CHANGES.txt

Release 4.1.0 - unreleased

   * A request can no longer supply openai-embedding-filter or
     jina-embedding-filter in a per-request metadata-filters list: 4.0.x
     built the filter from the request, endpoint and key included, and
     posted the document's text there. The flat per-request skipEmbedding
     override is unchanged. The two filters are deprecated; configure the
     endpoint under "engines" and bind it with a TEXT inference binding.
     Removal in 4.2.0 (TIKA-4889).

   * PDF: "pdf-parser": {"text": ...} says where a page's text comes from:
     EXTRACT, AUTO (default), EXTRACT_AND_OCR, OCR, or the new NONE, which
     writes no text from any source while pages are still rendered for
     annotators and PAGES inference bindings. The 4.0 "ocr": {"strategy"}
     spellings load as aliases and dump as "text"; OcrConfig.getStrategy()
     is removed, read PDFParserConfig.getText() (TIKA-4889).

   * Engines are closed with the config that loaded them: Engine is
     Closeable (no-op by default), EngineRegistry closes every engine once,
     and the pipes server closes the registry at shutdown, so an
     openai-embedding-engine or a VLM used as a text recognizer releases its
     HTTP client (TIKA-4889).
     
   * Per-request switch for text recognizers: {"parse-context":
     {"text-recognizers": {"enabled": false}}} runs no recognizer or annotator
     from the list for that parse, in every container, and OCR_ONLY no longer
     reports a missing engine for it. With the inference switch this is
     "everything off for one request". New docs page with the OCR and
     inference recipes (TIKA-4889).

   * New tika-eval-structure: a standalone tool that compares two sets of
     XHTML extracts block by block (paragraph order as Kendall tau, splits and
     merges, words one side glues that the other splits, artifact and untagged
     shares), per file and per producer, where token dice cannot see order
     (TIKA-4891).

   * Improve extraction of tagged PDFs(TIKA-4891).

   * TEXT input for inference bindings: a binding on TEXT with a chunker
     (markdown-chunker) embeds the text of every document in a tree in
     batched requests, after the parse and before the metadata filters; each
     chunk lands on its document with a text locator. Chunks now name the
     binding that wrote them as "producer". A PDF released as PAGES records
     tk:inference-released and is not embedded twice. The embedding filters
     keep working (TIKA-4889).
     
   * "pdf-parser": {"maxRenderedPages": N} bounds how many pages the
     RENDER_PAGES_BEFORE_PARSE and RENDER_PAGES_AT_PAGE_END image strategies
     render, counted from the first page and independent of maxPages: text
     extraction can cover the whole document while only the first page is
     rendered, as for a thumbnail (TIKA-4856).
   
   * PDF: a bookmark outline is now walked without recursion, and lists nest at
     most 50 deep with deeper items joining the deepest list (TIKA-4894).

   * A text recognizer is an engine: configure Tesseract, Tess4J or a VLM once
     under "engines" and name it from "text-recognizers" with
     {"engine": "<name>"} (plus _mime-include/_mime-exclude). The 4.0 form with
     the engine inline still works. An inference binding cannot use a
     recognizer for a task it does not fit (TIKA-4889).
     
   * PAGES input for inference bindings: pdf-parser.inference.input lists what
     the PDF releases (TEXT by default; PAGES renders every page for a PAGES
     binding). One render per page feeds OCR and the bindings, and each page's
     vector lands on the PDF with a page locator. A page render emitted as an
     embedded document is no longer an IMAGES unit. OCR_ONLY annotated an OCR'd
     page twice when pages were rendered as embedded documents; once now
     (TIKA-4889).

   * New "inference" config section: bind a named engine to an input (IMAGES)
     and tasks (embed), and every image document Tika parses is embedded in
     one request per document tree; vectors land on the document the image
     belongs to, as it appears in the output. Per request, "inference" in
     parse-context selects bindings or switches inference off.
     openai-image-embedding-parser keeps working, one request per image
     (TIKA-4889).

   * New "engines" config section: a map of names to inference engines, each
     an endpoint or local binary with its settings (openai-embedding-engine
     posts up to maxBatchSize images per request to an OpenAI-compatible
     /v1/embeddings). Engines are called by the inference bindings that
     follow; nothing calls them yet (TIKA-4889).

   * The image embedder writes a picture's vector onto the document the
     picture appears in (INLINE and RENDERING children) with an "embedded"
     locator naming the picture; attachments keep their own, and
     liftToParent: false keeps every picture's vector on the picture.
     RecursiveParserWrapper exposes the parent's Metadata to a child's parse
     as ParentMetadata, and ChunkTarget (tika-inference) resolves where a
     child's chunks land (TIKA-4888).
     
   * A PDF page rendered as an embedded document (imageStrategy RENDER_PAGES_*)
     is enriched once, by the PDF parser: the text recognizer where the OCR
     strategy says so, every other text-recognizers entry on every page, results on
     the PDF itself. The embedded copy is no longer enriched again by
     image-parser, so an image embedder no longer yields two vectors for a
     text-poor page and Tesseract no longer OCRs a page twice; NO_OCR now means
     no page OCR at all. ContentEnrichers gains scoped suspend and
     suspendRecognizers for containers that enrich their own renderings
     (TIKA-4887).

   * OCR engines and other content enrichers: name the engine in "text-recognizers"
     and every default parser stays loaded; "text-recognizers": [] turns enrichment
     off everywhere. With no list, the loader resolves one engine
     per media type from the loaded parsers at startup (an engine configured under
     "parsers" beats one default-parser found; last registered wins; an entry's
     _mime-include/_mime-exclude applies) and logs one line per effective engine;
     a collision is a WARN at startup, as are two text recognizers sharing a type
     (both run, their text is concatenated). An engine under "parsers" is a parser
     only for the types no other parser there claims; when every type is claimed,
     startup logs that it acts only as the enricher, or warns that it never runs.
     The image/ocr-* pseudo-types
     are retired: Content-Type never carries the prefix (extracted images get their
     real extension instead of .bin) and a _mime-include/_mime-exclude naming one
     fails config load with the real type in the message. image-parser now owns
     image/jp2, image/jpx and image/x-portable-pixmap and calls the OCR engine on
     them. VLM parsers gain "textRecognizer" (default true); set it false for a
     captioning prompt so AUTO OCR never replaces extracted text with the VLM's
     output (TIKA-4884).

   * For enricher authors: implement ContentEnricher (TextRecognizer for OCR) and
     advertise the real image types; an enricher the classpath supplied is never
     dispatched to, and one named under "parsers" only fills gaps. A third-party
     engine still advertising image/ocr-* is treated as a legacy text recognizer
     with a WARN until 5.0. An invoked enricher is recorded in tk:parsed-by (images)
     and tk:parsed-by-full-set (PDF page renders) as a dispatched parser would be.
     ContentEnrichers.resolve(Parser) is the load-time resolution.
     EmbeddedDocumentUtil.normalizeMediaType is deprecated (TIKA-4884).

   * The PDF parser's AUTO OCR strategy no longer emits a page's extracted text
     next to the OCR output that superseded it. Text that trips the AUTO verdict
     is held back and replaced by OCR, and released unchanged when OCR cannot run
     on that page: no engine on the classpath or in text-recognizers,
     maxPagesToOcr exhausted, or a timeout or error. AUTO without an engine now
     behaves as NO_OCR (TIKA-4883).

   * New TextRecognizer capability for content enrichers (tika-core): a parser
     implements it to say its output is the document's text rather than an
     annotation. The PDF parser's AUTO strategy supersedes extracted text only
     when a text recognizer is configured (Tesseract, Tess4J and the VLM parsers
     declare it; an image-embedding enricher does not) and only when the engine
     actually wrote text. Other enrichers still run on the rendered page
     (TIKA-4883).

   * tika-app's -m and --json wrote metadata at the SAX endDocument event
     thereby dropping keys added by a parser after the endDocument (TIKA-4885).
     
   * "text-recognizers": _mime-exclude now matches the real media type, so
     excluding image/tiff also drops a legacy image/ocr-tiff engine (it was
     a no-op). The zero-media-types check at config load asks the engine
     itself, so a _mime-include list no longer masks an unreachable one.
     VLM parsers implement Closeable and release their HTTP client
     (TIKA-4884).

   * tika-inference and tika-vlm are experimental: classes, config keys and
     metadata output may change in minor releases without deprecation or
     compatibility shims (TIKA-4884).

   * A missing OOXML relationship target no longer aborts the whole file:
     the threaded-comment and person lookups in xlsx and the page lookups in
     vsdx went straight to POI's getRelatedPart, whose unchecked
     IllegalArgumentException surfaced as "Error creating OOXML extractor" and
     dropped the text already extracted. They route through
     safeGetRelatedPart, as branch_3x already did (TIKA-4879).
     
   * RawTiffDetector rejects a BigTIFF directory offset near Long.MAX_VALUE
     instead of letting the bounds check overflow. Adding the entry-count
     size to such an offset wrapped negative and read as "already in the
     prefix", so a 16-byte file threw ArrayIndexOutOfBoundsException out of
     Detector.detect, which CompositeDetector does not catch: detection
     failed for the document and the remaining detectors never ran. Raw
     detection runs on every stream, so this was reachable from every entry
     point (TIKA-4861).
     
   * Entries of ODF, EPUB, GeoGebra, WACZ, XLZ and iWork containers, mbox
     messages and the AppleSingle data fork are re-opened from their
     container on rewind instead of cached: digesting rewinds every
     embedded document, and the cached copy cost heap for the whole entry
     and, past the cache budget or the 1 MB floor, a temp file. AppleSingle
     no longer spools its data fork to a temp file on every parse, and
     GeoGebra no longer spools every embedded picture to detect it
     (TIKA-4878).

   * A declared Content-Length is no longer treated as a measurement: the
     zip-bomb ratio counts only measured input bytes (a container-declared
     size on an embedded document could inflate its denominator), and a
     re-openable source no longer reserves cache budget or sizes its buffer
     from the declared length (a lying one could push a small payload to
     disk or churn the shared budget). Neither is in a release: the
     exposure arrived with TIKA-4868 and TIKA-4873 (TIKA-4878).

   * Embedded objects in Office documents are re-opened from their container
     instead of cached: every OOXML part (pictures, media, attachments), the
     OLE 2.0 package inside an OOXML part, the CONTENTS entry of an OLE 2.0
     object in a binary Office file, embedded objects in .ppt, XPS page
     images and Word EMF icons. Digesting an embedded document rewinds it;
     the cached copy that made possible cost heap for the whole object and,
     past the cache budget or the 1 MB floor, a temp file. The container
     hands the bytes back on demand, so neither is needed (TIKA-4878).

   * PDF attachments, PDF XMP packets, 3D on-instantiate scripts and PST
     attachments are re-opened from their document instead of cached when
     the embedded-document extractor rewinds them (digesting does, for every
     embedded document). The cached copy cost heap for the whole attachment
     and, past the cache budget or the 1 MB floor, a temp file; PDFBox and
     java-libpst hand the bytes back on demand (TIKA-4878).

   * tika-server: named configuration presets (TIKA-4856).

   * Temp files follow -Djava.io.tmpdir on the parent JVM (Tika, its
     libraries, and forks all honor it); TikaLoader fails at config load
     if it is unusable. pipes.tempDirectory is deprecated for removal in
     5.0: it only covered the forks. Do not use tmpfs: spool size is
     bounded by input, and an orphaned fork dir pins RAM (TIKA-4877).

   * A fork whose parent dies deletes its own temp dir; the parent
     surfaces a fork's hs_err log before every delete. Failure-path temp
     file leaks fixed in PDFBoxRenderer, PopplerRenderer, truncated RTF,
     and MarianTranslator (TIKA-4877).
     
   * tika-eval Profile/Compare speedups: single-pass URL/mail stripping
     replaces the bounded regexes in langdetect preprocessing (same output,
     17-290x faster on web text), the default H2 db URL sizes the page cache
     at a quarter of the heap clamped to [64MB, 1GB] (override with
     -Dtika.eval.h2.cacheSizeKb=<kb>), and the status log adds a last-interval
     docs-per-sec rate next to the cumulative average (TIKA-4875).
     
   * New "text-recognizers" config list (TIKA-4872): select the OCR engine
     ("tesseract-ocr-parser", "tess4j-parser", "openai-vlm-parser", ...) by
     name instead of by classpath registration of the image/ocr-* pseudo
     media types. Enrichers advertise real media types (legacy engines that
     still advertise image/ocr-* are mapped to the real type, so all are
     nameable) and are invoked by the image and PDF parsers rather than
     dispatched to by the composite, so an enricher no longer displaces the
     parser registered for the same type. Enricher selection uses the
     detected media type, captured before a parser can refine Content-Type.
     Every enricher matching a media type runs, in config order (e.g. an
     OCR engine then a VLM tagger for the same image), best-effort: one
     enricher's failure does not stop the others and is still reported;
     timeouts abort the chain. The list is authoritative: a media type no
     configured enricher matches gets no enrichment -- never a classpath
     engine that was not named -- and a named engine that reports no media
     types at load (missing binary, unreachable inference server) fails
     config load instead of going silently inert. With no
     "text-recognizers" configured, the legacy ocr-* dispatch applies
     unchanged; a WARN at config load now names colliding OCR engines and
     the winner. TesseractOCRParser's
     component name is pinned as "tesseract-ocr-parser".

   * Inference/OCR hardening (TIKA-4871): OpenAIVLMParser no longer
     auto-registers via SPI, matching its Claude/Gemini siblings; select
     it by name ("openai-vlm-parser") in config. Per-request parse-context
     config for the embedding filters now works and is validated:
     {"openai-embedding-filter": {"skipEmbedding": true}} (likewise
     "jina-embedding-filter") merges over the server config, and
     baseUrl/apiKey/model may not be changed at runtime. The embedding
     filters release their HTTP client resources on close(). Inline PDF
     page OCR now accumulates tk:chunks from every OCR'd page onto the
     parent document instead of keeping only the first page's.

   * Placeholder streams -- the empty stand-ins parsers hand parseEmbedded
     for content that is never extracted -- report an unknown length rather
     than their own zero, and the macro-failure entry is registered without
     parsing its sentinel (TIKA-4874).

   * TikaInputStream.hasReliableLength() distinguishes measured lengths
     from declared Content-Length hints, and one-shot streams now carry a
     declared length without spooling; detection sizes its magic read only
     from measured lengths, so a lying declared length can no longer
     truncate it (TIKA-4868).

   * ParseContext entries holding per-parse runtime state (ParseRecord,
     ParseTimeout, the parser-map cache) are skipped during serialization
     instead of failing as unregistered components (TIKA-4868).

   * Mojibuster's adaptive probe strips incrementally instead of
     re-stripping the whole buffer on every read (quadratic on tag-heavy
     pages); JunkDetector's Unicode block lookup uses a precomputed BMP
     table. Output unchanged (TIKA-4868).

   * Markdown rendering is another ~8x faster on large documents: a custom
     Text-node renderer emits unescaped spans in bulk instead of
     commonmark's per-character escape-check-and-append. Byte-identical
     output, guarded by a fast-vs-stock differential test (TIKA-4868).

   * ZipParser no longer re-decompresses an entry on every rewind when the
     entry uses a legacy compression method (implode, shrink, bzip2, ...):
     such entries replay from the budgeted cache instead of re-opening.
     An imploded 606KB zip drops from 610ms to 102ms (TIKA-4868).

   * More detection/dispatch savings: the message/rfc822 priority-45 magic
     is gated behind a lossless ':' scan of the first 30 bytes;
     CompositeParser caches the built type->parser map in the ParseContext
     so embedded documents reuse the container's map; MagicDetector
     precomputes a per-pattern first-byte table. Adds a RESOURCE_TIMING
     log, silenced by default in the shipped log4j2 configs (TIKA-4868).

   * WordExtractor (.doc) cleans each character run and tests paragraph
     blankness in single passes; ToMarkdownContentHandler collapses line
     breaks copy-free for clean runs. ~39% off a text-heavy 2MB .doc
     (TIKA-4868).

   * tika-server's raw-output endpoints (/tika, /tika/text, ...) carry the
     extracted content as raw UTF-8 bytes from the pipes worker to the HTTP
     response instead of a Smile-encoded string (9MB text: 115ms -> 75ms).
     Opt-in via the new content-bytes-config parse-context component, which
     moves CONTENT_ONLY passback content out of tk:content into
     EmitData.getContentBytes(); results routed to a regular Emitter get
     the content restored to the metadata (TIKA-4868).

   * Detection hot-path cleanups: MagicMatch resolves its detector via
     double-checked locking; glob patterns are compiled once at
     registration; MimeTypes.forName reads a ConcurrentHashMap (fixing an
     unsynchronized-read race) and indexes normalized keys; resource names
     containing spaces skip the URI-parse-by-exception; the magic-header
     buffer is sized by the stream's measured length instead of a fixed
     64KB; the Adobe Illustrator ranged regex is gated behind a literal
     scan. Detection results unchanged (TIKA-4868).

   * Magic detection is ~35% faster on unmatched (e.g. plain-text) input:
     range scans find first-byte candidates before running the full
     masked/case-folded compare (TIKA-4868).

   * CSVSniffer reads its detection window once into a shared buffer and
     runs every delimiter hypothesis against it; windows with no delimiter
     and no quote skip the scan outright. Results unchanged (TIKA-4868).

   * Pipes workers no longer stall between pre-parse and parse waiting
     for the client to acknowledge the intermediate-result frame; the
     ACK round trip now overlaps the parse. Adds per-request timing logs
     on org.apache.tika.pipes.timing.*, silenced by default in the shipped
     log4j2 configs; raise that logger to info to enable (TIKA-4868).

   * DefaultDetector honors CONTENT_TYPE_USER_OVERRIDE and
     CONTENT_TYPE_PARSER_OVERRIDE before running magic detection, matching
     CompositeDetector's contract. Removes the second full magic scan every
     pipes parse paid per document. Compat note: with either override set,
     DefaultDetector no longer lets a more specific magic result overrule
     it; parts whose parser declares a type from container headers (e.g.
     inline text/* mail parts) now report the declared type, and
     CONTENT_TYPE_MAGIC_DETECTED is not recorded when an override short
     circuits detection (TIKA-4868).

   * Markdown output is ~4x faster on large documents:
     ToMarkdownContentHandler now buffers the commonmark renderer's
     per-character writes instead of paying the synchronized
     Writer.write(int) cost for every character (TIKA-4868).
   * Embedded documents carry their size: ParsingEmbeddedDocumentExtractor
     sets Content-Length from the stream where the stream knows it and the
     parser did not say, which never spools to measure one, and the raw
     camera previews and the audio cover art set the length they read from
     the file (TIKA-4873).

   * AVIF images are parsed rather than only detected: HeifParser accepts
     image/avif, which is the same ISO-BMFF container, so dimensions, EXIF
     and XMP come out of it the way they do for HEIC (TIKA-4870).

   * The video of a Google/Android motion photo, appended after the image and
     described by the Motion Photo or MicroVideo XMP, is emitted as an
     ATTACHMENT embedded document named after what the file declares it to
     be. Nothing is emitted, and nothing is recorded, when the declared
     length does not fit the file or the bytes there are not recognized,
     which is what sharing a motion photo out of a gallery leaves behind
     (TIKA-4869).

   * Raster previews for the vector thumbnails of Office documents: the new
     poi-metafile-renderer draws EMF and WMF images through POI (a PNG of
     a configurable width; Word's bitmap-in-WMF thumbnails from the bitmap
     directly), EMFParser and WMFParser are RenderingParsers that emit the
     rendering as a RENDERING embedded document with "emf-parser" /
     "wmf-parser": {"renderImage": true, "renderWidth": 800}, off by
     default and restrictable to e.g. THUMBNAIL embedded documents with
     "renderOnlyEmbeddedResourceTypes", and OfficeParser emits the
     SummaryInformation thumbnail of the OLE2 formats (a WMF) as a THUMBNAIL
     embedded document, as the OOXML parsers do with the docProps thumbnail,
     switchable with "office-parser": {"extractThumbnail": false}
     (TIKA-4855).

   * Add "exception-reporting" parse-context config to redact and bound
     exception text in metadata, tika-server error bodies and pipes/grpc
     messages; FileSystemEmitter writes atomically (TIKA-4848).
     Compat notes: a truncated TSD envelope now records its read failure
     under tk:exception:embedded-stream-exception rather than
     tk:exception:embedded-exception; recordException and
     recordEmbeddedStreamException no longer strip a bare TikaException
     wrapper, so the first line of tk:exception:* values may name the
     wrapper (affects consumers keyed on that line, e.g. eval cause
     counts across the 4.1 boundary).
     
   * Audio cover art is emitted as a THUMBNAIL embedded document, like the
     preview image of the document container formats: the front cover (ID3
     APIC and FLAC/Vorbis picture type 3), else the first picture of type
     "Other" or unknown, else the first picture, and the first covr image
     of an MP4. Further pictures
     stay INLINE. Clients that looked for cover art as INLINE need to
     accept THUMBNAIL as well (TIKA-4850).

   * tika-grpc resolves its plugin-roots fallback against the install
     layout via DefaultPluginsDir instead of a working-directory-relative
     pf4j default, and a WARN names the resolved directory when no plugins
     directory exists (TIKA-4865).

   * The tika-server full and tika-grpc Docker images install fonts-noto-cjk:
     without any CJK face, PDFs using non-embedded CJK fonts render (and OCR)
     as .notdef boxes in every renderer, even though the images ship Japanese
     tesseract data (TIKA-4866).
     
   * The default plugins directory is resolved against the install layout
     (next to the jar, or next to its lib/ directory) and always as an
     absolute path, shared by tika-server, PipesForkParser and the async
     CLI; it no longer depends on the working directory (TIKA-4864).

   * tika-server error bodies (the 422/500 exception mapper, /meta/{field})
     now honor the exception-reporting policy; /meta/{field} returns the
     already-formatted container exception instead of re-wrapping it with
     server frames. Completes TIKA-4848 (TIKA-4848).

   * The exception-reporting policy now also governs the messages a pipes
     worker returns (fetch/emit/crash) and the container exception it
     records; part of TIKA-4848 step 3 (TIKA-4848).
     
   * Allow image compression settings in PDFBox-based renderer (TIKA-4862).

   * The tika-server full and tika-grpc Docker images set OMP_THREAD_LIMIT=1:
     to avoid oversubscribing the CPU under forked parse workers (TIKA-4863).

   * embedded-limits maxDepth counts embedding levels again instead of the
     parsers a parse passes through; with AutoDetectParser over DefaultParser
     every value above 1 used to stop one level early (TIKA-4857).

   * GeoGebraParser emits the icon of a tool (*.ggt, the macro's iconFile)
     as its THUMBNAIL embedded document; tool files have no thumbnail of
     their own (TIKA-4831).

   * Enum values in JSON configuration are matched case-insensitively, so
     "no_ocr" works as well as "NO_OCR"; the server docs used the lower-case
     form in their examples (TIKA-4859).

   * The preview image of iWork '09 packages (QuickLook/Thumbnail.jpg) and of
     iWork '18 packages (preview.jpg) is emitted as a THUMBNAIL embedded
     document, as it already was for iWork '13 (TIKA-4854).

   * Raw camera formats are detected by content: RawTiffDetector tells
     Nikon NEF/NRW, Pentax PEF/PTX, Sony ARW/SRF/SR2, Samsung SRW and Adobe
     DNG from a plain TIFF by their image directory (DNGVersion, the vendor
     Compression codes, or a CFA/LinearRaw image plus Make), and Fuji RAF,
     Panasonic RW2, Minolta MRW and the remaining Olympus ORF byte orders
     get magic entries. Streams without a file name used to be image/tiff.
     image/x-raw-samsung (*.srw) is new and parsed by RawTiffParser
     (TIKA-4861).

   * DWGReadParser emits the drawing's THUMBNAILIMAGE as a THUMBNAIL embedded
     document instead of INLINE (TIKA-4853).

