python-gphoto2 v2.4.0

python-gphoto2 is a comprehensive Python interface (or binding) to libgphoto2. It is built using SWIG to automatically generate the interface code. This gives direct access to nearly all the libgphoto2 functions, but sometimes in a rather un-Pythonic manner.

Other Python bindings to libgphoto2 are available: piggyphoto uses ctypes (included in standard Python installations) to interface to the library. The gphoto2 source tree includes some Python bindings which also use ctypes. gphoto2-cffi uses cffi.

Installation

Since python-gphoto2 version 2.3.1 "binary wheels" are provided for many Linux and MacOS computers. These include a recent version of the libgphoto2 libraries, and pre-built Python interface modules, which makes installation quick and easy. Use pip's --only-binary option to install one of these wheels:

$ pip3 install gphoto2 --user --only-binary :all:

If this fails it is most likely because none of the available wheels is compatible with your computer. In this case you must install some dependencies before installing python-gphoto2. See INSTALL.rst for more details.

Raspberry Pi

Binary wheels for the Raspberry Pi are available from piwheels. You need to install some system packages to use these:

$ sudo apt install libexif12 libgphoto2-6 libgphoto2-port12 libltdl7
$ pip3 install gphoto2 --user

Using python-gphoto2

The Python interface to libgphoto2 should allow you to do anything you could do in a C program. However, there are still bits missing and functions that cannot be called from Python. Let me know if you run into any problems.

The following paragraphs show how the Python interfaces differ from C. See the example programs for typical usage of the Python gphoto2 API.

"C" interface

These functions are as similar as possible to their libgphoto2 equivalents. Most of them return an error code which you must check.

Using SWIG to generate the Python interfaces automatically means that every function in libgphoto2 should be available to Python. You can show the documentation of a function with the pydoc command (or python -m pydoc if you installed gphoto2 with pip inside a virtual environment):

$ pydoc3 gphoto2.gp_camera_folder_list_files
Help on built-in function gp_camera_folder_list_files in gphoto2:

gphoto2.gp_camera_folder_list_files = gp_camera_folder_list_files(...)
    gp_camera_folder_list_files(camera, folder, context) -> int

    Parameters
    ----------
    camera: gphoto2.Camera
    folder: str
    context: gphoto2.Context (default=None)


    Lists the files in supplied `folder`.

    Parameters
    ----------
    * `camera` :
        a Camera
    * `folder` :
        a folder
    * `list` :
        a CameraList
    * `context` :
        a GPContext

    Returns
    -------
    a gphoto2 error code

    See also gphoto2.Camera.folder_list_files

The first part of this text is the function signature and parameter list generated by SWIG. (Note that context is optional - it's only needed if you need the callback functions that can be associated with a context.) The rest of the text is copied from the "doxygen" format documentation in the C source code. (The online API documentation shows how it is intended to look.) Note that this includes a list parameter that is not in the function signature. In C this is an "output" parameter, a concept that doesn't really exist in Python. The Python version of gp_camera_folder_list_files returns a sequence containing the integer error code and the list value.

Most of the libgphoto2 functions that use pointer parameters to return values in the C API have been adapted like this in the Python API. (Unfortunately I've not found a way to persuade SWIG to include this extra return value in the documentation. You should use pydoc to check the actual parameters expected by the Python function.)

For example, the C code:

#include "gphoto2.h"
int error;
Camera *camera;
error = gp_camera_new(&camera);
...
error = gp_camera_unref(camera);

has this Python equivalent:

import gphoto2 as gp
error, camera = gp.gp_camera_new()
...

Note that the gp_camera_unref() call is not needed. It is called automatically when the Python camera object is deleted.

Here is a complete example program (without any error checking):

import gphoto2 as gp
error, camera = gp.gp_camera_new()
error = gp.gp_camera_init(camera)
error, text = gp.gp_camera_get_summary(camera)
print('Summary')
print('=======')
print(text.text)
error = gp.gp_camera_exit(camera)

"Object oriented" interface

This is the preferred way to use libgphoto2 from Python. Most of the libgphoto2 functions have been added as methods of the appropriate GPhoto2 object. This allows GPhoto2 to be used in a more "Pythonic" style. For example, gp.gp_camera_init(camera) can be replaced by camera.init(). These methods also include error checking. If an error occurs they raise a Python GPhoto2Error exception.

The example program can be re-written as follows:

import gphoto2 as gp
camera = gp.Camera()
camera.init()
text = camera.get_summary()
print('Summary')
print('=======')
print(str(text))
camera.exit()

No additional error checking is required.

Error checking

Most of the libgphoto2 functions return an integer to indicate success or failure. The Python interface includes a check_result() function to check these values and raise a GPhoto2Error exception if an error occurs.

This function also removes the error code from lists such as that returned by gp_camera_new() in the example. Using this function the earlier example becomes:

import gphoto2 as gp
camera = gp.check_result(gp.gp_camera_new())
gp.check_result(gp.gp_camera_init(camera))
text = gp.check_result(gp.gp_camera_get_summary(camera))
print('Summary')
print('=======')
print(text.text)
gp.check_result(gp.gp_camera_exit(camera))

There may be some circumstances where you don't want an exception to be raised when some errors occur. You can "fine tune" the behaviour of the check_result() function by adjusting the error_severity variable:

import gphoto2 as gp
gp.error_severity[gp.GP_ERROR] = logging.WARNING
...

In this case a warning message will be logged (using Python's standard logging module) but no exception will be raised when a GP_ERROR error occurs. However, this is a "blanket" approach that treats all GP_ERROR errors the same. It is better to test for particular error conditions after particular operations, as described below.

