Skip to content
CLI Epub Reader
Branch: master
Clone or download
Latest commit ad444d8 Nov 17, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
LICENSE Create LICENSE Mar 22, 2019 Update Nov 17, 2019 Add separate logic for <pre> tags in HTMLtoLines Nov 17, 2019
screenshot.png [ImgBot] Optimize images Mar 22, 2019 Edit meta. May 5, 2019

$ epr License: MIT


Terminal/CLI Epub reader written in Python 3.7 with features:

  • Remembers last read file (just run epr without any argument)
  • Remembers last reading state for each file (per file saved state written to $HOME/.config/epr/config or $HOME/.epr respectively depending on availability)
  • Adjustable text area width
  • Adaptive to terminal resize
  • Supports EPUB3 (no audio support)
  • Secondary vim-like bindings
  • Supports opening images
  • Dark/Light colorscheme (depends on terminal color capability)


  • Minimum width: 22 cols
  • Supports regex search only
  • Supports only horizontal left-to-right text
  • Doesn't support hyperlinks
  • Some known issues mentioned below
  • Customizing keybindings & colorscheme done inside the source code itself (no separated config file)


  • curses (Linux) or windows-curses (Windows)


Clone this repo, tweak as much as you see fit, rename it to epr, make it executable and put it somewhere in PATH. Or simply (be careful if you already have package named epr installed):

$ pip3 install git+

Or via chocolatey (maintained by cybercatgurrl):

$ choco install epr

Quickly Read from History

Rather than invoking epr /path/to/file each time you are going to read, you might find it easier to do just epr STRINGS.


$ epr dumas count mont

If STRINGS is not any file, epr will choose from reading history, best matched path/to/file with those STRINGS. So, the more STRINGS given the more accurate it will find.

Run epr -r to show list of all reading history.

Opening an Image

Just hit o when [IMG:n] (n is any number) comes up on a page. If there's only one of those, it will automatically open the image using viewer, but if there are more than one, cursor will appear to help you choose which image then press RET to open it and q to cancel.


This is just a simple colorscheme involving foreground dan background color only, no syntax highlighting. You can cycle color between default terminal color, dark or light respectively by pressing c. You can also switch color to default, dark or light by pressing 0c, 1c or 2c respectively.

Customizing dark/light colorscheme needs to be done inside the source code by editing these lines:

# colorscheme
# DARK/LIGHT = (fg, bg)
# -1 is default terminal fg/bg
DARK = (252, 235)
LIGHT = (239, 223)

To see available values assigned to colors, you can run this one-liner on bash:

$ i=0; for j in {1..16}; do for k in {1..16}; do printf "\e[1;48;05;${i}m %03d \e[0m" $i; i=$((i+1)); done; echo; done

Vanilla or Markdown?


If you'd like to read epub in markdown format, which requires additional dependency: html2text, checkout markdown branch of this repo or simply:

$ pip3 install git+

Useful when you read more nonfiction reference epub (like manual or documentation) than fiction one.

Reading Progress Indicator

If you need reading progress indicator and you don't mind a little more startup time, checkout epy. It's just my fork of this epr, but with one extra feature: Reading Progress Percentage.


    epr             read last epub
    epr EPUBFILE    read EPUBFILE
    epr STRINGS     read matched STRINGS from history
    epr NUMBER      read file from history
                    with associated NUMBER

    -r              print reading history
    -d              dump epub
    -h, --help      print short, long help

Key Binding:
    Help            : ?
    Quit            : q
    Scroll down     : DOWN      j
    Scroll up       : UP        k
    Page down       : PGDN      RIGHT   SPC
    Page up         : PGUP      LEFT
    Next chapter    : n
    Prev chapter    : p
    Beginning of ch : HOME      g
    End of ch       : END       G
    Open image      : o
    Search          : /
    Next Occurence  : n
    Prev Occurence  : N
    Toggle width    : =
    Set width       : [count]=
    Shrink          : -
    Enlarge         : +
    ToC             : TAB       t
    Metadata        : m
    Switch colorsch : [default=0, dark=1, light=2]c

Known Issues

  1. Search function can't find occurences that span across multiple lines

    Only capable of finding pattern that span inside a single line, not sentence. So works more effectively for finding word or letter rather than long phrase or sentence.

    As workarounds, You can increase text area width to increase its reach or dump the content of epub using -d option, which will dump each paragraph into a single line separated by empty line (or lines depending on the epub), to be later piped into grep, rg etc. Pretty useful to find book quotes.


    # to get 1 paragraph before and after a paragraph containing "Overdue"
    $ epr -d the_girl_next_door.epub | grep Overdue -C 2
  2. Superscript and subscript displayed as ^{Superscript} and _{subscript}.

  3. Some TOC issues:

    • "-" chapters in TOC

      This happens because not every chapter file (inside some epubs) is given navigation points. Some epubs even won't let you navigate between chapter, thus you'll find all chapters named as "-" using epr for these kind of epubs.

    • Skipped chapters in TOC


      Table of Contents
          1. Title Page
          2. Chapter I
          3. Chapter V

      This happens because Chapter II to Chapter IV is probably in the same file with Chapter I, but in different sections, e. g. ch000.html#section1 and ch000.html#section2.

      But don't worry, you should not miss any part to read. This just won't let you navigate to some points using TOC.

    If you feel bothered by these 2 TOC issues, checkout branch commontoc or install via:

    $ pip3 install git+

    which will give you TOC behavior like in many common readers.

    NOTE: I'm not merging commontoc to master since master already does most of its job efficiently and (supposedly, I'm not doing any test) faster. It just navigates between files inside epub rather than pre-defined sections like commontoc does.


You can’t perform that action at this time.