   * EpubParser emits the cover image named by the OPF (the EPUB 3
     cover-image manifest property, or the EPUB 2 cover meta) as a THUMBNAIL
     embedded document (TIKA-4852).

   * RawTiffParser marks only the largest embedded JPEG preview as the
     THUMBNAIL embedded document; the smaller previews of the same image are
     INLINE images named image-N.jpg. Previously every preview was a
     THUMBNAIL, so a client had to compare them to find the representative
     one (TIKA-4851).

   * tika-eval: Profile/Compare accept the batch run's jsonl crash ledger
     (--pipesReport, -pa/-pb) and a run-info json (--runInfo, -ra/-rb), and
     read both from <extracts>/.run-info/ by default (refusing an ambiguous
     dir). containers gains pipes_status/pipes_message; a new run_info table
     records eval and batch provenance; reports and
     summary.md classify NO_EXTRACT_FILE by ledger status (CRASH, the raw
     status, NO_PIPES_RECORD, BATCH_WITHOUT_LEDGER, NO_PIPES_REPORT_SUPPLIED).
     Report on a db from an earlier tika-eval skips the reports it cannot run
     instead of aborting (TIKA-4847).

   * New file-system-jsonl-reporter pipes reporter (TIKA-4846).
   
   * Stop spooling OLE2 objects whose header over-reserves BAT capacity
     (TIKA-4845).

   * Add Micrometer reporting and opt-in endpoint for tika-server (TIKA-4839).
     
   * Improve spooling/decrease number of spills to disk (TIKA-4835).

   * Fixed a bug that made per-request (parse-context) configuration unusable
     for parsers that lock some config fields against caller modification --
     Tess4J, the VLM parsers and the OpenAI image-embedding parser. Any such
     config threw, including an empty one: the defaults were deep-copied
     through their own setters, which the runtime config overrides to reject
     caller input, so the copy tripped the parser's own guards before the
     caller's JSON was read. Locked fields are still rejected when a caller
     actually sets them. Configuration supplied at initialization time (the
     "parsers" section) was never affected (TIKA-4843).
     
   * OOXML parsers flag package parts that are unreachable through the OPC
     relationship graph: msoffice:has-unreferenced-parts (boolean) and
     msoffice:unreferenced-part-names. Purely structural (no bytes are
     inspected; content types come from [Content_Types].xml by extension), so
     expect false positives from tools that leave orphan parts behind. A hiding
     place a raw-ZIP scanner can still see, not a statement about what Tika
     parsed. Applies to Word, Excel, PowerPoint and Visio OOXML (including
     macro-enabled variants); XPS links content by markup rather than
     relationships and is not checked (TIKA-4837).
     
   * Shared pipes server (useSharedServer: true, not the default): a client whose
     in-flight parse was killed by another client's restart could restart the
     healthy replacement. ensureRunning holds its lock across the whole fork, so
     siblings cannot report a dead worker until after the replacement is up, and
     the pending-restart flag carried no process identity -- so a report about
     the process that just died was applied to its successor, which was then
     destroyed and re-forked. One worker death produced two restarts and a second
     round of destroyed in-flight work; under sustained concurrent load it
     sustained itself at one spurious restart per round, appearing as periodic
     unexplained worker churn and intermittent parse failures that succeed on
     retry. Each fork now carries a generation that clients capture when they
     connect and hand back with every report, and reports about a superseded
     process are dropped. Also fixed in shared mode: ensureRunning could fork a
     replacement after shutdown() that nothing owned and nothing would ever
     destroy, and an interrupt during process teardown left the process handle
     pointing at a killed process and leaked the temp directory. Affects 4.0.0
     and earlier (TIKA-4844).

   * tika-pipes: the cache memory budget (how much rewindable content a forked
     worker keeps in memory before spilling to disk; new since 4.0.0, which had
     no budget at all) defaults to a quarter of the fork's heap, so raising
     -Xmx raises it. It is one pool per forked JVM shared by all of its threads.
     -Dtika.pipes.cacheMemoryBudgetBytes in forkedJvmArgs overrides it (below
     the quarter-heap ceiling; <=0 disables); the fork logs the value and its
     source at startup. TikaInputStream.hasFile() now also reports content the
     stream cache spilled on its own, not only content a getPath() call put on
     disk; note getPath() may still have to drain the rest of the source into
     that file. TikaInputStream.toString() no longer forces a spill, so logging
     or debugger-inspecting a stream is side-effect-free.
     TikaInputStream.inMemoryContent(channel) gives a zero-copy read-only view
     of cached content for consumers that need random access. Digester
     gains digestSink(), a DigestSink that digests as it is written; nothing is
     written to the metadata unless the producer calls commit(), so any failure
     -- exception, Error, or a producer that closes the sink itself -- publishes
     no digest rather than a digest of the bytes that happened to arrive. A
     translator that claims a stream and writes nothing likewise publishes
     nothing: embedded PST mail items, whose translator is still a stub, no
     longer carry the digest of zero bytes (the same value for every one of
     them) and instead carry no digest at all. DigestHelper uses it for
     translated embedded streams, which no longer touch a temp file when the
     digester implements digestSink (all of Tika's do; one that only implements
     digest() still buffers).
     TemporaryResources.closeAll(Closeable...) closes every argument even when
     one throws unchecked; TemporaryResources, CachingSource, CachingInputStream
     and CompositeDigester use it (TIKA-4835).

   * Documentation: corrected a batch of pages and javadocs that contradicted
     the code. Notably: the ES/OpenSearch attachmentStrategy has no default
     (unset means embedded documents get neither the parent field nor the
     parent/child relation); Kafka's connectionsMaxIdleMs is passed to the
     producer, not ignored; jdbc queryTimeoutSeconds is applied only when > 0,
     so 0 does not mean "no limit"; the Solr emitter/iterator support only
     basic auth, not ntlm, and only when a userName is set; pipes-reporters
     silently loads zero reporters when given a JSON array, and
     pipes-iterator/pipes-reporters instances are built at config load rather
     than lazily; under CONTENT_ONLY only a parse-context filter replaces the
     built-in one, not the top-level metadata-filters chain;
     _mime-include/_mime-exclude also accept a bare string; Tess4J locks
     poolSize and maxImagePixels as well as the two paths; and pdf:trapped and
     xmp:pdf:Trapped are new 4.x keys rather than renames (3.x captured the
     flag only as pdf:docinfo:trapped and dropped the XMP value). Also
     corrected the config nesting shown in every pipes-plugin fetcher/emitter
     javadoc -- 23 of them, which had it inverted (the instance id is the
     outer key, the component name the inner) -- and removed references to a
     TesseractOCRConfig.properties file that 4.x does not load (TIKA-4842).

   * Pipes plugins no longer bundle their own Jackson: jackson-core, -databind
     and -annotations are provided by the host (tika-serialization) and the
     plugins parent pom now bans bundling them, so a mapper can cross the
     plugin boundary without a second copy of the Jackson classes (seven plugin
     zips shipped one). Plugin configuration JSON is parsed by one shared
     mapper, PluginJson (tika-plugins-core), which rejects unknown keys,
     numbers for enums and duplicate keys, and accepts
     // and /* */ comments; the 33 per-plugin *Config classes use it instead of
     their own bare ObjectMapper (TIKA-4840).

   * tika-server and tika-async-cli now start from a config that contains
     // or /* */ comments, as the configuration docs have always said they
     may. The main loader accepted them; the steps that re-read the user's
     file to merge in server/CLI overrides (ConfigMerger, ensurePluginRoots)
     used their own bare parser and refused the whole file; they now use the
     shared TikaObjectMapperFactory mapper (TIKA-4834).

   * The Kafka pipes iterator no longer stops at the first empty poll. A newly
     subscribed consumer spends its first poll(s) joining the group and returns
     empty even when the topic has a backlog, so the iterator could enqueue zero
     files and report success. It now waits for a partition assignment (bounded
     by the new assignmentTimeoutMs, default 30s) and requires a continuous quiet
     window (drainIdleMs, default 1s) before concluding the topic is drained.
     groupInitialRebalanceDelayMs is deprecated and no longer sent to the
     consumer: it is a broker setting that Kafka has always ignored (TIKA-4833).

   * Pipes IPC: carry inline document bytes as a raw binary field beside the
     tuple in the request envelope -- never inside the tuple or its
     ParseContext -- and disable Smile's 7-bit binary encoding. Tuple JSON
     serialized by 4.0.0 with an "inline-bytes" parse-context entry no longer
     loads; it is rejected with a tailored message (TIKA-4829).

   * Digesting embedded documents no longer buffers each embedded object to a
     temp file. Zip entries are re-read from the parent archive on rewind, and
     a new process-wide CacheMemoryBudget (seeded by the pipes forked server;
     default 256MB, clamped to a quarter of the fork's heap; tunable via
     -Dtika.pipes.cacheMemoryBudgetBytes in the config's forkedJvmArgs, <=0
     disables) lets embedded objects stay in memory past the per-object 1MB
     threshold. New public API on TikaInputStream: get(IOSupplier,...),
     enableRewind(CacheMemoryBudget), getSeekableByteChannel(). Zip/7z/epub/odf
     parsing and zip container detection now read through seekable channels, so
     after detection/parsing a TikaInputStream may no longer be file-backed
     (hasFile() false); getPath()/getFile() still work and spool on demand
     (TIKA-4828).
     
   * Pipes now carries the caller-supplied Content-Type across the worker's
     fresh-metadata boundary as a soft detection hint, so every forked-parse
     endpoint (/tika, /meta, /rmeta, /unpack, /async, /pipes, plus tika-grpc
     and embedded PipesForkParser) can route on a client Content-Type, not
     only on the filename. Detection keeps the hint only when it equals or
     specializes the content-detected type (e.g. refining image/tiff to
     image/x-canon-cr2); for bytes with no magic it can select any type,
     matching the routing power the filename already had. The
     CONTENT_TYPE_USER_OVERRIDE key is deliberately not carried, so the hint
     cannot force an unrelated type (TIKA-4825).

   * OneNote extraction now follows document order, omits superseded page
     revisions, sorts author metadata, extracts embedded object BLOBs, and
     bounds malformed-input recursion and file-derived allocations. Parse
     warnings and embedded relationship IDs are exposed in metadata. Malformed
     or truncated files that cannot be fully parsed, and files whose walk
     yields no content, now fall back to the legacy string dump instead of
     failing or returning empty output. The legacy MS-ONESTORE walker bounds
     its recursion (depth caps plus file-node-list and fragment-chain cycle
     guards) and now honors shouldParseEmbedded for embedded file data
   * PDF: extractFontNames threw NullPointerException on a page with no
     /Resources dictionary (TIKA-4842).

   * tika-server: opt-in Micrometer metrics reporting and endpoint
     (TIKA-4839).

   * Per-request (parse-context) config for parsers that lock fields
     (Tess4J, VLM, OpenAI image-embedding) threw even when empty; locked
     fields are still rejected when actually set (TIKA-4843).

   * OOXML: new msoffice:has-unreferenced-parts and
     msoffice:unreferenced-part-names flag package parts unreachable via the
     OPC relationship graph. Structural only, expect false positives; not
     applied to XPS (TIKA-4837).

   * Shared pipes server (useSharedServer: true): a worker death could trigger
     a second, spurious restart that killed the healthy replacement. Forks now
     carry a generation; stale reports are dropped. Also fixed: a fork after
     shutdown() that was never destroyed, and a temp-dir leak on interrupt
     during teardown (TIKA-4844).

   * tika-pipes cache memory budget defaults to a quarter of the fork heap;
     override with -Dtika.pipes.cacheMemoryBudgetBytes in forkedJvmArgs
     (<=0 disables). TikaInputStream: hasFile() also reports cache spills,
     toString() no longer spills, new inMemoryContent(channel). Digester gains
     digestSink(); a digest is published only on commit(), so failed or empty
     translations (e.g. stub PST items) publish no digest. New
     TemporaryResources.closeAll(Closeable...) (TIKA-4835).

   * Docs/javadocs reconciled with the code: ES/OpenSearch attachmentStrategy
     has no default; Kafka connectionsMaxIdleMs is honored; jdbc
     queryTimeoutSeconds 0 is not "no limit"; Solr basic auth only; pipes
     reporters/iterators are built at config load; Tess4J also locks poolSize
     and maxImagePixels; pdf:trapped is new, not renamed; plugin config
     nesting fixed in 23 javadocs (TIKA-4842).

   * Pipes plugins no longer bundle Jackson; the host provides it. Plugin
     config is parsed by a shared strict PluginJson mapper (rejects unknown
     and duplicate keys; accepts comments) (TIKA-4840).

   * tika-server and tika-async-cli accept // and /* */ comments in config
     during override merging, as documented (TIKA-4834).

   * Kafka pipes iterator no longer stops on the first empty poll; waits for
     partition assignment (assignmentTimeoutMs, 30s) and a quiet window
     (drainIdleMs, 1s). groupInitialRebalanceDelayMs is deprecated
     (TIKA-4833).

   * Pipes IPC carries inline bytes as a raw binary field, not in the tuple;
     Smile 7-bit binary encoding disabled. 4.0.0 tuples with an "inline-bytes"
     parse-context entry are rejected (TIKA-4829).

   * Digesting embedded documents no longer spools each to a temp file; a
     process-wide CacheMemoryBudget (default 256MB) keeps them in memory. New
     TikaInputStream API: get(IOSupplier,...), enableRewind(CacheMemoryBudget),
     getSeekableByteChannel(). Zip-family parsing and detection use seekable
     channels, so hasFile() may be false afterward; getPath() still spools on
     demand (TIKA-4828).

   * Pipes carries the client Content-Type into the forked worker as a
     detection hint for all forked endpoints; honored only when it equals or
     specializes the detected type, or when there is no magic. The
     user-override key is not carried (TIKA-4825).

   * OneNote: document-order extraction, superseded revisions omitted, embedded
     BLOBs extracted, warnings and relationship IDs in metadata, bounded
     recursion/allocation; malformed files fall back to the legacy string dump
     (TIKA-4814).

   * New GeoGebraParser for *.ggb/*.ggs/*.ggt: geogebra:* metadata, text and
     the thumbnail as a THUMBNAIL embedded document, with content-based
     detection. Previously typed application/zip with every entry as an
     attachment. *.ggs and *.ggp are new mime types; *.ggp is glob-only
     (TIKA-4831).

   * RawTiffParser extracts the camera-generated JPEG previews embedded in
     TIFF-based raw images (Nikon NEF/NRW, Sony ARW/SRF/SR2, Pentax PEF/PTX,
     Adobe DNG and Canon CR2, including BigTIFF DNG containers) as thumbnail
     embedded documents. image/x-raw-{nikon,sony,pentax,adobe} are now
     sub-classes of image/tiff, so a named NEF/ARW/PEF/DNG that used to detect
     as image/tiff (TiffParser, metadata only) now detects as image/x-raw-* and
     emits thumbnail-N.jpg attachments in /rmeta and /unpack; CR2 keeps its
     detection but also gains the attachments. Disable via
     "raw-tiff-parser": {"extractPreviews": false} (TIKA-4824).
   * RawTiffParser extracts embedded JPEG previews from NEF/NRW, ARW/SRF/SR2,
     PEF/PTX, DNG and CR2 as thumbnail embedded documents. image/x-raw-* are
     now subtypes of image/tiff, so named raw files detect as image/x-raw-*.
     Disable with "raw-tiff-parser": {"extractPreviews": false} (TIKA-4824).

Release 4.0.0 - 8/18/2026

  This section is the complete delta from 3.x. It includes everything first
  released in 4.0.0-alpha-1 and 4.0.0-beta-1; those sections below are stubs.

  Upgrading from 3.x? Start with the migration guides at
  https://tika.apache.org/docs -- "Migrating to Tika 4.x", "Migrating Tika
  Server to 4.x" and "Metadata Changes in Tika 4.x". They carry the detail
  and the code examples behind the summaries here.

  Important architectural change: parsing now runs in a forked process
  where possible. tika-server's endpoints, tika-app's -a/--async and -f/--fork,
  and tika-grpc all parse in forked, crash-isolated tika-pipes workers.
  Applications embedding Tika should consider getting the same isolation from
  PipesForkParser (tika-pipes-fork-parser) rather than parsing in-process with
  AutoDetectParser. Note that the project does not treat denial of service --
  memory exhaustion, CPU exhaustion, a crashed process -- as a security issue
  when files are parsed outside these isolated paths; see
  https://tika.apache.org/security-model.html.

  BREAKING CHANGES

  --- Platform, packaging, configuration and output format (everyone) ---

   * Tika 4.x requires Java 17 or later; 3.x built and ran on Java 11. All
     published artifacts are compiled with --release 17 (TIKA-4685).

   * tika-app and tika-server-standard now ship as zip distributions with an
     adjacent lib/ directory; the published jars are thin launchers and fail
     with NoClassDefFoundError if run on their own. This catches
     tika-server-standard hardest, because its jar is still on Maven Central:
     unzip the distribution and run from inside it (TIKA-4733).

   * tika-parsers-standard-package is now a pom, not a jar. Users must add
     <type>pom</type> in Maven or @pom in Gradle (TIKA-4712).

   * The default content handler is now Markdown. tika-app, tika-server (the
     /tika and /rmeta endpoints) and the async/pipes CLI emit Markdown content
     by default instead of XHTML/XML (plain text for the async CLI). Request
     the previous format explicitly: tika-app -x/--xml, the server's /tika/xml
     and /rmeta/xml paths, the async CLI --handler x (TIKA-4663).

   * Configuration moved from XML to JSON. TikaConfig and the
     org.apache.tika.config XML-configuration API are removed: TikaConfig,
     ConfigBase, Field, Param, ParamField, LoadErrorHandler,
     InitializableProblemHandler, TikaConfigSerializer and TikaTaskTimeout are
     gone. Use TikaLoader from tika-serialization. tika-app
     --convert-config-xml-to-json converts a 3.x parsers section as a starting
     point; every other section needs manual migration
     (TIKA-4544, TIKA-4545, TIKA-4553, TIKA-4565).

   * An unregistered component name in a default-parser, default-detector or
     default-encoding-detector "exclude" list now throws a TikaConfigException
     at config load instead of logging a WARN, so a 3.x config that named the
     component by class name or misspelled it now refuses to start. Use the
     registered name (e.g. "pdf-parser"); tika-app --list-parser-names prints
     them (TIKA-3268, TIKA-4808).

   * Metadata keys were renamed for consistency and provenance. Every
     Tika-asserted key now lives under a single tk: prefix, replacing 3.x's
     scattered X-TIKA:, tika:, tika_pg:, rendering:, signature: and
     imagereader: prefixes and bare names such as resourceName; names Tika
     coined inside format namespaces are kebab-cased (pdf:hasMarkedContent ->
     pdf:has-marked-content) while names from a file or an external standard
     keep their spelling; and open key families gained prefixes (audio:, ner:,
     envi:, ogg:streams-, grobid:, iso19115:, gdal:, geotopic:, mif:, idml:).
     Code using the TikaCoreProperties / TikaPagedText / Rendering constants is
     unaffected. Code that references keys by String has two paths: update the
     strings with the key-for-key tables in metadata-changes-4x.adoc, or turn
     on the compatibility filter below and migrate on your own schedule
     (TIKA-4816).

   * The opt-in legacy-key-migration-filter restores 3.x key spellings at the
     emit edge (default direction V4_TO_V3), so an unmigrated consumer keeps
     working against 4.x output; V3_TO_V4 maps 3.x names forward instead.
     tika-core bundles metadata-migration-3x-4x.json, the machine-readable
     rename/drop table (TIKA-4797).

   * The reserved tk: (and legacy X-TIKA:) namespace is now a trust boundary
     for String-keyed writes. Metadata#set/add(String, String) throw
     IllegalArgumentException on a reserved key instead of 3.x's silent
     success, where a document-controlled property named X-TIKA:Parsed-By could
     overwrite Tika's own value, and Property's public factories reject reserved
     names outright. Document- and tool-derived names now go through
     Metadata#add(KeyPrefix, String, String) -- append-only, skip-and-WARN on
     hostile names -- or its Instant overload for source-typed dates
     (TIKA-4816).

   * Metadata no longer implements CreativeCommons, Geographic, HttpHeaders,
     Message, ClimateForcast, TIFF or TikaMimeKeys: inherited constants move to
     their home interface, e.g. Metadata.CONTENT_TYPE becomes
     HttpHeaders.CONTENT_TYPE (now a Property, though the key string is
     unchanged). TikaMimeKeys and
     ClimateForcast are deleted outright; ClimateForecast (corrected spelling)
     replaces the latter, with its keys under cf: (TIKA-4816).

   * Other Metadata API changes: setAll(Properties) removed with no replacement
     -- it bypassed both the limiter and the reserved-key guard, so use
     putAll(Metadata) or individual set/add calls; PassthroughPrefix renamed
     KeyPrefix; the Property factories internalClosedChoise / internalOpenChoise
     / externalClosedChoise / externalOpenChoise renamed to ...Choice with no
     forwarders; the dead enum constants PropertyType.STRUCTURE and
     ValueType.{LOCALE, MIME_TYPE, PROPER_NAME, URL, XPATH} removed; package
     org.apache.tika.metadata.writefilter renamed to ...metadata.writelimiter.
     Metadata's serialVersionUID also changed, so a 3.x-serialized instance now
     fails with InvalidClassException instead of deserializing into an object
     that throws on first write (TIKA-4816).

  --- Java API (library integrators) ---

   * The core SPI signatures changed. Parser.parse takes a TikaInputStream
     instead of an InputStream (there is no InputStream overload),
     Detector.detect takes (TikaInputStream, Metadata, ParseContext), and
     EmbeddedDocumentExtractor's shouldParseEmbedded/parseEmbedded gained a
     ParseContext and take a TikaInputStream. Every third-party implementation
     must be updated; callers can wrap with TikaInputStream.get(...). The Tika
     facade still accepts an InputStream, but Tika.detect(InputStream, ...) no
     longer returns the caller's stream at its original position. The detector
     still resets the TikaInputStream it reads -- but that read-ahead is
     buffered inside an internal wrapper that detect() discards, so the
     caller's own stream comes back advanced. Pass a TikaInputStream you own
     (and rewind it), or re-open the source
     (TIKA-4399, TIKA-4541, TIKA-4569).

   * TikaInputStream no longer caches by default. A stream is consumed in
     passthrough mode unless enableRewind() is called at position 0;
     rewind()/getFile()/getPath() after reading without enableRewind() throw
     instead of silently spooling. A parser that read part of a stream and then
     asked for a file worked in 3.x and now fails. Digesters call
     enableRewind() themselves (TIKA-4618, TIKA-4623).

   * Parsing with a concrete parser (not AutoDetectParser) and an empty
     ParseContext no longer auto-generates an AutoDetectParser to handle
     embedded files: they are silently skipped, with no content and no
     exception. Nor does it auto-generate a Detector to identify them; they are
     reported as application/octet-stream instead. Set Parser.class and
     Detector.class in the ParseContext, or go through AutoDetectParser, which
     does this for you (TIKA-4819).

   * EmbeddedDocumentExtractorFactory and friends are removed;
     ParsingEmbeddedDocumentExtractor and UnpackExtractor are now stateless
     singletons (use INSTANCE) that take the enclosing ParseContext as a method
     parameter rather than capturing one at construction. Code that supplied a
     custom factory should bind an EmbeddedDocumentExtractor instance directly.
     EmbeddedDocumentUtil's instance API is likewise removed in favor of statics
     that take a ParseContext explicitly (TIKA-4819).

   * ParseContext configuration is now resolved per component instance rather
     than per config class, because a class-keyed write leaked one component's
     config to every other component binding the same config class. Two
     consequences: parseContext.get(SomeConfig.class) no longer returns a
     JSON-resolved config, so a third-party component following the
     PDFBoxRenderer pattern must be handed its config explicitly; and precedence
     is inverted -- a JSON config now beats a programmatic
     context.set(XConfig.class, ...), which used to win (TIKA-4808).

