 7. There are significant changes in this release.
Some of these may break backwards compatibility; we have included some new JVM properties
to aid in gradually adapting to these changes. See the
README file for a description of these properties.
There is a new capability that detects operations which change feature values used as keys
in sorted and set indexes; by default, the Framework now insures that the indexes
are properly updated when this happens. A new report can be generated to show the user when this is happening,
so they may improve their code.
This release supports a new JSON serialization of the CAS and of UIMA descriptors. This is
described in the reference manual.
Many performance improvements have been made.
The complete list of fixes is here.
|
23 October 2014
– Apache UIMA DUCC 1.1.0 released
|
UIMA DUCC 1.1.0 is a maintenance release that contains fixes and improvements over UIMA-DUCC 1.0.0.
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system providing tooling,
management, and scheduling facilities to automate the scale-out of applications written to the UIMA framework.
Core UIMA provides a generalized framework for applications that process unstructured information such as human
language, but does not provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster management of the resources.
DUCC defines a formal job model that closely maps to a standard UIMA pipeline. Around this job model DUCC
provides cluster management services to automate the scale-out of UIMA pipelines over computing clusters.
The complete list of fixes is here.
|
23 September 2014
– Apache UIMA Ruta 2.2.1 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
For a complete list of bugs and improvements included in this release please see
issue report.
The UIMA Ruta Workbench 2.2.1 requires a newer version of Eclipse (>4.0) due to the CAS Editor. It has been tested with Eclipse 4.3.2.
|
12 June 2014
– Apache uimaFIT 2.1.0 released
|
uimaFIT 2.1.0 is a bugfix release to uimaFIT 2.0.0. The reason that the minor version was
increased instead of the bugfix version is, that uima 2.1.0 requires Java 6 and uses the UIMA
Java SDK 2.6.0. Otherwise, this version should serve as a drop-in replacement for uimaFIT
2.0.0.
The complete list of fixes is here.
This version of uimaFIT requires Java version 6 or later.
|
12 May 2014
– Apache UIMA Java SDK 2.6.0 released
|
The UIMA framework has been enhanced to support large memory/ large multi-core machines,
with various improvements to synchronization, lock contention, and locality of reference to optimize
memory bandwidth in the presence of non-uniform memory access speeds. In addition, embedders of the framework
can now choose to run initialization of UIMA components that share a common Resource Manager
and/or common UIMA Context, on multiple threads, concurrently. CASes which are not being modified can
be accessed (read-only) by multiple threads, simultaneously. Along the way, many subtle bugs were fixed.
The complete list of fixes is here.
This version of UIMA requires Java version 6 or later.
|
15 April 2014
– Apache UIMA Ruta 2.2.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
The major changes in this release are:
UIMA Ruta Language and Analysis Engine:
- Major performance improvements (3-17 times faster in test use cases)
- Improved import type functionality and handling of ambiguous short names
- Support of block extensions for rule inference adaptions
- Options to determine where the next match should start
- Requires at least Java 6
- Many bug fixes
UIMA Ruta Workbench:
- Smaller improvements in many views
- Support of mixin Java/Ruta projects
- Many bug fixes
For a complete list of bugs and improvements included in this release please see
issue report.
|
30 January 2014
– Apache UIMA DUCC 1.0.0 released
|
This is the first release of the sandbox project UIMA DUCC.
DUCC stands for Distributed UIMA Cluster Computing. DUCC is a cluster management system providing tooling,
management, and scheduling facilities to automate the scale-out of applications written to the UIMA framework.
Core UIMA provides a generalized framework for applications that process unstructured information such as human
language, but does not provide a scale-out mechanism. UIMA-AS provides a scale-out mechanism to distribute UIMA
pipelines over a cluster of computing resources, but does not provide job or cluster management of the resources.
DUCC defines a formal job model that closely maps to a standard UIMA pipeline. Around this job model DUCC
provides cluster management services to automate the scale-out of UIMA pipelines over computing clusters.
|
14 November 2013
– Apache UIMA Asynchronous Scaleout (UIMA-AS) 2.4.2 released
|
UIMA AS 2.4.2 is a maintenance release that contains fixes and improvements over UIMA-AS 2.4.0.
Besides bug fixes, this release includes:
- Replaced ActiveMQ version 5.4.1 with version 5.6.0
- Implemented PEAR support
- Added client-side callbacks to notify an application where a CAS is being processed
- Refactored JMS Session expiration management to prevent broker OOMs
- Refactored UIMA-AS Async Aggregate to support processing multiple input CASes in parallel
- Refactored CAS Accounting code to prevent loosing CASes in complex/nested deployments
- Improved error handling and recovery
- Performance improvements
Please see the release notes on the downloads page, for a more complete summary
and a complete list of changes.
|
20 September 2013
– Apache UIMA Ruta 2.1.0 released
|
Apache UIMA Ruta is a rule-based script language supported by
Eclipse-based tooling. The language is designed to enable rapid
development of text processing applications within UIMA. A special
focus lies on the intuitive and flexible domain specific language for defining
patterns of annotations. The Eclipse-based tooling,
called the Apache UIMA Ruta Workbench, was created to support the
user and to facilitate every step when writing rules. Both
the rule language and the workbench integrate
smoothly with Apache UIMA.
