i-net PDFC - Help - i-net PDFC Core

In strict mode, images can now be detected as 'moved' even if they were moved outside the tolerance area.
Performance improvement for image comparison. The actual impact depends on the type of the images and the available heap memory.

Fixed Bugs

[parser.pdf] Fix for compressed encrypted streams where the encryption is set via the first page references table.
Improved statistical analysis for word spacing and fix for spaces in combined characters.

Improved white space detection. They are now determined heuristically by character distances instead of relying on the font settings. The algorithm is only used in the absence of native white space characters.
[parser.pdf] Added support for linearized PDFs with embedded reference streams

Fixed Bugs

[parser.pdf] Replaced color transparency for Type3-Fonts that caused the following images to be hidden

Added comparison of meta data in images.
Control elements are now compared as type 'text' instead of 'lines and shapes'.

Fixed Bugs

Annotation elements with no visual appearance like a 'popup' are now ignored by filters
Unicode 'combining characters' are now excluded from the white space detection. This improves the white space detection in e.g. Hebrew text which heavly relies on combined characters.
Negative text overhang is recognized and will update the location of the affected text chunk. This solves incorrect word bounds and white spaces.
[parser.pdf] Reduced redundancy when handling clips. This reduces the CPU and memory requirements for documents with very complex clip shapes.
[parser.pdf] Re-use and caching of 'Mask'-images in composite paint operations. This prevents OutOfMemory errors and reduces the storage size of comparisons.
[parser.pdf] Bidirectional or right-to-left text is now correctly mapped to it's originating page instead of page 1.
[parser.pdf] Added fallback to map glyph codes to system font characters. This may help to display special characters event if the mapping in the document is missing.
[parser.pdf] Fix for compressed encrypted streams where the encryption is set via the first page references table.

A new option is available to combine large text differences. This feature drastically reduces the unintended matches in large changes.
PDF Annotations which relate to some text, like strikethrough or link click areas, are now recognized as a property of the text instead of independent objects. As a result the comparison is much more accurate an can handle line breaks in such annotations.
For higher accuracy, the pixel tolerances in the "Verify Identitiy" comparison profile were reduced from 1 to 0.01.
The accuracy of the side-by-side page synchronization (export and GUI) was improved for pages with little to no differences.

PDF and PNG exports can now be configured to automatically select an output page size that exactly fits to the displayed content or pages. This reduces the blank space on exported pages and is the best option to view the PDFs/PNGs on screen devices.
Support for the hyperlink annotation is added to the document comparison.

Fixed Bugs

Public profiles are restored correctly during backup.
Subscript and Superscript are now correctly detected in any line of text.
Detailed comparison of images without clip, e.g. page background images
[parser.pdf] Changed decoding of 'WindAnsiEncoding' to exclude control characters like line feed.

New option for any type of PDF export to not replace external font references by default fonts. This keeps the export more alike similar to the original files even though font references are discoured for PDF.
Modified handling of overlapping text chunks: Fonts without space character will tolerate larger overlaps without splitting the chunk. Leading spaces will be ignored on overlap.
Added strong named assemblies for i-net PDFC .Net Bridge to the SDK
[parser.pdf] It is now optional to replace external font references by PDF native fonts. The replacement will cause a consistent result for any PDF viewer but may cause rendering artifacts.

Fixed Bugs

[parser.pdf] PDF SetColor commands with only one parameter will now set a grey scale color instead of being skipped

Tesseract plugins on Windows have been combined into one plugin
Batch comparison added to the .NET API
PDFCNunit: Support for multiple .Net Frameworks added
Whitespace calculation for small fonts improved
Footer position corrected in case the header is missing.
A font difference will now be detected in case the design font name is the same, but the actually used font differs due to a missing embedded font
JPEG2000 images will be added with original JPEG2000 data to the PDF export
The internal filter SOLVEFALSEREPLACE is now active even for legacy profiles. The filter has been improved and drastically reduces false positives in drawings an shapes with slight position differences due to rounding errors
It is now supported to compare annotations as well
"Across then down" will be used alwaysif a multi column report will be exported into an endless page output format
Performance optimization for high amount of graphic operations with blend mode
Image serialization disparity in the swapping mechanism occurred. This could lead to rare image differences in high load scenarios
NullPointerException occurred. To solve it we improved the handling of defective PDFs (missing PatternType parameter)
Incorrect differences occurred due to a zero-width non-joining character
A text with vertical font was not displayed and the position was wrong
Glyph was missing in rendering due to an inconsistent font entry in the source PDF
NullPointerExceptino occurred if the compared Docx files has no style information
Line height calculation für inconsistent fonts was incorrect
Text in Courier New font not being displayed in comparison result
Some parenthesis not displayed in comparison result because of index bug in font generation which leads to missing characters
Line height calculation was wrong for inconsistent fonts
Wrong difference details reported: Text in two table columns was moved to the same X,Y position
Rendering issues occurred for PDF file where the graphics states are not properly closed
Fixed incorrect white space detection in case the same font is embedded multiple times in a PDF file
ParserConfigPlugin extended for an option to deactivate the cmap from PDF files because a difference was not detected
[parser.pdf] Not embedded fonts which refer to external system fonts will no longer be replaced by PDF-native fonts. This avoids artifacts like overlapping words in a potential PDF export of the result.

Fixed Bugs

The comparison was aborted because of an "Index out of bounds" exception that occurred in case of a style change (e.g. font face) in words with decomposable Unicode characters.
[parser.pdf] Fixed white space issue when converting ligatures to distinct characters
[parser.pdf] 'Alternative algorithm' for unmapped glyphes according to PDF specification implemented. This may improve the comparison result and render quality for incomplete embedded Type0 fonts.

It is supported to password protect the exported PDF file containing the comparison result
Vastly improved drawing speed for document images, leading to a much faster page loading in the server and desktop application
PDF-parser optimized to use stroke instead of fill if possible, resulting in smaller exported PDF files and more precise rendering
Jpeg2000 encoded images supported
The font Dejvu-Sans is used as default font for font embedding in PDF export
Table comparison overhauled to compare the structure as well. Will now show added/removed cells, columns, rows and tables as well
Text location check in strict mode will now only check the position along the baseline of the text and no longer all bounds of each word. This is more robust if the font family or size is changed as well
Header/footer detection in documents with large line heights improved
Recognition of the text styles ''subscript'', ''superscript'' and ''strike through'' for both PDF and DocX
Improved table recognition for tables with header box, "cross tabs" and tables with only inner borders
Link-Verification now checks relative URLs for potential http-Protocol as well
Handling for incorrect PDF files improved, shapes will be ignored in that case
Multi column detection for plain two column layouts and heading + columns scenarios improved
Visibility calculation improved for shapes with mixed fill+stroke colors and redundant elements
'Compute actual visiblity' improved - it will now exclude text on filled background of equal color as well
Word detection and separation improved in case of chunking inside a numerical value
Tesseract OCR:
- The tesseract plugin now clear the tmp files after the usage
DOCX parser:
- Fallback added that is used in case of not supported pattern
- Improvements of the DOCX parser

Fixed Bugs

Performance issues occurred because of language detection and improved rendering features
Different errors occurred due to zero-height text chunks
Because of a rounding error on first page footer detection there could be incorrect differences on the first page
Table was detected on the first page while the header&footer filter was active
Text merge issue in the differences output text occurred. Some white spaces were missing
False differences occurred in case of duplicated objected in the source document
DOCX parser:
- Surplus numbering override levels and auto-color in numberings in tables
Cmap handling for TrueTypeFonts was incorrect. Additional brackets have been inserted between characters
Layout problems with centered tabs occurred
Multiple font issues occurred, e.g. with wrong font information in PDF for text with 0 width
Scroll bug in desktop application occurred at the end of the documents in strict comparison mode
Additional page border in single side export removed
Graphic state was not reset for exported comparison result

ArrayIndexOutOfBoundsException occurred due to style changes in parts of a word with ligature
Recignition of the text styles subscript, superscript and strike through for both PDF and DocX
Performance improvements for pages with >100k shape elements
Comparison Performance improved
It is now possible to detect style changes within a word
Multiple bugs with OCR comparison occurred like wrong detection of text in images
Clipping calculation for complex white shapes and stroked vertical/horizontal lines improved
The pattern filter can now be used to filter single characters or parts of words as well. The limitation to full words is now optional but enabled by default
Added font styles to the fonts of the exported PDF comparison result. This enables the reader to copy&paste the correct color and style for text elements from the exported PDF file
Underline recognition for boxed text improved
Greatly reduced heap usage and export size in case of recurring or tiled images in compared PDF documents
Improved space detection in case of negative kerning and justified paragraphs
Multi-Column filter improved: handling of transparent objects, recognition of row separators, narrow line spacing
Page range filter: It is now possible to remove last page(s) per document from the comparison
Font name normalization for 'TimesNewRomanPSMT' true type font improved
Table filter improved to detect more types of tables in compared PDF documents