   * ForkParser and the entire org.apache.tika.fork package are removed from
     tika-core. Out-of-process parsing is now provided by PipesForkParser in
     the new tika-pipes-fork-parser module -- the recommended parser for
     untrusted documents. tika-app's -f/--fork routes through it, and
     --fork-timeout is rejected rather than silently ignored
     (TIKA-4554, TIKA-4571, TIKA-4651).

   * Unified timeout model across the library, pipes and server: a total-task
     budget plus a progress/stall timeout, composed recursively over embedded
     documents. TikaTimeoutException is now a checked exception, and several
     parser/pipes config fields were renamed (*TimeoutSeconds / *TimeoutMs ->
     *TimeoutMillis, including a unit change for Tess4J) (TIKA-4813).

   * Parsers and detectors no longer expose bean setters/getters for their
     settings. Configuration moves to per-component *Config objects supplied
     through the ParseContext (e.g. GeoParserConfig, DWGParserConfig,
     AmazonTranscribeConfig, MagikaDetector/SiegfriedDetector configs)
     (TIKA-4758).

   * The encoding detectors moved out of parser packages into
     org.apache.tika.detect.* and into new tika-encoding-detector-* modules:
     org.apache.tika.parser.txt.{CharsetDetector,CharsetMatch,
     Icu4jEncodingDetector,UniversalEncodingDetector,BOMDetector,...} are now
     org.apache.tika.detect.icu4j.*, org.apache.tika.detect.universal.* and
     org.apache.tika.detect.BOMDetector, and
     org.apache.tika.parser.html.HtmlEncodingDetector is now
     org.apache.tika.detect.html.HtmlEncodingDetector.
     NonDetectingEncodingDetector is removed (TIKA-4685, TIKA-4720).

   * MetadataListFilter has been renamed MetadataFilter, and the 3.x
     MetadataFilter has been removed (TIKA-4546).

   * API changes in the EmbeddedStreamTranslator (TIKA-4518), and
     DigestingParser is removed (TIKA-4607).

   * BasicContentHandlerFactory.parseHandlerType now throws
     IllegalArgumentException for an unrecognized handler name instead of
     silently returning the supplied default (TIKA-4809).

  --- tika-server ---

   * All parsing now runs out-of-process through tika-pipes. /tika, /rmeta,
     /meta, /unpack, /detect, /pipes and /async share a fixed pool of
     numClients forked worker JVMs (default derived from host cores), so a
     parser crash, OOM or timeout no longer takes down the server. The cost is
     a sizing decision 3.x never asked of you: numClients is both the server's
     concurrency ceiling and its CPU/memory footprint, and each fork's heap is
     set with pipes.forkedJvmArgs (e.g. -Xmx1g), not the server JVM's. Size
     both deliberately; see the cpu-sizing docs (TIKA-4809).

   * Capability flags are default-deny and split in two. enableUnsecureFeatures
     no longer exists -- a config still carrying it fails to start with an
     "Unrecognized field" error -- and is replaced by allowPipes (gates /pipes
     and /async) and allowPerRequestConfig (gates the /config endpoints and the
     multipart config part). /status is no longer gated and is enabled simply
     by listing it under endpoints. tika-grpc gains
     the same allowPerRequestConfig flag plus allowComponentModifications, which
     gates runtime Save/Delete of fetchers and pipes iterators (TIKA-4764).

   * Endpoints removed: /translate/* (unusable as shipped), /tika/main and
     /tika/form/main (Boilerpipe; use /tika/text), and the /tika/form family.
     The 3.x /tika/config and /tika/form/config forms are replaced by the
     /tika/config* multipart POSTs, which require allowPerRequestConfig
     (TIKA-4809).

   * Endpoints collapsed: /detect/stream is now /detect, and /language/stream
     and /language/string are both /language. Behavior changed with the rename:
     /detect now runs in the fork pool, so it can return 429, 503 or 413, and a
     failure reading the body is a 500 where 3.x returned 200 with
     application/octet-stream as if detection had succeeded; /language caps
     input at the first 100,000 characters and uses the default LanguageDetector
     on the classpath, where 3.x pinned Optimaize (TIKA-4809).

   * Output-format routing on /tika changed. The bare /tika endpoint returns
     Markdown (was XHTML); use /tika/xml for XHTML. /tika/text is body-only
     again, as in 3.x. /tika/json and /tika/config/json default to the server
     default (markdown) rather than hardcoded plain text. The Accept header no
     longer selects the output format -- 3.x routed bare /tika among plain
     text, HTML and XHTML by Accept (nondeterministically for */*); now the
     path names the format. An unrecognized handler name in the path is a 400
     listing the valid types, instead of silently falling back to the default
     (TIKA-4663, TIKA-4809).

   * Per-request configuration headers are removed, and are now silently
     ignored if sent: writeLimit, throwOnWriteLimitReached,
     maxEmbeddedResources/maxEmbeddedCount, X-Tika-Handler and the meta_*
     metadata-injection family. The limits move to parse-context
     (output-limits.writeLimit, output-limits.throwOnWriteLimit,
     embedded-limits.maxCount); X-Tika-Handler becomes an explicit handler path;
     meta_* has no replacement, and with per-request config off by default a
     caller can no longer bound the output of a single request. The
     X-Tika-OCR* and X-Tika-PDF* families were removed earlier in the 4.x line
     (TIKA-4809).

   * Caller errors now map to accurate HTTP status codes instead of always
     returning 200 or 500. A saturated worker pool returns 429, a
     crashed/timed-out/OOM worker returns 503, an unknown or reserved
     fetcher/emitter or bad handler returns 400, and an over-limit body returns
     413; the 429 and 503 responses carry a Retry-After header. Error bodies are
     now JSON ({"status":"TIMEOUT"}, with a message field when one is
     available) where 3.x returned plain text such as "Parse failed: TIMEOUT"
     (TIKA-4809).

   * The raw /tika family's 422 responses carry the extracted content only; the
     exception is no longer appended to the body -- use /rmeta for the
     structured exception (TIKA-4809).

   * /meta now runs through the same pipes-backed parser as the other
     extraction endpoints, so it gains their crash isolation and their error
     handling: a container exception comes back as 200 with
     tk:exception:container-exception instead of 500, and /meta/{field} returns
     422 instead of 500 or 400. A request with no Accept header now returns
     JSON; 3.x returned CSV, still available via Accept: text/csv. /meta also no
     longer returns a language field -- it parses with the ignore handler, so
     there is no text to detect from; configure a language-detection metadata
     filter and use /rmeta or /tika/json instead (TIKA-4809).

   * /async requires an object body {"tuples":[...]} instead of a bare JSON
     array, validates fetcher/emitter ids at POST time (400), rejects a batch
     larger than the queue's total capacity with 400 instead of throttling it,
     and one bad tuple no longer stops the async workers. /pipes returns the
     same JSON body as /tika/rmeta/unpack -- {"status":<RESULT_STATUS>,
     "message":...} -- instead of a /pipes-only {"status":"ok"|"process_crash"}
     shape, returns 400 with the reason for a malformed request body, and
     rejects emit strategies other than EMIT_ALL, whose passed-back data the
     /pipes response cannot carry (TIKA-4809).

   * Many server config keys were removed or renamed (logLevel, idBase,
     digest, returnStackTrace, port ranges, the spawn-child options, ...), and
     an unrecognized key now fails startup with an error naming it; see
     migrating-tika-server-4x.adoc for the key-by-key migration. One change no
     startup error will flag: taskTimeoutMillis is now
     parse-context.timeout-limits.totalTaskTimeoutMillis, and its default grew
     from 5 minutes to 1 hour (TIKA-4809, TIKA-4813).

   * Request bodies are now capped by maxRequestSizeBytes, defaulting to 1 GiB;
     larger requests are rejected with 413, including over-limit chunked
     uploads, which previously surfaced as an empty 500 (TIKA-4809).

   * The 'endpoints' allowlist now also gates SPI-provided resources; a
     discovered resource binds only when its root endpoint is enabled
     (TIKA-4809).

   * Fetcher-based streaming is removed: the InputStreamFactory pattern for
     fetching documents via the fetcherName/fetchKey headers is gone, and all
     documents now go through the pipes infrastructure. The no-op
     -a/--pluginsConfig flag is removed and now fails option parsing, --help
     exits 0, and the tika-server-client module is removed (TIKA-4809).

  --- tika-pipes and tika-grpc ---

   * tika-pipes implementation modules are now pf4j plugins, reorganized by
     resource (tika-pipes-solr) vs task (tika-pipes-fetcher-solr). Core classes
     moved to tika-pipes-core, and the file-system components moved out of it
     into their own tika-pipes-file-system plugin
     (TIKA-4334, TIKA-4519, TIKA-4543).

   * FetchEmitTuple JSON now names the per-tuple parse context "parse-context"
     (was "parseContext") and rejects unknown tuple fields with an error naming
     the field (TIKA-4809).

   * The pipes config keys staleFetcherTimeoutSeconds and
     staleFetcherDelaySeconds have been removed; a config still carrying them
     fails startup (TIKA-4809).

   * TimeoutLimits: progressTimeoutMillis of 0 combined with a positive
     totalTaskTimeoutMillis is now rejected at config load; it would kill
     every task immediately (TIKA-4809).

   * The http-fetcher now verifies TLS certificates and hostnames by default;
     set verifySsl:false to opt out (TIKA-4809).

   * SolrJ moves from 8.11.4 to 10.0.0; the Solr fetcher, emitter and pipes
     iterator no longer support Solr 8 (TIKA-4789).

   * tika-grpc: the generated Java classes moved from package org.apache.tika
     to org.apache.tika.pipes.grpc.proto, so every generated type moves and Java
     gRPC clients must update their imports. This is a source break only: the
     proto package ("tika") and the service name ("Tika") are unchanged, so the
     wire protocol is identical and clients in other languages are unaffected
     (TIKA-4808).

  --- tika-app and tika-eval-app ---

   * tika-app's batch mode is gone. The -bc/batch directory-to-directory command
     line (backed by the removed tika-batch module) has no successor flag; use
     -a/--async, which runs the same work through tika-pipes
     (TIKA-4333, TIKA-4340).

   * tika-core/tika-app: NetworkParser and tika-app's -c/--client=<uri>
     network-client mode were removed -- they dispatched raw sockets to an
     arbitrary user-supplied host with no auth or TLS. Use tika-server
     instead (TIKA-4808).

   * tika-eval-app's command line changed: the FileProfile sub-command is
     removed, the -bc batch-config option is gone, and extract directories are
     now named with -e/--extracts (Profile) and -a/--extractsA + -b/--extractsB
     (Compare); -i/--inputDir, -d/--db, -c/--config, -n/--numWorkers and
     -m/--maxExtractLength replace the 3.x spellings
     (TIKA-4342, TIKA-4450, TIKA-4452, TIKA-4507).

   * tika-app: the inline short forms -eX (output encoding) and -pX (document
     password) were removed from standard mode; use --encoding=X and
     --password=X. Prefix-matching them silently swallowed single-dash long
     names, so every single-dash long name is now rejected with a message
     naming the two-dash form (TIKA-4808).

  --- Parser, detector and output behavior ---

   * The tika-langdetect-tika module is removed (TikaLanguageDetector,
     LanguageIdentifier, LanguageProfile, LanguageProfilerBuilder,
     ProfilingWriter). tika-app, tika-server and tika-eval now bundle the new
     CharSoup detector (tika-langdetect-charsoup) instead of
     tika-langdetect-optimaize, so the language reported by default changes
     (TIKA-4662).

   * PDF: extractIncrementalUpdateInfo now defaults to true (was false), so
     every PDF parse emits pdf:incremental-update-count and related keys
     without configuration. parseIncrementalUpdates remains false
     (TIKA-4354, TIKA-4358).

   * Audio cover art is now extracted as embedded documents from MP3 (ID3v2
     APIC/PIC), MP4 (covr), Vorbis and FLAC. Embedded-document counts and
     /rmeta list lengths for audio files change (TIKA-4801).

   * The DOM-based OOXML extractors are removed (XWPFWordExtractorDecorator,
     XSLFPowerPointExtractorDecorator, POIXMLTextExtractorDecorator,
     XPSTextExtractor) and with them the OfficeParserConfig keys
     useSAXDocxExtractor and useSAXPptxExtractor. The SAX extractors are the
     only implementation (TIKA-4692, TIKA-4708).

   * The MuPDF renderer is removed (org.apache.tika.renderer.pdf.mutool); PDF
     page rendering for OCR now uses the PDFBox renderer or the new
     PopplerRenderer (TIKA-4664).

   * The legacy ExternalParser is removed; external parsers now require explicit
     JSON configuration. CompositeExternalParser and ExternalParsersFactory,
     which loaded tika-external-parsers.xml definitions from the classpath
     automatically, are gone (TIKA-4707).

   * Headers are no longer injected into the body/content of MSG files
     (TIKA-4345). Please open a ticket if you need this behavior across email
     formats.

  --- Removed modules and classes ---

   * Removed modules with no direct replacement: tika-batch (TIKA-4333),
     tika-dl (TIKA-4499), the advanced media module (TIKA-4500), tika-fuzzing
     (TIKA-4506), tika-age-recogniser (TIKA-4343), tika-server-eval
     (TIKA-4555), the dotnet bindings (TIKA-4332) and snaps deployment
     (TIKA-4502).

   * Removed parsers and classes with no direct replacement:
     SentimentAnalysisParser (TIKA-4574); ObjectRecognitionParser and the
     Tensorflow recognisers/captioners; PooledTimeSeriesParser;
     org.apache.tika.parser.pdf.AccessChecker (replaced by
     PDFParserConfig.AccessCheckMode); and the jempbox-based JempboxExtractor /
     XMPMetadataExtractor / pdf.xmpschemas.* classes superseded by the unified
     XMP extractor (TIKA-4775).

   * Smaller removals: ExceptionUtils.trimMessage moved from tika-core to
     tika-eval-core; HttpClientFactory's inert redirect-host allowlist
     accessors (get/setAllowedHostsForRedirect) are gone (TIKA-4809); and
     org.apache.tika.utils.{RereadableInputStream,AnnotationUtils},
     org.apache.tika.io.{IOUtils,InputStreamFactory},
     org.apache.tika.sax.DIFContentHandler,
     org.apache.tika.parser.{AutoDetectParserFactory,ParserFactory} and
     org.apache.tika.parser.internal.Activator are removed.

  NEW FEATURES

   * tika-pipes gains three parse modes: NO_PARSE (detect only, no parse),
     CONTENT_ONLY (emitters write raw content, no metadata envelope) and
     UNPACK (write embedded bytes out). All three are wired through tika-app
     and tika-server (TIKA-4631, TIKA-4637, TIKA-4656).

   * New tika-pipes plugins: an Elasticsearch emitter (TIKA-4672), an
     Atlassian JWT fetcher (TIKA-4604), Google Drive, Microsoft Graph, Azure
     Blob, JSON and HTTP plugins, and an Apache Ignite ConfigStore for
     runtime fetcher/emitter configuration (TIKA-4583, TIKA-4587, TIKA-4598).

   * New inference and OCR modules: tika-inference and tika-vlm add
     vision-language-model parsers (Claude, Gemini, OpenAI) that emit
     vlm:prompt-tokens / vlm:completion-tokens, and
     tika-parser-tess4j-module adds in-process Tesseract OCR
     (TIKA-4665, TIKA-4666, TIKA-4667, TIKA-4690).

   * Unified XMP extraction across containers; adds HEIF/HEIC and WebP XMP,
     including Samsung/Google Motion Photo (TIKA-4775).

   * tika-app and tika-server can load extra jars (additional EncodingDetectors,
     Parsers, etc.) from the directory named by the -Dtika.extras.dir system
     property, without repackaging the application. Off by default; the
     directory is a trusted code location whose contents run with full process
     privileges. The jars are forwarded onto forked pipes/server workers too,
     so they are available where parsing happens (TIKA-4755).

   * A Markdown parser with structured, lossless XHTML output, complementing
     the Markdown content handler (TIKA-4770).

   * Content-based detection of ASN.1/DER crypto containers, enabled by
     default: the PKCS#7/CMS, PKCS#12 and RFC 5544 timestamped-data families
     gained magic where 3.x had globs only, and Pkcs7Parser refines the
     smime-type on the output Content-Type at parse time. Files that detected
     as application/octet-stream in 3.x may now detect as a crypto type; see
     configuration/detectors.adoc for the opt-in detect-time refinement
     (TIKA-1997, TIKA-2856).

   * New detection: Android binary XML (application/vnd.android.axml,
     TIKA-4747) and Frictionless Data packages (TIKA-4643); improved mp3/aac
     (TIKA-4612) and grib (TIKA-4655) detection.

  OTHER CHANGES

   * Dependency upgrades since 4.0.0-beta-1, including Jetty 12.1.12, CXF
     4.2.3 and SolrJ 10.0.0 (TIKA-4327).

   * Release artifacts are now channel-specific. Maven Central gets slim
     per-module jars (plus pom, sources and javadoc); the Apache dist area
     gets runnable zip distributions (tika-app, tika-server-standard,
     tika-eval-app) and drop-in pf4j plugin zips; Docker Hub gets ready-to-run
     images. Fat/shaded artifacts no longer go to Maven Central (TIKA-4733).

   * The charset, junk-text and language detection stack was rewritten:
     language-aware charset detection, a universal junk detector, wider
     Unicode handling, the new CharSoup language detector, and more efficient
     common-token lookups via bloom filters (TIKA-4662, TIKA-4671, TIKA-4675,
     TIKA-4691, TIKA-4719, TIKA-4731, TIKA-4745, TIKA-4754, TIKA-4810).

   * Pipes now carries small documents to the forked worker inside the request
     instead of writing them to disk first. Content at or below the new
     pipes.maxInlineBytes (default 10 MB) rides in the request and is served in
     the worker by the reserved __bytes fetcher, touching no disk; larger
     content is written out once as before, and a stream already backed by a
     file keeps its file. Set maxInlineBytes to 0 to spool every non-empty body
     (TIKA-4808).

   * PipesClient/PipesServer IPC now enforces a configurable payload limit
     (pipes.maxIpcPayloadBytes, default 100 MB) in both directions. Results
     that exceed the limit return PAYLOAD_LIMIT_EXCEEDED instead of causing
     heap exhaustion; crash messages are also size-capped (TIKA-4793).

   * tika-server requests now carry only their own parse-context entries to the
     forked worker, which supplies the config defaults itself. Previously the
     server sent its config's parse-context with every request, so the worker
     treated the operator's timeout-limits as caller input and clamped them at
     pipes.maxTotalTaskTimeoutMillis (TIKA-4808).

   * tika-grpc now routes fetchAndParse through the PipesParser client pool
     instead of one shared single-threaded PipesClient: concurrent calls no
     longer crash the worker, pipes.numClients and (for the first time)
     pipes.useSharedServer take effect, and pool saturation surfaces in-band as
     CLIENT_UNAVAILABLE_WITHIN_MS. An interrupted call recycles its worker, so a
     pooled client cannot go back to the queue dirty (TIKA-4815).
     fetchAndParseServerSideStreaming now completes the call after delivering
     its reply, instead of leaving the client waiting forever (TIKA-4804).

   * New audio/video metadata: audio:bitrate, audio:is-variable-bitrate,
     audio:has-drm, audio:channels, video:frame-rate, video:bitrate and MP4
     sample size; ID3 TCOP and Vorbis COPYRIGHT map to xmpDM:copyright, EXIF
     GPS altitude maps to geo:alt, and the presentation start of delayed
     QuickTime timed-metadata tracks is exposed (TIKA-4777, TIKA-4779,
     TIKA-4780, TIKA-4781, TIKA-4800, TIKA-4802).

   * Parsing and robustness fixes across formats: CHM (TIKA-4783), ID3 UTF-16
     (TIKA-4784), MPEG2/2.5 Layer III frame sizing (TIKA-4791), .doc empty
     comments (TIKA-4718), OOXML hyphenation and field-code hyperlinks
     (TIKA-4646, TIKA-4683), RTF attachments in HTML decapsulation
     (TIKA-4710), image extraction (TIKA-4736), embedded-file extension
     calculation (TIKA-4808), and general media-file robustness (TIKA-4812).
     MAPI properties no longer overwrite better-fitting Dublin Core terms
     (TIKA-4806). Embedded-file naming was streamlined (TIKA-4689).

   * PDFs whose %PDF- header is preceded by a print-composition job ticket are
     no longer detected as text/x-matlab: up to 50 %% comment or blank lines may
     now precede it, extending the TIKA-3328 rule past its 512-byte reach
     (TIKA-4782).

   * MagicDetector now compiles its regular expression once, in the
     constructor, instead of recompiling it on every match (TIKA-4796).

   * tika-eval-core is no longer published as a fat jar (TIKA-4414) and
     tika-grpc no longer shades gRPC (TIKA-4709).

   * Fix concurrency bug in TikaToXMP (TIKA-4393).


Release 4.0.0-beta-1 - 6/29/2026

  Prerelease. Its changes are folded into the 4.0.0 section above.


Release 4.0.0-alpha-1 - 5/4/2026

  Prerelease. Its changes are folded into the 4.0.0 section above.


Release 3.3.0 - 3/18/2026

  * Switch to poi-ooxml-full (TIKA-4563).

  * Users need to add "allowAbsolutePaths=true" for the FileSystemFetcher to fetch
    an absolute path (TIKA-4649).

  * Add a markdown option for content handlers (TIKA-4563).

  * Improve zip parsing (TIKA-4650).

  * Add detection of compressed bmp (TIKA-4511).

  * Allow per file timeouts in tika-pipes (TIKA-4497).

  * Add matroska detector (TIKA-1180).

  * Allow multiple values for many Dublin Core keys (TIKA-4466).

  * Extract macros by default in tika-app's commandline and gui (TIKA-4472).

  * Improve extraction of Javascript from PDFs (TIKA-4465).

Release 3.2.3 - 9/11/2025

  * Allow backwards compatibility with versions of commons-compress before 1.28.0 (TIKA-4469).

  * Fix XFA parsing within PDFs when woodstox is on the classpath as in tika-server (TIKA-4482).

  * Dependency updates.


Release 3.2.2 - 8/6/2025

  * Fix for CVE-2025-54988.

  * Improve detection of encrypted ODT files (TIKA-4459).

  * Dependency updates (TIKA-4455).

Release 3.2.1 - 6/26/2025

  * Fix POIFSContainerDetector regression when wrapping an InputStream in
    a TikaInputStream (TIKA-4441).

  * Important bug fix for zip-based detection on a non-TikaInputStream (TIKA-4424).

  * Improve text extraction from EMF (TIKA-4432).

  * Dependency updates (TIKA-4421).

Release 3.2.0 - 05/21/2025

  * Detect inline images in MSG files (TIKA-4391).

  * Improve extraction of metadata in MSG files (TIKA-4381).

  * Fix concurrency bug in TikaToXMP (TIKA-4393).

  * Fix potential GDAL deadlock (TIKA-4385).

  * Improve extraction of properties from msg files (TIKA-4381).

  * Include internal attachment path in tika-eval reports (TIKA-4374).

  * Upgrade jsoup to 1.20.1 with workaround for change in self-closing tag behavior (TIKA-4419).

  * Upgrade dependencies (TIKA-4379).

Release 3.1.0 - 01/28/25

   * Allow users to turn off the injection of some headers into the content stream of MSG
     files (TIKA-4345).

   * Add a wrapper for Google's magika detector (TIKA-4344).

   * Add support for MachO via Alexey Pelykh (TIKA-4309).

   * Add logic to inject spaces in XPS files based on font widths via Ruairidh Williamson (TIKA-4315).

   * Mime type "application/json" is now a sub class of "text/javascript" not "application/javascript" (TIKA-4336)

   * Remove tagsoup from the project entirely. Note that
     some of the tags produced by the SourceCodeParser are slightly different (TIKA-4338)

Release 3.0.0 - 10/15/2024

   * Fix regression in TextAndCSVParser (TIKA-4278).