The major changes in this release are:
UIMA Ruta Language and Analysis Engine:
- Combinations of rule elements that specify no sequential constraint
- Rule elements with inlined rules, which are interpreted as "actions"
or "conditions"
- Dot notation for accessing features, also available as matching conditions
- Implicit actions and conditions
- New analysis engines for CAS manipulation
- Direct import of uimaFIT analysis engines without descriptor
- RutaEngine is now a uimaFIT component
- Manual specification of the start anchor for a rule match
- Many bug fixes
UIMA Ruta Workbench:
- Compatible with uimaj 2.4.2 and Eclipse Kepler
- New framework for constraint-driven evaluation
- Two new rule learning algorithms for the TextRuler framework
- New views, e.g., Annotation Check and Html viewer
- Combinations with java projects
- Open CAS Editor at Eclipse startup causes less problems
- Many bug fixes
For a complete list of bugs and improvements included in this release please see
issue report.
|
31 August 2013
– Apache uimaFIT 2.0.0 released
|
Apache uimaFIT™ is a library that
facilitates the building of Apache UIMA components, the programmatic use of Apache UIMA
analysis pipelines, and their testing.
uimaFIT employs Java annotations to integrate UIMA meta data directly into the source
code, allowing for less lines of code and better refactorability then traditional, XML
descriptor-based UIMA projects. It is capable of automatically detecting meta data,
e.g. type system information, from the classpath. Convenience methods are provided for
constructing components, pipelines, and for accessing annotations.
This is the first Apache release of the project previously known simply as "uimaFIT".
The major changes in this release are:
- Migration to Apache Foundation as part of the Apache UIMA project.
- New Maven coordinates (groupId: org.apache.uima; artifactId: uimafit-core)
- New uimaFIT Maven plugin for automatically generating uimaFIT meta data in component
classes and UIMA descriptors from uimaFIT-annotated component classes
- New module for backwards compatibility with uimaFIT 1.4.x-based components
- Consolidated documentation in HTML and PDF format
- Moved CPE support to separate dependency
(groupId: org.apache.uima; artifactId: uimafit-cpe)
- Improved component life-cycle handling
- New support for setting multi-valued parameters with simple single
values (no need to wrap them in a collection or array anymore)
- Improved support for external resources, including support for
multi-valued external resources
Note on compatibility and migration: Apache uimaFIT 2.0.0 is not a drop-in replacement
for uimaFIT 1.4.x, although, it was tried to keep the migration effort minimal. Please
refer to the
migration guide for all the details on incompatible changes between uimaFIT 1.4.0 and
Apache uimaFIT 2.0.0.
Additionally, there have been more bug fixes and minor improvements. For a full list of
the changes, please refer to the
issue report.
|
26 July 2013
– Apache UIMA Java SDK 2.4.1 released
|
This release only has the Apache UIMA™ Java SDK, and does not yet have the other
components. It is backwards compatible with previous releases, except some classes and interfaces
which have additional methods supporting new capabilities and features.
Besides some bug fixes, this release includes:
- Documentation of binary serialization.
- New kinds of compressed binary serialization that compress the data significantly
and one form that supports unequal source/target type systems
- A new facility called External Parameter Overrides for specifying parameter settings for annotators
that uses properties files and is independent of the annotator hierarchy
- CasCopier enhancements to allow copying one view to a different view.
- Additional options to restrict JCasGen operation to generating just those types that are defined in a project,
excluding other types that are imported from other projects
- A new Maven plugin that runs JCasGen (see tools documentation for how to configure and use this)
- a new ability to preserve white space (indentation) when parsing XML descriptors; this is now used in the
Component Descriptor Editor (CDE), to preserve indentation when editing an existing descriptor.
- Performance and space improvements, including some bulk methods for efficiently removing Feature Structures from Indexes
- The CDE supports 3 new things: the preserving of existing white-space in XML descriptors,
External Parameter Overrides, and a configuration option to restrict JCasGen to just those
types defined in the project.
- Enhancements to the DocumentAnalyzer utility to support reading CASes in various formats
Please see the release notes on the downloads page, for a more complete summary
and a complete list of changes.
|
29 May 2013
– Apache UIMA Ruta 2.0.1 released
|
Apache UIMA Ruta is a rule-based script language supported by Eclipse-based tooling.