i-net PDFC Server

The permission Comparison API has been removed. It is now covered by the combination of the Modify and run comparisons and Web API permissions (requires the Web API Plugin to be enabled).
The plugin name ocr.tesseract.winavx changed to ocr.tesseract.installed
The plugin name ocr.tesseract changed to ocr.tesseract.windows
*Task Planner: * The action "Print" is only supported for job export formats PDF and Images
Rendering Performance improved with a better font metrics caching
New notifications are now directly shown in the web client when the OS notifications are disabled or not possible
SameSite=Lax Attribute set for login cookies
Changes of heap memory, language, country and VM arguments will work with a server restart from the web interface. Before a service restart was required
Note added to configuration property "Restrict Permissions" in dialog "User & Groups" because when global permissions are not restricted then all users have administrative access!
Web API:
- Result parameter for comparison info added (timestamp for create and modify)
- Option to add header information added (for example for authentication) for remote PDF request
- New parameter exportbydiff for the requests result/pdf and result/report. If this parameter is set to true the result will be returned only in case of differences
- Compare Command:
  - can receive additional (or exclusive) JSON file to configure PDF documents, e.g. to set their passwords.
  - JSON may contain Request Headers for HTTP/S request to load PDF files, e.g. to authenticate the request
  - Option to add header information (for example for authentication) for remote PDF request added. (Upload pdf via URL)
Maintenance: It is now possible to restore backups that were not made with the current version. The backup is checked for whether it is compatible with the current version, and if so, it is able to be restored
Tesseract OCR:
- Plugin ocr.tesseract.installed can be configured, e.g. setting the runtime path to Tesseract
- Plugin ocr.tesseract.windows is included only in the Windows and the SDK installer. If using the plugin fails, a Windows System Tesseract Installation is required (see plugin ocr.tesseract.installed)
Task Planner: The file format(PDF,XLSX,ODS,RTF) can now be selected for export type 'report' in the Task Planner module

Web API: NullPointerException occurred when using not existing profile name. The post request handling without filename has been fixed and a better message for not exist profil name/id will be displayed
NullPointerException occurred during "Loading Comparisons" in the i-net PDFC server Maintenance
Fix for content-disposition header when uploading multipart content using the comparison web api
System.Net.WebException: The remote server returned an error - occurred with PDFCNunit
It was not possible to login if a localhost URL was used in Chrome browser to open the web interface
Parameterized streams were evaluated in the wrong order by the comparison Web API
Temporary errors (Extenal visible URL '...' was not validated) occurred during validation of Private Cloud License
Comparison profiles (shared or not) was no longer accessible by the API users (not administrative users)
Exception "Object doesn't support or method 'append'" occurred in web interface during comparison in Internet Explorer 11

i-net PDFC GUI

Comparison result details displayed in a column on the left side of the documents
Configuration dialog added to modify setting like Logging and Tesseract
Native high DPI support, e.g. for Mac Retina displays or screen scaling on Windows

Position of the desktop application was sometimes wrong or it was not displayed on a computer with two or more screens
A high number of difference will no longer block the desktop application for some time once the comparison has finished
Desktop application on macOS does not show the comparison state in the icon, if no comparison is started
The size of the maximized window of the Desktop application was wrong with Java 11.0.8 or Java 15 or newer
MSI installer set i-net PDFC desktop application as default application for DOCX, PDF and TXT

Fixed Bugs

Chinese content not aligned well on right side if Justified was used
Bug with word wrapping occurred that was wider than the field (without a space), surrogate characters and font replacement
NullPointerException occurred due to a missing 'Name' attribute in an optional content group
Performance optimized regarding shapes
Multiple columns not detected because of too close separator lines
Font rendering issue in PDF layers occurred. Error in WebGUI: begin -1, end 6, length 6
Docx Parser:
- Missing text in chunked field elements occurred
- Bugfix for separating words within footnote. The footnote sign shouldn't stay alone
- Numbering with jumps (e.g. from a to c) if empty lines are formatted as numbering
- Word separation at the end of the line was wrong
- Excel table was embedded instead of an icon for the link
- Unnecessary page breaks after a parapraph
- Elements disappeard when using automatic header&footer filter for DOCX documents
- Rendering for symbol and symbol font was incorrect
- Text positioning incorrect with centered tabs
- Rendering of embedded EMF in DOCX was incorrect
- Link (element in structured document tags) in DOCX document was not dispalyed
The rendering of embedded EMF in DOCX was incorrect
Charcaters were drawn apart in the comparison result if fixed size fonts without width table data was used
NoSuchElementException occurred in case of a style change in a word with a ligature
Recognition of stacked white elements on document background color
Text selection/mapping of ligatures and custom glyphes was wrong in exported PDF files
NullPointerException occurred during comparison of DocX documents because the table border could not read correctly
Invalid Character after export occurred, bugfix for embedding PDF descendant fonts with different parent fonts
Corrupt text elements in PDF occurred in case a Form element was re-used on several pages
Incorrectly encoded texts occurred when using a Type1C font: '' was replaced by '•'
Regression in version 20.4: The rendering with Blendmode and CMYK Colorspace changed back to the behaviour in i-net PDFC version 6. (This combination is currently not supported)
Handling of hex-encoded resource entry keys was incorrect, leading to missing resources and images
IllegalStateException: glyph size is not defined occurred because of invalid value in the HEAD table of embedded TrueType fonts
Incorrect 'PageBreakAfter' in DocX files occurred which lead to additional empty pages
Occasional incorrect vertical offset occurred during Docx comparison when using keep-together for paragraphs
Wrong word splits occurred due to invalid intermediate space characters combined with ligatures
'Image Error' occurred in comparison result caused by an incorrect number of color channels in a masked image
SDK: The header footer was not displayed in the comparison result if -highlights headerfooter was set in the command line

Security Fixes

Security Fix: Open Redirect Vulnerability occurred (CVE-2020-28150)
Security Fix: Jetty CVE-2020-27216
- In Eclipse Jetty versions 1.0 thru 9.4.32.v20200930, 10.0.0.alpha1 thru 10.0.0.beta2, and 11.0.0.alpha1 thru 11.0.0.beta2O, on Unix like systems, the system's temporary directory is shared between all users on that system. A collocated user can observe the process of creating a temporary sub directory in the shared temporary directory and race to complete the creation of the temporary subdirectory. If the attacker wins the race then they will have read and write permission to the subdirectory used to unpack web applications, including their WEB-INF/lib jar files and JSP files. If any code is ever executed out of this temporary directory, this can lead to a local privilege escalation vulnerability
Possible JavaScript injections prevented
Security Fix for CVE-2020-13692
- PostgreSQL JDBC Driver (aka PgJDBC) before 42.2.13 allows XXE

PDF export optimized to reduce average size by 10 to 50%
Overhauled 'invisible objects' filter to compute the actual visibility of all elements in the document
New filter added to use the (optional) PDF structure information to improve the comparison result
Handling for stacked sdt* elements in docx improved
Multi-Column filter improved
- Detection of outlined tables and recognition of columns by embedded images
- Column mapping for shapes and images in multi column layout
- Column detection near very large text chunks
Bug in ASCII stream decode. The memory consumption of SMasks was optimized
Optimized Ascent calculation for TT-Fonts in PDF files
Support for languages with word breaks in non latin text improved, for example Chinese
Password protected PDF files: Support for PDF 1.7-extended and PDF 2.0 decryption algorithms added
The differences for image detail comparison improved
Improved block recognition and ordering in rotated text sequences
New fall back to use the glyph code in case of missing unicode mapping

i-net PDFC Server

Login of Members of Windows group Guest is possible
Private Cloud License added
New collector option "Cross compare all files per folder" added to the TaskPlanner batch comparison to cross compare each file in a folder with each reference file in another folder
WebAPI - comparison result parameter (timestamp for create and modify) added to comparison info
Task Planner: New collector option in the TaskPlanner batch comparison to compare each file with each reference file

i-net PFDC Desktop Application

Drag&Drop document to the view area to compare them (in addition to the header buttons)
New control to show a list of all differences with the option to confirm & hide certain differences
Preview of the first page of selected documents
Annotations sometimes not transparent
Comparison result details displayed in a column on the left side of the documents

Fixed Bugs

StackOverflowError with DOCX documents occurred
Invsible paragraphs in table cells occurred where the 'framePr' property was set
Wrong location and clipping of annotations in PDF documents
Referring to file locations in web API calls is limited to administrators
OCR (tesseract) does not work on Debian
Invalid characters occurred in comparison result export. This version contains a bugfix for embedding PDF descendant fonts with different parent fonts
Image was not displayed because of unsupported imagemask
Error "missing initial moveto in path definition" occurred

Security Fixes

Fixed XXE vulnerability for authenticated users with privileges to ad-hoc reporting or remote designer (CVE-2020-12684)
Fixed multiple XSS vulnerabilities (login was not required).
Fixed a path traversal vulnerability which allowed access to files within the installation folder and its sub-folders