Release 3.0.0-BETA2 - 07/09/2024

   BREAKING CHANGES

   * Updated PST parser to use standard Message metadata keys and improved
     handling of embedded files (TIKA-4248).

   * Convenience methods for XML readers were moved from ParseContext to
     XMLReaderUtils (TIKA-4259).

   Other Changes

   * Add GRPC server (TIKA-4181).

   * Improved configurability in tika-pipes (TIKA-4243).

   * Add optional PST parser based on libpst/readpst (TIKA-4250).

Release 3.0.0-BETA - 12/01/2023

   BREAKING CHANGES

   * Require Java 11 (TIKA-4128).

   * The boilerpipe handler has been moved to the tika-handler-boiler-pipe
     package (TIKA-4138).

   * We've migrated HTML parsing to the JSoup parser instead of TagSoup. If
     you have a custom configuration on the HTMLParser, you'll need to change
     that to o.a.t.p.html.JSoupParser (TIKA-1599).

   * Removed xerces2 as a dependency (TIKA-4135).

   * tika-core now has a scope of "provided" in most non-app modules (TIKA-4191).

   * Tika will look for "custom-mimetypes.xml" directly on the classpath, NOT
     under "/org/apache/tika/mime/". (TIKA-4147).

   * Return media type "text/javascript" instead of "application/javascript"
     to follow RFC-9239. (TIKA-4119).

   Other Changes/Updates

   * Improve detection of sqlite3-based file formats (TIKA-4187).

   * Upgrade PDFBox to 3.0.1 (TIKA-3347)

   * Deprecated AbstractParser for removal in 4.x (TIKA-4132).

   * Fix bug in DateUtils that stripped timezone information from
     incoming Calendar objects (TIKA-4126).

   * The InputStreamDigester now calculates stream length (TIKA-4016).

Release 2.9.0 - 8/23/2023

   * With user configuration, the PDFParser can now throw an EncryptedDocumentException
     for Microsoft IRM PDF containers with encrypted payloads. Separately,
     the PDFParser now throws an EncryptedDocumentException instead of an IOException
     if the security handler cannot be found (TIKA-4082).

   * Fix bug that led to duplicate extraction of macros from some OLE2 containers (TIKA-4116).

   * Parse iframe's srcdoc as an embedded file (TIKA-3109).

   * Add detection of warc.gz as a specialization of gz and parse as if a standard WARC (TIKA-4048).

   * Allow users to modify the attachment limit size in the /unpack resource (TIKA-4039)

   * Fixed write limit bug in RecursiveParserWrapper (TIKA-4055).

   * Add mime detection for many files with thanks to Gregory Lepore (TIKA-3992).

   * Fixed iWork 13 keynote detection on files with wrong extension (TIKA-4111).

Release 2.8.0 - 5/11/2023

   * Enable counting and/or parsing of incremental updates in PDFs.  This
     is an experimental feature and may change in later releases (TIKA-4017).

   * Fixed bug that prevented the the loading of CompositeExternalParser in tika-app and
     tika-server-standard. This parser will call exiftool and ffmpeg if those are installed, as was
     the behavior in Tika 1.x. Exclude org.apache.tika.parser.external.CompositeExternalParser
     if you do not want this behavior (TIKA-4022).

   * Removed the shading of tika-parsers-standard-module (TIKA-4038).

   * Enable optional extraction of file system metadata in FileSystemFetcher (TIKA-4035).

   * Allow pretty printing in FileSystemEmitter (TIKA-4034).

   * Add detection for and a new mime type for older postscript-based
     Adobe Illustrator "application/illustrator+ps" files (TIKA-3971).

   * Add magic detection for canon raw file types: crw, cr2 and cr3 (TIKA-3991).

   * Add detection for ONIX message files (TIKA-4011).

   * Add detection and a parser for ActiveMime files (TIKA-3987).

   * Add extraction of rendition layout value and version from Epub (TIKA-4013).

   * Improve embedded file extraction from PDFs (TIKA-4012).

   * Improve metadata extraction from WARCs (TIKA-4018).

   * Update to PDFBox 2.0.28 (TIKA-4016).

   * Users may now avoid the ZeroByteFileException via a
     setting on the AutoDetectParserConfig (TIKA-3976).

   * Fix bug in closing <a> elements in the presence of <b> elements
     in RTF files (TIKA-3972).

   * Improve extraction of embedded file names in .docx (TIKA-3968).

   * Normalize author, title, subject and description to their Dublin Core
     properties in the HTMLParser (TIKA-3963).


Release 2.7.0 - 1/31/2023

   * Add SVG detection for svg files that lack the xml header (TIKA-3308).

   * Migrate to a live fork of Universal Charset Detector (TIKA-3213).

   * Improve handling of text-based attachments inside .eml files (TIKA-3959).

   * Add tika-parser-nlp-package to release artifacts (TIKA-3958).

   * Remove need for <params/> element in classes that extend ConfigBase (TIKA-3946).

   * Add X-TIKA:embedded_id_path to ensure unique embedded file paths (TIKA-3942).

   * Fix bug that prevented digests when the fallback/EmptyParser
     was called (TIKA-3939).

   * Remove log4j 1.2.x (and slf4j-log4j12 which now redirects to slf4j-reload4j) from
     all modules (TIKA-3935).

   * Upgrade mime4j to 0.8.9 (TIKA-3950).

   * Refactor date parsing for emails (TIKA-3957)

   * Upgrade to Bouncy Castle 1.71 and jdk18on jars (TIKA-3933).

   * Add a JDBCPipesReporter (TIKA-3931).

   * Add multivalued field strategy option in jdbc-emitter (TIKA-3930).
     Default is now 'concatenate' with ', ' as the delimiter.

   * Downgrade logging in PipesClient for each parse from info to debug.

Release 2.6.0 - 11/3/2022

   * Add optional Siegfried detector (TIKA-3901).

   * Move OverrideDetector's functionality to the CompositeDetector (TIKA-3904).

   * The FileCommandDetector has been refactored to have the same
     behavior as the Siegfried detector; see setUseMime in the javadoc (TIKA-3902).

   * Fix bug in OpenSearch emitter that prevented upserts on
     documents with embedded files (TIKA-3882).

   * Extract PDF actions and triggers into the file's metadata (TIKA-3887).

   * Add a tika-async-cli module (TIKA-3885).

   * Fetch keys sent via headers to tika server are now URL decoded (TIKA-3864).


Release 2.5.0 - 09/30/2022

   * Improved extraction of PDF subset info for PDF/UA, PDF/VT, and PDF/X.
     NOTE: we no longer append PDF/A information, e.g. 'version="A-1b"'
     to the 'dc:format'. Users must now get that information from the
     'pdfa:PDFVersion' key or from 'pdfaid:conformance'
     and 'pdfaid:part' (TIKA-3844).

   * Avoid infinite loop in bookmark extraction from PDFs (TIKA-3832).

   * Upgraded to slf4j 2.0.1 (TIKA-3842).

   * Added upsert option for the OpenSearch emitter (TIKA-3855).

   * Extract PDF signature information at the document level
     into the metadata (TIKA-3852).

   * Enable configuration of digests via AutoDetectParserConfig (TIKA-3853).

   * Use commons-io byte array streams via PJ Fanning (TIKA-3843).

   * Upgrade to PDFBox 2.0.27 (TIKA-3866).

   * Upgrade to JempBox 1.8.17 (TIKA-3856).

   * Add extraction of ODF version from ODF files (TIKA-3840).

   * tika-parser-html-commons (BoilerPipeHandler) is no longer a
     a dependency of tika-parser-html-module. tika-app and tika-server-standard
     have added a dependency on tika-parser-html-commons.  However,
     users who are managing custom dependencies and who want the BoilerPipeHandler
     will have to now include the tika-parser-html-commons dependency
     (TIKA-1484).

   * Add unrar as an optional parser (TIKA-3800).

   * Refactor FuzzingCLI to use PipesParser (TIKA-3799).

   * ServiceLoader's loadServiceProviders() now guarantees
     unique classes (TIKA-3797).

   * Fix bug that prevented setting of includeHeadersAndFooters
     for xls, xlsx, doc and docx via tika-config (TIKA-3796).

   * Fix bug that prevented specification of rendered image type
     via http header in the PDFParser (TIKA-3794).

   * Fix bug causing some Exif dates to be decoded wrongly on
     timezones different than UTC (TIKA-3815).

   * Numerous dependency upgrades (TIKA-3795).

   * Add ALPHA-level initial releases of JDBCEmitter,
     FileSystemStatusReporter and OpenSearchPipesReporter.
     These may have breaking changes in subsequent releases.

Release 2.4.1 - 06/14/2022

   * Implement bulk upload in the OpenSearch emitter (TIKA-3791).

   * Implement tika-server client via pipes mode (TIKA-3790).

   * Custom embedded parsers and EmbeddedDocumentHandlers
     can now add metadata to the container file's
     metadata (TIKA-3789).

   * Record embedded file exceptions in the container
     file's metadata (TIKA-3788).

   * Allow continuation of parsing after write limit has
     been reached (TIKA-3787).

   * Allow pass-through of 'Content-Length' header to metadata
     in TikaResource (TIKA-3786).

   * Add embedded depth to profiles tables in tika-eval (TIKA-3775).

   * Add stop() method to TikaServerCli so that it can be run
     with Apache Commons Daemon (TIKA-1570).

   * Fixed bug in ordering of Parsers during service loading (TIKA-3750).

   * Users can expand system properties from the forking
     process into forked tika-server processes (TIKA-3748).

   * Fix a few files being wrongly detected as EML (TIKA-3771).

   * Fix ignoreCharsets param of Icu4jEncodingDetector (TIKA-3774).

Release 2.4.0 - 04/23/2022

   * NOTE: To save on resources, we no longer include the
     deeplearning4j dependencies in the tika-dl jar. The dependencies for the
     tika-dl package must be provided by users.  See:
     https://github.com/apache/tika/blob/main/tika-parsers/tika-parsers-ml/tika-dl/pom.xml
     for the dependencies that must be provided at run-time (TIKA-3676).

   * NOTE: Added prefix "dwg-custom:" to DWG custom metadata properties (TIKA-3731).

   * Add initial, BETA-grade TLS encryption option for tika-server;
     configuration may change in future releases (TIKA-3719).

   * Allow specification of fetcherName and fetchKey via query parameters
     in request URI in tika-server (TIKA-3714).

   * Add basic parsers for WARC and WACZ in tika-parsers-standard (TIKA-3697).

   * Add MetadataWriteFilter capability to improve memory profile in
     Metadata objects (TIKA-3695).

   * Allow configurability of the ContentHandlerDecorator used
     by the AutoDetectParser (TIKA-3723).

   * Allow configurability of the EmbeddedDocumentExtractor used
     by the AutoDetectParser (TIKA-3711).

   * Add detection for Frictionless Data packages and WACZ (TIKA-3696).

   * Add detection for DGN files with gratitude and credit
     to Steven Frew's tika-dgn-detector (TIKA-3721).

   * Add parser for metadata from DGN 8 files via Dan Coldrick (TIKA-3721).

   * Add a fetcher and emitter for Azure blob storage (TIKA-3707).

   * Add detection for files encrypted by Microsoft's Rights Management Service
     (TIKA-3666).

   * Fixed regression in 2.3.0 that led to more embedded filenames
     than appropriate being written to the content (TIKA-3711).

   * tika-server now clones forking process' environment variables
     into forked process (TIKA-3715).

   * Add an optional /eval endpoint for tika-eval profile or compare
     capabilities in tika-server (TIKA-3689).

   * Add a Parsed-By-Full-Set metadata item to record all parsers that processed
     a file (TIKA-3716).

   * Add metadata filters for Optimaize and OpenNLP language detectors (TIKA-3717).

   * Upgrade to PDFBox 2.0.26 (TIKA-3726).

   * Upgrade deeplearning4j to 1.0.0-M2 (TIKA-3458 and PR#527).

   * Various dependency upgrades, including POI, dl4j, gson, jackson,
     twelvemonkeys, log4j2 and others (TIKA-3675 and many PRs from dependabot).

   * Switch cipher from ECB to GCM in HttpClientFactory (TIKA-3724).

Release 2.3.0 - 02/02/2022

   * Upgrade to Apache POI 5.2.0. This is the first upgrade to POI
     5.x and represents a major refactoring. Users may experience
     significantly more logging (TIKA-3164).

   * Upgrade to log4j2 2.17.1 (TIKA-3638).

   * Improve consistency in reporting package-entry divs across
     all parsers for embedded files (TIKA-3644). This leads
     to some more text (embedded file names) in files with
     many embedded attachments.

   * Improve configuration of maps as params for parsers in
     TikaConfig (TIKA-3645).

   * Improve identification of iWorks 13 files and add parsing
     for thumbnails, some metadata and attachments (TIKA-3634).
     Skip handling of .iwa files, which are not yet supported.

   * Limit the default in-memory processing (maxMainMemoryBytes) in
     the PDFParser to 512MB as in the 1.x branch (TIKA-3642).

   * Added IDML Parser from 1.x series to 2.x series (TIKA-3188).

   * Extract annotation types and subtypes for PDFs into metadata (TIKA-3653).

   * Add metadata value for PDFs that contain 3D annotations (TIKA-3653).

   * Add parser for Translation Memory eXchange (TMX) files (TIKA-3660).

   * Add Bill of Materials (Maven BOM) for centralized module version management (TIKA-3367).


Release 2.2.1 - 12/19/2021

   * Fix multithreading bug for ooxml files (TIKA-3627).

   * Upgrade log4j to 2.17.0 (TIKA-3625).

   * Upgrade to PDFBox 2.0.25 (TIKA-3622)

   * Fix bug that prevented metadata keys in the UnpackerResource
     in tika-server (TIKA-3624).

   * Upgrade log4j to 2.16.0 (TIKA-3623)

Release 2.2.0 - 12/13/2021

   * Add support for OneNote files downloaded from O365 (TIKA-3446).

   * Fix logic bug in PipesServer that prevented concatenation of
     content from attachments (TIKA-3609).

   * Improve extraction of embedded files from MSOffice files created
     by non-Microsoft tools (TIKA-3526).

   * Added back ability to ignore load errors in TikaConfig (TIKA-3575).

   * Make SecureContentHandler and other parameters configurable in
     AutoDetectParser programmatically and via tika-config.xml (TIKA-3594).

   * Fix default logging in tika-app in batch mode (TIKA-3589).

   * Fix bug that prevented specifying a config with the long
     --config= option in tika-app in batch mode (TIKA-3589).

   * Fix thread starvation after numerous restarts in
     PipesClient (TIKA-3588).

   * Fix race condition when starting multiple forked
     servers on multiple ports (TIKA-3586).

   * Add timeout per task to be configured via headers
     for tika-server's legacy endpoints /tika and /rmeta.
     Note that this timeout greater than taskTimeoutMillis (TIKA-3582).

   * Add metadata item for whether or not a PDF has a collection/
     is a Portfolio PDF (TIKA-3579).

   * Add detection of ESRI Layer files (TIKA-3570).

   * Add detection of JPEG XL, MARC, ICC profiles, NES-ROM file types
     (TIKA-3562 and TIKA-3563)

   * Remove duplicate "subject" metadata keys that were intended
     for backwards compatibility within 1.x only (TIKA-3564).

   * Fix Open Office mime types to be subclasses of application/zip
     and no longer require OPCPackageDetector-last ordering of zip
     detectors (TIKA-3556).

   * Improve robustness and features of the httpfetcher (TIKA-3543)

   * Add optional fetch ranges to FetchEmitTuple to allow range fetching from,
        e.g. http or s3 (TIKA-3542).

   * Exclude dependencies on jsoup and ehcache in ucar grib/cdm (TIKA-3003).


Release 2.1.0 - 08/18/2021

   MAJOR CHANGES in 2.1.0:

   * Improved packaging for tika-parsers-extended. Use the tika-parser-scientific-package and
     tika-parser-sqlite3-package artifacts if you want fat jars with dependencies. (TIKA-3510)

   * Tika app writes UTF-8 when an encoding is not specified; the legacy behavior
     was UTF-8 on Mac OS, but System default on other OSs (TIKA-3515).

   * Change the default rendering strategy for PDFs from NO_TEXT to ALL (TIKA-3520).

   Other changes:

   * Fixed bug that pointed to the wrong tessdata directory if the user specified
     a tesseract path but not also a tessdata path (TIKA-3518).

   * Fixed bug in Icu4j's encoding detector where it would return non-standard
     names for charsets, e.g. IBM424_rtl is now returned as IBM424 (TIKA-3516).

   * Add a simple UrlFetcher in tika-core as a basic alternative
     to tika-fetcher-http (TIKA-3527).

   * Add tika-pipes support for Google Cloud Storage (TIKA-3524).

   * Fix markup ordering errors in xhtml output for ODT files (TIKA-2242).

   * Fix serialization of embedded docs in OpenSearch emitter
     and fix embedded documents not being indexed in some use
     cases in the Solr emitter (TIKA-3490).

   * Add pipesClientId system property to PipesServer so that each
     forked process can log to its own logger (TIKA-3480).

   * Add DateNormalizingMetadataFilter let users ensure that all dates
     emitted to Solr/OpenSearch are in UTC. Users can configure which
     timezone they'd like to use in cases where the file format does
     not store a timezone (TIKA-3496).

   * Breaking change in the Solr and OpenSearch emitters. To achieve
     the SKIP or CONCATENATE attachment strategy, modify the
     parseMode in the pipesiterators or in the FetchEmitTuple (TIKA-3494).

Release 2.0.0 - 07/07/2021

   * Cleanup of fetcher integration with tika-server.

   * Update dependencies.

Release 2.0.0-BETA - 05/19/2021

   * Refactor pipes module for resilience

   * Add transcribe capability (TIKA-94).

Release 2.0.0-ALPHA - 01/13/2021

   BREAKING CHANGES in 2.0.0
   * General
     * OCR is now triggered automatically for PDFs if tesseract
       is on the user's path see (https://cwiki.apache.org/confluence/display/TIKA/TikaOCR#TikaOCR-disable-ocr)
       for how to disable OCR.
     * We upgraded from log4j to log4j2 in tika-app, tika-server and anywhere else
       we used to use log4j.
     * By default, when rendering a page for OCR, the PDFParser does not render glyphs/text.
     * Removed deprecated Metadata keys/properties (TIKA-1974).
     * Removed deprecated PDFPreflightParser (TIKA-3437).
     * Removed dangerous calls to read an inputstream or convert to bytes
       without specifying a charset
     * Parsers can be configured via tika-config.xml on instantiation.
       We have moved away from configuration via .properties files because
       of confusion among users.  This affects the PDFParser, TesseractOCRParser
       and the StringsParser.
     * Changed namespaces of translator implementations (o.a.t.language.translate.impl) to avoid
       split-package with tika-core

   * tika-parsers
     * The parser modules have been broken into three main modules:
        tika-parsers-standard, tika-parsers-extended and tika-parsers-ml.
        Users may now need to add tika-parsers-extended's
        tika-parser-scientific-module or tika-parser-sqlite3-module to tika-app and
        tika-server to include parsers that used to be included by default
        (for example: envi, gdal, grib, isatab, netcdf, sqlite3).
     * PDFParser -- a) see above on OCR. b) This parser no longer warns if the jpeg2000
       dependency is not included. Tika now relies on PDFBox to log an error if a jpeg2000
       image should be processed but can't because the required external dependency is
       not available.  See https://pdfbox.apache.org/2.0/dependencies.html#jai-image-io
       for the non-ASF-2.0-compatible jpeg2000 library.
     * CompressorParser -- users must add the com.github.luben:zstd-jni dependency to
       the classpath to process zstd files.  This is an optional library that is no longer bundled
       in tika-parsers-standard-package because it contains native libs.
     * ChmParser was moved to org.apache.tika.parser.microsoft.chm
     * RTFParser was moved to org.apache.tika.parser.microsoft.rtf
     * We are now using non-shaded versions of xmpcore with namespaces com.adobe.internal.*
       vs com.adobe.*.

   * tika-app
     * See above on default inclusion of only tika-parsers-standard.

   * tika-server
     * tika-server now by default forks a process to isolate the parsing
       in the forked process (this was called the -spawnChild option
       in tika-1.x).  Clients must now expect that tika-server
       will restart on OOM, timeouts, crashes or after parsing a
       large number of files.  When this happens tika-server will restand and not
       receive connections for brief periods.  The less robust, legacy behavior
       of not forking a process is available with "-noFork"=
     * Most of tika-server's legacy configuration via the commandline has been moved
       into configuration via a tika-config.xml file.
     * tika-server's "enableFileUrl" has been removed in favor of a FileSystemFetcher.
     * tika-server's /metadata endpoint requires tika-server-standard to write XMP/rdf output.
       This output is not available in tika-server-core.
     * In tika-server, for those parsers that can be configured per parse via a config object
       passed in through the ParseContext, the config object will only update those fields
       that the user has modified.  The config object will no longer
       fully reset all settings to the default settings per parse.
       This has a more intuitive "update the base/configured settings" with
       what has been changed in the config object.

  * tika-eval
    * tika-eval's default profile and comparison reports no longer include tag reports.
      Users can get the report configs that include tags (*-tags.xml):
      https://github.com/apache/tika/tree/main/tika-eval/tika-eval-app/src/main/resources

Release 1.27 - 06/30/2021

   * Migrate MP4 parsing to Drew Noakes' metadata-extractor (TIKA-3459).
     To revert to legacy parser turn off NoakesMP4Parser and turn on MP4Parser
     via tika-config.xml.

   * Prevent rare infinite loop in tika-server's -spawnChild mode
     when restart fails because of failure to bind to the port (TIKA-3441).

   * Improve likelihood that tesseract will not be orphaned on
     jvm restart in tika-server (TIKA-3441).

   * Deprecate experimental PDFPreflightParser (TIKA-3437).

   * Apply encoding detection to zip entry names via Ryan421 (TIKA-3374).

   * Add json output for /tika endpoint in tika-server (TIKA-3352).

   * Tika's PDFParser should use the underlying file if one is passed in
     via a TikaInputStream (TIKA-3350)

Release 1.26 - 03/24/2021

   * Fix thread safety bug in OpenOffice parser (TIKA-3334).

   * The "writeLimit" header now pertains to the combined characters
     written per container document (and embedded documents) in the /rmeta
     endpoint in tika-server (TIKA-3325); it no longer functions only
     per container or embedded document.

   * Extract more embedded files in PDFs by recursively processing the
     embedded file tree (TIKA-3332).

   * Allow for case insensitive headers for configuration of the PDFParser
     and the TesseractOCRParser in tika-server via Subhajit Das (TIKA-3320).

   * Improve detection and parsing of XPS files (TIKA-3316).

   * General dependency upgrades (TIKA-3244).

   * Great optimization in ForkParser (TIKA-3237).

   * Fix parsing of emails attached to other emails in PST files (TIKA-3004).

   * MP3 parser should output the xmpDM:duration metadata as seconds not
     milliseconds, consistent with the other Audio and Video parsers (TIKA-3318).

   * MP4 parser check if any of the Compatible Brands match when identifying
     the subtype (TIKA-3310).

Release 1.25 - 11/25/2020

   * Fix inconsistent license in xmpcore (TIKA-3204).

   * General upgrades including some dependencies with
     recently found security vulnerabilities (TIKA-3119).

   * Add detection and a parser for flat ODF files (TIKA-3159).

   * Add extraction of macros from ODF files  (TIKA-3161).

   * Add mime detection for hprof and hprof text files (TIKA-3144).

   * Add TextSignature and TextProfileSignature to tika-eval (TIKA-3145 and TIKA-3146)

   * Create a metadata filter to trigger tika-eval stats post parsing (TIKA-3140)

   * Add a configurable metadata-filter for the RecursiveParserWrapper (TIKA-3137)

   * Parameterize writeLimit and maxEmbeddedResources for RecursiveParserWrapper
     in tika-server (TIKA-3133)

   * Add status endpoint to tika-server (TIKA-3129).

