Fixed #275, #276
PdfTokenizer: Fixed free-after-use after failing to parse content while reading literal tokens
PdfFont: Improved heursitic for word spacing
PdfDifferenceEncoding: Fixed handling of ligatures in AGL character names
PdfXMPPAcket: Make GetDescription() const correct
PdfMemDocument: Fixed upgrade to PDF2.0 in a incremental update

Version 1.0.1

Fixed #265, #264
Fixed several issues related to use of nullable<std::unique_ptr<T>>
XMP: Fixed removing extension from extension bag
XMP: Fix double inserting pdfuaid schema
Text extraction fixes: improved word spacing heuristic, space trimming in ligatures, text in XObject form with non identity matrix
Fixed pkg-config for shared built library

Version 1.0.0

Added support for Type1, CFF and OpenType CFF font subsetting
Added support for Type3 font subsetting (no encoding subsetting, yet)
Implemented full "Adobe Glyph List" specification (https://github.com/adobe-type-tools/agl-specification) for text extraction and glyph selection
Implemented full Type1, TrueType font glyph selection
Many fixes in TrueType legacy subsetting
Added support for PDF/UA preserving when adding annotations/form fields
Improved PDF/A preserving (eg. when PDF/UA level is set as well in the XMP metadata)
Added support for automatically rotating imported images drawn on a PdfPainter when a orientation is detected in the source image codec metadata. Currently supported on Tiff images only (Jpeg exif metadata support pending)
Added high-level signing API, see PdfSignerCMS and TestSignature1 test case
Added support for signing encrypted documents
Added support for preserving encryption among savings
Removed Libidn dependency, default to AESV3R6 encryption
Added support for predefined CMap(s) for improved CJK text extraction
Added much better CMake integration
Added raw rectangle corners retrieval in PdfAnnotation, PdfPage
PdfDocument: Added GetFieldsIterator()
PdfPage: Added GetFieldsIterator()
PdfSignature: Added TryGetPreviousRevision()
PdfCanvas: Added CopyContentsTo()
FileStreamDevice now uses again C stdio for better performance
PdfName:
- Optimized for struct size and construction from string const literal
- Added PdfName operator""_nm(const char*, size_t)
PdfString:
- Optimized for struct size
- Added std::string&& constructor
PdfVariant: Optimized for accessing PdfString, PdfName and PdfReference
Reviewed PdfFileSpec, PdfAction, PdfDestination API and their usage in PdfOutlineItem, PdfOutlines, PdfAnnotationActionBase, PdfAnnotationLink, PdfAnnotationFileAttachment
Reviewed PdfExtension API
Reviewed PdfNameTree, renamed to PdfNameTrees and added PdfNameTree to pick specific trees with typed element
Reviewed PdfExtGState
Reviewed PdfTilingPattern, PdfShadingPattern, PdfFunction: the API now exposes the full capabilities of the PDF specification
PdfEncrypt is now stateless: added PdfEncryptContext as a separate state context and used as argument in PdfEncrypt methods
Set PdfSignature to have correct /ByteRange and /Contents after signing with PoDoFo::SignDocument
Added PdfNames and moved all known names there from PdfName
PdfPageCollection: Methods creating pages now takes PdfPageSize or default inferred size from doc
Fixed PdfStreamedDocument, see #88
Tons of API improvements (see https://github.com/podofo/podofo/blob/master/API-MIGRATION.md "API-MIGRATION.md")
Tons of other bug fixes