Automatic header and footer detection for documents with large line spacing improved
Strict line comparison no longer detects a line size change if the size and orientation of the line is vastly different. Instead a replace or removed/new will be displayed as this is more likely
Hyphen detection in multi column layouts improved
Strict comparison for curves takes into account the overall offset of the shape when checking the curve points
Word boundary detection improved when kerning is simulated by a negative text chunk offset
Boundary detection for inserted/removed sentences and paragraphs improved
Word detection in case of stacked text fragments improved
Plugin to compare PDF files in the order of the document internal structure is available on-demand. This may produce better results for complex documents from the same generator application
Words with no stroke or fill color are now covered by the 'Ignore invisible elements' filter
"Simplify special characters" enabled by default in "Continuous Document" profile
Method ''Engine.Prompt(string,string)'' added to the .NET process bridge

i-net PDFC Server

Data protection:
- New option to removed all unnamed comparisons as soon as all viewers are disconnected
- It is possible to configure the time after which inactive users will be disconnected so that the unnamed comparisons of inactive users will be removed
- New Permission added to define whether users are allowed to create/modify/run comparisons. Without this permission the user can only view published comparisons
- New publishing mode added to publish comparisons only temporarily
Usage of the Web-API for micro services is now available out of the box
Drastically reduced size of exported PDFs in case there are re-used images in the compared files
The "External visible URL" being used for the EHLO command when connecting to an SMTP server
PDF export optimized to reduce average size of the result by 10 to 50%
Any subdomain (like *.example.com) on "Allowed Cross Origins" supported
Web API: Parameter comments added to the PDF export function
Task Planner: It is now supported to set "Delete previous results after X days" for a file action. With this property it is possible for example to delete old backups

Unexpected result with webapi compare request occurred
Multiple values in the property "Other VM Arguments" in configuration dialog "Web Server" were not supported
Timeout occurred during long running backup job in Task Planner
Because of a corrupt file the following exception occurred in Maintenance / Comparisons: java.io.EOFException: Unexpected end of JSON data

Fixed Bugs

Rare false positives occurred for edges of bordering similar shapes, in case the shapes were slightly moved in one of the documents
Kerning handling in RTL bidirectional context
The markup for an underline difference was too long
Exception due to inconsistent path definitions in PDF files
Incorrect unicode characters in case the character names in the font are inconsistent
IndexOutOfBoundsException occurred in comparison core algorithm
Incorrect unicode characters occurred in case the character names in the font are inconsistent
Differences like "Font Style was changed from bold to plain" occurred although the font style was not changed
Memory optimization for PDF files with a lot of shapes
Rare false positives occurred around punctuation
Page rotation was wrong in case of negative angles
Incorrect glyph mapping occurred. It prevented some characters from beeing rendered
Comparison result detail was wrong if it contains .......
Bug with Type1C Font occurred. Difference result details was replaced by ..........
ArrayIndexOutOfBoundsException occurred, incorrectly sorted NameTrees in PDF document will now be handled correctly
Bold style recognition for embedded fonts that duplicate default fonts was wrong
IllegalArgumentException: Color parameter outside of expected range: Alpha
Bold style recognition for embedded fonts that duplicate default fonts was wrong
DocX files: Paragraphs were invisible in table cells where the 'framePr' property was set
White space detection issue in case tabulators are encoded as space with kerning
NoninvertibleTransformException: Determinant is 0 occurred because of a zero-scaled shader clip
NullPointerException occurred in rendering if the PDF explicitly sets the text transformation to zero
IllegalStateException: Fatal error: Index out of text occurred
NullPointerException occurred sporadically during batch comparison
Rare incorrect horizontal offset in print and print preview
Issue in the report generation occurred after changing visibilities
IllegalStateException: Repositioning of a d is not allowed occurred
java.lang.NullPointerException occurred with Docx files
The Windows Tesseract plugin has been improved to better recognise a single number
Difference Marker not displayed because of NullPointerException

Security Fixes

Critical Security Update for Help Plugin (CVE-2020-11431)
Fixed XXE vulnerability for authenticated users with privileges to ad-hoc reporting or remote designer (CVE-2020-12684)
Fixed multiple XSS vulnerabilities (login was not required).
Fixed a path traversal vulnerability which allowed access to files within the installation folder and its sub-folders

Improvement of the differences accumulation - the comparer will no longer return ambiguous groups in case of repeating content
Improved comparison of subscript and superscript text. This reduces false positives due to mixed baselines in tables
Improved support for layers (Optional Content Groups) - covers shapes and images now as well
New text comparison features to detect changes in stylistic spaces and line breaks
Added a fallback for the windows tesseract plugin for the case that both tesseract engines didn't work
Tesseract plugin: error handling and documentationn improved, OCR sample added to the .NET samples in the SDK
White space detection in case of ambiguous kerning data improved

i-net PDFC Server

A restart in the server interface triggers a restart of all nodes if database persistence (MongoDb, Redis) is used
Event log contains a "node" column if it runs with database persistence (MongoDb, Redis)
Let's Encrypt certificate requests now work with multiple server and database persistence
Type1CFont use now a mapping because of difference: Text "–" was replaced by "+-"

Fixed Bugs

Configuration data lost if MongoDB was used for persistence with multiple i-net PDFC server instances
Synchronization of cached user data, groups, task planner, maintenance data between multiple nodes was incorrect if using database persistence (MongoDb, Redis)
There was a random missing image in columns detected by the 'Multi column' filter
Placeholders in error message "Tesseract couldnt find the language file {0} in path {1}." were not replaced
Detection of styles and naming for custom Type1C fonts was incorrect
NoSuchElementException occurred in the continuous difference generator
Default Preview size was wrong in the Desktop Application

Desktop application now shows a notification if the license will expire in a few days
Filter plugin "Link verification" added. With the new filter it is possible to verify external hyperlinks. Also added new difference types for per-document validation errors
Corrupt PDF-character mappings are now passed to the comparer as well. The OCR filter can be used to reconstruct the original character mapping
Invisible elements detection improved, will now check stroke width as well
"Hide pages without differences" now ignores hidden differences types as well
Underline detection improved to not mistake background lines for underlines.
New differences type 'ValidationError' to mark errors that only affect one document, e.g. generated by the "Link verification" plugin

i-net PDFC Server

Notification for low disk space added
Web API will start now a plain text comparison, if the URL contains unknown content instead of PDF, text or image files
The webserver can be configured to send addtional header fields with HTTP responses to, e.g. enforce HSTS or provide custom server information to the web client
New file associations for the batch comparison job in the Task Planner added
The 'Visibilities' settings now enables you to mark up filtered elements and change colors
Event Log views added to Statistics app in report server interface
Multiple LDAP server on authentication supported as fallback
Account id of the user added to the "stored data" view
A master account will be created after a valid login using the master password even if the setting "Create new User" is deactivated
New feature "stay logged in". After login, each user will remain logged in until they log out. After 28 days, they will be automatically logged out. It is also possible to delete user sessions in the "User and Groups" module, if you have the permission to access this module
Login Sessions displayed in the User details
Account id of the user added to the "stored data" view
Task Planner: It is now supported to select a preconfigured datasource for a database series. The user defined JDBC settings are still possible

Fixed Bugs

Configuration data lost if MongoDB was used for persistence with multiple i-net PDFC server instances
Delete option for user profiles was missing
Infinite loop occurred when using i-net PDFC in an iFrame
Invalid differences due to ZERO WIDTH SPACE characters occurred
Fixed white space calculation of the text filter, now equal to the text selection

Bounding box calculation for Type3 fonts improved
Improved performance when running comparisons without filters that have to be executed ahead of the actual comparison (currently OCR and CMAP filter)
Optimized comparison result for consecutively numbered paragraph headings
Switched from word to line comparison for rotated text blocks
Underline detection improved to not mistake background lines for underlines

i-net PDFC Server

New user/group permission to publish comparisons. With this permission it is possible to hide the "Publish" button in the web interface
Generation of needless text selection data in the server improved
Reduced I/O load in the User Manager and comparison meta data
Improved support when OCR cannot initialize
Domain license supports a single host also without reverse DNS lookup
Task Planner: Format options added to the PDF export in the Comparison Job dialog

i-net PDFC SDK

Command line parameter for landscape export added
JsonPresenter added to .Net Process Bridge
Improved performance and memory requirements for very large batch comparisons

Fixed Bugs

OCR Profile added regex filter for "..."
Different styles for type1c fonts occurred without any changes
Problem with duplicate fonts occurred. Some characters were written on top of each other in the comparison result
Vertical lines that are in the background of the compared PDF files are displayed above the text in the comparison result
Missing differences due to incorrectly detected numbering changes
White space calculation of the text filter was wrong
Font style recognition improved
Zoom notification was displayed two times in the Desktop Application
Maintenance: The following error has occurred in Internet Explorer while restoring a backup: "Object doesn't support property or method 'includes'"
Task Planner: NullPointerException occurred in DefaultSetting.setColor
SDK: Invalid characters in the export names of recursive batch jobs occurred

Security Fixes

Critical Security Update for Help Plugin (CVE-2020-11431)
Fixed XXE vulnerability for authenticated users with privileges to ad-hoc reporting or remote designer (CVE-2020-12684)
Fixed multiple XSS vulnerabilities (login was not required).
Fixed a path traversal vulnerability which allowed access to files within the installation folder and its sub-folders

New image parser to read images and ZIP-archives of images
Table filter can now recognize repeating headers even if there are some differences
Improved disctinction between replacements and adds/removes in the comparison result
Annotation issues in the Chrome PDF viewer occurred
Decoding of HTML encoded actual content in PDF files

i-net PDFC Server

Export button added to the web GUI. Using this button it is possible to create the comparison report in the select formats XLS, XLSX, ODS and RTF in addition to PDF
Comparison quotas can now be set per user group and per user as well
Support for SSL certificate in PEM format added
The Task Planner Web API can now export and import tasks (single or multiple). The exported task definitions can be used to manually create one-off task definitions to quickly schedule and execute tasks
Filter settings added to the WebAPI
Language selection for OCR
For all quota values in the configuration: The value for 'unlimited' is now -1 instead of 0. The values will automatically be adjusted by the setup if required
Option "Only pages with differences" is now available
A differences limit can be defined by the user to avoid slowing down the web browser with massive differences. By default this limit is set to 1000
The PDF report of the command line and desktop version can now be generated in the web application as well
Users can administer their own account data and login settings
Search feature in the help improved. The help tree will be displayed and help pages with search results are highlighted in the help tree
The output format for the report can now be configured

