Releases: CentreForDigitalHumanities/Textcavator
5.30.0
What's Changed
- Improved text analysis for multiple languages (#2015)
- Optimisation for term frequency and ngram visualisations (#2010)
- Improved tokenisation in wordcloud visualisation (#2024)
- Dependency updates (#2056 , #2043 , #2044 , #2040 , #2041 , #2042, #2045 , #2049 , #2050 , #2051 , #2052 , #2053 , #2054 , #2055 , #2031 , #1995 , #2036 , #2035 , #2037 , #2034 , #2039 , #2038 )
Changes to server configuration
The improvements to text analysis (#2015) do not require any changes to the server configuration or corpus definitions, but for existing corpora, they will only take effect if you re-index the corpus.
Full Changelog: 5.29.0...5.30.0
5.29.0
5.28.0
What's changed
- Changes to the naming scheme for Elasticsearch indices (#1962)
- Wait for corpus refresh when saving custom corpus (#2004)
- Consistency between CSV sniffer and CSV reader (#2006)
- Clarify required fields in corpus form (#2007)
- Various fixes to corpus form (#2013)
- Better support for Chinese/Japanese/Korean data (#2012)
- Security updates (#1997 , #2005 , #2008 , #2014)
Changes to server configuration: There are some small changes to the naming scheme for Elasticsearch indices (see above). This only affects newly created corpora.
Full Changelog: 5.27.1...5.28.0
5.27.1
Fixed
- Explictly type wordcloud data to prevent build failure
Contributors: @JeltevanBoheemen
Full Changelog: 5.27.0...5.27.1
5.27.0
Corpora
- Extended European Union Parliament corpus (#1987)
Custom corpus upload
- Remove and add fields (#1984)
New features
- Add collocation counting (#1970)
Internal changes
- Change Django corpora setting format (#1866)
- Update SAML development settings (#1982)
- Update production server URL (#1974)
- Improve term frequency efficiency (#1912)
Bugfixes
- Fix multiline csv documentation (#1963)
- Fix visualisation select if corpus has no visualisation (#1969)
- Fix filter multiselect layout (#1972)
- Disable form options on active corpus (#1981)
- Fix document position in search popup (#1994)
Dependency upgrades
- Update dependencies (#1976, #1977, https://github.com/CentreForDigitalHumanities/Textcavator/pull/, #1979, #1980, #1985, #1986, #1988, #1989)
Contributors: @lukavdplas
Full Changelog: 5.26.1...5.27.0
5.26.1
What's Changed
- Fixes to whitespace and characters in document contents (#1959)
Full Changelog: 5.26.0...5.26.1
5.26.0
What's Changed
New corpora
- The Punch Historical Archive (#1960)
Changed
- Better descriptions for recent Gale corpora (#1961)
Contributors: @JeltevanBoheemen, @bbonf
Full Changelog: 5.25.0...5.26.0
5.25.0
What's Changed
Name change
- The application has been renamed from I-Analyzer to Textcavator (#1926, #1936, #1952, #1923, #1929). Read more about this change in this announcement.
New corpora
Internal changes
- Improvements and fixes for custom corpus functionality (#1930, #1918, #1919, #1868, #1921)
- Test improvements (#1931, #1949)
- Restructure frontend files (#1904)
- Extended corpus categories (#1916)
- Fix named entity icons (#1935)
- Improve simple query string parse (#1905)
- Outfactor menu dropdown (#1902)
Dependency and security updates
Contributors: @bbonf, @JeltevanBoheemen, @lukavdplas, @Meesch,
Full Changelog: 5.24.0...5.25.0
5.24.0
New corpora
- 19th Century US Newspapers (#1840)
- Course Description Corpus for the AI Dashboard group (#1898)
- Illustrated London News (#1862)
- Swerik dataset (#1897)
Internal changes
Changes to server configuration:
- Backend must now use Python 3.12 instead of 3.9
Contributors: @bbonf, @Meesch, @lukavdplas
Full Changelog: 5.23.0...5.24.0
5.23.0
What's Changed
- Support URL fields in corpus form (#1863)
- Updated date range in Troonredes corpus (#1888)
- Validate sample size for term frequency requests (#1883)
- Fixed date filter initialisation (#1887)
Contributors: @JeltevanBoheemen, @lukavdplas
Full Changelog: 5.22.1...5.23.0