SetaPDF-Extractor

Rev. 1508 to 1597

SetaPDF-Extractor Component

Feature

• Added setKeepIntersectingSpaces() and getKeepIntersectingSpaces() methods to control behavior of overlapping white signs.

Change

• Moved SetaPDF_Extractor_TextItem::_getFontBBoxVector() to SetaPDF_Core_Font class in Core.

Bugfix

• Do not process text if no font is registered in graphic state in all strategies extending the Glpyh strategy.

Tweak

• Optimized some if-logic in the Plain strategy.
• Added SetaPDF_Extractor_Sorter_Baseline::(set|get)BaselineThreshold() methods to allow modification of the threshold value.
• Optimized handling of inline images.
• Optimized interpretation of obsolete empty strings (recurring spaces) in all strategies that inherit the Glyph strategy (all but the Plain text strategy).

SetaPDF-Core Component

Feature

• Added TempStream writer class which uses buffering and a temporary stream internally. Both combined result in best CPU and memory usage.
• Updated Http writer to extend the new TempStream writer class to reduce CPU and memory usage.
• Added support for multiple quadrilaterals in setQuadPoints() method in all annoation classes supporting quad points.
• Added SetaPDF_Core_Canvas_Draw::polygon() method.

Bugfix

• Fixed fallback logic in Flate filter class.
• Ignore invalid references in AcroForm data and handle them as they would not exist.
• Allow DateTimeInterface instances in set*Date() methods in Annotation classes.
• Handle missing encoding information in Type0 font instead of throwing a fatal error.
• Handle empty Keyword array in XmpHelper class.
• Resolve values of FontBBox array in several font classes (instead of accessing it directly).
• Fixed access to undefined annotation appearance streams.
• Added compatibility for PHP 8.
• Handle invalid /Parent values in form field structures.

Tweak

• Recreate FirstChar or LastChar entries in faulty Type1 and TrueType fonts (if possible).
• Added support for DateTimeInterface in SetaPDF_Core_DataStructure_Date class.
• Return color space instance in SetaPDF_Core_TransparencyGroup::getColorSpace() instead of raw value.
• Fixed too many recursions in gif frame handling.
• Micro optimization in pdf parser classes.
• Ignore empty strings or streams in AES encrypted documents.
• Moved SetaPDF_Core_Font::getFontBBox($recalculate=true) to own method (recalculateFontBBox).
• Optimized performance of implementations of SetaPDF_Core_Font_FontInterface::getFontBBox().
• Optimized performance of SetaPDF_Core_Geometry Rectangle and Vector.
• Ensure that SetaPDF_Core_Document_Info::getDictionary(true) resolves to, or creates a dictionary if needed.
• Speed optimization for parsing of JPEG files.
• Optimized SetaPDF_Core_Document_Catalog::setOpenAction() to accept "null" as parameter which removes the OpenAction entry from the catalog dictionary.
• Optimized handling of custom lists in SetaPDF_Core_Font_Glyph_List.
• Optimized handling of the /Difference array in Simple and Type3 fonts.
• Reduce extending page boundaries to their intersection with the media box.
• Optimized SetaPDF_Core_Type_Dictionary_Helper::resolveAttribute() to ignore not found references.

Demos

• Completely rewritten with a structured GUI and hosted on GitHub now.