Fonts not correct embedded in comparison result export when using 'only pages with differences'

i-net PDFC Desktop Application

The 'Visibilities' tab now enables you to mark up filtered elements
New application-internal help browser with an extended content
Slider values for header/footer filter updated
The export panel allows the save the comparison result as XLS, XLSX, ODS and RTF in addition to PDF

Problem with filechooser on Linux
ClassCastException occurred after repeated start/stop of the comparison
Different sized pages were centered incorrectly

i-net PDFC SDK

Filter settings added to the web API
New API-method for to retrieve the presenter errors, if any
Command Line parameter "-parameters <File>" added, to configurate the command line via a file
Command Line parameters "-loglevel", "-logfile" and "-logmaxerror" added
New optional plugin to export the result as JSON summary using the command line interface
The profile properties LOG_FILE, LOG_LEVEL and MAX_ERRORS_PER_FILE will be ignored since they modify the global configuration. Use the command line switches loglevel, logmaxerror and logfile instead
The profile properties EXPORT_PDF_ALWAYS and the commandline switch exportonlyondifferences will be removed soon use instead the commandline switch exportAlways (Default is exportAlways false) or for SDK the Settings.EXPORT.ONLY_ON_DIFFERENCES
The profile properties CREATE_DIFFIMAGES, CREATE_ORIGIMAGES, CREATE_XORIMAGES will be removed soon use instead the commandline switch x, o, f, s and i or for SDK the Settings.EXPORT
The profile properties IMAGE_SCALE_FACTOR will be ignored since they modify the global configuration. Use the command line switches ImageScaleFactor instead or for SDK the Settings.IMAGE_SCALE_FACTOR
Removed API components that were deprecated with i-net PDFC 4.0
If command line parameter exportOnlyOnDifferences is active, the comparison report will be created only if no difference available
Some properties are depreaced now: CREATE_DIFFIMAGES, CREATE_XORIMAGES, CREATE_ORIGIMAGES, IMAGE_SCALE_FACTOR, EXPORT_PDF_ALWAYS, LOG_LEVEL, LOG_FILE, MAX_ERRORS_PER_FILE

If exportOnlyOnDifferences was set to true, it will not export the comparison report (-e) in case of no differences
WebAPI returns an error, if a PDF was specified that requires a password
Error occurred if the compare.bat was used with more than 9 parameters
Performance leak in batch comparisons occurred. Reason was a potential file leak due to equal fonts in different documents

Fixed Bugs

Fixed hyphen + soft hyphen duplication, improved handling for bogus space characters that were used instead of soft hyphens
Word/line break issues occurred when comparing text files
PDF generation was interrupted in case of unsupported gradient fills
WebAPI shows an error if a document requires a password

New filter plugin to select the pages to be compared
Detailed image comparison added. It can be used to show details of differences in images. The difference details will be displayed only if less than 50% of the image are different
New command line parameter to select the format of the generated report
Images rendered by Type3 fonts will no longer be compared unless only image comparison is active
OTF handling improved
Handling with pattern type1 improved
Basic table filter improved in case of styled cell borders
Table filter improved to recognize extended repeating table headers, including a new difference tolerance
Filter plugin added to ignore specified areas in PDF files from comparison

i-net PDFC Server

Redesigned and streamlined entry page and application frame
There no longer is a /remote context. All remote modules have been moved one level up. So it's now /comparison instead of /remote/comparison for instance
The TaskPlanner jobs will no longer clean up the 'normal' user comparisons.
Web API: Redirected URL was wrong

Web interface don't show error message, if the uploaded document can not be read
Exception: "org.bouncycastle.asn1.pkcs.PrivateKeyInfo cannot be cast to org.bouncycastle.openssl.PEMKeyPair" occurred
WebSocket connections has been closed when using the guest account
If the comparison result contains no differences then PDF export created a PDF file with 0 KB file size
WebAPI: Specified PDF files has been loaded mutliple times for one comparison
WebAPI: It was not possible to delete a comparison using web API after an error has occurred

i-net PDFC Desktop Application

The desktop application now uses the native file chooser of the system. This file chooser provides more features and better performance compared to the custom implementation of previous versions
The 'Visibilities' tab now provides switches to visualize the effects of each filter plugin
Difference popup now shows intial more (10) elements
New export formats for the report: XLS, ODS and RTF

Difference popup was not updated after the visibility settings has been changed
Header / Footer selection was not visible after a comparison. It was necessary to disable and enable the header/footer filter in the visibility settings
"IllegalArgumentException: setSelectedIndex: 0 out of bounds" occurred during start of application if no paper format was available
Kerning issue occurred in comparison result and PDF export
Text in PDF file was not displayed

i-net PDFC SDK

Simplified SDK structure to provide better insight of the product
The C# implementation based on IKVM is deprecated. It was replaced with the ProcessBridge
Missing export settings for .NET process bridge added

Comparison report could not be created if i-net PDFC Server is used
It was not possible to disable Difference Marker using API
The execution of the compare.bat with more than 9 parameters throws an error
It was not possible to disable the difference marker using PDFCVisibility.ALLOUTLINEMARKER_VISIBILITY
java.io.FileNotFoundException occurred while creating the report as XLSX file
Font in the comparison report was wrong if command line parameters -p was used together with -e

Fixed Bugs

Empty characters were partly replaced through not-defined characters
Colors in images were inverted
The location of differences markers for overlines was incorrect
ArrayIndexOutOfBoundsException: Array index out of range: 508 occurred. Only glyphs having valid indexes will be proceeded during AWT font generation
Adobe built-in font processing was incorrect
The reading of Lookup tables for index color space was incorrect
Word wrap problem in table cells occurred due to changes in the split position of a word or word group
False positives could occur if a text column and a drawing were aligned side by side
NullPointerException occurred while comparing two files
Repeating table headers will now marked as difference if the whole table was added or removed
Method setScaleToFitPageHeight was removed from the .NET API. It can be set using property "PDFCVisibility.SCALETOPAGEHEIGHT_VISIBILITY" and pdfComparer.SetSettings(settings)
Command line parameter -exportOnlyOnDifferences was not working
Because of encoding problem characters was not displayed correctly in difference details popup
Error "Start point cannot equal endpoint" occurred
Improved table header recognition and reduced false negatives in case of very similar tables
Fixes and improvements to compare and select rotated text lines
Whitespace detection issue with Chinese characters
"BadPaddingException: Given final block not properly padded" occurred due to a corrupt stream
Shape matching in strict mode improved
ArrayIndexOutOfBoundsException in PDF parser occurred due to zero-length text replacement
Annotation issue in the Chrome PDF viewer
IndexOutOfBoundsException: Index 0 out-of-bounds for length 0 occurred in table filter
ArrayIndexOutOfBoundsException: -1 occurred
Bounding box calculation for Type3 fonts improved
ArrayIndexOutOfBoundsException: 6 at com.sun.pdfview.colorspace.IndexedColor occurred
IllegalPathStateException: missing initial moveto in path definition - occurred due to inconsistent path definitions in PDF files

Security Fixes

Critical Security Update for Help Plugin (CVE-2020-11431)
Fixed XXE vulnerability for authenticated users with privileges to ad-hoc reporting or remote designer (CVE-2020-12684)
Fixed multiple XSS vulnerabilities (login was not required).
Fixed a path traversal vulnerability which allowed access to files within the installation folder and its sub-folders

New filter "Optimize Basic Tables" to greatly improve the comparison of tables with visual borders
- Supports merged cells
- Supports repeating table headers
- Is inactive by default, has to be activated by using a custom comparison profile
New command line switch "exportOnlyOnDifferences" to export PDFs only in case of differences
Differences are now joined per text line, creating a much more compact and readable result
The JAR file inetslf4j.jar was renamed into inetloggeradapter.jar. It also contains an adapter for Commons Logging
Optimized multiple column filter: text in drawings or fragmented page layouts will not longer be mistaken for separate columns
Page alignment when printing 'scale to fit page' with 'collapse equal pages' improved