Version 0.10.5

Fix #191, #197, #201, #212, #233, #241, #251, #252, #253
PdfParser: Fixed stack overflow parsing documents with many XRef stream updates
PdfFont: Fixed GetBoundingBox() retrival
PdfFontMetricsObject: Fixed reading /FontBBox
PdfEncodingFactory: Fixed parsing of limits with /FirstChar equals to /LastChar
PdfFontMetricsStandard14: Fixed parsing /Widths
PdfMetadata: Fixed missing init ensure for SetAuthor()
PdfTokenizer: Fixed character escaping when reading strings
PdfPageCollection: Fix memory leak in RemovePageAt
Compilation and linking fixes in various conditions
PdfFontManager: Fixed GetOrCreateFontFromBuffer stealing memory
PdfPageCollection: Disable copy/assignment
PdfPage_TextExtraction: Fix decodeString with no font
Fix eating of non-space chars in SplitTextAsLines
Fix FreeType segfault race condition
PdfCheckBox: Fixed IsChecked()
PdfParser: Uncondtionally try to read XRef stream in all PDFs that doesn't have a cross reference section

Version 0.10.4

Fixes #161, #162, #167, #183, merges #157
StandardStreamDevice: Fixed seek() in case of iostream/fstream
PdfWriter: Fixed computing the doc identifier with a wrong buffer
PdfPainter: Fix SetCurrentMatrix() to really update CTM
Fixed compilation in mingw < 12
PdfCIDToGIDMap: Fixed map reading
PdfPainter: Fixed offset on multiline text if text is not left aligned

Version 0.10.3

Fixed big performance regression introduced in 0.10, see #108
Fixed data loss with encrypted documents, see #99
Fixed compilation with VS2022 >= 17.8
Fixed compilation using libxml >= 2.12.0

Version 0.10.2

Security related bugfixes #76, #89, #96
Some compilation and test fixes

Version 0.10.1

Security bugfixes, #66, #67, #69, #70, #71, #72
Rewritten PdfPageCollection for performance
PdfCMapEncoding: Fix parsing some invalid CMap(s) supported by Acrobat
PdfXRefStreamParserObject: Fixed handling of invalid XRef stream entries
Support compilation of the library header (not the library itself) with C++20

Version 0.10.0

PdfPage/PdfAnnotationCollection/PdfAnnotation: Now functions with rect input assume it to be using the canonical coordinate system with no rotation
PdfImage: Added support for CYMK jpeg
PdfParser: Cleaned FindToken2 -> FindTokenBackward
Renamed base source folder -> main
PdfPainter: Revamped API, added full state inspection with current point, added added PdfPainterTextContext to handle text object operations Use it with PdfPainter::Text instance member. Added PdfContentStreamOperators low level interface for PdfPainter
PdfFontMetrics: Added FilePath/FaceIndex for debugging, when available
PdfFont: Renamed GetStringLength() overloads with PdfString to GetEncodedStringLength()
PdfFontManager: Renamed GetFont() -> SearchFont() Re-Added better GetOrCreateFont() from file/buffer
PdfEncrypt: Cleaned factory methods
Added PdfArray::FindAtAs(), PdfArray::FindAtAsSafe(), PdfArray::TryFindAtAs(), PdfArray::GetAtAs(), PdfArray::GetAtAsSafe(), PdfArray::TryGetAtAs()
Added PdfDictionary::FindKeyAsSafe() and PdfDictionary::TryFindKeyAs()
PdfDictionary::AddKeyIndirect/PdfArray::AddKeyIndirect accepts a reference
PdfAnnotation/PdfField API review
PdfDate: Introduced PdfDate::LocalNow() and PdfDate::UtcNow() and default constructor is epoch time instead
Renamed PdfDocument::GetNameTree() -> GetNames()
PdfObject: Flate compress on write objects that have no filters
PdfMemDocument does collect garbage by default when saving
PdfField/PdfAnntation: Fully reworked the hierarchy and added proper fields ownership
Added PdfField::GetParent(), PdfField::GetChildren()
PdfImage: Cleaned/reviewed/fixed SetData()/SetDataRaw()
Renamed PdfPageTree -> PdfPageCollection
Added XMP metadata reading/saving. Added PdfMetadata class
Added text extraction API
Review I/O API: Merged InputDevice/OutputDevice into StreamDevice. New hierarchy deriving StreamDevice
Reviewed PdfObjectStream API: added streaming operations, GetInputStream(), GetOutputStream(). Renamed GetFilteredCopy() -> GetUnwrappedCopy()/UnwrapTo(). They only unwrap non media filters (see PdfImage::DecodeTo for media ones). Added proper copy and move assignment operators
PdfImage: Added DecodeTo(pixelFormat)

