Skip to content

Latest commit

 

History

History
149 lines (128 loc) · 4.57 KB

CHANGES.md

File metadata and controls

149 lines (128 loc) · 4.57 KB

Change log

0.6.0

Initial version since version policy is in force.

0.7.0

  • Updated and fixed pom.xml (maven release plugin, https repos)
  • Module PESVReader: is more forgiving about missing columns

0.7.1

  • Module RDFProjector: added parameters rdfFormat and language

0.8.0

  • Fixed:
    • Embedded resource reader.plan
    • Documentation in command-line (embedded resource alvisnlp-doc2ansi.xslt)
    • Documentation online
  • Added embedded resource yatea.plan
  • Added current version in the online documentation
  • Refactored release script

0.8.1

  • Fixed:
    • launch script now interprets correctly arguments protected by quotes that contains spaces
    • release script correctly replaces versions in POM files

0.9.0

  • Removed support for alvisnlp-rest
  • Fixed
    • online documentation
    • function after
    • module MergeSections: relations
    • dependencies updates
  • New
    • command line option -maxMmapSize for systems with CONFIG_STRICT_DEVMEM
    • module LinguaLID: language identification
    • module OpenNLPDocumentCategorizerTrain, OpenNLPDocumentCategorizer: document classification
    • module CompareFeatures: compare two features of specified elements
    • module FasttextClassifierLabel, FasttextClassifierTrain: document classification
    • library hash
    • str:find
  • Improved
    • temp dir cleaning: if -cleanTmp is set each module cleans after processing, prevents filesystem bloating
    • pre-process checking: more complete, skip inactive and not mentioned in select parameters
    • module CompareElements: new similarity functions
    • module TabularReader: reads true CSV, handle header line
    • module StanfordParser: more recent models, language parameter
    • module XMLReader: a:inline also copy attributes
    • modules OBOMapper and OBOProjector: added parameter altPathFeatures
  • Improved documentation
    • refctored module docs
    • modules TomapTrain and TomapProjector
    • modules WapitiTrain and WapitiLabel
    • module PatternMatcher
    • module RemoveOverlaps
    • module YateaTermsProjector

0.10.0

  • Definitive name is AlvisNLP
  • Removed
    • support for OpenMinTeD
    • modules XMIImport, XMIExport
    • module BioLG
    • module CopyFeature
    • module RunProlog
    • module FillDB
    • module EnrichedDocumentWriter
    • module ProminentConceptReporter
    • module XMLWriter2ForINIST
    • module AlvisREPrepareCrossValidation
    • module AlvisDBIndexer
    • module ADBWriter
    • modules AlvisAEReader2, ExportCadixeJSON
    • module EnjuParser2
    • modules AttestedTermsProjector, TyDIProjector, YateaProjector, ElementProjector2, SimpleProjector2
    • module FileMapper2
    • modules SelectingElementClassifier, TaggingElementClassifier, TrainingElementClassifier
    • module Script
    • module WhatsWrongExport
    • module AnimalReader
    • support for custom entities in plans
    • old plan syntax
    • dumping from plan
    • miscellanous deprecated parameters
  • Fixed
    • documentation
    • bug in TabularExport, create output directory if it does not exist
    • dump framework
    • bug in XMLWriter, will not crash if attribute name is illegal
  • New
    • Python script interoperability framework
    • module PythonScript: generic Python script module
    • module Stanza: Stanza pipeline
    • module REBERTPredict: relation extraction with BERT
    • functions xafter, xbefore, xinside, xoutside, xoverlapping, xspan: same as no-x counterparts but includes context element
    • parameter FileMapper#headerLine
    • parameter QuickHTML#documentTitle
    • command-line option -analysisFile
    • module CoreNLP
    • module MultiRegExp
    • module JsonExport
    • library ctx
    • module TagTogReader
    • parameter TikaReader#baseNameId
  • Improved
    • compiled trie works around mmap limitations
    • support for snippet examples in modules documentation
    • homogenize parameter names: source for readers, feature/layer/section/relation/arg names
    • homogenize dependency parser parameter names
    • API: replace obsoleteUseInstead method with @deprecated
    • cleaned code a bit
    • install AlvisNLP in a dedicated directory
    • updated dependencies versions

0.10.1

  • Added
    • REBERTPredict#ensembleModels
  • Improved
    • QuickHTML
    • check that output-feed parameters do not exist

0.10.2

  • Improved

    • memory management when deleting elements
    • alias syntax see #169
    • code simplification
    • test for QuickHTML
    • RDFExport supports type and language for literals
    • Shell exports
  • Added

    • PythonScript#outputFile
    • AlvisAEReader#databasePropsFile replaces database access parameters
  • Fixed

    • bug in MergeSections
    • bug in HttpServer
    • bug in overlapping()
    • bug in parameter conversion of user functions