Fixed Bugs

Several minor fixes and improvements for Shape normalization, Line merging and matching
Some special characters have been displayed incorrect
java.lang.ArrayIndexOutOfBoundsException: 37 - occurred
java.lang.IllegalArgumentException occurred because of wrong end parameter
Incorrectly detected word breaks due to ligatures occurred. Tolerance for overlapping characters in the same word increased
Solved IndexOutOfBoundsException in the header&footer filter that could occur for 4-page documents
Overlapping and redundant text chunks
IllegalArgumentException: Keyframe fractions must be increasing: 1.0
Incorrectly located differences marker for overlines
Some texts written with Type 3 fonts were not displayed
Some comparisons terminated if a lot of huge PDF files compared in a batch job
Clipping error that was caused by subscript or superscript text
Some diacritical marks (akut and grave) were mistaken
IllegalArgumentException: invalid key length: 128 for algorithm version 1 - occurred. No exception will be thrown in the case of contradictory encryption parameters
Difference occurred because one of the compared PDF files contains the same text multiple times on top of each other
Because of wrong spacing in CJK text chunks, it could occur that Chinese text was marked as different although it is the same text
Clip/alignment of italic right aligned text with a large italic angle was incorrect
Another whitespace detection issue occurred

i-net PDFC Server

No cipher suite error with HTTPS connections occurred
The web API redirect to the web application now creates preview images as well and will no longer cause an additional inactive user to be displayed
It was possible to override comparison profil by switch to profile editing

i-net PDFC Desktop Application

Tooltip for header/footer slider was not displayed

i-net PDFC SDK

Image information where lost if the .NET API was used. Instead of the image a Black area was displayed

The class com.inet.pdfc.results.ResultModel now implements the Closable interface. We highly recommend to call close the instance as soon as the result is processed. This will prevent the backing store to proceed swapping the document data and clear all resource immediately
The class com.inet.pdfc.config.PdfSource no longer exposes the method getContent(). Since the PdfSource classes has to be Serializable, it cannot provide access to the actual content as the content may no longer be available after deserialization. It's generally discouraged to access the document model directly. But if required, please use the readDocument() method if the com.inet.pdfc.plugin.DocumentReader class instead. Please note that the document model itself is not public API and may be modified in future versions
The constructor ExceptionData( Throwable ) in class com.inet.pdfc.error.ExceptionData was removed
Tolerance for splitted images added to prevent false image replacement differences
White space detection improved in case of customized word spacing
Setting a Server or Plus license with admin privileges will now activate these licenses for all user accounts on the system. This restores the version 3 behavior for such license types
Properties of Presenters are now defined via a com.inet.pdfc.config.Settings instance instead of dedicated setters. This generalizes the configuration of all presenter types
Removed com.inet.pdfc.results.diffimage.FileDiffImageWriter and com.inet.pdfc.results.diffimage.IDiffImageWriter since they were deprecated as of version 3 and no longer in use

i-net PDFC Server

New Feature TaskPlanner: This feature can be used to schedule comparisons and save or send the results
- New remote/web module and plugin
- Operable per user if TaskPlanner permissions are granted
- Supports single and batch comparisons
- Results can be stored to the file system, to the persistence or sent by email
- Extendable by plugins
- Can be triggered by web API

Because of an tolerance error during difference calculation the following error occurred: The line 'line(xxx)' has been modified: Shape fill color was changed from '(xxx, xxx, xxx)' to '(xxx, xxx, xxx)
Image background was displayed in reverse color in PDF comparison result file
java.util.EmptyStackException occurred due to missing start of Marked Content
"Unknown command: PS" exception occurred
Some currency characters could not be recognized
Wrong font color comparison results occurred
IllegalArgumentException: Raster ByteInterleavedRaster: width = 1024 height = 1024 #numDataElements 1 dataOff[0] = 0 is incompatible with ColorModel
UnsupportedOperationException: shading type 4 with function attribute is not supported
Encoding for buildin fonts was incorrect
Arabic characters in the exported comparison result file not readable because images for character glyphs have used inverted colors
Invalid differences due to incorrectly merged line shapes
IndexOutOfBoundsException in PDFDrawTextCmds occurred
Some characters could not be displayed because of not embedded fonts
Differences in ImageMask objects not found, e.g. in signatures
Outlines of some characters was incorrect
ArrayIndexOutOfBoundsException occurred due to empty text
Characters in composite font have not been displayed due to errors in unicode mapping
IllegalArgumentException: Radius must be greater than zero - occurred because of special case of radial shading
UnsupportedOperationException: ps command: div occurred
The following difference could occur with one page PDF files: "the page ‘page #1’ was added"
Additional differences with Chinese characters occurred
Rotated PDF files are sometimes not displayed correctly
Line swapping detection was wrong

i-net PDFC Desktop Application

It is now easy to change the color of difference markers for PDF documents in the Visibilities tab. Also the color of the blue difference markers between the PDF documents can be changed or they can be disabled
It is now possible to disable the "Comments" in the PDF export of the comparison result
It is now possible to enable/disable the header/footer filter in the Visibilities tab of the Desktop application

i-net PDFC SDK

Command line switch "exportOnlyOnDifferences" added. If it is set, then the PDF comparison result file will be created only in case of differences

The service scripts from the SDK have been renamed to represent their action, e.g. startRecovery.sh
The 'differences' folder creation in batch mode on slow drives could result in missing result files if multiple comparison result files should be written at the same time

Fixed Bugs

i-net PDFC Server

Web GUI application
- The i-net PDFC web application can be used to compare PDF and text files online. It's features and visualization are similar to the desktop application
- Supports all comparison features of i-net PDFC as well as the PDF export
- Comparisons can be stored and published
The Web API component is a RESTful interface to execute and manage comparisons and configurations
- Can execute comparisons and return the result in different formats with a singe HTTP request
- Supports all comparison and management features of the WebGUI
- Can access and manage the comparison of the WebGUI
JUnit test client - a slim client library to compare and verify documents using an i-net PDFC server in the local network
Configuration plugin - to configure all server features, permissions and user settings remotely
Statistics plugin - a plugin to monitor the server status and manage the executed compare jobs
Maintenance plugin - an extension to backup & restore the server state as well as cleaning the user storages
Full support for right-to-left writing context (e.g. Arabic)
Document synchronization was completely overhauled. This affects the interactive GUIs as well as the PDF export. It greatly improves the alignment of differences
Improved PDF export, reduces the filesize
New API sample project at [[https://github.com/i-net-software/pdfc-samples]]
New API methods to get the page numbers for differences
Streamlined comparison pipeline - resulting in a smaller memory requirement and a more responsive comparison
New text parser plugin to compare any type of text files, can be compared to PDF documents as well
The configuration keys MODULES, CONTINUOUS_COMPARE_TYPES are deprecated and replaced by COMPARE_TYPES
The configuration keys NORMALIZERS, CONTINUOUS_FILTERS are deprecated and replaced by FILTER
The configuration key INVISIBLEELEMENTS_HIDE_ROTATION is deprecated and replaced by the filter key HIDEROTATEDTEXT
The configuration key COMPARE_TEXT_STYLES was extended to specify the attributes to compare like SIZE or COLOR

i-net PDFC GUI

Configurations are now stored and managed as comparison profiles. A base set of profiles is provided with the application
The 'Report' tab has been merged into the 'Export & Print' tab
Improved zoom and page controls

Window states will be saved and restored after restart and compare
Label of the disabled "RegEx" filter button was black on a black background

Fixed Bugs

Word recognition in case of overlapping white spaces
The regular expressions were not able to match rotated text lines
IllegalArgumentException: Comparison method violates its general contract
Some characters have not been not displayed because processing of Type1 font was incorrect
Seldomly occurring endless loop in the comparer
Some currency characters could not be recognized
PdfcException: PDF parser error: First item in dictionary must be a /Name.
Property "EXPORT_PDF_ALWAYS" ignored if command line interface was used.
False positive font color differences occurred because of wrong font color comparison.

v3.5.253 (January 09, 2017)

Cleanup of temporary files improved.

v3.5.249 (January 05, 2017)

Help page has not been opened in i-net PDFC GUI on macOS.

v3.5.236 (December 23, 2016)

Comparison of tensor product mesh gradients improved.

v3.5.177 (October 25, 2016)

The DifferencesImagePresenter will only create a folder for the difference images of a pair of PDF files if there were difference.

v3.5.163 (October 11, 2016)

Command line parameter -e does not work together with -c in the same command line.
Space detection by using the kerning data optimized.

v3.5.136 (September 14, 2016)

The number of false differences reduced when comparing drawings.

v3.5.115 (August 24, 2016)

Improvement for redundant elements.
Improvement for false positives in case of text in drawings.

v3.5.86 (July 26, 2016)

Positioning of markers in one page export optimized.

v3.5.81 (July 21, 2016)

Performance improvements, especially for batch comparisons with a lot of PDF files.
It is now possible to use some mixed CID Fonts which actually not according to PDF Reference.
Word boundaries detection further improved.

v3.5.60 (June 30, 2016)

Misspelled word in comparison result report.
"ERROR - unsupported value type: ..." removed from log output.

v3.5.46 (June 16, 2016)

New implementation of the cache to provide sporadic NullPointerException and other exceptions.

v3.5.38 (June 08, 2016)

Open Type Fonts will be identified, even if they are declared as other font type (Type1C).
Parameters to detect word boundaries has been optimized.

v3.2.310 (February 05, 2016)

Processing of white spaces in regular expression filter has been improved.
The text line recognition for rotated text blocks has been improved.
Memory usage of exported PDF comparison result optimized for PDF files with large number of images.
The default configuration will now be founds alongside the PDF files as well.

v3.2.278 (January 04, 2016)

Processing of white spaces in regular expression filter has been improved.

v3.2.233 (November 20, 2015)

CCITT decoding algorithm replaced.

v3.2.218 (November 05, 2015)

JPX / Jpeg2000 encoded images are not supported. Therefore they are replaced with a warning image now.
Don't abort comparison on JPX and JP2 images.
Support for case insensitive comparison added.

