A tiny CLI to perform common operations on HedgeDoc (the largest open-source fork of HackMD).
For more background, see the initial discussion on the hackmd codimd repo.
There is an alternative, TypeScript-based CodiMD CLI for hackmdio/codimd
maintained by the HackMD team here: https://github.com/hackmdio/codimd-cli.
(it may or may not be compatible with the hedgedoc/hedgedoc
server)
- A HedgeDoc server running somewhere
curl
(install viaapt install curl
orbrew install curl
on Mac)wget
(install viaapt install wget
orbrew install wget
on Mac)jq
(install viaapt install jq
orbrew install jq
on Mac)
Clone the repository.
$ git clone https://github.com/hedgedoc/cli
Enter the folder with the script.
$ cd cli/bin
Optionally symlink the hedgedoc script somewhere into your $PATH to make it globally accessible. Otherwise you will have to provide the path to the script manually. This command might need admin rights (sudo)!
$ ln -s $PWD/hedgedoc /usr/local/bin/hedgedoc
Check if the new command exists. You should see the documentation
$ hedgedoc
Set HEDGEDOC_SERVER
environment variable to your server's URL. It defaults to http://127.0.0.1:3000
Do this once on the command line or persist it in .profile
and/or .bashrc
.
$ export HEDGEDOC_SERVER='https://hedgedoc.example.com'
If you added the variable to .profile
or .bashrc
, re-open the terminal to read the new variable.
Test your configuration by creating a new note with FREELY access and no login required. You will receive the generated <note_id>
for the document like 3jXcabSfSNesbH6KT72ieg
.
Caution: You won't have the right to delete the new document if not authenticated!
$ echo "# HedgeDoc!" > test.md
$ hedgedoc import test.md
Check for the document in the browser by concatenating the
address of your server and the <note_id>
.
It's not necessary to authenticate against the server in order to make use of hedgedoc-cli
. But without authentication you won't have access to the non-FREELY documents and everything that's accessible behind the login.
Authentication with a cookie is so far the only way if you login with GitLab and the like. Use browser extensions like Get cookies.txt to store the cookie in key.conf
.
Possible you have many lines in key.conf
. You only need the line with connect.sid
followed by a long hash!
Optionally add the HEDGEDOC_COOKIES_FILE environment variable to specify where cookies will be stored. It defaults to ~/.config/hedgedoc/key.conf
# You can put this in .profile and/or .bashrc, too.
$ export HEDGEDOC_COOKIES_FILE=~/.config/hedgedoc/key.conf
$ hedgedoc login --email [email protected] p4sW0rD # takes an email and password as optional args
$ hedgedoc login --email # or pass them via stdin prompt instead
$ hedgedoc login --ldap username p4sW0rD # takes a username and a password as optional args
$ hedgedoc login --ldap # or pass them via stdin prompt instead
If your authentication method is set up correctly the following commands will work.
$ hedgedoc profile
You are logged in to a HedgeDoc server.
HEDGEDOC_SERVER=https://hedgedoc.example.com
HEDGEDOC_COOKIES_FILE=/Users/someuser/.config/hedgedoc/key.conf
USER_NAME=alice
USER_ID=abc93e9b-bf57-490f-a4c6-0d7a842b7cd4
USER_PHOTO=https://cdn.libravatar.org/avatar/ba8b1ebe25440cd38748639eebdc8eaf?s=96
$ hedgedoc history
ID Title
fCbvG5pdSYOLobNN1SDUhg Example-note-title
...
$ hedgedoc logout
Your hedgedoc auth session cookie is written to $HEDGEDOC_COOKIES_FILE
(which defaults to ~/.config/hedgedoc/key.conf
).