Version 0.9.22 (pdfmm)

Fixed serialization of strings with non ASCII PdfDocEncoding characters
Removed PdfLocaleImbue
PdfEncrypt: Removed PdfReference state. Added PdfStatefulEncrypt
Removed use of std::ostringstream. Added efficient outstringstream
Added PdfMath functionalities (matrix transformations and so on)

Version 0.9.21 (pdfmm)

Fixed serialization of UTF-16BE strings
More lenient PdfDate parsing

Version 0.9.20 (pdfmm)

The project is now a C++17 library
Added move semantics for PdfVariant, PdfObject, PdfArray, PdfDictionary
Improved XRefStream support, added support in incremental saves
Many fixes in save incremental object/generation number incrementing
PdfString backed with UTF-8 storage
PdfName backed with UTF-8 storage
Brand new PdfEncoding class with support for both /Encoding and /ToUnicode, more complete Unicode support
Added a PdfDynamicEncoding class that creates a custom CID encoding based on actual used glyphs used
Automatic creation of CIDMap and /ToUnicode
Added PdfSigner class and SignDocument()
Added PdfFontType1Encoding, which support Type1 implicit encoding
Added support for PDF 2.0 UTF-8 strings (untested)
Added indirect iteration for PdfArray/PdfDictionary (see GetIndirectIterator methods)
Added PdfDocument::GetPdfALevel()
Added PDFA preserving writing
Refactored/Reviewed PdfInputDevice: versions that take buffer do not copy it (use istringviewstream)
Added font replacement facility PdfFont::TryCreateFontSubstitute()
Added standard14 fonts embedding, with font programs from PDFium
Reviewed PdfXObject hierarchy, added PdfXObjectForm, PdfXObjectPostScript
Added PdfTextState and use it to compute string widths in PdfFont
Improved PdfDocEncoding to expose conversion utf8 conversion facilities
PdfParser: Support also files with whitespace offset before magic start
PdfObject auto ownership
PdfContents: create on demand /Contents. First create a single stream, after array
Improved IsDirty handling: less dirty bit sets
Added PdfPostScriptTokenizer that as better general support for PostScript
PdfDictionary: Reviewed/convert GetKey -> FindKey
PdfDictionary: Reviewed GetKeyAs methods
Fixed hundreds of warnings. No warnings left in tested builds
FontConfigManager: better handling with custom configurations
Removed PdfMutex. Used std::mutex where necessary
Datatypte: removed PdfDataType::HexString
Removed support for old compilers (MSVC6, hpux, borland, turbo...)
Added better endian swap functions
Removed auto_ptr usage
Removed pdf_int/pdf_uint types
Removed use of pdf_long/long types
Removed use of ptrdiff_t
Removed unistring and ugly string conversion code. Moved to utfcpp
Reviewed PdfObject::GetNumber/PdfObject::GetReal (strict/lenient)
Remove PdfObject inheritance on PdfVariant
Reviewed PdfVariant/PdfObject/PdfArray/PdfDictionary equality/disequality operators
Simplified copyright headers
Object copy constructor must copy also stream
PdfElement: GetDocument(), GetObject() refs
PdfIndirectObjectList: GetParentDocument() -> GetDocument()
Remove PdfSignOutputDevice::SetSignatureSize(size)
Moved inline code to .cpp
Remove comments on overrides
Cleaned CMakeFiles (removed custom Find<>.cmake)
NULL -> nullptr
Use std::shared_ptr in PdfFontMetrics, PdfEncoding in PdfFont
Cleaned PdfFontCache (renamed PdfFontManager), removed font functions from PdfDocument
PdfArray::FindAt() return ref
Simplified license headers
Reviewed PdfPageTree and PdfPageTreeCache API
Reviewed most int vs unsigned indexing
Remove all hungarian notation
Sanitize code style
Removed all const char* and passed to string/string_view
Renamed PdfFontCache -> PdfFontManager
Renamed PdfVecObjects -> PdfIndirectObjectList
Renamed PdfNamesTree -> PdfNameTree
Renamed PdfPagesTree -> PdfPageTree
Renamed PdfPagesTreeCache -> PdfPageTreeCache
PdfObject: removed GetIndirectObject(), MustGetIndirectKey (must use PdfDictionary now)
Remove PdfMemoryManagement.cpp, Removed podofo_new, podofo_free
Add chars type for char array storage/buffering which inerits string
Review pointer vs ref parameters/return types
Removed PdfRefCountedBuffer, PdfRefCountedInputDevice, PdfRefCountedOutputDevice
Removed PdfMemoryManagement and all C style malloc/free usage
Refactored PdfOutputDevice (PdfMemoryOutputDevice/PdfFileOutputDevice/etc.)
Clean PdfError, remove wchar_t
Remove printf, snprintf
Renamed PdfElement-> {PdfDictionaryElement|PdfArrayElement} that respectively have GetDictionary(), GetArray()
Reviewed PdfWriteFlags, added PdfSaveOptions