The GPhoto2Error exception object has two attributes that may be useful in an exception handler. GPhoto2Error.code stores the integer error generated by the library function and GPhoto2Error.string stores the corresponding error message.

For example, to wait for a user to connect a camera you could do something like this:

import gphoto2 as gp
...
print('Please connect and switch on your camera')
while True:
    try:
        camera.init()
    except gp.GPhoto2Error as ex:
        if ex.code == gp.GP_ERROR_MODEL_NOT_FOUND:
            # no camera, try again in 2 seconds
            time.sleep(2)
            continue
        # some other error we can't handle here
        raise
    # operation completed successfully so exit loop
    break
# continue with rest of program
...

When just calling a single function like this, it's probably easier to test the error value directly instead of using Python exceptions:

import gphoto2 as gp
...
print('Please connect and switch on your camera')
while True:
    error = gp.gp_camera_init(camera)
    if error >= gp.GP_OK:
        # operation completed successfully so exit loop
        break
    if error != gp.GP_ERROR_MODEL_NOT_FOUND:
        # some other error we can't handle here
        raise gp.GPhoto2Error(error)
    # no camera, try again in 2 seconds
    time.sleep(2)
# continue with rest of program
...

Logging

The libgphoto2 library includes functions (such as gp_log()) to output messages from its various functions. These messages are mostly used for debugging purposes, and it can be helpful to see them when using libgphoto2 from Python. The Python interface includes a use_python_logging() function to connect libgphoto2 logging to the standard Python logging system. If you want to see the messages you should call use_python_logging() near the start of your program, as shown in the examples. In normal use you probably don't want to see these messages (libgphoto2 is rather verbose) so this could be controlled by a "verbose" or "debug" option in your application.

The libgphoto2 logging messages have four possible severity levels, each of which is mapped to a suitable Python logging severity. You can override this mapping by passing your own to use_python_logging():

import logging
import gphoto2 as gp
...
callback_obj = gp.check_result(gp.use_python_logging(mapping={
    gp.GP_LOG_ERROR   : logging.INFO,
    gp.GP_LOG_DEBUG   : logging.DEBUG,
    gp.GP_LOG_VERBOSE : logging.DEBUG - 3,
    gp.GP_LOG_DATA    : logging.DEBUG - 6}))
...

If you prefer to use your own logging system you can define a logging callback function in Python. The function must take 3 or 4 parameters: level, domain, string and an optional data. The data parameter allows you to pass some user data to your callback function (e.g. to log which thread an error occurred in): The callback function is installed with gp_log_add_func:

import gphoto2 as gp
...
def callback(level, domain, string, data=None):
    print('Callback: level =', level, ', domain =', domain, ', string =', string, 'data =', data)
...
callback_obj1 = gp.check_result(gp.gp_log_add_func(gp.GP_LOG_VERBOSE, callback))
callback_obj2 = gp.check_result(gp.gp_log_add_func(gp.GP_LOG_VERBOSE, callback, 123))
...

What to do if you have a problem

If you find a problem in the Python gphoto2 interface (e.g. a segfault, a missing function, or a function without a usable return value) then please report it on the GitHub "issues" page (https://github.com/jim-easterbrook/python-gphoto2/issues) or email [email protected].

If your problem is more general, e.g. difficulty with capturing multiple images, then try doing what you want to do with the gphoto2 command line program. If the problem persists then it might be worth asking on the gphoto-user mailing list. Another reader of the mailing list may have the same camera model and already know what to do.

Notes on some gphoto2 functions

gp_log_add_func / use_python_logging

Since python-gphoto2 version 2.0.0 these functions return a sequence containing an error code and an object storing details of the callback. The callback is automatically uninstalled when this object is deleted.

In earlier versions of python-gphoto2 these functions return an integer id that must be passed to gp_log_remove_func to uninstall the callback.

gp_context_set_idle_func / gp_context_set_progress_funcs / etc.

These functions are only usable since python-gphoto2 version 1.9.0. They return a Python object which your program must store until the callback(s) are no longer required. Deleting the returned object cancels the callback(s), so there is no need to do this yourself. See the context_with_callbacks.py example for a convenient way to do this.

gp_file_get_data_and_size / CameraFile.get_data_and_size

Since python-gphoto2 version 2.4.0 these functions return a Python memoryview object. Prior to that they returned a FileData object that supports the buffer protocol so its data can be made accessible to Python by using a memoryview object. This allows the data to be used without copying. See the copy-data.py example for typical usage.

Note that if the CameraFile object is deleted, or another function (such as gp_file_set_data_and_size or gp_file_open) changes the CameraFile's data, then this object will be invalidated and you will probably get a segmentation fault.

gp_file_set_data_and_size / CameraFile.set_data_and_size

Since python-gphoto2 version 2.1.0 these functions accept any bytes-like object. In earlier versions of python-gphoto2 these functions required a string and its length, and didn't work correctly anyway.

gp_camera_file_read / Camera.file_read

The buf parameter can be any Python object that exposes a writeable buffer interface. This allows you to read a file directly into a Python object without additional copying. See the copy-chunks.py example which uses memoryview to expose a bytearray.

gp_camera_wait_for_event / Camera.wait_for_event

These functions return both the event type and the event data. The data you get depends on the type. GP_EVENT_FILE_ADDED and GP_EVENT_FOLDER_ADDED events return a CameraFilePath, others return None or a text string.

Licence

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.