Apache Tika 1.13
The most notable changes in Tika 1.13 over the previous release are:
- Major changes to the PDFParser including upgrade to PDFBox 2.0.1 (TIKA-1285, TIKA-1959). This include the classic sequential parser is no longer available, Tiff files are no longer extracted by default, Some truncated/corrupted files that had some content extracted with 1.8.x may have no content extracted in 2.0.x.
- The MIT-NLP Information Extraction (MITIE) Named Entity Recognition (NER) system is now supported in Tika (TIKA-1913, Github-108).
- Tika now supports the use of the Yandex translation service (TIKA-1943, Github-106).
- Tika now uses NER to extract scientific measurements from text using either GROBID Quantities which uses conditional random fields and NLTK which uses regular expressesions (TIKA-1917, Github-104).
- Fixed JournalParser to handle null responses from GROBID and to log a message (TIKA-1925).
- Refactored Language Detector into tika-landetect module, added default N-Gram implementation}}, Optimaize LangDetector and MIT Text.jl implementation (TIKA-1872, TIKA-1696, TIKA-1723).
- Extract metadata from MP4 videos whether or not the PooledTimeSeries parser is available via Aditya Dhulipala (TIKA-1844).
- Fix NPE when trying to get embedded image identifier in WordParser (TIKA-1956).
- Improvements to MIME database for detection of Scientific and other formats present in the TREC-DD-Polar dataset (TIKA-1881, Github-85, TIKA-1883, TIKA-1884, TIKA-1886, TIKA-1882).
- LinkContentHandler now extracts links from script tags via Joseph Naegele (TIKA-1937).
- Handle per page IOExceptions more robustly in PDFParser (TIKA-1948).
- Upgrade commons-compress to 1.11 (TIKA-1949).
- Add detection for embedded MSChart.Graph files (TIKA-1033).
- Fix NPE in Sqlite parser from Nick C (TIKA-1927).
- Fix NPE in Open Document parser from Nick C (TIKA-1916).
- Upgrade mp4parser's isoparser to 1.1.7 (TIKA-1924 and TIKA-1931).
- Upgrade BouncyCastle to 1.54 (TIKA-1923).
- Upgrade Jackcess to 2.1.3 (TIKA-1922).
- Upgrade Drew Noakes' metadata-extractor to 2.8.1 (TIKA-1921).
- Upgrade Gson in tika-serialization to 2.6.2 (TIKA-1920).
- Upgrade commons-cli in tika-batch to 1.3.1 (TIKA-1919).
- Add XMPMM support to PDFParser and JpegParser via Jempbox (TIKA-1894).
- Move serialization of TikaConfig to tika-core and enable dumping of the config file via tika-app (TIKA-1657).
- Tika now incorporates the Natural Language Toolkit (NLTK) from the Python community as an option for Named Entity Recognition (TIKA-1876).
- Add support for XFA extraction via Pascal Essiembre (TIKA-1857).
- Upgrade to sqlite-jdbc 3.8.11.2 (TIKA-1861). NOTE: this dependency is still scopeprovided/scope. You need to include this dependency in order to parse sqlite files.
- Upgrade to POI 3.15-beta1 (TIKA-1895).
- Upgrade to Jackson 2.7.1 (TIKA-1869).
- Upgrade to Apache SIS 0.6 (TIKA-1878).
- RichTextContentHandler moved from the Server package to Core (TIKA-1870).
- Added ZeroSizeFileDetector to support application/x-zerovalue via Adesh Gupta (TIKA-1885).
- Addition of types information to Grobid quantities parser via Can Menekse (TIKA-1965).
The following people have contributed to Tika 1.13 by submitting or commenting on the issues resolved in this release:
- Adesh Gupta
- Ajay Kumar Loganathan Ravichandran
- Alessandro De Angelis
- Avinash
- Ayesha Hasan
- Can Menekse
- Chris A. Mattmann
- Dave Meikle
- Franco Catto
- Hendy Irawan
- Ian Williams
- Jeremy Anderson
- John Patrick
- Jorge Spinsanti
- Joseph Naegele
- Ken Krugler
- Lewis John McGibbney
- Luca Moretti
- Manali Shah
- Manisha Kampasi
- Mark Duske
- Namitha Sanjeeva Ganiga
- Nandan Chandrashekar
- Nick Burch
- Nick C
- Pascal Essiembre
- Paul Ramirez
- Prasad Nagaraj Subramanya
- Ramit Wadhwa
- Ray Gauss II
- Sergey Beryozkin
- Shawn Johnson
- Steffen Netz
- Suman Kashyap
- Thamme Gowda N
- Tim Allison
- Trevor Lewis
- Yash Tanna
- aoeu
- kostali
See https://s.apache.org/BaRc for more details on these contributions.