   * Remove whitelist/blacklist terminology (TIKA-3120)

   * Add detection for parquet files (TIKA-3115).

   * Add detection and parsing for bplist (TIKA-3104).

   * Enable metadata value filtering for RecursiveParserWrapper (TIKA-3137)

   * Add a basic parser for plist files based on com.googlecode.plist:dd-plist (TIKA-3104).

   * Read hyperlinked images from ODT files (TIKA-3156).

   * Updated GrobidRESTParser to use new API location (TIKA-3191).

   * Add FileProfiler to tika-eval (TIKA-3216).

   * Add status endpoint to tika-server (TIKA-3129).

   * Improved handling of zip files with STORED entries with
     data descriptor (TIKA-3196).

   * Add parsers for XLZ, IDML and MIF (TIKA-2976, TIKA-3188 and TIKA-3189).

   * Add the beginnings of a format-aware fuzzing module (TIKA-3083).

   * Add wrapper for Linux 'file' command for mime detection (TIKA-3215).

   * Added ability to skip parsing of embedded files in Tika Server (TIKA-3227).

Release 1.24.1 - 4/17/2020

   * Allow gzip compression of input and output streams for tika-server (TIKA-3073).

Release 1.24 - 3/11/2019

   * Add scripts to run tika-server as a service via Eric Pugh,
    and add these scripts and jar as a new artifact in the release (TIKA-3010).

   * Upgrade Drew Noakes' metadata-extractor (TIKA-2952).

   * Enable optional extraction of structural tags in PDFs (alpha-grade) (TIKA-3026).

   * Tika app's --extract mode now outputs to STDOUT (TIKA-3035).

   * Add an optional Preflight parser for PDFs (TIKA-3055).

   * Improve detection of some zip-based formats (TIKA-3057).

   * Upgrade metadata-extractor to 2.13.0 (TIKA-2952).

   * Upgrade to POI 4.1.2 (TIKA-3047).

   * Extract XMP from PSD files (TIKA-3050).

   * Added XMLProfiler as an optional parser to profile XFA and XMP
     in PDFs (TIKA-3045).

   * Extract inline images that rely on the DCT filter from PDFs (TIKA-3041).

   * Upgrade to PDFBox 2.0.19 (TIKA-3033).

   * Fix bug in ASM parser configuration (TIKA-2992).

   * Upgrade to java-libpst 0.9.3 (TIKA-2546).

   * Fixed XLIFF12Parser failures with ToXMLHandler (TIKA-3014).

Release 1.23 - 12/02/2019

   * NOTE: The PDFParser now relies on OCRDPI to render page images when
     users configure OCR on rendered page images. This will have the effect
     of increasing rendered image size (TIKA-2624).

   * NOTE: tika-server no longer returns 415 for file types for which there
     is no parser.

   * Fix bug in AUTO OCR strategy in the PDFParser (TIKA-3002).

   * Fix incorrect height and width metadata extraction from JPEG images (TIKA-2630).

   * Upgrade to POI 4.1.1 (TIKA-2851).

   * Upgrade to PDFBox 2.0.17 (TIKA-2951).

   * Ensure that the PDFParser respects custom configuration of Tesseract
     from tika-config.xml via Eric Pugh (TIKA-2970).

   * Add parser for XLIFF v1.2 files (TIKA-2975).

   * Add mime type detection support for WebAssembly (TIKA-2894),
     HEIF / HEIC images (TIKA-2942), Digilite FDF (TIKA-2988);
     and xml-root detection for XFDF (TIKA-2990) and XDP (TIKA-2989).

   * Add an XLZ Parser (TIKA-2976).

   * Fix deadlock with ForkParser when InputStream throws IOException (TIKA-2892).

Release 1.22 - 07/29/2019

   * NOTE: tika-server no longer hard-codes the HtmlParser to handle
     XML files (TIKA-2910).  Users must now configure that behavior
     via a tika-config.xml file.

   * NOTE: Known regression: PDFBOX-4587 -- PDF passwords with codepoints
     between 0xF000 and 0XF0000 will cause an exception.

   * Add parser for HWP v5 files via SooMyung Lee (soomyung) and
     JinSup Kim (ddoleye) (TIKA-2909).

   * Fix order of closing streams to avoid "Failed to close temporary resource"
     exception in TesseractOCRParser (TIKA-2908).

   * Improve AutoDetectReader performance by caching encoding
     detector (TIKA-1568).

   * Prevent RTFParser from outputting illegal tag combinations (TIKA-2889).

   * Fix RereadableInputStream to release all resources (TIKA-2903).

   * Implement custom language identifier in the tika-eval module based on
     OpenNLP's language detector; add 18 languages and add common words
     lists for all 121 languages (TIKA-2790).

   * Fix NPE in MimeTypesReader.releaseParser() via Eamonn Saunders (TIKA-2896).

   * Fix RTFParser to extract more content (TIKA-2883).

   * Add clientSubmitTime to the metadata extracted from PST files (TIKA-2898).

   * Improve StreamingZipContainerDetector for xltx, xltm and
     several other file formats (TIKA-2886).

Release 1.21 - 05/14/2019

   * Add optional AUTO mode to OCR'ing of PDFs.  If tesseract is installed
     and on the path, and this option is selected programmatically
     or via TikaConfig(), the PDFParser will use heuristics to decide
     whether or not to run OCR per page on PDFs. (TIKA-2749)

   * The ZipContainerDetector's default behavior was changed to run
     streaming detection up to its markLimit.  Users can get the
     legacy behavior (spool-to-file/rely-on-underlying-file-in-TikaInputStream)
     by setting markLimit=-1. The POIFSContainerDetector requires an underlying file;
     it will try to spool the file to disk; if the file's length is > markLimit,
     it will not attempt detection; set markLimit to -1 for legacy behavior (TIKA-2849).

   * Upgrade PDFBox to 2.0.14 (TIKA-2834).

   * Add CSV detection and replace TXTParser with TextAndCSVParser;
     users can turn off CSV detection by excluding the TextAndCSVParser
     and adding back the TXTParser via tika-config (TIKA-2833).

   * Add a CSVParser.  CSV detection is currently based solely on filename
     and/or information conveyed via Metadata (TIKA-2826).

   * General upgrades: asm, bouncycastle, commons-codec, commons-lang3, cxf,
     guava, h2, httpcomponents, jackcess, junrar, Lucene, mime4j, opennlp, parso,
     sqlite-jdbc (provided), zstd-jni (provided) (TIKA-2824)

   * Bundle xerces2 with tika-parsers (TIKA-2802).

   * Upgrade jaxb to 2.3.2 (TIKA-2819).

   * Upgrade jackson to 2.9.8 (TIKA-2717).

   * Update tika-eval's common tokens lists (TIKA-2822).

   * Handle bad tags in tika-eval more robustly (TIKA-2810).

   * Add reports for tags in tika-eval (TIKA-2809).

   * Extract text from SDT element within textboxes in .docx files (TIKA-2807).

   * Try to handle truncated OOXML files more robustly (TIKA-2765).

Release 1.20 - 12/17/2018

   * Upgrade to POI 4.0.1 (TIKA-2751).

   * Integrate/parameterize new angles handling in
     PDFBox (TIKA-2779).

   * Upgrade to PDFBox 2.0.13 (TIKA-2788).

   * Prevent content within <style/> and <script/> elements
     to be written in the ToTextContentHandler (TIKA-2550).

   * Switch child to parent communication to a shared memory-mapped
     file in tika-server's -spawnChild mode.

   * Fix bug in tika-server when run in legacy mode (not -spawnChild)
     that caused it to return 503 on documents submitted after
     it hit an OutOfMemoryError (TIKA-2776).

   * Upgrade jaxb-runtime and javax.activation (TIKA-2778).

   * tika-app in batch mode now requires an interrupt or
     kill signal to the parent process to stop the parent
     and the child processes (TIKA-2780).

   * Bulk upgrade of dependencies (TIKA-2775).

   * Improve language id efficiency in tika-eval (TIKA-2777).

   * Upgrade sqlite "provided" dependency to 3.25.2 (TIKA-2773).

   * Remove duplication of notes in PPT slides (TIKA-2735)

   * Use -javaHome or $JAVA_HOME (if they exist) when
     spawning child in tika-server's -spawnChild mode.

   * Fixed closing of styles around Hyperlinks in Word Parser
     Contributed by Ronan O'Sullivan (TIKA-2599).

Release 1.19.1 - 10/4/2018

   * Update PDFBox to 2.0.12, jempbox to 1.8.16
     and jbig2 to 3.0.2 (TIKA-2745).

   * Fix regression in parser for MP3 files (TIKA-2730).

   * Updated Python Dependency Check for TesseractOCR (TIKA-2740).

   * Improve SAXParser robustness (TIKA-2727).

   * Remove dependency on slf4j-log4j12 by upgrading jmatio (TIKA-2742).

   * Replace com.sun.xml.bind:jaxb-impl and jaxb-core with
     org.glassfish.jaxb:jaxb-runtime and jaxb-core (TIKA-2743)

Release 1.19 - 9/14/2018

   * Require Java 8 (TIKA-2679).

   * Enable building with Java 11 (TIKA-2668)

   * Add an option to make tika-server robust against infinite loops,
     OOMs, and memory leaks (TIKA-2725).

   * Allow configuration of the Tesseract parser via the standard
     tika-config.xml options (TIKA-2705).

   * Improve handling of empty cells across table-based
     formats (TIKA-2479).

   * Add a Standards compliant HTML encoding detector
     via Gerard Bouchar (TIKA-2673).

   * Improved XML parsing -- limited default entity expansions to 20.
     To raise this limit, add -Djdk.xml.entityExpansionLimit=XXX to
     your commandline.

   * Mime magic improvements for Olympus RAW (TIKA-2658), interpreted
     server-side languages via HTTP (TIKA-2648), MHTML (TIKA-2723)

   * Add absolute timeout to ForkParser rather than testing
     for active (TIKA-2656).

   * Make the RecursiveParserWrapper work with the ForkParser (TIKA-2655).

   * Allow the ForkParser to specify a directory containing tika-app.jar
     for use by the ForkServer.  This allows users to keep most of the
     parser dependencies out of their code; and it allows for an easy
     addition of optional jars for Parser dependencies,
     such as the xerial sqlite jar (TIKA-2653).

   * Use a pool for SAXParsers and DOMBuilders rather than creating
     a new parser/builder for every parse.
     For better performance, set XMLReaderUtils.setPoolSize() to the
     number of threads you're using with Tika (TIKA-2645).

   * Add the RecursiveParserWrapperHandler to improve the RecursiveParserWrapper
     API slightly (TIKA-2644).

   * Upgraded to Commons-Compress 1.18 (TIKA-2707).

   * Upgraded to Apache POI 4.0.0 (TIKA-2552).

   * Upgraded to Apache PDFBox 2.0.11 (TIKA-2681).

   * Upgraded to deeplearning4j 1.0.0-beta2 (TIKA-2672).

   * Upgraded jmatio to 1.4 (TIKA-2667)

   * Upgraded Apache Lucene to 7.4.0 in tika-eval and tika-examples (TIKA-2695).

   * Upgraded junrar to 1.0.1 (TIKA-2664).

   * Numerous other upgrades (TIKA-2692).

   * Excluded Spring as a transitive dependency (TIKA-2721).

Release 1.18 - 4/20/2018

   * Upgrade jackson to 2.9.5 (TIKA-2634).

   * Add support for brotli (TIKA-2621).

   * Upgrade PDFBox to 2.0.9 and include new jbig2-imageio
     from org.apache.pdfbox (TIKA-2579 and TIKA-2607).

   * Support for TIFF images in PDF files (TIKA-2338)

   * Detection of full encrypted 7z files (TIKA-2568)

   * Various new mimes and typo fixes in tika-mimetypes.xml
     via Andreas Meier (TIKA-2527).

   * Revert to listenForAllRecords=false in ExcelExtractor
     via Grigoriy Alekseev (TIKA-2590)

   * Add workaround to identify TIFFs that might confuse
     commons-compress's tar detection via Daniel Schmidt
     (TIKA-2591)

   * Ignore non-IANA supported charsets in HTML meta-headers
     during charset detection in HTMLEncodingDetector
     via Andreas Meier (TIKA-2592)

   * Add detection and parsing of zstd (if user provides
     com.github.luben:zstd-jni) via Andreas Meier (TIKA-2576)

   * Allow for RFC822 detection for files starting with "dkim-"
     and/or "x-" via Andreas Meier (TIKA-2578 and TIKA-2587)

   * Extract xlsx files embedded in OLE objects within PPT and PPTX
     via Brian McColgan (TIKA-2588).

   * Extract files embedded in HTML and javascript inside HTML
     that are stored in the Data URI scheme (TIKA-2563).

   * Extract text from grouped text boxes in PPT (TIKA-2569).

   * Extract language metadata item from PDF files via Matt Sheppard (TIKA-2559)

   * RFC822 with multipart/mixed, first text element should be treated
     as the main body of the email, not an attachment (TIKA-2547).

   * Swap out com.tdunning:json for com.github.openjson:openjson to avoid
     jar conflicts (TIKA-2556).

   * No longer hardcode HtmlParser for XML files in tika-server (TIKA-2551).

   * Require Java 8 (TIKA-2553).

   * Add a parser for XPS (TIKA-2524).

   * Mime magic for Dolby Digital AC3 and EAC3 files

   * Fixed bug where TesseractOCRParser ignores configured ImageMagickPath,
     and set rotation script to ignore Python warnings (TIKA-2509)

   * Upgrade geo-apis to 3.0.1 (TIKA-2535)

   * Mime definition and magic improvements for text-based programming
     and config formats (TIKA-2554, TIKA-2567, TIKA-1141)

   * Added local Docker image build using dockerfile-maven-plugin to allow
     images to be built from source (TIKA-1518).

   * Support for SAS7BDAT data files (TIKA-2462)

   * Handle .epub files using .htm rather than .html extensions for the
     embedded contents (TIKA-1288)

   * Mime magic for ACES Images (TIKA-2628) and DPX Images (TIKA-2629)

   * For sparse XLSX and XLSB files, always output missing cells to
     the left of filled ones (matching XLS), and optionally output
     missing rows on all 3 formats if requested via the
     OfficeParserContext (TIKA-2479)

Release 1.17 - 12/8/2017

  ***NOTE: THIS IS THE LAST VERSION OF TIKA THAT WILL RUN
     ON Java 7.  The next versions will require Java 8***

  * Fix thread-safety in ChmExtractor (TIKA-2519).

  * Upgrade cxf to 3.0.16 (TIKA-2516).

  * Allow users to configure maxMainMemoryBytes for PDFs via shrike (PR-213).

  * Extract underline and strikethrough in docx (TIKA-2347 and TIKA-2512).

  * Cache TikaConfig in EmbeddedDocumentUtil for better performance
    in documents with large number of attachments (TIKA-2511).

  * Extract media files from ooxml (TIKA-2510).

  * Standardize the way the Image and Video captioning
    dockers and extraction work (TIKA-2400, GitHub-208)

  * Upgrade to xmpcore 5.1.3 (TIKA-2034).

  * Upgrade to metadata-extractor 2.10.1 (TIKA-2486).

  * Upgrade to OpenNLP 1.8.3 (TIKA-2502).

  * Upgrade to Jackson 2.9.2 (TIKA-2501).

  * Catch potential NPE in getting InputStream for attachments
    in PST file (TIKA-2488).

  * Upgrade to PDFBox 2.0.8 (TIKA-2489).

  * Allow configuration of markLimit in EncodingDetectors
    via tika-config.xml (TIKA-2485).

  * RFC822Parser now selects the best alternative for
    multipart/alternative body components.  This aligns with the
    behavior of the OutlookParser (TIKA-2478).  Users can select
    legacy behavior via the "extractAllAlternatives" parameter
    in the RFC822 parser definition in tika-config.xml.

  * Narrow mime detection for ms-owner files and add detection
    for .nls files (TIKA-2469).

  * Fix bug in CharsetDetector that led to different detected charsets
    depending on whether user setText with a byte[] or an InputStream
    via Sean Story (TIKA-2475).

  * Remove JAXB for easier use with Java 9 via Robert Munteanu (TIKA-2466).

  * Upgrade to POI 3.17 (TIKA-2429).

  * Enabling extraction of standard references from text (TIKA-2449).

  * Load external custom mimetypes XML from system property
    tika.custom-mimetypes (TIKA-2460).

  * Extract number of tiffs in a multi-page tiff (TIKA-2451).

  * Fix detection of emails extracted from mbox (TIKA-2456).

  * Add OverrideDetector and allow PSTParser to specify body content type
    as text or html -- to avoid incorrect auto-detection of
    rfc/mbox, etc. (TIKA-2454)

  * AutoDetectParser throws ZeroByteFileException for zero-byte files after
    detection on the file extension (TIKA-2450).

  * Extract phonetic runs in docx with experimental SAX parser (TIKA-2448).

  * Extract phonetic runs from xls and allow users to turn off extraction
    of phonetic runs in both xls and xlsx (TIKA-2440).

  * OOXML locale should be set by POI's LocaleUtil not Locale.getDefault().
    Fix unit tests to be robust against different locales in OOXML
    and ExcelParser (TIKA-2438).

  * Upgrade to PDFBox 2.0.7 (TIKA-2431).

  * Tika now has support for automatic image captioning, that
    combines Computer Vision and Natural Language Processing to
    automatically generate a readable caption for an image
    (TIKA-2262, TIKA-2355, TIKA-2402, Gh-198, Gh-196, Gh-189).

  * Add TestCorruptedFiles to allow devs to test parsers against
    corrupted input files (TIKA-2430).

  * Correct Mimetype definition for Windows batch files (CMD and BAT)
    which are the same (TIKA-2445)

  * PSDParser memory use improvements (TIKA-2447)

  * Add underline extraction from Word documents (doc/docx) via Stuart Hendren
    as well as strikethrough extraction in docx (TIKA-2347, GitHub-173)

  * Corrected Tesseract OCR rotation.py script and made it a configurable
    option via Peter Weiss (TIKA-2385)

Release 1.16 - 7/7/2017

  * Exclude jj2000 from edu.ucar grip to avoid potential
    license conflicts with ASL 2.0

  * Add Age recognition using Ensemble model for Linear regression
    and Apache OpenNLP Maximum Entropy. Tika can now detect age from
    text (TIKA-1988).

  * Add Tika Deep Learning support for the VGG16 model for
    Very Deep Convolutional Networks for Large-Scale Image Recognition.
    Now Tika supports both Inception v3/v4 and VGG16 based image
    recognition (TIKA-2298).

  * Extract macros from PPT (TIKA-2089).

  * Extract absolute path for last saved location when available
    in .xlsx and .xlsb (TIKA-2335).

  * Rename SentimentParser to SentimentAnalysisParser to
    prevent conflict with dependency (TIKA-2368).

  * tika-app now extracts inline images in PDFs by
    default, and it includes a warning to users that this is not the
    default behavior elsewhere in Tika (TIKA-2374).

  * Allow configurability of warnings for problems during
    parser initialization (TIKA-2389).

  * Upgrade to Jackcess 2.1.8 (TIKA-2380).

  * Upgrade to POI 3.17-beta1 (TIKA-2336).

  * Remove non-ASL-2.0-compatible org.json (TIKA-1804).

  * Allow extraction of <script> elements in HTML as embedded "MACRO".
    Users must turn this on via TikaConfig (TIKA-2391).

  * Allow users to turn off extraction of headers and footers
    from .doc, .docx, .xls, .xlsx, .xlsb (TIKA-2362)

  * Extract text from charts in .docx, .pptx, .xlsx and .xlsb
    (TIKA-2254).

  * Extract text from diagrams in .docx, .pptx, .xlsx and .xlsb
    (TIKA-1945).

  * Fix bug in tika-server that led to an attempt to close the
    input stream twice (TIKA-2384).

  * Enable base32 encoding of digests and enable BouncyCastle implementations
    of digest algorithms (TIKA-2386).

  * Add snap builds to codebase (TIKA-2401)

  * Canonical Mimetype of WAVE audio changed to match RFC 2361 defined
    version, audio/vnd.wave, older audio/x-wav remains as an alias

  * Upgrade "provided" xerial to 3.19.3 (TIKA-2412).

  * Upgrade Gson to 2.8.1 (TIKA-2414).

  * Upgrade mime4j to 0.8.1 (TIKA-2413).

  * Mime magic improvements for GraphViz (TIKA-2422), HTML files which
    claim to be XML but aren't quite valid XML (TIKA-2419) and QuickTime
    / MP4 (TIKA-2418)

Release 1.15 - 05/23/2017

  * Tika now has a module for Deep Learning powered by the
    DL4J toolkit. The initial included model is for InceptionV3
    and so using this module, natively in Java, Tika can use
    Deep learning for metadata/text extraction from Images using
    the power of the Inception model (Github-165).

  * A new parser for sentiment analysis using a categorical
    (multi-class, anry, sad, neutral, like, love) and binary
    (positive/negative) was added leveraging the USC data
    science work (TIKA-2016).

  * Tika now has the ability to automatically detect objects in videos,
    using OpenCV and Tensorflow (TIKA-2322).

  * Change default behavior to parse embedded documents even if the user
    forgets to specify a Parser.class in the ParseContext (TIKA-2096).
    Users who wish to parse only the container document should set
    an EmptyParser as the Parser.class in the ParseContext.

  * Change default behavior of Office Parsers to _not_ extract
    Macros.  User needs to setExtractMacros to "true" (TIKA-2302).

  * Added tika-eval module (TIKA-1332).

  * Unified logging across Tika: SLF4J as logging API, Apache Log4j as
    implementation with JCL and JUL bridges in standalone tools like
    tika-app, tika-batch and tika-server (TIKA-2245).

  * Add parser for XLSB files (TIKA-1195).

  * Add parsers for EMF/WMF files (TIKA-2246/TIKA-2247).

  * Add parsers for WordPerfect and QuattroPro (.qpw) files.
    Contributed by Pascal Essiembre (TIKA-1946 and TIKA-2228).

  * Add experimental SAX parser for .pptx files. To select this parser,
    set useSAXPptxExtractor(true) on OfficeParserConfig (TIKA-2210).

  * Add experimental SAX parser for .docx files. To select this parser,
    set useSAXDocxExtractor(true) on OfficeParserConfig (TIKA-1321, TIKA-2191).

  * Add mime detection and parser for Word 2006ML format (TIKA-2179).

  * Bug fix for WordPerfect via Pascal Essiembre (TIKA-2352).

  * Added "text-main" equivalent option to tika-server via
    /tika/main (TIKA-2343).

  * Enabled configuration of the EncodingDetector used by
    parsers that extend AbstractEncodingDetectorParser (TIKA-2273).

  * Prevent easily preventable OOMs for both detection and parsing
    of some compression formats (TIKA-2330).

  * Extract images and thumbnails from ODT via Sam Bayer (TIKA-2295).

  * Fix potential NPE in FeedParser via Julien Nioche (TIKA-2269).

  * Official mime types for BMP, EMF and WMF have been registered with
    IANA, so switch to these (image/bmp image/emf image/wmf) (TIKA-2250)

  * Be more parsimonious with BufferedInputStreams via Josh Hight
    (TIKA-2244).

  * Enable handling of hyphenated language codes in TesseractOCRParser
    via Graham Russell (TIKA-2231).

  * Improve style tags in ODT (TIKA-2242).

  * Add container detection for embedded MSEquation files (TIKA-2238).

  * Add parsing of JBIG2 and extraction of JBIG2 from PDFs when
    required dependencies are added to class path by user.
    Contributed by Pascal Essiembre (TIKA-2232).

  * Mime magic for the OneNote family (.one / .onetoc / .onepkg), no parser
    (TIKA-2224).

  * Add configurability of "preserve-interword-spacing" to
    TesseractOCRParser (TIKA-2190).

  * Upgrade to PDFBox 2.0.6 and JempBox 1.8.13 (TIKA-2209/TIKA-2236/TIKA-2361).

  * Refactor MockParser to consolidate service loading
    and mime types into tika-core/src/test (TIKA-2195).