Old PoDoFo ChangeLog

Version 0.7

Fixed PdfPainter::ArcTo
Fixed crash in FlateDecode and LZWDecode predictor functions
Fixed writing of unicode PdfStrings with brackets (will be escaped now)
Fixed encoding of bytes in PdfName
Added many new unit tests
Added methods to free object memory of objects that are not needed
anymore (These objects will be re-read from disk if they are needed again)
Fixed a crash when appending PDFs
Added unicode support on Win32 (wchar_t* constructors and methods
where apropriate, e.g. file handling)
Fixed a memory leak in PdfStream
Small optimizations in various places
Fixed DCTDecode filter for CMYK images
PdfReference is now immutable, which allows for various optimizations
Fixed PdfInputDevice::Read to return now the correct number of bytes read
Fixed a memory leak in PdfImmediateWriter/PdfStreamedDocument
Fixed several minor parsing issues
Fixed adding text to existing pages
Added Lua5.1 plan file support to podofoimpose
Several fixes in creating XObjects from pages
Fixed a possible crash in PdfNamesTree
Added support for setting colors in PdfAnnotations
Added advanced text drawing support to PdfPainter
Fixed parsing of Type1 fonts
Fixed deletion of PdfAnnations (fixes a memory leak)

Version 0.5

Added support for Embedded Files (annotations & named objects)
Added support for ExtGStates when drawing initially only supports basic transparency
Fixed reading values from nametrees
Added support for named destinations
Fixed a memory leak in PdfDestination::GetPage
Pages do not know their page number inside of the document
Fixed reading PdfActions from PDF files
Moved filter implementations to PdfFiltersPrivate.h
Added PdfFilter::CanEncode and PdfFilter::CanDecode
Simpliefid PoDoFos handling of XRef tables

Version 0.4

PdfImage now supports creating an image stream from a "raw bitmap" which can also be optionally Flate compressed
Added some new Page-related methods to PdfDocument:
- Append - append one document to another
- InsertPages - insert a range of pages from one document to another
- DeletePages - delete a range of pages
Added new tool podofomerge for merging two PDFs together
Added methods to get & set a document's PageMode
Added methods to set a document's FullScreen mode
Added methods to set all the various ViewerPreferences for a document
Added methods to set the document's PageLayout
Added Outline support modified podofoinfo to display them, if present
Added a PdfDestination class
Added PdfNamesTree class for handling the global named objects modified podofoinfo to display them, if present
PdfPainter can draw bezier curves
Added XCode project for building on Mac OS X fixed up conditionals in font code to enable building on MacOSX - but no font loading, YET
Added support for writing linearized PDF
Added support for garbage collection of unused objects
Simplified PdfObject constructor
Improved annotation support
Added support to encode names and various name testcases
Fixed ascent and descent of fonts
Improved PdfImage API
Added support for the creation of file identifiers which makes PoDoFo created PDF files work in more different PDF workflows
PdfImage optionally takes ownership of buffers
Fixed a major parser bug: The parser expected all objects in object streams to be of type dictionary.