v3.2.204 (October 22, 2015)

Add complete filename to the header of the exported PDF file containing comparison result.

v3.2.170 (September 18, 2015)

The i-net PDFC GUI re-displays the red comparison result panel by moving the mouse to the result panel position.

v3.2.139 (August 18, 2015)

New onProgressUpdate API method added for presenters to get progress updates.

v3.2.134 (August 13, 2015)

New onProgressUpdate method added to Presenter to get the current progress.
Parameter /verysilent for setup and uninstall now runs the applications silently without any messages or dialogs.

v3.2.113 (July 23, 2015)

More reporting log messages hidden.

v3.2.107 (July 17, 2015)

New command line parameters added to use all export types, switch visible types and highlights.
Reporting log messages hidden and redirected to the logger of i-net PDFC.
Cache handling optimized to reduce OutOfMemoryErrors with Java 8.

v3.2.64 (June 04, 2015)

MULTI_COLUMN filter improved. Significantly better results for regular three and two column layouts.
Implemented soft-hyphen handling for tagged PDFs.
Color transformation defined in Adobe Application segment will be used for converting cmyk jpeg images.
Optimized white space detection.

v3.1.190 (April 09, 2015)

Default visibility for layers / optional content implemented.

v3.1.176 (March 26, 2015)

The command line tool prints the full absolute path after ''Scanning <filename>''.

v3.1.148 (February 26, 2015)

Handling for '...' in table of contents optimized.
A better error message will be displayed if the PDF files contains an corrupt image.
Optimized internal handling of JPG images. Increases performance for PDF files with large JPG images.
Special cases of text spacing occurred.
Property EXPORT_PDF_ALWAYS added to the config.xml.

v3.1.120 (January 29, 2015)

New GUI control to reset the GUI to it's startup state (clear documents and result).
Support for embedded OpenType/CFF fonts.
Support for very large CID Fonts, Type 2.

v3.1.77 (December 17, 2014)

Extended font name mapping to compare embedded and built-in fonts.

v3.1.57 (November 27, 2014)

When exporting graphical comparison to PDF, PNG or printer (-p commandline), PDF file pairs without differences will no longer cause an export.
New Visibilities option: "Only pages with differences" added to the GUI. Pages which do not have differences (with current visibility filter) will be minimized and will only show their page number. It affects the output in the GUI and the printing, PNG and PDF export of the comparison result. In the command line, option -m or -minimize can be used.
- Command line parameter "-m" added. Only in conjunction with -p: causes pages in the graphical PDF-Export which have no differences to be minimized.
Reduced memory requirements for large PDF files and PDF files with large images.
The file commons-logging.properties removed from the PDFC.jar because it is not used.

v3.0.441 (August 25, 2014)

Word break detection optimized

v3.0.393 (July 08, 2014)

While you scroll through the pages, the page number will be displayed on the right side of the GUI.

v3.0.373 (June 18, 2014)

Modified font engine to return equal font metrics on all operating systems an Java runtime types. This may affect the comparison result if a very low TEXT_ALIGN_RATIO is configured.

v3.0.295 (April 01, 2014)

Configuration property INVISIBLEELEMENTS_HIDE_ROTATION added. With this property it is possible to filter rotated text (vertical text) as well.

v3.0.246 (February 11, 2014)

Version 3.0.235 was delivered as a beta version.

v3.0.214 (January 10, 2014)

New features in the i-net PDFC GUI:
- Export the comparison result in a PDF file or print it on a printer
- Display annotations in the PDF files
- Search within the PDF file(s) is possible
- Continuous Zoom of the PDF files is possible
- Meta Data added to the visibilities filter
- Tabs layout has been redesigned
- Tabs Search, Annotations, Export / Print added
- Comparison result view:
  - With a rigth click on a difference marker it is possible to get detailled information about the difference, copy the text, ignore selected text, ignore selected difference
- Export/Import of the configuration, useful to use it with i-net PDFC command line / API
It is now possible to save the comparison result in PDF file using -p command line parameter
A color difference tolerance for text and shapes can now be defined using the configuration property TOLERANCE_COLOR
Legacy API deprecated, will be removed in Version 4.0
Batchrunner, IReportResult, IPageResult, IDifference is now deprecated
Batchrunner.setLoader is not used anymore, the new threading architecture requires full control over the parser process
Complete new API with extended capabilities
New Filter to normalize visually equals characters added. This is useful especially for ORC or PDF-printer generated PDF documents
Improved Space recognition for the 'Text by pattern' Filter
PS commands gt, ge, lt and le supported
Support for special type of Adobe-encoded Jpeg images
Default language for the comparison report was German. Now it is English.

Fixed Bugs

v3.5.268 (January 24, 2017)

Incorrect space detection in italic text occurred.

v3.5.247 (January 03, 2017)

'Unrecoverable stack overflow' occurred on zero dash stroke.

v3.5.236 (December 23, 2016)

'Space ranges not defined' error occurred.

v3.5.234 (December 21, 2016)

IllegalArgumentException: argument type mismatch - occurred if i-net PDFC GUI was used on Mac OSX and the help button was clicked.

v3.5.207 (November 24, 2016)

Whitespace detection for CID fonts was incorrect.
IllegalArgumentException: Unknown encoding: Identity-H occurred.
NullPointerException occurred during start of the GUI.
Potential endless loop in the memory cleanup thread occurred.

v3.5.194 (November 11, 2016)

Some texts were not displayed in a PDF/A document.

v3.5.192 (November 09, 2016)

NullPointerException occurred during start of the GUI if there was no icon available for a file.
Some difference details has not been displayed in the exported PDF comparison result, if the export properties: Portrait orientation, Only right and Scale to page height were enabled.

v3.5.177 (October 25, 2016)

Some subscript or raised characters were incorrect positioned.

v3.5.163 (October 11, 2016)

Some texts were not displayed in the comparison result.
Rounding errors have been taken as location differences.
Some characters have not been displayed in the comparison result because of changed character to unicode mapping.
If there are empty text elements the comparer ignored any shape in between these empty text elements.

v3.5.137 (September 15, 2016)

RuntimeException: Transformation matrix is non-invertable - occurred.

v3.5.136 (September 14, 2016)

IllegalStateException: PDF stream contains invalid characters after endstream - occurred.
ArgumentException: Parameter is not valid - occurred if .NET API was used.

v3.5.115 (August 24, 2016)

Bugfix for distributed rotated text.

v3.5.107 (August 16, 2016)

Text was not visible in comparison result because the bounding boxes of the text elements have negative widths.

v3.5.81 (July 21, 2016)

Corrupt PDF export occurred in multiple thread environments.
Filled shapes with width or height of 0 were not filtered properly.

v3.5.52 (June 22, 2016)

Small lines were falsely classified as invisible.

v3.5.38 (June 08, 2016)

Comparison result could contain additional white spaces.
Elements which are marked with "Rendering-Mode=Ignore" were not ignored during comparison.
A bug with images with a specific combination of grid and color model occurred.
Wrong text recognition due to corrupt unicode mapping in font occurred.
Problem in PDF parser occurred for TJ command combined with char spacing.
RuntimeException: "Transformation matrix is non-invertable" occurred.
Text was not displayed because of NullPointerException.

v3.2.336 (March 02, 2016)

PDFParseException: "First item in dictionary must be a /Name" occurred if PDF metadata object was corrupt.
Problem with rotated text occurred.
Display problems with some ligatures occurred.
Exception was not thrown to calling function of i-net PDFC API.

v3.2.320 (February 15, 2016)

java.lang.IllegalStateException: java.io.EOFException occurred.

v3.2.310 (February 05, 2016)

IllegalArgumentException: "Dimensions (width=2147483647 height=1) are too large" occurred.
Fonts in the exported PDF comparison result were different to the original PDF files.

v3.2.247 (December 04, 2015)

Type 3 fonts couldn't be displayed.
Very narrow rectangles has been recognized as line.

v3.2.233 (November 20, 2015)

NullPointerException because of invalid ICC profile occurred.

v3.2.218 (November 05, 2015)

IllegalArgumentException: Bad command (15) occurred.
Fix for invisible annotations/reviews.
Fix for hexadecimal encoded font names.
Glyph transformation in CID fonts with postscript layout were incorrect.

v3.2.204 (October 22, 2015)

Text was misplaced, if the compared PDF files were created with dvipdf.

v3.2.201 (October 19, 2015)

Because of threading problems incorrect image comparison results occurred, if .NET API was used.

v3.2.195 (October 13, 2015)

The following warning occurred if command line interface was used: log4j:WARN No appenders could be found for logger.

v3.2.190 (October 08, 2015)

Thread bug could produce invalid text differences if the GUI was used to compare PDF files.

v3.2.182 (September 30, 2015)

Quote characters %%"%% and %%'%% were not compared correctly. Unicode values for characters in Type1 font corrected.

v3.2.170 (September 18, 2015)

ArrayIndexOutOfBoundsException in Latin character mapper has occurred.
IllegalStateException occurred: Could not read from stream.
If the export option "Scale to page height" was enabled, then some difference marker were missing.
ArrayIndexOutOfBoundsException occurred if the sub type of font type 0 has not been correct declared.

v3.2.160 (September 08, 2015)

Some characters like apostrophes not displayed in the comparison result.