  * Enabled extraction of embedded objects from headers, footers,
    footnotes, endnotes and comments in legacy .docx parser (TIKA-2192).

  * Allow extraction of PDActions (including Javascript) from
    PDFs (TIKA-2090).  This is turned off by default.  Users
    must setExtractActions(true) on the PDFParserConfig.

  * Change default behavior in experimental .docx parser to ignore
    deleted text to align with .doc (TIKA-2187).

  * Upgrade to POI 3.16 (TIKA-2116, TIKA-2181, TIKA-2329).

  * Allow configuration of timeout for ForkParser (TIKA-2170).

  * Add extraction of .jpx inline images from PDFs when required
    dependencies are added by user to class path (TIKA-2175).

  * Add .jpx, .jp2, .ppm to formats handled by Tesseract (TIKA-2174).

  * Upgrade SQLite "provided" dependency to 3.16.1 (TIKA-2334).

  * Update Apache CXF version to 3.0.12 (TIKA-2292).

  * Add Lingo24 Language Detector (TIKA-2297).

  * Further mime magic for WebVTT (TIKA-1772)

  * Extend support for increased PSM options up to 13 for modern
    versions of Tesseract (TIKA-2357).

  * Prevent potential resource leak by closing TrueTypeFont
    via Cameron Rollheiser (TIKA-2370).

Release 1.14 - 10/19/2016

  * Extract all headers from MSG/RFC822 (TIKA-2122).

  * Upgrade metadata-extractor to 2.9.1 (TIKA-2113).

  * Extract PDF DocInfo metadata into separate keys to prevent
    overwriting by XMP metadata (TIKA-2057).

  * Re-enable fileUrl for tika-server (TIKA-2081).  If you choose,
    to use this feature, beware of the security vulnerabilities!
    See: https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2015-3271

  * Add Tesseract's hOCR output format as an option, via Eric Pugh
    (TIKA-2093)

  * Extract macros from MSOffice files (TIKA-2069).

  * Maintain passed-in mime in TXTParser (TIKA-2047).

  * Upgrade to POI.3-15 (TIKA-2013).

  * Upgrade to PDFBox 2.0.3 (TIKA-2051).

  * Fix hyperlinks with formatting in DOC and DOCX (TIKA-1255
    and TIKA-2078)

  * Tika now is integrated with the Tensorflow library from Google
    and it can use its Inception v3 image classification model to
    identify objects in images (TIKA-1993).

  * Parser configuration is now type-safe and parameters for parsers
    can have assigned types (TIKA-1508, TIKA-1986).

  * Prevent OOM/permanent hang on some corrupt CHM files (TIKA-2040).

  * Upgrade ICU4J charset detection components to fix multithreading
    bug (TIKA-2041).

  * Upgrade to Jackcess 2.1.4 (TIKA-2039).

  * Maintain more significant digits in cells of "General" format
    in XLS and XLSX (TIKA-2025).

  * Avoid mark/reset issues when extracting or detecting embedded resources
    in RFC822 emails (TIKA-2037).

  * Improving accuracy of Tesseract for better extraction of numeric
    and alphanumeric text from images (TIKA-2021, TIKA-2031).

  * Improve extraction of embedded documents from PPT, PPTX and XLSX
    (TIKA-2026).

  * Add parser for applefile (AppleSingle) (TIKA-2022).

  * Add mime types, mime magic and/or globs for:
     * Endnote Import File (TIKA-2011)
     * DJVU files (TIKA-2009)
     * MS Owner File (TIKA-2008)
     * Windows Media Metafile (TIKA-2004)
     * iCal and vCalendar (TIKA-2006)
     * MBOX (TIKA-2042)
     * Stata DTA (TIKA-2064)

  * Add configurable maximum threshold for number of events extracted
    from the XMP Media Management Schema in JempboxExtractor (TIKA-1999).

  * Integrate TesseractOCR with full page image rendering for PDFs (TIKA-1994).

  * Add mime detection via Nick C and parser for DBF files (TIKA-1513).

  * Add mime detection and parsers for MSOffice 2003 XML Word
    and Excel formats (TIKA-1958).

  * Extract hyperlinks from PPT, PPTX, XSLX (TIKA-1454).

  * Upgrade to Commons Compress 1.12 (supports progress on TIKA-1358)

Release 1.13 - 05/08/2016

  * Upgrade to PDFBox 2.0.1 (TIKA-1285/TIKA-1959).
    MAJOR CHANGES in PDFParser:
    * The classic sequential parser is no longer available.
    * Tiff files are no longer extracted by default.  See
      https://pdfbox.apache.org/2.0/dependencies.html#optional-components
      for optional components to process Tiff files.
    * Some truncated/corrupted files that had some content extracted
      with 1.8.x may have no content extracted in 2.0.x (see TIKA-1912).

  * 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 Lang
    Detector 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 <scope>provided</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).

Release 1.12 - 01/24/2016

  * Support for iFrames and element link extraction is provided in
    the link Content Handler (TIKA-1835).

  * Slide notes are now linked to the slide XHTML in the PPT output
    (TIKA-1840).

  * JSON tests in Tika server were updated to remove impossible casts
    (Github-73).

  * Fix bug in GeoTopicParser where NER is reused instead of instantiated
    with each request (TIKA-1834).

  * Upgrade rome to 1.5.1 && Downgrade Rome dependency to 0.9 to avoid
    nasty NPE (TIKA-1820, TIKA-1516)

  * The NamedEntityParser was enhanced to generate text content
    in addition to metadata (TIKA-1815, TIKA-1816).

  * A significant speed-up is made to the GeoTopicParser by
    using the new REST server capabilities from Lucene Geo
    Gazetteer (TIKA-1803).

  * A parser to compute motion properties in Videos, e.g.,
    Histogram of Oriented Gradients and Histogram of Optical Flows
    using the Pooled Time Series algorithm, was added (TIKA-1798).

  * Provide NamedEntityParser which exposes Named Entity Recognition
    from OpenNLP and Stanford NER providers (TIKA-1787, GitHub-61,
    GitHub-62).

  * Allow XHTMLContentHandler to pass attributes of html element
    via Markus Jelsma (TIKA-1782).

  * Fix regression with spacing in PPT via Andreas Beeker (TIKA-1777).

  * Tika Facade parse methods for Path and File added which take a
    Metadata object, to mirror the existing InputStream one (GitHub-60)

  * GeoParser fix for loading the NER model from a jar file (TIKA-1791)


Release 1.11 - 10/18/2015

  * Java7 API support for allowing java.nio.file.Path as method arguments
    was added to Tika and to ParsingReader, TikaFileTypeDetector, and to
    Tika Config (TIKA-1745, TIKA-1746, TIKA-1751).

  * MIME support was added for WebVTT: The Web Video Text Tracks Format
    files (TIKA-1772).

  * MIME magic improved to ensure emails detected as message/rfc822
    (TIKA-1771).

  * Upgrade to Jackcess Encrypt 2.1.1 to avoid binary incompatibility
    with Bouncy Castle (TIKA-1736).

  * Make div and other markup more consistent between PPT and
    PPTX (TIKA-1755).

  * Parse multiple authors from MSOffice's semi-colon delimited
    author field (TIKA-1765).

  * Include CTAKESConfig.properties within tika-parsers resources
    by default (TIKA-1741).

  * Prevent infinite recursion when processing inline images
    in PDF files by limiting extraction of duplicate images
    within the same page (TIKA-1742).

  * Upgrade to POI 3.13-final (via Andreas Beeker) (TIKA-1707).

  * Upgraded tika-batch to use Path throughout (TIKA-1747 and
    (TIKA-1754).

  * Upgraded to Path in TikaInputStream (via Yaniv Kunda) (TIKA-1744).

  * Changed default content handler type for "/rmeta" in tika-server
    to "xml" to align with "-J" option in tika-app.
    Clients can now specify handler types via PathParam. (TIKA-1716).

  * The fantastic GROBID (or Grobid) GeneRation Of BIbliographic Data
    for machine learning from PDF files is now integrated as a
    Tika parser (TIKA-1699, TIKA-1712).

  * The ability to specify the Tesseract Config Path was added
    to the OCR Parser (TIKA-1703).

  * Upgraded to ASM 5.0.4 (TIKA-1705).

  * Corrected Tika Config XML detector definition explicit loading
    of MimeTypes (TIKA-1708)

  * In Tika Parsers, Batch, Server, App and Examples, use Apache
    Commons IO instead of inlined ex-Commons classes, and the Java 7
    Standard Charset definitions (TIKA-1710)

  * Upgraded to Commons Compress 1.10, which enables zlib compressed
    archives support (TIKA-1718)


Release 1.10 - 8/1/2015

  * Tika Config XML can now be used to create composite detectors,
    and exclude detectors that DefaultDetector would otherwise
    have used. This brings support in-line with Parsers. (TIKA-1702)

  * Reverted to legacy sort order of parsers that was
    mistakenly reversed in Tika 1.9 (TIKA-1689).

  * Upgrade to POI 3.13-beta1 (TIKA-1667).

  * Upgrade to PDFBox 1.8.10 (TIKA-1588).

  * MimeTypes now tries to find a registered type with and
    without parameters (TIKA-1692).

  * Added more robust error handling for encoding detection
    of .MSG files (TIKA-1238).

  * Fixed bug in Tika's use of the Jackcess parser that
    prevented reading of v97 Access files (TIKA-1681).

  * Upgrade xerial.org's sqlite-jdbc to 3.8.10.1. NOTE:
    as of Tika 1.9, this jar is "provided." Make sure
    to upgrade your provided jar! (TIKA-1687).

  * Add header/footer extraction to xls (via Aeham Abushwashi)
    (TIKA-1400).

  * Drop the source file name from the embedded file path in
    RecursiveParserWrapper's "X-TIKA:embedded_resource_path"
    (TIKA-1673).

  * Upgraded to Java 7 (TIKA-1536).

  * Non-standards compliant emails are now correctly detected
    as message/rfc822 (TIKA-1602).

  * Added parser for MS Access files via Jackcess. Many thanks
    to Health Market Science, Brian O'Neill and James Ahlborn
    for relicensing Jackcess to Apache v2! (TIKA-1601)

  * GDALParser now correctly sets "nitf" as a supported
    MediaType (TIKA-1664).

  * Added DigestingParser to calculate digest hashes
    and record them in metadata. Integrated with
    tika-app and tika-server (TIKA-1663).

  * Fixed ZipContainerDetector to detect all IPA files
    (TIKA-1659).


Release 1.9 - 6/6/2015

  * The ability to use the cTAKES clinical text
    knowledge extraction system for biomedical data is
    now included as a Tika parser (TIKA-1645, TIKA-1642).

  * Tika-server allows a user to specify the Tika config
    from the command line (TIKA-1652, TIKA-1426).

  * Matlab file detection has been improved (TIKA-1634).

  * The EXIFTool was added as an External parser
    (TIKA-1639).

  * If FFMPEG is installed and on the PATH, it is a
    usable Parser in Tika now (TIKA-1510).

  * Fixes have been applied to the ExternalParser to make
    it functional (TIKA-1638).

  * Tika service loading can now be more verbose with the
    org.apache.tika.service.error.warn system property (TIKA-1636).

  * Tika Server now allows for metadata extraction from remote
    URLs and in addition it outputs the detected language as a
    metadata field (TIKA-1625).

  * OUTPUT_FILE_TOKEN not being replaced in ExternalParser
    contributed by Pascal Essiembre (TIKA-1620).

  * Tika REST server now supports language identification
    (TIKA-1622).

  * All of the example code from the Tika in Action book has
    been donated to Tika and added to tika-examples (TIKA-1562).

  * Tika server now logs errors determining ContentDisposition
    (TIKA-1621).

  * An algorithm for using Byte Histogram frequencies to construct
    a Neural Network and to perform MIME detection was added
    (TIKA-1582).

  * A Bayesian algorithm for MIME detection by probabilistic
    means was added (TIKA-1517).

  * Tika now incorporates the Apache Spatial Information
    System capability of parsing Geographic ISO 19139
    files (TIKA-443). It can also detect those files as
    well.

  * Update the MimeTypes code to support inheritance
    (TIKA-1535).

  * Provide ability to parse and identify Global Change
    Master Directory Interchange Format (GCMD DIF)
    scientific data files (TIKA-1532).

  * Improvements to detect CBOR files by extension (TIKA-1610).

  * Change xerial.org's sqlite-jdbc jar to "provided" (TIKA-1511).
    Users will now need to add sqlite-jdbc to their classpath for
    the Sqlite3Parser to work.

  * ExternalParser.check now catches (suppresses) SecurityException
    and returns false, so it's OK to run Tika with a security policy
    that does not allow execution of external processes (TIKA-1628).

Release 1.8 - 4/13/2015

  * Fix null pointer when processing ODT footer styles (TIKA-1600).

  * Upgrade to com.drewnoakes' metadata-extractor to 2.0 and
    add parser for webp metadata (TIKA-1594).

  * Duration extracted from MP3s with no ID3 tags (TIKA-1589).

  * Upgraded to PDFBox 1.8.9 (TIKA-1575).

  * Tika now supports the IsaTab data standard for bioinformatics
    both in terms of MIME identification and in terms of parsing
    (TIKA-1580).

  * Tika server can now enable CORS requests with the command line
    "--cors" or "-C" option (TIKA-1586).

  * Update jhighlight dependency to avoid using LGPL license. Thank
    @kkrugler for his great contribution (TIKA-1581).

  * Updated HDF and NetCDF parsers to output file version in
    metadata (TIKA-1578 and TIKA-1579).

  * Upgraded to POI 3.12-beta1 (TIKA-1531).

  * Added tika-batch module for directory to directory batch
    processing.  This is a new, experimental capability, and the API will
    likely change in future releases (TIKA-1330).

  * Translator.translate() Exceptions are now restricted to
    TikaException and IOException (TIKA-1416).

  * Tika now supports MIME detection for Microsoft Extended
    Makefiles (EMF) (TIKA-1554).

  * Tika has improved delineation in XML and HTML MIME detection
    (TIKA-1365).

  * Upgraded the Drew Noakes metadata-extractor to version 2.7.2
    (TIKA-1576).

  * Added basic style support for ODF documents, contributed by
    Axel Dörfler (TIKA-1063).

  * Move Tika server resources and writers to separate
    org.apache.tika.server.resource and writer packages (TIKA-1564).

  * Upgrade UCAR dependencies to 4.5.5 (TIKA-1571).

  * Fix Paths in Tika server welcome page (TIKA-1567).

  * Fixed infinite recursion while parsing some PDFs (TIKA-1038).

  * XHTMLContentHandler now properly passes along body attributes,
    contributed by Markus Jelsma (TIKA-995).

  * TikaCLI option --compare-file-magic to report mime types known to
    the file(1) tool but not known / fully known to Tika.

  * MediaTypeRegistry support for returning known child types.

  * Support for excluding certain Parsers from being
    used by DefaultParser via the Tika Config file, using the new
    parser-exclude tag (TIKA-1558).

  * Detect Global Change Master Directory (GCMD) Directory
    Interchange Format (DIF) files (TIKA-1561).

  * Tika's JAX-RS server can now return stacktraces for
    parse exceptions (TIKA-1323).

  * Added MockParser for testing handling of exceptions, errors
    and hangs in code that uses parsers (TIKA-1553).

  * The ForkParser service removed from Activator. Rollback of (TIKA-1354).

  * Increased the speed of language identification by
    a factor of two -- contributed by Toke Eskildsen (TIKA-1549).

  * Added parser for Sqlite3 db files. Some users will need to
    exclude the dependency on xerial.org's sqlite-jdbc because
    it contains native libs (TIKA-1511).

  * Use POST instead of PUT for tika-server form methods
    (TIKA-1547).

  * A basic wrapper around the UNIX file command was
    added to extract Strings. In addition a parse to
    handle Strings parsing from octet-streams using Latin1
    charsets as added (TIKA-1541, TIKA-1483).

  * Add test files and detection mechanism for Gridded
    Binary (GRIB) files (TIKA-1539).

  * The RAR parser was updated to handle Chinese characters
    using the functionality provided by allowing encoding to
    be used within ZipArchiveInputStream (TIKA-936).

  * Fix out of memory error in surefire plugin (TIKA-1537).

  * Build a parser to extract data from GRIB formats (TIKA-1423).

  * Upgrade to Commons Compress 1.9 (TIKA-1534).

  * Include media duration in metadata parsed by MP4Parser (TIKA-1530).

  * Support password protected 7zip files (using a PasswordProvider,
    in keeping with the other password supporting formats) (TIKA-1521).

  * Password protected Zip files should not trigger an exception (TIKA-1028).

Release 1.7 - 1/9/2015

  * Fixed resource leak in OutlookPSTParser that caused TikaException
    when invoked via AutoDetectParser on Windows (TIKA-1506).

  * HTML tags are properly stripped from content by FeedParser
    (TIKA-1500).

  * Tika Server support for selecting a single metadata key;
    wrapped MetadataEP into MetadataResource (TIKA-1499).

  * Tika Server support for JSON and XMP views of metadata (TIKA-1497).

  * Tika Parent uses dependency management to keep duplicate
    dependencies in different modules the same version (TIKA-1384).

  * Upgraded slf4j to version 1.7.7 (TIKA-1496).

  * Tika Server support for RecursiveParserWrapper's JSON output
    (endpoint=rmeta) equivalent to (TIKA-1451's) -J option
    in tika-app (TIKA-1498).

  * Tika Server support for providing the password for files on a
    per-request basis through the Password http header (TIKA-1494).

  * Simple support for the BPG (Better Portable Graphics) image format
    (TIKA-1491, TIKA-1495).

  * Prevent exceptions from being thrown for some malformed
    mp3 files (TIKA-1218).

  * Reformat pom.xml files to use two spaces per indent (TIKA-1475).

  * Fix warning of slf4j logger on Tika Server startup (TIKA-1472).

  * Tika CLI and GUI now have option to view JSON rendering of output
    of RecursiveParserWrapper (TIKA-1451).

  * Tika now integrates the Geospatial Data Abstraction Library
    (GDAL) for parsing hundreds of geospatial formats (TIKA-605,
    TIKA-1503).

  * ExternalParsers can now use Regexs to specify dynamic keys
   (TIKA-1441).

  * Thread safety issues in ImageMetadataExtractor were resolved
    (TIKA-1369).

  * The ForkParser service is now registered in Activator
    (TIKA-1354).

  * The Rome Library was upgraded to version 1.5 (TIKA-1435).

  * Add markup for files embedded in PDFs (TIKA-1427).

  * Extract files embedded in annotations in PDFS (TIKA-1433).

  * Upgrade to PDFBox 1.8.8 (TIKA-1419, TIKA-1442).

  * Add RecursiveParserWrapper (aka Jukka's and Nick's)
    RecursiveMetadataParser (TIKA-1329)

  * Add example for how to dump TikaConfig to XML (TIKA-1418).

  * Allow users to specify a tika config file for tika-app (TIKA-1426).

  * PackageParser includes the last-modified date from the archive
    in the metadata, when handling embedded entries (TIKA-1246)

  * Created a new Tesseract OCR Parser to extract text from images.
    Requires installation of Tesseract before use (TIKA-93).

  * Basic parser for older Excel formats, such as Excel 4, 5 and 95,
    which can get simple text, and metadata for Excel 5+95 (TIKA-1490)


Release 1.6 - 08/31/2014

  * Parse output should indicate which Parser was actually used
    (TIKA-674).

  * Use the forbidden-apis Maven plugin to check for unsafe Java
    operations (TIKA-1387).

  * Created an ExternalTranslator class to interface with command
    line Translators (TIKA-1385).

  * Created a MosesTranslator as a subclass of ExternalTranslator
    that calls the Moses Decoder machine translation program (TIKA-1385).

  * Created the tika-example module. It will have examples of how to
    use the main Tika interfaces (TIKA-1390).

  * Upgraded to Commons Compress 1.8.1 (TIKA-1275).

  * Upgraded to POI 3.11-beta1 (TIKA-1380).

  * Tika now extracts SDTCell content from tables in .docx files (TIKA-1317).

  * Tika now supports detection of the Persian/Farsi language.
    (TIKA-1337)

  * The Tika Detector interface is now exposed through the JAX-RS
    server (TIKA-1336, TIKA-1336).

  * Tika now has support for parsing binary Matlab files as part of
    our larger effort to increase the number of scientific data formats
    supported. (TIKA-1327)

  * The Tika Server URLs for the unpacker resources have been changed,
    to bring them under a common prefix (TIKA-1324). The mapping is
    /unpacker/{id} -> /unpack/{id}
    /all/{id}      -> /unpack/all/{id}

  * Added module and core Tika interface for translating text between
    languages and added a default implementation that call's Microsoft's
    translate service (TIKA-1319)

  * Added an Translator implementation that calls Lingo24's Premium
    Machine Translation API (TIKA-1381)

  * Made RTFParser's list handling slightly more robust against corrupt
    list metadata (TIKA-1305)

  * Fixed bug in CLI json output (TIKA-1291/TIKA-1310)

  * Added ability to turn off image extraction from PDFs (TIKA-1294).
    Users must now turn on this capability via the PDFParserConfig.

  * Upgrade to PDFBox 1.8.6 (TIKA-1290, TIKA-1231, TIKA-1233, TIKA-1352)

  * Zip Container Detection for DWFX and XPS formats, which are OPC
    based (TIKA-1204, TIKA-1221)

  * Added a user facing welcome page to the Tika Server, which
    says what it is, and a very brief summary of what is available.
    (TIKA-1269)

  * Added Tika Server endpoints to list the available mime types,
    Parsers and Detectors, similar to the --list-<foo> methods on
    the Tika CLI App (TIKA-1270)

  * Improvements to NetCDF and HDF parsing to mimic the output of
    ncdump and extract text dimensions and spatial and variable
    information from scientific data files (TIKA-1265)

  * Extract attachments from RTF files (TIKA-1010)

  * Support Outlook Personal Folders File Format *.pst (TIKA-623)

  * Added mime entries for additional Ogg based formats (TIKA-1259)

  * Updated the Ogg Vorbis plugin to v0.4, which adds detection for a wider
    range of Ogg formats, and parsers for more Ogg Audio ones (TIKA-1113)

  * PDF: Images in PDF documents can now be extracted as embedded resources.
    (TIKA-1268)

  * Fixed RuntimeException thrown for certain Word Documents (TIKA-1251).

  * CLI: TikaCLI now has another option: --list-parser-details-apt, which outputs
    the list of supported parsers in APT format. This is used to generate the list
    on the formats page (TIKA-411).

Release 1.5 - 02/04/2014

  * Fixed bug in handling of embedded file processing in PDFs (TIKA-1228).

  * Added SourceCodeParser to support java, Groovy, C++ files (TIKA-1224).

  * Updated Tika Server to support multipart/form-data payloads (TIKA-1198).

  * Updated Tika Server to CXF 2.7.8 (TIKA-1197).

  * Updated Tika Server to accept requests over wildcard addresses (TIKA-1196).

  * Added option to use alternate NonSequentialPDFParser (TIKA-1201).

  * Content from PDF AcroForms is now extracted (TIKA-973).

  * Fixed invalid asterisks from master slide in PPT (TIKA-1171).

  * Added test cases to confirm handling of auto-date in PPT and PPTX (TIKA-817).

  * Text from tables in PPT files is once again extracted correctly (TIKA-1076).

  * Text is extracted from text boxes in XLSX (TIKA-1100).

  * Tika no longer hangs when processing Excel files with custom fraction format (TIKA-1132).

  * Disconcerting stacktrace from missing beans no longer printed for some DOCX files (TIKA-792).

  * Upgraded POI to 3.10-beta2 (TIKA-1173).

  * Upgraded PDFBox to 1.8.4 (TIKA-1230).

  * Made HtmlEncodingDetector more flexible in finding meta
    header charset (TIKA-1001).

  * Added sanitized test HTML file for local file test (TIKA-1139).