Version 0.3

TOTAL revamp of PdfObject & PdfVariant to enable clean/consistent object handling; Added new PdfDocument object - new high level object for reading & writing a PDF doc;
TOTAL revamp of the PdfDocument, PdfWriter & PdfParser relationship PdfDocument is now hub for both reading and writing a document it holds the PdfVecObjects - the others just reference them.
TOTAL revamp of PdfPainter
- now uses PDF coordinates - UserUnits from bottom/left added PdfPainterMM for mm-based coords
- supports user-specified float precision and writes out floats in an optimal manner
- supports "appending" mode for drawing on existing documents
Improved handling of the /Info dict for both reading and writing PDFs;
Added new test app - podofopdfinfo, which will be used to dump metadata, etc. from a PDF;
Added PdfError::DebugMessage() as the official way to write out debugging info; updated all other debugging msgs to use this;
Added PdfError::DebugEnabled() to enable/disable display of debug messages;
Added tracking of file size in PdfParser;
Minor tweak to Linearization handling - to enable getting the status from a doc;
Added getting GetPdfVersionAsString() to PdfWriter;
Added new info/object getting methods to PdfDocument:
- bool IsLinearized();
- size_t FileSize();
- PdfObject* GetStructTreeRoot();
- PdfObject* GetMetadata();
- PdfObject* GetOutlines();
- PdfObject* GetAcroForm().
Updated pdfinfo & podofopdfinfo to call the new PdfDocument methods;
Added PdfDictionary and PdfArray classes;
Added new PdfPagesTree (inside of PdfDocument.cpp) for handling walking a /Pages tree;
Added new GetPageCount() method to PdfDocument;
Modifications to PdfPage to attach it to a PdfDocument & construct from a PdfObject;
Added new Legal and A3 standard page sizes;
Changed page coordinates to be PDF unit-based instead of 1/1000mm;
Changed PdfRect to use PDF units and also use bottom instead of top;
Added ability to go between PdfRect and PdfArray & also get string version of a PdfRect;
Added support for PdfPage to return all the standard boxes (Media, Crop, etc);
Added support for fetching inherited values from pages (eg. boxes, rotation, etc.)
Added more methods to PdfPage:
- GetRotation();
- GetNumAnnots().
Use Exceptions now instead of error codes;
Removed Init from PdfOutputDevice;
Removed Init from PdfParser;
Added LZW Filter support;
Added PdfElement as base class for PdfAction, PdfAnnotation and
PdfPage;
Fixed podofoimageextract, podofotxt2pdf and podofopdfinfo;
Removed PdfSimpleWriter in favour of PdfDocument;
Headers are now installed in includedir/podofo/;
Added a new WatermarkTest demonstrates how to read an existing PDF and draw on each page.

Version 0.2

Improved Documentation;
Added SetInformation for additional error information to PdfError;
Fixed the underline color of text;
Introduced PdfReference class;
Fixed PdfStream::GetFilteredCopy;
Improved handling of DecodeParms for filters;
Fixed PDF files with more than one DecodeParms dictionary in one object;
Added on demand loading of objects to the PdfParser;
Ported to windows by Leonard Rosenthol;
On demand loading of objects is now the default;
Refactored PdfFilter interface so that filters are cached;
Fixed multiple connected XRef tables through /Prev keys in the trailer;
Fixed a number of compiler warnings;
Replaced char*'s with std::strings in a number of classes;
Added std::ostream support to PdfOutputDevice;
Improvements to the ImageExtractor tool;
Refactored PdfVariant so that it is easier to use.

Version 0.1 (11 June 2006)

Initial release.