Get annotated PDFs from your Sony PRS-T1 ereader
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.



The PRSAnnots project exists to help you get annotated PDF files off of your Sony PRS-T1 or PRS-T2 ereader. These annotated versions include both the freehand and highlighted annotations you made on the reader. To do this, PRSAnnots provides several programs as well as a Python package to help you write your own.

PRSAnnots is alpha software. There are many rough corners to cut yourself on, so please be careful. But rest assured it does very little to the ereader itself, so it's unlikely for that to be broken.


PRSAnnots provides two command-line programs and a simple GUI script to help manage your annotated PDF files.

prsam: The PRS Annotation Manager

The most featureful program, the annotation manager manages a library of PDF files on your reader. It notices when the annotations of a given file have changed, and creates an annotated PDF file on your computer with your up-to-date notes. Here is some basic usage:

prsam init /media/READER

Initialize a library for a reader that is mounted at /media/READER.

prsam add path/to/file.pdf

Copy file.pdf to the reader, and add it to the library.

prsam add path/to/twocolumn.pdf -d 2x2

Cut up each page of twocolumn.pdf into two columns and two rows, so it's easier to read on the reader. When synced back to the computer, the PDF will be reassembled into full pages, with annotations in the correct spots.

prsam import download/file.pdf path/on/computer/

Add to the library file.pdf, which is already in the download directory of the reader. If file.pdf already has annotations, it will be synced to path/on/computer/file.annot.pdf.

prsam import --all directory/

Add all annotated PDFs on the reader to the library. They will all be synced to directory/ on your computer.

prsam sync

Sync all PDFs with annotations changed since the last sync.

prsam --help

Gives information on the various commands you can give prsam. This information is also available online.

prsam-tk: A graphical Annotation Manager

prsam-tk provides a simple graphic interface to the annotation manager. It provides a window where you can select PDF files to add to the library, and it will help initialize a library if none exists. It lacks many of the features in prsam. But it uses the same library, so you can switch between them as you like.

getannotations: The simple, do-it-now approach

If you just want to get an annotated PDF file without messing with a library manager, run this script with the mount point of your ereader as the argument:

getannotations /path/to/ereader

You'll be presented with a list of all PDF files with annotations on the ereader; select the one you want and optionally specify the filename to save it as. The script will exit after creating the annotated PDF file.

Python Package

The heavy-lifting of dealing with the annotated PDF files has been pulled out of the programs above and put into the prsannots package. My hope is that this will allow the creation of better GUIs, for normal people who aren't comfortable living on the command line.

Here are the modules provided:

manager Provides the high-level library management of prsam and prsam-tk.
generic Provides abstractions for the reader, its books, and their annotations. Planned to be useful for all Sony ereaders, but this has not been tested.
prst1 Fills in the PRS-T1 and -T2 specific parts of generic.
pdfannotation Create PDF annotations and add them to PDF files.
pagetext Get the location of each character on a PDF page, and get bounding boxes that contain certain characters.
pdfdice Cut the pages of a PDF file into subpages.
pdfcontent Adds content to PDF pages. Includes a very simple SVG-to- PDF converter.


You'll need some version of Python 2. Thus far, PRSAnnots has been tested on 2.6 and 2.7, but it should also work on 2.5. (Can someone confirm this?) It will not work on 2.4 and earlier, and it is not compatible with Python 3.

In addition to the standard library, it requires the pyPdf (>= 1.13) and PDFMiner libraries. For Debian-based systems, the former is provided by the python-pypdf package. (Note that Debian prior to Wheezy and Ubuntu prior to 11.10 have a too-old version of pyPdf.) Both are available in PyPI. Depending on your installation method, these may be installed for you.

PRSAnnots has been developed and tested in Linux, but it should work on any other operating system that can meet the above requirements. I wouldn't be surprised if there were path or encoding problems on Windows; please let me know and they should be fixable.


The best way to get prsannots is by cloning the git repository:

git clone git://

Alternatively, you can download and unpack the tarball or zipball. All of the scripts may be run without installation, assuming you have satisfied the requirements.

You may install PRSAnnots with the script:

python install

to install it globally on your system. (You may need to be root.) Or, to install it in your home directory:

python install --home=~

Since Python package distribution is a mess, I've tried to make sure will work with distutils, setuptools, distribute, and pip. If you use any of the last three, the dependencies should be installed automatically if they are needed. If you're not sure what distribution systems you have installed, just run It will report at the end if there are missing dependencies for you to install by hand.

Alternative Windows Installer

We've put together a Windows installer that should take care of all of dependencies for you. You need to have Python installed first; here are some appropriate installers. Then run the PRSAnnots installer. It seems to run okay as a normal user, but you'll need to run it as an Administrator in order to get the uninstaller to work. (Note that you may get a warning from your SmartScreen filter that the installer is "not commonly downloaded and not signed." I believe this basically mean that Microsoft has no idea whether the installer is dangerous; it's not saying that something's wrong.) The scripts are given appropriate file extensions and placed in <Python root>\Scripts (C:\Python27\Scripts, if you accept the defaults). If you can help make a better Windows installer, please let us know.


0.2.2 (2014/1/10):
A few bug fixes and an update of the PDFMiner API.
0.2.1 (2012/11/17):
Support for the PRS-T2, and a few minor bug fixes.
0.2 (2012/10/21):
Adds cropping PDFs, previewing cropped and diced files, better handling of PDF annotations, and a number of bug fixes.
0.1.1 (2012/4/29):
Fixes some Windows path issues.
0.1 (2012/4/28):
Initial release.


PRSAnnots is being developed on GitHub. Check out that site for updated versions. Please report bugs and feature requests to the Github bug tracker.


Annotation type:
Both freehand and highlight annotations are supported. Text notes attached to highlighted annotations are supported, but drawings are not. The difficulty is in figuring out how to represent such notes in the PDF file.
Device support:

Right now, only the Sony PRS-T1 and -T2 are supported, because that's what the author has. From what I can tell, other Sony readers have similar schemes for their annotations, but store the information differently. Adding support for these readers is hopefully as simple as producing an altered version of

I don't know how similarly other brands behave, but I'm happy to provide what assistance I can in trying to make them work.

Sync speed:
Syncing PDFs may take a while (tens of seconds for short PDFs with few annotations). This should be sped up, but I haven't figured out where the bottleneck is yet. In the meantime, please be patient.
PDF viewers:
The freehand annotations get written directly on the PDF file, and should be viewable in any PDF viewer. Highlight annotations use the PDF annotation mechanism, which is supported by only some PDF viewers. We ensure that all notations are visible in Acrobat Reader. Other viewers (notably Evince and Xpdf) do not display the highlight annotations. For these, a "fake highlighting" mode can be enabled, adding a text annotation that is more widely supported. (Use prsam config --fake-highlight-on.)
microSD cards:
PRSAnnots only works with the files on the reader's main memory by default. You may tell it to work with the files on the microSD card instead by specifying the mount point of the card instead of the mount point of the reader. For getannotations, this works pretty well. For prsam(-tk), it doesn't. There is no way currently to deal with all the files at once.


The wiki has some information on using PRSAnnots, and the Google Group is a good place to ask questions and make comments.

PRSAnnots has been written (thus far) by Robert Schroll ( Feel free to get in touch with questions and comments.