  * Fixed bug that prevented attachments within a PDF from being processed
    if the PDF itself was an attachment (TIKA-1124).

  * Text from paragraph-level structured document tags in DOCX files is now extracted (TIKA-1130).

  * RTF: Fixed ArrayIndexOutOfBoundsException when parsing list override (TIKA-1192).

  * CLI: TikaCLI now escapes invalid filename characters as hex
    characters (TIKA-1078).

Release 1.4 - 06/15/2013

  * Removed a test HTML file with a poorly chosen GPL text in it (TIKA-1129).

  * Improvements to tika-server to allow it to produce text/html and
    text/xml content (TIKA-1126, TIKA-1127).

  * Improvements were made to the Compressor Parser to handle g'zipped files
    that require the decompressConcatenated option set to true (TIKA-1096).

  * Addressed a typographic error that was preventing from detection of
    awk files (TIKA-1081).

  * Added a new end-point to Tika's JAX-RS REST server that only detects
    the media-type based on a small portion of the document submitted
   (TIKA-1047).

  * RTF: Ordered and unordered lists are now extracted (TIKA-1062).

  * MP3: Audio duration is now extracted (TIKA-991)

  * Java .class files: upgraded from ASM 3.1 to ASM 4.1 for parsing
    the Java bytecodes (TIKA-1053).

  * Mime Types: Definitions extended to optionally include Link (URL) and
    UTI, along with details for several common formats (TIKA-1012 / TIKA-1083)

  * Exceptions when parsing OLE10 embedded documents, when parsing
    summary information from Office documents, and when saving
    embedded documennts in TikaCLI are now logged instead
    of aborting extraction (TIKA-1074)

  * MS Word: line tabular character is now replaced with newline
    (TIKA-1128)

  * XML: ElementMetadataHandlers can now optionally accept duplicate
    and empty values (TIKA-1133)

Release 1.3 - 01/19/2013

  * Mimetype definitions added for more common programming languages,
    including common extensions, but not magic patterns. (TIKA-1055)

  * MS Word: When a Word (.doc) document contains embedded files or
    links to external documents, Tika now places a <div
    class="embedded" id="_XXX"/> placeholder into the XHTML so you can
    see where in the main text the embedded document occurred
    (TIKA-956, TIKA-1019).  Embedded Wordpad/RTF documents are now
    recognized (TIKA-982).

  * PDF: Text from pop-up annotations is now extracted (TIKA-981).
    Text from bookmarks is now extracted (TIKA-1035).

  * PKCS7: Detached signatures no longer through NullPointerException
    (TIKA-986).

  * iWork: The chart name for charts embedded in numbers documents is
    now extracted (TIKA-918).

  * CLI: TikaCLI -m now handles multi-valued metadata keys correctly
    (previously it only printed the first value).  (TIKA-920)

  * MS Word (.docx): When a Word (.docx) document contains embedded
    files, Tika now places a <div class="embedded" id="XXX"/> into the
    XHTML so you can see where in the main text the embedded document
    occurred.  The id (rId) is included in the Metadata of each
    embedded document as the new Metadata.EMBEDDED_RELATIONSHIP_ID
    key, and TikaCLI prepends the rId (if present) onto the filename
    it extracts (TIKA-989).  Fixed NullPointerException when style is
    null (TIKA-1006).  Text inside text boxes is now extracted
    (TIKA-1005).

  * RTF: Page, word, character count and creation date metadata are
    now extracted for RTF documents (TIKA-999).

  * MS PowerPoint (.pptx): When a PowerPoint (.pptx) document contains
    embedded files, Tika now places a <div class="embedded" id="XXX"/> into the
    XHTML so you can see where in the main text the embedded document
    occurred.  The id (rId) is included in the Metadata of each
    embedded document as the new Metadata.EMBEDDED_RELATIONSHIP_ID
    key, and TikaCLI prepends the rId (if present) onto the filename
    it extracts (TIKA-997, TIKA-1032).

  * MS PowerPoint (.ppt): When a PowerPoint (.ppt) document contains
    embedded files, Tika now places a <div class="embedded" id="XXX"/> into the
    XHTML so you can see where in the main text the embedded document
    occurred (TIKA-1025).  Text from the master slide is now extracted
    (TIKA-712).

  * MHTML: fixed Null charset name exception when a mime part has an
    unrecognized charset (TIKA-1011).

  * MP3: if an ID3 tag was encoded in UTF-16 with only the BOM then on
    certain JVMs this would incorrectly extract the BOM as the tag's
    value (TIKA-1024).

  * ZIP: placeholders (<div class="embedded" id="<entry name>"/>) are
    now left in the XHTML so you can see where each archive member
    appears (TIKA-1036). TikaCLI would hit FileNotFoundException when
    extracting files that were under sub-directories from a ZIP
    archive, because it failed to create the parent directories first
    (TIKA-1031).

  * XML: a space character is now added before each element
    (TIKA-1048)

Release 1.2 - 07/10/2012
---------------------------------

  * Tika's JAX-RS based Network server now is based on Apache CXF,
    which is available in Maven Central and now allows the server
    module to be packaged and included in our release
    (TIKA-593, TIKA-901).

  * Tika: parseToString now lets you specify the max string length
    per-call, in addition to per-Tika-instance. (TIKA-870)

  * Tika now has the ability to detect FITS (Flexible Image Transport System)
    files (TIKA-874).

  * Images: Fixed file handle leak in ImageParser. (TIKA-875)

  * iWork: Comments in Pages files are now extracted (TIKA-907).
    Headers, footers and footnotes in Pages files are now extracted
    (TIKA-906).  Don't throw NullPointerException on passsword
    protected iWork files, even though we can't parse their contents
    yet (TIKA-903).  Text extracted from Keynote text boxes and bullet
    points no longer runs together (TIKA-910). Also extract text for
    Pages documents created in layout mode (TIKA-904).  Table names
    are now extracted in Numbers documents (TIKA-924).  Content added
    to master slides is also extracted (TIKA-923).

  * Archive and compression formats: The Commons Compress dependency was
    upgraded from 1.3 to 1.4.1. With this change Tika can now parse also
    Unix dump archives and documents compressed using the XZ and Pack200
    compression formats. (TIKA-932)

  * KML: Tika now has basic support for Keyhole Markup Language documents
    (KML and KMZ) used by tools like Google Earth. See also
    http://www.opengeospatial.org/standards/kml/. (TIKA-941)

  * CLI: You can now use the TIKA_PASSWORD environment variable or the
    --password=X command line option to specify the password that Tika CLI
    should use for opening encrypted documents (TIKA-943).

  * Character encodings: Tika's character encoding detection mechanism was
    improved by adding integration to the juniversalchardet library that
    implements Mozilla's universal charset detection algorithm. The slower
    ICU4J algorithms are still used as a fallback thanks to their wider
    coverage of custom character encodings. (TIKA-322, TIKA-471)

  * Charset parameter: Related to the character encoding improvements
    mentioned above, Tika now returns the detected character encoding as
    a "charset" parameter of the content type metadata field for text/plain
    and text/html documents. For example, instead of just "text/plain", the
    returned content type will be something like "text/plain; charset=UTF-8"
    for a UTF-8 encoded text document. Character encoding information is still
    present also in the content encoding metadata field for backwards
    compatibility, but that field should be considered deprecated. (TIKA-431)

  * Extraction of embedded resources from OLE2 Office Documents, where
    the resource isn't another office document, has been fixed (TIKA-948)

Release 1.1 - 3/7/2012
---------------------------------

 * Link Extraction: The rel attribute is now extracted from
   links per the LinkConteHandler. (TIKA-824)

 * MP3: Fixed handling of UTF-16 (two byte) ID3v2 tags (previously
   the last character in a UTF-16 tag could be corrupted) (TIKA-793)

 * Performance: Loading of the default media type registry is now
   significantly faster. (TIKA-780)

 * PDF: Allow controlling whether overlapping duplicated text should
   be removed.  Disabling this (the default) can give big
   speedups to text extraction and may workaround cases where
   non-duplicated characters were incorrectly removed (TIKA-767).
   Allow controlling whether text tokens should be sorted by their x/y
   position before extracting text (TIKA-612); this is necessary for
   certain PDFs.  Fixed cases where too many </p> tags appear in the
   XHTML output, causing NPE when opening some PDFs with the GUI
   (TIKA-778).

 * RTF: Fixed case where a font change would result in processing
   bytes in the wrong font's charset, producing bogus text output
   (TIKA-777).  Don't output whitespace in ignored group states,
   avoiding excessive whitespace output (TIKA-781).  Binary embedded
   content (using \bin control word) is now skipped correctly;
   previously it could cause the parser to incorrectly extract binary
   content as text (TIKA-782).

 * CLI: New TikaCLI option "--list-detectors", which displays the
   mimetype detectors that are available, similar to the existing
   "--list-parsers" option for parsers. (TIKA-785).

 * Detectors: The order of detectors, as supplied via the service
   registry loader, is now controlled. User supplied detectors are
   prefered, then Tika detectors (such as the container aware ones),
   and finally the core Tika MimeTypes is used as a backup. This
   allows for specific, detailed detectors to take preference over
   the default mime magic + filename detector. (TIKA-786)

 * Microsoft Project (MPP): Filetype detection has been fixed,
   and basic metadata (but no text) is now extracted. (TIKA-789)

 * Outlook: fixed NullPointerException in TikaGUI when messages with
   embedded RTF or HTML content were filtered (TIKA-801).

 * Ogg Vorbis and FLAC: Parser added for Ogg Vorbis and FLAC audio
   files, which extract audio metadata and tags (TIKA-747)

 * MP4: Improved mime magic detection for MP4 based formats (including
   QuickTime, MP4 Video and Audio, and 3GPP) (TIKA-851)

 * MP4: Basic metadata extracting parser for MP4 files added, which includes
   limited audio and video metadata, along with the iTunes media metadata
   (such as Artist and Title) (TIKA-852)

 * Document Passwords: A new ParseContext object, PasswordProvider,
   has been added. This provides a way to supply the password for
   a document during processing. Currently, only password protected
   PDFs and Microsoft OOXML Files are supported. (TIKA-850)

Release 1.0 - 11/4/2011
---------------------------------

The most notable changes in Tika 1.0 over previous releases are:

 * API: All methods, classes and interfaces that were marked as
   deprecated in Tika 0.10 have been removed to clean up the API
   (TIKA-703). You may need to adjust and recompile client code
   accordingly. The declared OSGi package versions are now 1.0, and
   will thus not resolve for client bundles that still refer to 0.x
   versions (TIKA-565).

 * Configuration: The context class loader of the current thread is
   no longer used as the default for loading configured parser and
   detector classes. You can still pass an explicit class loader
   to the configuration mechanism to get the previous behaviour.
   (TIKA-565)

 * OSGi: The tika-core bundle will now automatically pick up and use
   any available Parser and Detector services when deployed to an OSGi
   environment. The tika-parsers bundle provides such services based on
   for all the supported file formats for which the upstream parser library
   is available. If you don't want to track all the parser libraries as
   separate OSGi bundles, you can use the tika-bundle bundle that packages
   tika-parsers together with all its upstream dependencies. (TIKA-565)

 * RTF: Hyperlinks in RTF documents are now extracted as an <a
   href=...>...</a> element (TIKA-632). The RTF parser is also now
   more robust when encountering too many closing {'s vs. opening {'s
   (TIKA-733).

 * MS Word: From Word (.doc) documents we now extract optional hyphen
   as Unicode zero-width space (U+200B), and non-breaking hyphen as
   Unicode non-breaking hyphen (U+2011). (TIKA-711)

 * Outlook: Tika can now process also attachments in Outlook messages.
   (TIKA-396)

 * MS Office: Performance of extracting embedded office docs was improved.
   (TIKA-753)

 * PDF: The PDF parser now extracts paragraphs within each page
   (TIKA-742) and  can now optionally extract text from PDF
   annotations (TIKA-738). There's also an option to enable (the
   default) or disable auto-space insertion (TIKA-724).

 * Language detection: Tika can now detect Belarusian, Catalan,
   Esperanto, Galician, Lithuanian (TIKA-582), Romanian, Slovak,
   Slovenian, and Ukrainian (TIKA-681).

 * Java: Tika no longer ships retrotranslated Java 1.4 binaries along
   with the normal ones that work with Java 5 and higher. (TIKA-744)

 * OpenOffice documents: header/footer text is now extracted for text,
   presentation and spreadsheet documents (TIKA-736)

Tika 1.0 relies on the following set of major dependencies (generated using
mvn dependency:tree from tika-parsers):

   org.apache.tika:tika-parsers:bundle:1.0
   +- org.apache.tika:tika-core:jar:1.0:compile
   +- edu.ucar:netcdf:jar:4.2-min:compile
   |  \- org.slf4j:slf4j-api:jar:1.5.6:compile
   +- org.apache.james:apache-mime4j-core:jar:0.7:compile
   +- org.apache.james:apache-mime4j-dom:jar:0.7:compile
   +- org.apache.commons:commons-compress:jar:1.3:compile
   +- commons-codec:commons-codec:jar:1.5:compile
   +- org.apache.pdfbox:pdfbox:jar:1.6.0:compile
   |  +- org.apache.pdfbox:fontbox:jar:1.6.0:compile
   |  +- org.apache.pdfbox:jempbox:jar:1.6.0:compile
   |  \- commons-logging:commons-logging:jar:1.1.1:compile
   +- org.bouncycastle:bcmail-jdk15:jar:1.45:compile
   +- org.bouncycastle:bcprov-jdk15:jar:1.45:compile
   +- org.apache.poi:poi:jar:3.8-beta4:compile
   +- org.apache.poi:poi-scratchpad:jar:3.8-beta4:compile
   +- org.apache.poi:poi-ooxml:jar:3.8-beta4:compile
   |  +- org.apache.poi:poi-ooxml-schemas:jar:3.8-beta4:compile
   |  |  \- org.apache.xmlbeans:xmlbeans:jar:2.3.0:compile
   |  \- dom4j:dom4j:jar:1.6.1:compile
   +- org.apache.geronimo.specs:geronimo-stax-api_1.0_spec:jar:1.0.1:compile
   +- org.ccil.cowan.tagsoup:tagsoup:jar:1.2.1:compile
   +- asm:asm:jar:3.1:compile
   +- com.drewnoakes:metadata-extractor:jar:2.4.0-beta-1:compile
   +- de.l3s.boilerpipe:boilerpipe:jar:1.1.0:compile
   +- rome:rome:jar:0.9:compile
      \- jdom:jdom:jar:1.0:compile

The following people have contributed to Tika 1.0 by submitting or commenting
on the issues resolved in this release:

Andrzej Bialecki
Antoni Mylka
Benson Margulies
Chris A. Mattmann
Cristian Vat
Dave Meikle
David Smiley
Dennis Adler
Erik Hetzner
Ingo Renner
Jeremias Maerki
Jeremy Anderson
Jeroen van Vianen
John Bartak
Jukka Zitting
Julien Nioche
Ken Krugler
Mark Butler
Maxim Valyanskiy
Michael Bryant
Michael McCandless
Nick Burch
Pablo Queixalos
Uwe Schindler
Žygimantas Medelis


See http://s.apache.org/Zk6 for more details on these contributions.


Release 0.10 - 09/25/2011
-------------------------

The most notable changes in Tika 0.10 over previous releases are:

 * A parser for CHM help files was added. (TIKA-245)

 * TIKA-698: Invalid characters are now replaced with the Unicode
   replacement character (U+FFFD), whereas before such characters were
   replaced with spaces, so you may need to change your processing of
   Tika's output to now handle U+FFFD.

 * The RTF parser was rewritten to perform its own direct shallow
   parse of the RTF content, instead of using RTFEditorKit from
   javax.swing.  This fixes several issues in the old parser,
   including doubling of Unicode characters in certain cases
   (TIKA-683), exceptions on mal-formed RTF docs (TIKA-666), and
   missing text from some elements (header/footer, hyperlinks,
   footnotes, text inside pictures).

 * Handling of temporary files within Tika was much improved
   (TIKA-701, TIKA-654, TIKA-645, TIKA-153)

 * The Tika GUI got a facelift and some extra features (TIKA-635)

 * The apache-mime4j dependency of the email message parser was upgraded
   from version 0.6 to 0.7 (TIKA-716). The parser also now accepts a
   MimeConfig object in the ParseContext as configuration (TIKA-640).

Tika 0.10 relies on the following set of major dependencies (generated using
mvn dependency:tree from tika-parsers):

   org.apache.tika:tika-parsers:bundle:0.10
   +- org.apache.tika:tika-core:jar:0.10:compile
   +- edu.ucar:netcdf:jar:4.2-min:compile
   |  \- org.slf4j:slf4j-api:jar:1.5.6:compile
   +- org.apache.james:apache-mime4j-core:jar:0.7:compile
   +- org.apache.james:apache-mime4j-dom:jar:0.7:compile
   +- org.apache.commons:commons-compress:jar:1.1:compile
   +- commons-codec:commons-codec:jar:1.4:compile
   +- org.apache.pdfbox:pdfbox:jar:1.6.0:compile
   |  +- org.apache.pdfbox:fontbox:jar:1.6.0:compile
   |  +- org.apache.pdfbox:jempbox:jar:1.6.0:compile
   |  \- commons-logging:commons-logging:jar:1.1.1:compile
   +- org.bouncycastle:bcmail-jdk15:jar:1.45:compile
   +- org.bouncycastle:bcprov-jdk15:jar:1.45:compile
   +- org.apache.poi:poi:jar:3.8-beta4:compile
   +- org.apache.poi:poi-scratchpad:jar:3.8-beta4:compile
   +- org.apache.poi:poi-ooxml:jar:3.8-beta4:compile
   |  +- org.apache.poi:poi-ooxml-schemas:jar:3.8-beta4:compile
   |  |  \- org.apache.xmlbeans:xmlbeans:jar:2.3.0:compile
   |  \- dom4j:dom4j:jar:1.6.1:compile
   +- org.apache.geronimo.specs:geronimo-stax-api_1.0_spec:jar:1.0.1:compile
   +- org.ccil.cowan.tagsoup:tagsoup:jar:1.2.1:compile
   +- asm:asm:jar:3.1:compile
   +- com.drewnoakes:metadata-extractor:jar:2.4.0-beta-1:compile
   +- de.l3s.boilerpipe:boilerpipe:jar:1.1.0:compile
   +- rome:rome:jar:0.9:compile
      \- jdom:jdom:jar:1.0:compile

The following people have contributed to Tika 0.10 by submitting or commenting
on the issues resolved in this release:

   Alain Viret
   Alex Ott
   Alexander Chow
   Andreas Kemkes
   Andrew Khoury
   Babak Farhang
   Benjamin Douglas
   Benson Margulies
   Chris A. Mattmann
   chris hudson
   Chris Lott
   Cristian Vat
   Curt Arnold
   Cynthia L Wong
   Dave Brosius
   David Benson
   Enrico Donelli
   Erik Hetzner
   Erna de Groot
   Gabriele Columbro
   Gavin
   Geoff Jarrad
   Gregory Kanevsky
   gunter rombauts
   Henning Gross
   Henri Bergius
   Ingo Renner
   Ingo Wiarda
   Izaak Alpert
   Jan H√∏ydahl
   Jens Wilmer
   Jeremy Anderson
   Joseph Vychtrle
   Joshua Turner
   Jukka Zitting
   Julien Nioche
   Karl Heinz Marbaise
   Ken Krugler
   Kostya Gribov
   Luciano Leggieri
   Mads Hansen
   Mark Butler
   Matt Sheppard
   Maxim Valyanskiy
   Michael McCandless
   Michael Pisula
   Murad Shahid
   Nick Burch
   Oleg Tikhonov
   Pablo Queixalos
   Paul Jakubik
   Raimund Merkert
   Rajiv Kumar
   Robert Trickey
   Sami Siren
   samraj
   Selva Ganesan
   Sjoerd Smeets
   Stephen Duncan Jr
   Tran Nam Quang
   Uwe Schindler
   Vitaliy Filippov

See http://s.apache.org/vR for more details on these contributions.


Release 0.9 - 02/13/2011
------------------------

The most notable changes in Tika 0.9 over previous releases are:

 * A critical bugfix preventing metadata from printing to the
   command line when the underlying Parser didn't generate
   XHTML output was fixed. (TIKA-596)

 * The 0.8 version of Tika included a NetCDF jar file that pulled
   in tremendous amounts of redundant dependencies. This has
   been addressed in Tika 0.9 by republishing a minimal NetCDF
   jar and changing Tika to depend on that. (TIKA-556)

 * MIME detection for iWork, and OpenXML documents has been
   improved. (TIKA-533, TIKA-562, TIKA-588)

 * A critical backwards incompatible bug in PDF parsing that
   was introduced in Tika 0.8 has been fixed. (TIKA-548)

 * Support for forked parsing in separate processes was added.
   (TIKA-416)

 * Tika's language identifier now supports the Lithuanian
   language. (TIKA-582)

Tika 0.9 relies on the following set of major dependencies (generated using
mvn dependency:tree from tika-parsers):

   org.apache.tika:tika-parsers:bundle:0.9
   +- org.apache.tika:tika-core:jar:0.9:compile
   +- edu.ucar:netcdf:jar:4.2-min:compile
   |  \- org.slf4j:slf4j-api:jar:1.5.6:compile
   +- commons-httpclient:commons-httpclient:jar:3.1:compile
   |  +- commons-logging:commons-logging:jar:1.1.1:compile (version managed from 1.0.4)
   |  \- commons-codec:commons-codec:jar:1.2:compile
   +- org.apache.james:apache-mime4j:jar:0.6:compile
   +- org.apache.commons:commons-compress:jar:1.1:compile
   +- org.apache.pdfbox:pdfbox:jar:1.4.0:compile
   |  +- org.apache.pdfbox:fontbox:jar:1.4.0:compile
   |  \- org.apache.pdfbox:jempbox:jar:1.4.0:compile
   +- org.bouncycastle:bcmail-jdk15:jar:1.45:compile
   +- org.bouncycastle:bcprov-jdk15:jar:1.45:compile
   +- org.apache.poi:poi:jar:3.7:compile
   +- org.apache.poi:poi-scratchpad:jar:3.7:compile
   +- org.apache.poi:poi-ooxml:jar:3.7:compile
   |  +- org.apache.poi:poi-ooxml-schemas:jar:3.7:compile
   |  |  \- org.apache.xmlbeans:xmlbeans:jar:2.3.0:compile
   |  \- dom4j:dom4j:jar:1.6.1:compile
   +- org.apache.geronimo.specs:geronimo-stax-api_1.0_spec:jar:1.0.1:compile
   +- org.ccil.cowan.tagsoup:tagsoup:jar:1.2:compile
   +- asm:asm:jar:3.1:compile
   +- com.drewnoakes:metadata-extractor:jar:2.4.0-beta-1:compile
   +- de.l3s.boilerpipe:boilerpipe:jar:1.1.0:compile
   +- rome:rome:jar:0.9:compile
      \- jdom:jdom:jar:1.0:compile

The following people have contributed to Tika 0.9 by submitting or commenting
on the issues resolved in this release:

   Alex Skochin
   Alexander Chow
   Antoine L.
   Antoni Mylka
   Benjamin Douglas
   Benson Margulies
   Chris A. Mattmann
   Cristian Vat
   Cyriel Vringer
   David Benson
   Erik Hetzner
   Gabriel Miklos
   Geoff Jarrad
   Jukka Zitting
   Ken Krugler
   Kostya Gribov
   Leszek Piotrowicz
   Martijn van Groningen
   Maxim Valyanskiy
   Michel Tremblay
   Nick Burch
   paul
   Paul Pearcy
   Peter van Raamsdonk
   Piotr Bartosiewicz
   Reinhard Schwab
   Scott Severtson
   Shinsuke Sugaya
   Staffan Olsson
   Steve Kearns
   Tom Klonikowski
   ≈Ωygimantas Medelis

See http://s.apache.org/qi for more details on these contributions.


