What is it?
Converts HTML pages to images (gif, jpg, png, ps, etc..)
html2jpg requires Mozilla (only some versions) to work. It's
pretty kludgy. I needed a solution and couldn't find one, so
maybe this is useful to you. If you have KDE, I recommend
- Easy to automate.
- Attempts to remove borders, menus, scrollbars, etc..
- Supports many image types (even postscript!)
- Very kludgy and unlikely to work!
- Perl, which kicks ass
- convert from ImageMagick
- xwininfo, xwd [standard unix utilities]
(or else modify find_window() for your browser)
From the usage:
- We don't know when the page is finished loading
- Rule of thumb for finding subwindow is guesswork!
If you're getting the wrong window, comment out call to subwindow()
- Browser geometry ignored if browser doesn't create a new window
(Such as opera in "mdi" mode)
- Dependent on xwd command and output of xwininfo and ...
- Browser can't be iconified or partially off-screen
- Only gets portion of html displayed in browser
- Requires mozilla browser - update find_window() for other browsers
- Opens up a bunch o' windows in your browser and leaves them there
- Might not even pick the right window... (the opera version is
better about this, but it is more version dependent)
This software is essentially free, but please read my
Please read the
It's a single perl script.
I also had it working with an older version of opera, but that's
not up to date: html2jpg.opera
See the usage for command-line options.
html2jpg supports all the image types that convert supports.
This also means you can specify the type for stdout with a file like: -o gif:-
If the script is renamed (or linked to) as "html2<type>" - then the default
image type will be <type>.
(Example: html2png, html2gif, html2jpg, html2ps, ...)
The script does the following:
This relies on a bunch of unix utilities working just the way they do on my system.
Caveat Emptor. :)
- Loads your url in a browser (with -remote, and all the bugs that has..)
- Sleeps for a few seconds (because we don't know when the page is loaded!)
- Tries to find the browser window in the Xwindow tree
- Tries to find the proper subwindow (so we can ignore menus, window frames, scroll bars..)
- Takes a window dump and converts it to the appropriate image type.
It's just a perl script. No install required.
See the CHANGELOG