Overview
The PDF service is a NIRVA external service that provides functionality for processing PDF files:
Current version 2.28
Downloads
- Crop
- Clip
- Move
- Copy
- Merge
- Concatenation
- Text extraction
- Text insert
- Rotation
Current version 2.28
Downloads
Updates
Version 2.28
Version 2.27
Version 2.26
Version 2.25
Version 2.24
Version 2.23
Version 2.22
Version 2.21
Version 2.20
Version 2.19
Version 2.18
Version 2.17
Version 2.16
Version 2.15
Version 2.14
Version 2.13
Version 2.12
Version 2.11
Version 2.10
Version 2.09
Version 2.08
Version 2.07
Version 2.06
Version 2.05
Version 2.04
Version 2.03
Version 2.02
Version 2.01
Version 2.00
- Compatibility with Nirva server 4.8.000: The management of variables has changed with the server 4.8.000. It is possible now to define the character used to recognize a variable in a Nirva command. This can be changed at application level (in application.dsc file) and at command level (parameter NV_VAR_IDENT). If you don't change the default character (#) you don't need to update the service. If you change it it is recommended to update the service otherwise some configuration screens may not work properly.
Version 2.27
- Technical bug correction in GET_TEXT.
Version 2.26
- Technical bug correction in GET_TEXT and SET_TEXT. The service was not detecting properly the end of an inline image data.
- Linux compatibility: we compiled with older versions of glibc (2.11.3) in order to be compatible with more linux versions.
Version 2.25
- Technical bug corrections in GET_TEXT command. The service was indefinitely looping when there was an indirect ressource in a Xobject.
Version 2.24
- Technical bug corrections GET_TEXT command.
Version 2.23
- Technical bug corrections in SET_TEXT and GET_TEXT commands.
Version 2.22
- Technical bug corrections in SET_TEXT and GET_TEXT commands.
Version 2.21
- Bug correction in the GET_TEXT command: the command was not accessing local xobject resources correctly. If there was a local resource with the same name than a global resource, the global resource was used instead of the local one. This may produce some infinite loop in rare situations.
Version 2.20
- Minor bug correction for linux versions: the internal time functions were not thread safe. We didn't see any impact until now for this bug.
Version 2.19
- GET_TEXT: 2 technical bug corrections: Get text, some text parameters were not correctly saved and restored when running an xobject (q and Q pdf commands). In objects streams, the offset table may have pairs separated by a carriage return instead of a space.
- The GET_TEXT command now returns an error when the requested page doesn't exist.
Version 2.18
- Get_text/Set_text: bug correction. Some commands were producing blank pages when there was some space before a stream data.
Version 2.17
- Very technical bug. Occurs in very rare cases. May produce a blank page. Cause: a buffer that is not large enough.
- Get_text: this is also a very technical issue that occurs when a stream has a double encoding. The first decoder was correctly called but one byte was missing before sending the result to the second decoder.
Version 2.16
- It was not possible to open files when the startxref line was not 2000 characters from the end of the file. This was producing an error when trying to open the pdf file.
- The GET_TEXT command is now able to decode complete unicode data. In order to decode unicode data, Nirva must be started in unicode mode (-u option).
Version 2.15
- CONCAT command: bug correction when the number of files to concatenate is too big
Version 2.14
- GET_TEXT bug correction in space detection.
Version 2.13
- GET_TEXT doesn't detect all unicode characters. In some situations, the GET_TEXT command cannot convert unicode characters.
- GET_TEXT doesn't work after a MERGE command. The GET_TEXT command doesn't work fine when the text is inside an overlay (after a merge command).
Version 2.12
- Get_text: bug in X sorting algorithm of the text blocks. This can produce a server crash under linux (error under windows)
Version 2.11
- Get_text: The GET_TEXT command doesn't detect when the same text is written 2 times at the same place on some PDF files
- Stream object: when the source pdf contains several objects in the same stream object, the output pdf is bad formed (second bug on same issue.
Version 2.10
- The pdf version is changed with some commands to 1.4. The PDF service should keep the version of the original file
Version 2.09
- Set_text: doesn't work properly with some PDFs using transparency.
Version 2.08
- Get_text: problem in x coordinates so in some cases, the order of recognized character was bad.
- Xobject name generation: when we use 2 times a merge, there is a possibility to have 2 Xobjects with the same name. This may result in an infinite loop when trying to use the get_text command on such file.
- Null pointer: We added a test for a null pointer in case that can occur only with a bad formed pdf.
- Stream object: when the source pdf contains several objects in the same stream object, the output pdf is bad formed.
Version 2.07
- In some situations, the GET_TEXT command correctly gets characters from pdf file but inverts their order. For example when reading an address, the zip code is read after the town while it's in the correct order in the pdf file.
- The issue is corrected in version 2.07. The order of characters in the pdf file may differ of what is displayed on the screen. We have added a character sort based on character position for all characters of a single line.
Version 2.06
- Linux 64 bits version (x64).
- Linux 32 bits version has been recompiled using libstd6 instead of libstd5.
Version 2.05
- The SET_TEXT command doesn't work in the version 2.04. There is no error message but the text is not inserted in the PDF.
- Windows 64 bits version (x64).
Version 2.04
- Bug correction in GET_PAGE command that produced an exception with some PDF files.
Version 2.03
- Bug correction in SET_TEXT when using an external font on unix platforms. There was an issue with the font name detection turning the font name into uppercase.
Version 2.02
- In some situations the GET_TEXT command was not detecting new lines.
Version 2.01
- In the SET_TEXT command, the space between lines is different for the first line. Also some lines may be truncated at the end.
Version 2.00
- Removing license control.