Release 0.8 - 11/07/2010
------------------------

The most notable changes in Tika 0.8 over previous releases are:

 * Language identification is now dynamically configurable,
   managed via a config file loaded from the classpath. (TIKA-490)

 * Tika now supports parsing Feeds by wrapping the underlying
   Rome library. (TIKA-466)

 * A quick-start guide for Tika parsing was contributed. (TIKA-464)

 * An approach for plumbing through XHTML attributes was added. (TIKA-379)

 * Media type hierarchy information is now taken into account when
   selecting the best parser for a given input document. (TIKA-298)

 * Support for parsing common scientific data formats including netCDF
   and HDF4/5 was added (TIKA-400 and TIKA-399).

 * Unit tests for Windows have been fixed, allowing TestParsers
   to complete. (TIKA-398)

Tika 0.8 relies on the following set of major dependencies (generated using
mvn dependency:tree from tika-parsers):

   org.apache.tika:tika-parsers:bundle:0.8
   +- org.apache.tika:tika-core:jar:0.8:compile
   +- edu.ucar:netcdf:jar:4.2:compile
   |  \- org.slf4j:slf4j-api:jar:1.5.6:compile
   +- commons-httpclient:commons-httpclient:jar:3.1:compile
   |  +- commons-logging:commons-logging:jar:1.1.1:compile (version managed from 1.0.4)
   |  \- commons-codec:commons-codec:jar:1.2:compile
   +- org.apache.commons:commons-compress:jar:1.1:compile
   +- org.apache.pdfbox:pdfbox:jar:1.3.1:compile
   |  +- org.apache.pdfbox:fontbox:jar:1.3.1:compile
   |  \- org.apache.pdfbox:jempbox:jar:1.3.1:compile
   +- org.bouncycastle:bcmail-jdk15:jar:1.45:compile
   +- org.bouncycastle:bcprov-jdk15:jar:1.45:compile
   +- org.apache.poi:poi:jar:3.7:compile
   +- org.apache.poi:poi-scratchpad:jar:3.7:compile
   +- org.apache.poi:poi-ooxml:jar:3.7:compile
   |  +- org.apache.poi:poi-ooxml-schemas:jar:3.7:compile
   |  |  \- org.apache.xmlbeans:xmlbeans:jar:2.3.0:compile
   |  \- dom4j:dom4j:jar:1.6.1:compile
   +- org.apache.geronimo.specs:geronimo-stax-api_1.0_spec:jar:1.0.1:compile
   +- org.ccil.cowan.tagsoup:tagsoup:jar:1.2:compile
   +- asm:asm:jar:3.1:compile
   +- com.drewnoakes:metadata-extractor:jar:2.4.0-beta-1:compile
   +- de.l3s.boilerpipe:boilerpipe:jar:1.1.0:compile
   +- rome:rome:jar:0.9:compile
      \- jdom:jdom:jar:1.0:compile

The following people have contributed to Tika 0.8 by submitting or commenting
on the issues resolved in this release:

   Łukasz Wiktor
   Adam Wilmer
   Alex Baranau
   Alex Ott
   André Ricardo
   Andrey Barhatov
   Andrey Sidorenko
   Antoni Mylka
   Arturo Beltran
   Attila Kir√°ly
   Brad Greenlee
   Bruno Dumon
   Chris A. Mattmann
   Chris Bamford
   Christophe Gourmelon
   Dave Meikle
   David Weekly
   Dmitry Kuzmenko
   Erik Hetzner
   Geoff Jarrad
   Gerd Bremer
   Grant Ingersoll
   Jan H√∏ydahl
   Jean-Philippe Ricard
   Jeremias Maerki
   Joao Garcia
   Jukka Zitting
   Julien Nioche
   Ken Krugler
   Liam O'Boyle
   Mads Hansen
   Marcel May
   Markus Goldbach
   Martijn van Groningen
   Maxim Valyanskiy
   Mike Hays
   Miroslav Pokorny
   Nick Burch
   Otis Gospodnetic
   Peter van Raamsdonk
   Peter Wolanin
   Peter_Lenahan@ibi.com
   Piotr Bartosiewicz
   Radek
   Rajiv Kumar
   Reinhard Schwab
   rick cameron
   Robert Muir
   Sanjeev Rao
   Simon Tyler
   Sjoerd Smeets
   Slavomir Varchula
   Staffan Olsson
   Tom De Leu
   Uwe Schindler
   Victor Kazakov

See http://s.apache.org/ab0 for more details on these contributions.


Release 0.7 - 3/31/2010
-----------------------

The most notable changes in Tika 0.7 over previous releases are:

 * MP3 file parsing was improved, including Channel and SampleRate
   extraction and ID3v2 support (TIKA-368, TIKA-372). Further, audio
   parsing mime detection was also improved for the MIDI format. (TIKA-199)

 * Tika no longer relies on X11 for its RTF parsing functionality. (TIKA-386)

 * A Thread-safe bug in the AutoDetectParser was discovered and
   addressed. (TIKA-374)

 * Upgrade to PDFBox 1.0.0. The new PDFBox version improves PDF parsing
   performance and fixes a number of text extraction issues. (TIKA-380)

The following people have contributed to Tika 0.7 by submitting or commenting
on the issues resolved in this release:

   Adam Rauch
   Benson Margulies
   Brett S.
   Chris A. Mattmann
   Daan de Wit
   Dave Meikle
   Durville
   Ingo Renner
   Jukka Zitting
   Ken Krugler
   Kenny Neal
   Markus Goldbach
   Maxim Valyanskiy
   Nick Burch
   Sami Siren
   Uwe Schindler

See http://tinyurl.com/yklopby for more details on these contributions.


Release 0.6 - 01/20/2010
------------------------

The most notable changes in Tika 0.6 over the previous release are:

 * Mime-type detection for HTML (and all types) has been improved, allowing malformed
   HTML files and those HTML files that require a bit more observed content
   before the type is properly detected, are now correctly identified by
   the AutoDetectParser. (TIKA-327, TIKA-357, TIKA-366, TIKA-367)

 * Tika now has an additional OSGi bundle packaging that includes all the
   required parser libraries. This bundle package makes it easy to use all
   Tika features in an OSGi environment. (TIKA-340, TIKA-342)

 * The Apache POI dependency used for parsing Microsoft Office file formats
   has been upgraded to version 3.6. The most visible improvement in this
   version is the notably reduced ooxml jar file size. The tika-app jar size
   is now down to 15MB from the 25MB in Tika 0.5. (TIKA-353)

 * Handling of character encoding information in input metadata and HTML
   <meta> tags has been improved. When no applicable encoding information is
   available, the encoding is detected by looking at the input data.
   (TIKA-332, TIKA-334, TIKA-335, TIKA-341)

 * Some document types like Excel spreadsheets contain content like
   numbers or formulas whose exact text format depends on the current locale.
   So far Tika has used the platform default locale in such cases, but
   clients can now explicitly specify the locale by passing a Locale instance
   in the parse context. (TIKA-125)

 * The default text output encoding of the tika-app jar is now UTF-8
   when running on Mac OS X. This is because the default encoding used
   by Java is not compatible with the console application in Mac OS X.
   On all other platforms the text output from tika-app still uses
   the platform default encoding. (TIKA-324)

 * A flash video (video/x-flv) parser has been added. (TIKA-328)

 * The handling of Number and Date cell formatting within the Microsoft Excel
   documents has been added. This include currencies, percentages and
   scientific formats. (TIKA-103)

The following people have contributed to Tika 0.6 by submitting or commenting
on the issues resolved in this release:

   Andrzej Bialecki
   Bertrand Delacretaz
   Chris A. Mattmann
   Dave Meikle
   Erik Hetzner
   Felix Meschberger
   Jukka Zitting
   Julien Nioche
   Ken Krugler
   Luke Nezda
   Maxim Valyanskiy
   Niall Pemberton
   Peter Wolanin
   Piotr B.
   Sami Siren
   Yuan-Fang Li

See http://tinyurl.com/yc3dk67 for more details on these contributions.


Release 0.5 - 11/14/2009
------------------------

The most notable changes in Tika 0.5 over the previous release are:

 * Improved RDF/OWL mime detection using both MIME magic as well as
   pattern matching (TIKA-309)

 * An org.apache.tika.Tika facade class has been added to simplify common
   text extraction and type detection use cases. (TIKA-269)

 * A new parse context argument was added to the Parser.parse() method.
   This context map can be used to pass things like a delegate parser or
   other settings to the parsing process. The previous parse() method
   signature has been deprecated and will be removed in Tika 1.0. (TIKA-275)

 * A simple ngram-based language detection mechanism has been added along
   with predefined language profiles for 18 languages. (TIKA-209)

 * The media type registry in Tika was synchronized with the MIME type
   configuration in the Apache HTTP Server. Tika now knows about 1274
   different media types and can detect 672 of those using 927 file
   extension and 280 magic byte patterns. (TIKA-285)

 * Tika now uses the Apache PDFBox version 0.8.0-incubating for parsing PDF
   documents. This version is notably better than the 0.7.3 release used
   earlier. (TIKA-158)

The following people have contributed to Tika 0.5 by submitting or commenting
on the issues resolved in this release:

   Alex Baranov
   Bart Hanssens
   Benson Margulies
   Chris A. Mattmann
   Daan de Wit
   Erik Hetzner
   Frank Hellwig
   Jeff Cadow
   Joachim Zittmayr
   Jukka Zitting
   Julien Nioche
   Ken Krugler
   Maxim Valyanskiy
   MRIT64
   Paul Borgermans
   Piotr B.
   Robert Newson
   Sascha Szott
   Ted Dunning
   Thilo Goetz
   Uwe Schindler
   Yuan-Fang Li

See http://tinyurl.com/yl9prwp for more details on these contributions.


Release 0.4 - 07/14/2009
------------------------

The most notable changes in Tika 0.4 over the previous release are:

  * Tika has been split to three different components for increased
    modularity. The tika-core component contains the key interfaces and
    core functionality of Tika, tika-parsers contains all the adapters
    to external parser libraries, and tika-app bundles everything together
    in a single executable jar file. (TIKA-219)

  * All the three Tika components are packaged as OSGi bundles. (TIKA-228)

  * Tika now uses the new Commons Compress library for improved support
    of compression and packaging formats like gzip, bzip2, tar, cpio,
    ar, zip and jar. (TIKA-204)

  * The memory use of parsing Excel sheets with lots of numbers
    has been considerably reduced. (TIKA-211)

  * The AutoDetectParser now has basic protection against "zip bomb"
    attacks, where a specially crafted input document can expand to
    practically infinite amount of output text. (TIKA-216)

  * The ParsingReader class can now use a thread pool or a more complex
    execution model (java.util.concurrent.Executor) for the background
    parsing task. (TIKA-215)

  * Automatic type detection of text- and XML-based documents has been
    improved. (TIKA-225)

  * Charset detection functionality from the ICU4J library was inlined
    in Tika to avoid the dependency to the large ICU4J jar. (TIKA-229)

  * Composite parsers like the AutoDetectParser now make sure that any
    RuntimeExceptions, IOExceptions or SAXExceptions unrelated to the given
    document stream or content handler are converted to TikaExceptions
    before being passed to the client. (TIKA-198, TIKA-237)

The following people have contributed to Tika 0.4 by submitting or commenting
on the issues resolved in this release:

   Chris A. Mattmann
   Daan de Wit
   Dave Meikle
   David Weekly
   Jeremias Maerki
   Jonathan Koren
   Jukka Zitting
   Karl Heinz Marbaise
   Keith R. Bennett
   Maxim Valyanskiy
   Niall Pemberton
   Robert Burrell Donkin
   Sami Siren
   Siddharth Gargate
   Uwe Schindler

See http://tinyurl.com/mgv9o3 for more details on these contributions.


Release 0.3 - 03/09/2009
------------------------

The most notable changes in Tika 0.3 over the previous release are:

 * Tika now supports mime type glob patterns specified using
   standard JDK 1.4 (and beyond) syntax via the isregex attribute
   on the glob tag. See:

     http://java.sun.com/j2se/1.4.2/docs/api/java/util/regex/Pattern.html

   for more information. (TIKA-194)

 * Tika now supports the Office Open XML format used by
   Microsoft Office 2007. (TIKA-152)

 * All the metadata keys for Microsoft Office document properties are now
   included as constants in the MSOffice interface. Clients should use
   these constants instead of the raw string values to refer to specific
   metadata items. (TIKA-186)

 * Automatic detection of document types in Tika has been improved.
   For example Tika can now detect plain text just by looking at the first
   few bytes of the document. (TIKA-154)

 * Tika now disables the loading of all external entities in XML files
   that it parses as input documents. This improves security and avoids
   problems with potentially broken references. (TIKA-185)

 * Tika now replaces all invalid XML characters in the extracted text
   content with spaces. This prevents problems when output from Tika
   is processed with XML tools. (TIKA-180)

 * The Tika CLI now correctly flushes its buffers when invoked with the
   --text argument. This prevents the end of the text output from being
   lost. (TIKA-179)

 * Embedded text in MIDI files is now extracted. For example many karaoke
   files contain song lyrics embedded as MIDI text.

 * The text content of Microsoft Outlook message files no longer appears as
   multiple copies in the extracted text. (TIKA-197)

 * The ParsingReader class now makes most document metadata available
   already before any of the extracted text is consumed. This makes it
   easier for example to construct Lucene Document instances that contain
   both extracted text and metadata. (TIKA-203)

See http://tinyurl.com/tika-0-3-changes for a list of all changes in Tika 0.3.

The following people have contributed to Tika 0.3 by submitting or commenting
on the issues resolved in this release:

   Andrzej Rusin
   Chris A. Mattmann
   Dave Meikle
   Georger Ara√∫jo
   Guillermo Arribas
   Jonathan Koren
   Jukka Zitting
   Karl Heinz Marbaise
   Kumar Raja Jana
   Paul Borgermans
   Peter Becker
   Sébastien Michel
   Uwe Schindler

See http://tinyurl.com/tika-0-3-contributions for more details on
these contributions.


Release 0.2 - 12/04/2008
------------------------

1.  TIKA-109 - WordParser fails on some Word files (Dave Meikle)

2.  TIKA-105 - Excel parser implementation based on POI's Event API
               (Niall Pemberton)

3.  TIKA-116 - Streaming parser for OpenDocument files (Jukka Zitting)

4.  TIKA-117 - Drop JDOM and Jaxen dependencies (Jukka Zitting)

5.  TIKA-115 - Tika package with all the dependencies (Jukka Zitting)

6.  TIKA-97  - Tika GUI (Jukka Zitting)

7.  TIKA-96  - Tika CLI (Jukka Zitting)

8.  TIKA-112 - Use Commons IO 1.4 (Jukka Zitting)

9.  TIKA-127 - Add support for Visio files (Jukka Zitting)

10. TIKA-129 - node() support for the streaming XPath utility (Jukka Zitting)

11. TIKA-130 - self-or-descendant axis does not match self in streaming XPath
               (Jukka Zitting)

12. TIKA-131 - Lazy XHTML prefix generation (Jukka Zitting)

13. TIKA-128 - HTML parser should produce XHTML SAX events (Jukka Zitting)

14. TIKA-133 - TeeContentHandler constructor should use varargs (Jukka Zitting)

15. TIKA-132 - Refactor Excel extractor to parse per sheet and add
               hyperlink support (Niall Pemberton)

16. TIKA-134 - mvn package does not produce packages for bin/src
               (Karl Heinz Marbaise)

17. TIKA-138 - Ignore HTML style and script content (Jukka Zitting)

18. TIKA-113 - Metadata (such as title) should not be part of content
               (Jukka Zitting)

19. TIKA-139 - Add a composite parser (Jukka Zitting)

20. TIKA-142 - Include application/xhtml+xml as valid mime type for XMLParser
               (mattmann)

21. TIKA-143 - Add ParsingReader (Jukka Zitting)

22. TIKA-144 - Upgrade nekohtml dependency (Jukka Zitting)

23. TIKA-145 - Separate NOTICEs and LICENSEs for binary and source packages
               (Jukka Zitting)

24. TIKA-146 - Upgrade to POI 3.1 (Jukka Zitting)

25. TIKA-99  - Support external parser programs (Jukka Zitting)

26. TIKA-149 - Parser for Zip files (Dave Meikle & Jukka Zitting)

27. TIKA-150 - Parser for tar files (Jukka Zitting)

28. TIKA-151 - Stream compression support (Jukka Zitting)

29. TIKA-156 - Some MIME magic patterns are ignored by MimeTypes
               (Jukka Zitting)

30. TIKA-155 - Java class file parser (Dave Brosius & Jukka Zitting)

31. TIKA-108 - New Tika logos (Yongqian Li & Jukka Zitting)

32. TIKA-120 - Add support for retrieving ID3 tags from MP3 files
               (Dave Meikle & Jukka Zitting)

33. TIKA-54  - Outlook msg parser
               (Rida Benjelloun, Dave Meikle & Jukka Zitting)

34. TIKA-114 - PDFParser : Getting content of the document using
               "writer.ToString ()" , some words are stuck together
               (Dave Meikle)

35. TIKA-161 - Enable PMD reports (Jukka Zitting)

36. TIKA-159 - Add support for parsing basic audio types: wav, aiff, au, midi
               (Sami Siren)

37. TIKA-140 - HTML parser unable to extract text
               (Julien Nioche & Jukka Zitting)

38. TIKA-163 - GUI does not support drag and drop in Gnome or KDE (Dave Meikle)

39. TIKA-166 - Update HTMLParser to parse contents of meta tags (Dave Meikle)

40. TIKA-164 - Upgrade of the nekohtml dependency to 1.9.9 (Jukka Zitting)

41. TIKA-165 - Upgrade of the ICU4J dependency to version 3.8 (Jukka Zitting)

42. TIKA-172 - New Open Document Parser that emits structured XHTML content
               (Uwe Schindler & Jukka Zitting)

43. TIKA-175 - Retrotranslate Tika for use in Java 1.4 environments (Jukka Zitting)

44. TIKA-177 - Improvements to build instruction in README (Chris Hostetter & Jukka Zitting)

45. TIKA-171 - New ContentHandler for plain text output that has no problem with
               missing white space after XHTML block tags (Uwe Schindler & Jukka Zitting)


Release 0.1-incubating - 12/27/2007
-----------------------------------

1. TIKA-5 - Port Metadata Framework from Nutch (mattmann)

2. TIKA-11 - Consolidate test classes into a src/test/java directory tree (mattmann)

3. TIKA-15 - Utils.print does not print a Content having no value (jukka)

4. TIKA-19 - org.apache.tika.TestParsers fails (bdelacretaz)

5. TIKA-16 - Issues with data files used for testing by TestParsers (bdelacretaz)

6. TIKA-14 - MimeTypeUtils.getMimeType() returns the default mime type for
             .odt (Open Office) file (bdelacretaz)

7. TIKA-12 - Add URL capability to MimeTypesUtils (jukka)

8. TIKA-13 - Fix obsolete package names in config.xml (siren)

9. TIKA-10 - Remove MimeInfoException catch clauses and import from TestParsers (siren)

10. TIKA-8 - Replaced the jmimeinfo dependency with a trivial mime type detector (jukka)

11. TIKA-7 - Added the Lius Lite code. Added missing dependencies to POM (jukka)

12. TIKA-18 - "Office" interface should be renamed "MSOffice" (mattmann)

13. TIKA-23 - Decouple Parser from ParserConfig (jukka)

14. TIKA-6 - Port Nutch (or better) MimeType detection system into Tika (J. Charron & mattmann)

15. TIKA-25 - Removed hardcoded reference to C:\oo.xml in OpenOfficeParser (K. Bennett & jukka)

16. TIKA-17 - Need to support URL's for input resources. (K. Bennett & mattmann)

17. TIKA-22 - Remove @author tags from the java source (mattmann)

18. TIKA-21 - Simplified configuration code (jukka)

19. TIKA-17 - Rename all "Lius" classes to be "Tika" classes (jukka)

20. TIKA-30 - Added utility constructors to TikaConfig (K. Bennett & jukka)

21. TIKA-28 - Rename config.xml to tika-config.xml or similar (mattmann)

22. TIKA-26 - Use Map<String, Content> instead of List<Content> (jukka)

23. TIKA-31 - protected Parser.parse(InputStream stream,
              Iterable<Content> contents) (jukka & K. Bennett)

24. TIKA-36 - A convenience method for getting a document's content's text
              would be helpful (K. Bennett & mattmann)

25. TIKA-33 - Stateless parsers (jukka)

26. TIKA-38 - TXTParser adds a space to the content it reads from a file (K. Bennett & ridabenjelloun)

27. TIKA-35 - Extract MsOffice properties, use RereadableInputStream devloped by K. Bennett (ridabenjelloun & K. Bennett)

28. TIKA-39 - Excel parsing improvements (siren & ridabenjelloun)

29. TIKA-34 - Provide a method that will return a default configuration
              (TikaConfig) (K. Bennett & mattmann)

30. TIKA-42 - Content class needs (String, String, String) constructor (K. Bennett)

31. TIKA-43 - Parser interface (jukka)

32. TIKA-47 - Remove TikaLogger (jukka)

33. TIKA-46 - Use Metadata in Parser (jukka & mattmann)

34. TIKA-48 - Merge MS Extractors and Parsers (jukka)

35. TIKA-45 - RereadableInputStream needs to be able to read to
              the end of the original stream on first rewind. (K. Bennett)

36. TIKA-41 - Resource files occur twice in jar file. (jukka)

37. TIKA-49 - Some files have old-style license headers, fixed (Robert Burrell Donkin & bdelacretaz)

38. TIKA-51 - Leftover temp files after running Tika tests, fixed (bdelacretaz)

39. TIKA-40 - Tika needs to support diverse character encodings (jukka)

40. TIKA-55 - ParseUtils.getParser() method variants should have consistent parameter orders
              (K. Bennett)

41. TIKA-52 - RereadableInputStream needs to support not closing the input stream it wraps.
              (K. Bennett via bdelacretaz)

42. TIKA-53 - XHTML SAX events from parsers (jukka)

43. TIKA-57 - Rename org.apache.tika.ms to org.apache.tika.parser.ms (jukka)

44. TIKA-62 - Use TikaConfig.getDefaultConfig() instead of a hardcoded
              config path in TestParsers (jukka)

45. TIKA-58 - Replace jtidy html parser with nekohtml based parser (siren)

46. TIKA-60 - Rename Microsoft parser classes (jukka)

47. TIKA-63 - Avoid multiple passes over the input stream in Microsoft parsers
              (jukka)

48. TIKA-66 - Use Java 5 features in org.apache.tika.mime (jukka)

49. TIKA-56 - Mime type detection fails with upper case file extensions such as "PDF"
             (mattmann)

50. TIKA-65 - Add encode detection support for HTML parser (siren)

51. TIKA-68 - Add dummy parser classes to be used as sentinels (jukka)

52. TIKA-67 - Add an auto-detecting Parser implementation (jukka)

53. TIKA-70 - Better MIME information for the Open Document formats (jukka)

54. TIKA-71 - Remove ParserConfig and ParserFactory (jukka)

55. TIKA-83 - Create a org.apache.tika.sax package for SAX utilities (jukka)

56. TIKA-84 - Add MimeTypes.getMimeType(InputStream) (jukka)

57. TIKA-85 - Add glob patterns from the ASF svn:eol-style documentation (jukka)

58. TIKA-100 - Structured PDF parsing (jukka)

59. TIKA-101 - Improve site and build (mattmann)

60. TIKA-102 - Parser implementations loading a large amount of content
               into a single String could be problematic (Niall Pemberton)

61. TIKA-107 - Remove use of assertions for argument checking (Niall Pemberton)

62. TIKA-104 - Add utility methods to throw IOException with the caused
               intialized (jukka & Niall Pemberton)

63. TIKA-106 - Remove dependency on Jakarta ORO - use JDK 1.4 Regex
               (Niall Pemberton)

64. TIKA-111 - Missing license headers (jukka)

65. TIKA-112 - XMLParser improvement (ridabenjelloun)