The language is designed to enable rapid development of text processing applications within UIMA.
For a complete list of bugs and improvements included in this release please see
release-notes.
|
20 November 2012
– Apache UIMA C++ SDK 2.4.0 released
|
This is the first release of UIMA C++ as a top level project.
Notable updates included in this release: standardization of Linux build process and packaging, updates to ActiveMQ service wrapper to support failover,
many bug fixes.
For a complete list of bugs and improvements included in this release please see
release-notes.
|
15 November 2012
– Apache UIMA Asynchronous Scaleout (UIMA-AS) 2.4.0 released
|
UIMA AS 2.4.0 is a maintenance release that contains fixes and
improvements over UIMA AS 2.3.1.
Notable updates included in this release: added support for programmatic generation of deployment descriptors,
user supplied CR can now be used by UIMA-AS client, per AE performance breakdown is collected and returned to
a client to expose bottlenecks, extended callback API to include info where a CAS is being processed (ip,pid,thread id),
improved stability and error handling, many bug fixes.
For a complete list of bugs and improvements included in this release please see
jira-report.
|
07 December 2011
– Apache UIMA Java SDK 2.4.0 released
|
This release only has the Apache UIMA™ Java SDK, and does not yet have the other
components. It is backwards compatible with previous releases, except for an API change in the
JMX tooling, and a redesign of the Eclipse plugin for editing CASes.
Besides some bug fixes, there is some tooling enhancement. A new Eclipse plugin hooks into Eclipse's "launching" capability to support launching
annotators more easily from Eclipse; see the Tools manual for information.
The Cas Editor Eclipse plugin has been substantially enhanced; please see the Tools manual
for more information.
Please see the release notes on the downloads page, for a more complete summary
and a complete list of changes.
|
29 August 2011
– UIMA Addons 2.3.1 Released
|
This release includes most of the projects in the Sandbox. Those projects there
that were ready to be released were moved to a new SVN top node in the
UIMA project, called "addons".
The addons package contains 2 new annotators: Solrcas (for storing CAS objects into an
Apache Solr instance), and AlchemyAPIAnnotator (wraps alchemyapi.com services).
Several fixes and improvements were made to other annotators.
Please see the release notes on the downloads page, for a more complete summary
and a complete list of changes.
|
22 March 2011
– Apache UIMA AS 2.3.1 released
|
UIMA AS 2.3.1 is a maintenance release that contains fixes and improvements over UIMA AS 2.3.0.
Notable updates included in this release: upgrade to ActiveMQ 5.4.1, inclusion of base UIMA
2.3.1 jars, improved stability and error handling, many bug fixes.
For a complete list of bugs and improvements included in this release please see
jira-report.
|
10 December 2010
– Apache UIMA Java SDK 2.3.1 released
|
This release only has the Apache UIMA™ Java SDK, and does not yet have the other
components. It is backwards compatible with previous releases, except
that Java 5 is now the minimum Java level required.
This is the first release as a top level project.
Not many changes are in this release; included are
some fixes for edge cases involving Results Specification,
a performance fix for the Component Descriptor Editor editing things having
remote components, and several improvements to the CAS editor.
Please see the release notes on the downloads page, for a more complete summary
and a complete list of changes.
|
18 March 2010
– Apache UIMA graduates from the incubator
|
At the ASF board meeting held on 17 March 2010, the Board unanimously approved the proposed resolution to
establish the Apache UIMA Project as a top level Apache project. This signals the successful completion
of the incubation process and the beginning of the next phase of Apache UIMA existence.
Becoming a top level project entails setting up some new infrastructure, and moving things.
It will take some time to accomplish this move; in the meanwhile, we'll be keeping this website up to date.
|
26 January 2010
– UIMA 2.3.0-incubating Released
|
This release includes all 4 major components:
the UIMA Java SDK, UIMACPP, UIMA-AS (Asynchronous Scalout capability),
and a growing set of UIMA Addons. It is backwards compatable with previous releases, except
that Java 5 is now the minimum Java level required.
The add-ons package contains many new components and annotators.
The UIMA-AS (Asynchronous Scaleout)
framework is extensively enhanced with much more support for error/failure recovery, driven by feedback
from actual use in several large scale deployments. The base framework now supports Java 5 generics, and
is enhanced to make it even more
light-weight and efficient; for example, it now supports a new network serialization format
for communicating with remote
annotators using a "delta-CAS" - limiting the response sent to just those items which have changed.
Please see the release notes on the downloads page, for a more complete summary
and a complete list of changes.
|
07 December 2006
– UIMA Sandbox established
|
The UIMA sandbox is a workspace that is open to all UIMA committers and developers who would
like to contribute code and join the UIMA developer community on Apache. Details about the
sandbox can be found here.
|
|
|