v3.2.156 (September 04, 2015)

Pages in the PDF file containing the comparison result are overlapping sometimes.
Some characters having varying code by different encodings were not drawn.

v3.2.153 (September 01, 2015)

If the visibility property "Only pages with differences" was enabled (-m parameter in command line) and the PDF files contained pages with different page widths, then the right page was cut off on the right side of the PDF file containing the comparison result.

v3.2.139 (August 18, 2015)

ArrayIndexOutOfBoundsException occurred if the PDF files contains invisible singular spaces with no width.

v3.2.134 (August 13, 2015)

NullPointerException occurred because the optional content dictionary of the PDF file was empty.

v3.2.113 (July 23, 2015)

Texts that are rendered with positive and negative font sizes were automatically evaluated as different.
RuntimeException: Transformation matrix is non-invertable: AffineTransform ... occurred.

v3.2.107 (July 17, 2015)

Error by decryption process occurred because of a NullPointerException. Padding errors occurred in encrypted PDF files.
Comparison does not show differences if one PDF file contains invalid characters. Command line parameters added.
Comparison result export ignores the checkbox for "Only Left" and "Only Right" if the overlay mode was used in the comparison result view.
Comparison result export does not split resulting pages within PDF file pages, if "Scale to page height" is enabled.
Because of a NullPointerException some text was not displayed in the comparison result.

v3.2.64 (June 04, 2015)

Fixed issue with nested form fields
Fixed non-invertable transformation in PDF export
OutOfMemoryError with occurred if .NET API was used.

v3.2.48 (May 19, 2015)

Image were displayed in inverse colors.

v3.2.44 (May 15, 2015)

Wrong scroll position when switching from overlay to side-by-side mode.
Fixed scroll position list when enabling/disabling 'hide similiar pages'.

v3.2.37 (May 08, 2015)

During search in more complex PDF files, the following exception occurs: IllegalArgumentException: Comparison method violates its general contract!.

v3.2.34 (May 05, 2015)

Decimal encoding of type 3 fonts was incorrect.
InvalidMarkException occurred due to inconsistent true type fonts embedded in pdf documents.
ArrayIndexOutOfBoundsException occurred due to kerning incorrections.
Glyph Mapping for certain CID Type 0 fonts was incorrect.
Whitespace detection for CID fonts was incorrect.
Modified name of folder for image export, fixed zip filenameof PNG export.

v3.1.176 (March 26, 2015)

IllegalStateException occurred: Invalid character code byte count: 1.
IllegalStateException occurred: The checkbox is not initialized.
ArrayIndexOutOfBoundsException in PDFHintTable.
The method SetTypeVisible was missing in the .Net API of the ResultModel object.

v3.1.148 (February 26, 2015)

Differences where displayed in upper / lower case although the words were displayed in uppercase.

v3.1.126 (February 04, 2015)

BufferUnderflowException occurred because of inconsistent CID-Type2 font.
NullPointerException in LZWDecode occurred.

v3.1.120 (January 29, 2015)

Fix for negative font sizes in PDF files.

v3.1.100 (January 09, 2015)

Broken links and images in the online help.
RuntimeException: "Transformation matrix is non-invertable" occurred.

v3.1.82 (December 22, 2014)

IllegalStateException: procedure is already running - has occurred.

v3.1.77 (December 17, 2014)

PDFParseException: Bad character in Hex String occurred because of Unix linebreaks in the PDF XRef table.
NullPointerException occurred if the command line parameter '-m' was used in combination with parameter '-i'.

v3.1.57 (November 27, 2014)

IllegalArgumentException: Bad size for Maxp table occurred.
FileNotFoundException: Could not find file ''c:\windows\system32\inetsrv\DifferenceColors.properties'' occurred if i-net PDFC was running from a service in IIS.
Scrollbars were not displayed correctly if the PDF files contains pages in portrait and landscape orientation.
The GUI has crashed on Mac OS X 10.7.
.Net environment:
- Font issue occurred which could lead to missing PDF difference export files in multi-thread and .Net environments.
- Could not load file or assembly 'IKVM.OpenJDK.Management, Version=7.4.5196.0
- Could not load file or assembly 'IKVM.OpenJDK.XML.XPath, Version=7.4.5196.0
Markers were incorrectly displayed on rotated pages.
OutOfMemoryError occurred if the PDF files contains a lot of images.
PDFParseException: "content reference not found" occurred.
ANSI-Encoding in certain TrueType fonts was wrong.
Annotations were not displayed correctly.

v3.0.462 (September 15, 2014)

When exporting graphical comparison to PDF, PNG or Printer (-p commandline), PDF-pairs without differences will no longer cause an export.

v3.0.445 (August 29, 2014)

Font issue occurred under .NET. An embedded font was not embedded in the PDF file containing the comparison result.
RuntimeException occurred: Error by decryption process

v3.0.441 (August 25, 2014)

The following exception occurred: PDFParseException: Unknown command: ...

v3.0.427 (August 11, 2014)

The following exception occurred during creation of the compariosn report: "com.inet.lib.json.h" with message "Field 'breakPoints' not found.

v3.0.417 (August 01, 2014)

Position and orientation of an annotation was incorrect.

v3.0.414 (July 29, 2014)

RuntimeException: Error by decryption process occurred.
ArrayIndexOutOfBoundsException: 253 occurred.
NumberFormatException: "For input string: "0/.notdef" in Type1 fonts occurred.

v3.0.393 (July 08, 2014)

IllegalArgumentException: Unknown encoding: ZapfDingbatsEncoding occurred. Character encoding ZapfDingbatsEncoding is now supported.
PDFParseException: Data format exception:incorrect header check. PDF files that are encrypted with AES algorithm with default 128 bit keys can be compared now.

v3.0.350 (May 26, 2014)

PDFParseException: xref part is corrupted occurred.

v3.0.337 (May 13, 2014)

ArrayIndexOutOfBoundsException in some Type 7 Shaders occurred.
Character from CIDFontType2 (Japanese fonts) could not be recognized if parent font does not define Unicode mapping.

v3.0.263 (February 28, 2014)

ArrayIndexOutOfBoundsException occurred.
If the .NET API of i-net PDFC was used then the following exception occurred: "A generic error occurred in GDI+".

v3.0.256 (February 21, 2014)

NullPointerException occurred.

v3.0.254 (February 19, 2014)

ClassCastException occurred in strict comparison mode on PDF control elements.
The following exception occurred with the .NET edition: cli.System.Runtime.InteropServices.ExternalException: A generic error occurred in GDI+.

v3.0.246 (February 11, 2014)

Activated / non activated checkbox was not detected as difference.

v3.0.235 (January 31, 2014)

Files with upper case suffix were not compared in batch compare mode.
Specific text differences were not found.
Chinese, Japanese and Korean characters have not been rendered correctly and an ArrayIndexOutOfBoundException has occurred if the PDF files containing such characters.
ProfileDataException with invalid ICC profile occurred.

v3.0.214 (January 10, 2014)

PDFParseException: Stream ended inappropriately occurred

v2.5.155 (November 12, 2013)

Text clipping will now be only active if the filter 'Invisible Elements' is in use.

v2.5.130 (October 18, 2013)

Solid red boxes has been displayed over images in the PDF file.

v2.5.64 (August 13, 2013)

Java and .Net edition of i-net PDFC installed with the same setup.

v2.5.7 (June 17, 2013)

new text filter which accepts plain text and regular expressions to exclude certain text from the comparison
major performance improvements for the internal and GUI rendering process
report can now be generated with the command line tool as well
GUI: cycle through differences the with left and right cursor keys
improved scrolling, doesn't jump when zooming anymore
the gui now logs to a file

v2.2.14 (April 15, 2013)

Performance optimization in parser.
Improvements in memory usage to prevent OutOfMemory errors.
It is now supported to scroll with the left and rigth key through the differences.
The GUI executable now accepts two parameters which will preset the first and second file of the GUI.
CheckBoxes in PDF files are now compared.
UnsupportedOperationException: ps command dup
RuntimeException: invalid reference: null
IllegalArgumentException: Unknown encoding: SymbolSetEncoding
ArrayIndexOutOfBoundsException
BufferUnderflowException
PDFParseException: Data format exception:incorrect header check

v2.1.38 (January 18, 2013)

Comparison report was improved

v2.1.7 (December 18, 2012)

EXPORT of a comparison report as PDF
API class IDifference has a new type "TYPE_STYLE_MODIFIED" to get the distinguish between content and style modifications
NullPointerException in PDFFontEncoding occurred
PDF parser improvements for:
type 6 and 7 shaders are now supported
support for embedded CIDMaps
unicode mapping to character arrays
improved support of CMYK colors

v2.0.161 (November 09, 2012)

Command line parameter -generateinfo added

v2.0.143 (October 22, 2012)

Optimization of the Multi-Column filter for single column scenarios - produces less columns boxes
Image comparison for images with different internal resolution is now possible

v2.0 (August 27, 2012)

Streamlined licensing process, include offline mode
Invisible elements filter
Performance optimization in parser and result view
Several bug fixes to conform strict and loose compare mode

v2.0 Beta (July 31, 2012)

