VK conversations exporter.
Necessity to export VK conversations completely, i.e., with all attachments; there is an official data export tool, but it produces an archive which contains the text only [as on 2023].
- Suitable for regular dialogs as well as group conversations.
- Supports various attachment types: photos/images, voice messages, stickers.
- The history is exported in two formats: plain text and HTML with pagination.
- All supported attachments are downloaded, saved to the disk and referenced in HTML pages as local resources with relative paths, which allows browsing the exported history even without working Internet connection.
- Automatic authorization at
vk.comusing cookies loaded from a specified browser. Supported browsers (in a format accepted by--browseroption):chrome,firefox,safari,chromium,edge,opera,vivaldi,brave.
pipx install vkimexp
vkimexp [OPTIONS] PEERS...
PEER should be VK ID of a person or a conversation in question (several PEERs can be provided at once). To find PEER of a person, open this page: https://vk.com/im and select the required dialog, and then his/her VK ID will appear in the address bar like this:
https://vk.com/im?sel=1234567890
where 1234567890 is a numeric ID in question. Use this number as PEER, e.g. for a person with VK ID 1234567890 the command is:
vkimexp 1234567890
For group conversations there is no VK ID in the URL, as they are identified differently, by their index. Nevertheless, take this number (together with 'c'!) and provide it as is, the application will figure out VK ID of a conversation by itself:
https://vk.com/im?sel=c195 => vkimexp c195
-b, --browser NAME Browser to load cookies from (process is automatic). [default: chrome]
-v, --verbose Print more details.
--help Show this message and exit.
The application first makes a probe query to VK API (as would the regular browser do) and receives total size of the current conversation. Then it makes series of similar queries to fetch all the message data in chunks, and also downloads all the attachments it can detect.
Note that if the application discovers that an attachment has been already downloaded, it will immediately skip the unnecessary downloading action and just go to the next one.
The results are saved into a directory relative to the application installation, and the path is printed in the last
line at the successful task completion. Default path is ../../out/<PEER>/.
-
Make sure you are logged into
vk.combefore launching the app. It's recommended to open or refreshvk.compage just before the start to ensure that cookies are fresh (if they expire, the export will fail to start).Some system configurations require to set up an environment variable
XDG_CURRENT_DESKTOP=GNOMEfor the app, so that the cookie extraction library could correctly identify the system keyring and, subsequently, to decrypt the cookies stored by browser. That can solve the cookie loading issue for Ubuntu 20, for example. -
Try Firefox.
-
Try Google Chrome.
Either of those almost certainly will work, and sometimes they both do *gasp*.
-
Run the application with
-voption for more verbose output and see if the debugging information contains any unexpected exceptions or potenital clues. Also take a look at application logs in the output directory for the same reason.
If all attempts to figure out what's going on are unsuccessful, feel free to open an issue. Attaching the logs or verbose output beforehand could simplify the task incredibly.
- Option that disables cookie autoload and lets the user to specify all cookies manually.
- A way to customize output directory.
--color/--no-colorformatting control options.- Parallel (multi-threaded) attachment downloads.
- Options to select which attachment types to process/ignore.
- GitHub Actions packaging automation.
- Fixed crashes on "Access Denied" and similar attachment errors.
- Cosmetic progress printer updates.
- Output HTML suboptimizations.
- Two extraction methods in a conjunction instead of one.
- Embedded CSS styles.
- Full rewrite.
- Prototype.