Entirely new continuous comparison engine to find modifications in large documents. This engine focuses on content changes rather than on exact location of each element. It's configurable by several new configuration properties.
Whole new comparison GUI for both comparison modes.
New result visualizations in the GUI for both modes
Visual configuration
Support for PDF form elements added
New paged mode normalizer 'HEADER_FOOTER' to exclude fixed header and footer areas
New module property 'COMPARE_TEXT_STYLES' to compare the text styles as well
New module property 'TEXT_ALIGN_RATIO' to better compensate alignment differences due to different PDF generators
New values for the 'CREATE_DIFFIMAGES' property to create only images for the left or right page
Support for Colored Tiling Patterns added

Fixed Bugs

v2.5.155 (November 12, 2013)

Decryption of PDF files with user permissions access is extended to standard encryption filters with revision 4 and version 4. IllegalArgumentException: Parsen of encrypted pdf files is not supported occurred in this case.

v2.5.135 (October 23, 2013)

IllegalArgumentException: Unknown encoding: StandardEncoding occurred.

v2.5.130 (October 18, 2013)

UnsupportedOperationException: ps command: add

v2.5.113 (October 01, 2013)

"PDFParseException: Unsupported shader type: 4" occurred.

v2.5.98 (September 16, 2013)

UnsupportedClassVersionError occurred if Java API or command line version was used with Java 6.

v2.5.92 (September 10, 2013)

'Illegal Capacity' exception occurred randomly for pages with a large amount of lines.

v2.5.78 (August 27, 2013)

CompareTwoFilesWithCustomHandler - Differences directory was deleted after comparison.
NullPointerException occurred, if a required value in the font description was missing.

v2.5.64 (August 13, 2013)

java.awt.color.CMMException: Invalid profile data occurred.
Difference description contains wrong page number if there is a text replacement with equal bounds but on different pages.
File name in the difference summary on the console was limited.
Text concatenation error occurred sometimes near punctuation marks.

v2.5.7 (June 17, 2013)

several fixes for embedded font, especially Type1C

v2.2.14 (April 15, 2013)

Following exceptions fixed:
- UnsupportedOperationException: ps command dup
- RuntimeException: invalid reference: null
- IllegalArgumentException: Unknown encoding: SymbolSetEncoding
- ArrayIndexOutOfBoundsException
- BufferUnderflowException
- PDFParseException: Data format exception:incorrect header check
It was not possible to compare PDF files on the command line using runPDFC.bat
Scrolling behavior in the GUI was wrong if the compared PDF files had different lengths.

v2.1.38 (January 18, 2013)

PDFParseException: Unknown command: Qq
Differences in watermark of PDF files were not highlighted
Various bugs in the GUI were fixed

v2.1.7 (December 18, 2012)

NullPointerException in PDFFontEncoding occurred

v2.0.161 (November 09, 2012)

BufferUnderflowException solved
IllegalStateException: "invalid dimension of return values" solved
French accents are now displayed correctly
Content of the compared PDF files were not displayed in the i-net PDFC GUI for some landscape pages

v2.0.143 (October 22, 2012)

Fix for scaled font sizes and font colors
Fix for Type3-Fonts with embedded images
Fix for the PDF 'carriage return' command

v2.0 (August 27, 2012)

Several bug fixes to conform strict and loose compare mode

v2.0 Beta (July 31, 2012)

Fix for 'div zero' error in pattern type 1
Fix for PDFs with corrupt kerning data
Fix for ParserException due to incorrect CCITT image decoding

v1.14 (February 29, 2012)

New Image scaling property 'IMAGE_SCALE_FACTOR' to scale all image output
New text module property 'TOLERANCE_TEXT_LOCATION' to verify text identity

v1.10 (November 21, 2011)

API enhancements: Classes in package "diffimage" added.
Command line parameter -x and -r added.

v1.09 (July 29, 2011)

Equal text passages marked as different.

v1.08 (June 10, 2011)

The international format for date-time is now used for error messages.
The batch file runPDFC.bat (or shell script runPDFC.sh) can now be launched from any directory.
The log level "OFF" produces no output.
The log level list is reduced. The following log levels can be used: OFF, ERROR, WARN, INFO, ALL.

v1.07 (May 09, 2011)

API revised
Properties "USE_HD_CACHE_FOR_PAGE_IMAGES" and "USE_MEMORY_CACHE_FOR_PAGE_IMAGES" replaced with "CACHE_FOR_PAGE_IMAGES".
API documentation added
Java code samples added
Documentation enhanced
The default encryption of pdf files is taken into account for additional elements such as functions, color spaces and hint tables.
Image comparision has been improved through corrected image cache.
Java 5 supported.

v1.06 (Feb 23, 2011)

Rectangles in difference images sometimes appeared a few pixels below the point at which they should appear.

v1.05 (Jan 26, 2011)

It is now supported to compare PDF files created with FastReports.

v1.04 (Dec 09, 2010)

Using hard disk or memory cache is now possible.
Improved image comparison: the pixel values of images are now compared too.

v1.01 (Mar 23, 2010)

Improved chart detection and comparison.

v1.0 (Feb 25, 2010)

Initial release.

Fixed Bugs

v1.15 (March 16, 2012)

BufferUnderflowException occurred

v1.14 (February 29, 2012)

UnsupportedOperationException: ps command: roll

v1.13 (February 10, 2012)

PDFParseException: Unsupported function type: 4

v1.12 (February 02, 2012)

Bug: ArrayIndexOutOfBoundsException occurred during processing of some Type1C fonts.
Bug: ArrayIndexOutOfBoundsException occurred by processing masked images.

v1.11 (November 28, 2011)

Bug: "IllegalArgumentException: space ranges not defined" occurred during the file comparison.

v1.10 (November 21, 2011)

Bug: IllegalArgumentException occurred during the file comparison.
Bug: Endless loop occurred during the file comparison.
The height and width of the text difference mark box were not correct calculated for large font size.

v1.09 (July 29, 2011)

Differences in several PDF/A documents were not found.
NumberFormatException: For input string: "xxx" occurred.

v1.08 (June 10, 2011)

Bug in ASCIIHexDecode was fixed. Some texts encoded with this encoding could not be decoded.
Text scaling was corrected in order to calculate the proper height of text-difference marker.
Differences on single PDF pages not found because of an error while reading the PDF comments in the file.

v1.07 (May 09, 2011)

Differences in PDF files were not found, because some characters from CFF fonts were not shown.
Character codes using MacRomanEncoding could not be compared.
Not embedded CID-fonts could not be used to build the page difference images.
Problems with PDF/A files occurred. If the key-length for default encryption is not set, it will be defined on the basis of security handler version.

v1.05 (Jan 26, 2011)

No differences were found between PDF files because of error by font reading.
ClassCastException occurred during the comparison of PDF files.
The width array length during string processing is now limited by the string length to avoid IndexOutOfBoundsExceptions.

v1.03 (Nov 30, 2010)

Fixed a problem regarding the reading of compressed objects and xref streams.
Fixed a bug with default width for CID-fonts type 0.
Fixed a bug with ToUnicode font map ranges for CID-fonts.

v1.02 (Jul 08, 2010)

Fixed a bug with inlined DCT-encoded images in a PDF.
Fixed a bug regarding Unicode special characters not being read correctly.
Fixed a problem regarding the reading of embedded True Type text.

v1.01 (Mar 23, 2010)

Fixed problems when identical shapes occurred more than once.

i-net PDFC CoreActive

Migration Information

Migration to Version 24.10

Migration to Version 23.4

Migration to Version 20.0

Changes in program structure

Behavioural changes

Migration to Version 5.0

Changelog Information

Changes in version 24.10

Fixed Bugs

Changes in version 24.4

Fixed Bugs

Changes in version 23.10

Fixed Bugs

Changes in version 23.4

Changes in version 22.10

Fixed Bugs

Changes in version 22.4

Fixed Bugs

Changes in version 21.10

Fixed Bugs

Changes in version 21.4

Fixed Bugs

Changes in version 20.10

i-net PDFC Server

i-net PDFC GUI

Fixed Bugs

Security Fixes

Changes in version 20.4

i-net PDFC Server

i-net PFDC Desktop Application

Fixed Bugs

Security Fixes

Changes in version 6.2

i-net PDFC Server

Fixed Bugs

Security Fixes

Changes in version 6.1

i-net PDFC Server

Fixed Bugs

Changes in version 6.0

i-net PDFC Server

Fixed Bugs

Changes in version 5.1

i-net PDFC Server

i-net PDFC SDK

Fixed Bugs

Security Fixes

Changes in version 5.0

i-net PDFC Server

i-net PDFC Desktop Application

i-net PDFC SDK

Fixed Bugs

Changes in version 4.3

i-net PDFC Server

i-net PDFC Desktop Application

i-net PDFC SDK

Fixed Bugs

Security Fixes

Changes in version 4.2

Fixed Bugs

i-net PDFC Server

i-net PDFC Desktop Application

i-net PDFC SDK

Changes in version 4.1

i-net PDFC Server

i-net PDFC Desktop Application

i-net PDFC SDK

Fixed Bugs

Changes in version 4.0

i-net PDFC Server

i-net PDFC GUI

Fixed Bugs

Changes in version 3.0

v3.5.253 (January 09, 2017)

v3.5.249 (January 05, 2017)

v3.5.236 (December 23, 2016)

v3.5.177 (October 25, 2016)

v3.5.163 (October 11, 2016)