You may need to log in again if:
- your session expired
- the hedgedoc server was restarted (which force-expires all sessions as a side-effect)
- the is
$HEDGEDOC_COOKIES_FILE
deleted, moved, or becomes unreadable byhedgedoc-cli
$ hedgedoc import <input_path> [note_id] # takes a local path to a text file, and an optional note id for the new note
qhmNmwmxSmK1H2oJmkKBQQ # returns <note_id> on success
You can open the new note on the server by going to $HEDGEDOC_SERVER/<note_id>
.
The optional note_id
is only available on servers with allowFreeURL
enabled.
Check the documentation
for more information.
$ hedgedoc publish qhmNmwmxSmK1H2oJmkKBQQ # takes a <note_id>
S1ok9no3f # returns public note id
You can open the published note on the server by going to $HEDGEDOC_SERVER/s/<public_note_id>
.
$ hedgedoc export --pdf qhmNmwmxSmK1H2oJmkKBQQ # takes a <note_id>, outputs to <note_id>.pdf by default
$ hedgedoc export --md qhmNmwmxSmK1H2oJmkKBQQ my_note.md # or you can specify an output path explicitly
$ hedgedoc export --html qhmNmwmxSmK1H2oJmkKBQQ
$ hedgedoc export --slides qhmNmwmxSmK1H2oJmkKBQQ my_slides.zip
These server endpoints are used by this project and can be unstable and undocumented, but may be of use if you're developing your own projects that need API access to HedgeDoc.
https://<hedgedoc_server>/login
https://<hedgedoc_server>/logout
https://<hedgedoc_server>/me
https://<hedgedoc_server>/history
(requires auth)https://<hedgedoc_server>/new
https://<hedgedoc_server>/new/<note_id>
https://<hedgedoc_server>/<note_id>/publish
https://<hedgedoc_server>/<note_id>/download
https://<hedgedoc_server>/<note_id>/pdf
https://<hedgedoc_server>/<note_id>/slide
We'd love a PR for any one of these commands!
hedgedoc edit <note_id> < new_content.md
hedgedoc inviteuser <email_to_invite>
hedgedoc chmod <permissions> <note_id>
hedgedoc chown <user> <note_id>
hedgedoc delete <note_id>
hedgedoc list --all
list all notes on the server by id: titlehedgedoc list <user_email>
list notes for a given user by id: titlehedgedoc search <query>
find a note ids matching a given query
If you want to build something with this CLI, here are some ideas that the community has requested:
Import/Export:
- Tests for HedgeDoc: write some easy tests with the hedgedoc-cli
- Saving to local filesystem: this is easy with
./hedgedoc export
- Export to PDF:
./hedgedoc export --pdf <note_id> <output.pdf>
bam. - Arbitrary file upload: we should support this with
./hedgedoc import <file.extension>
- Export slides as PDF & hackmdio/codimd#545: we should add this to
./hedgedoc export --slides
once it's ready - Show all files in database:
./hedgedoc list --all
- Import/export file in GitHub repository:
wget <file from github> > file.md; ./hedgedoc import file.md
- API to download HTML: works already
./hedgedoc export --html <note_id> note.html
- CLI for adding new note: can be documented in the README with
hackmd
command - "Takeout" feature: ability to download all user data with
./hedgedoc export
Permission management:
- Support permission of invitee only:
./hedgedoc inviteuser --permissions r <user_email>
could work by firstchmod
andchown
ing the note, then sending an email invite to that note - Find the notes by the owner:
./hedgedoc list <user_email>
works as a rudimentary API to do this form the command line - Add user administration: setup a script that creates all the users & their notes from .md files on disk,
chown
&chmod
s them to their proper permissions, then invites all the users to join. Would be possible to run this on a timer to do regularly as well.
Sync backends:
- Import/Export with Github: trigger
./hedgedoc import|export
script via timer or webhook upon github file modification - Import/Export with Evernote: same deal as github/dropbox sync, write a script with import/export
- Support for ownCloud: same deal as github/dropbox sync, write a script with import/export
- Support auto sync with dropbox: write a script that does a full
import
andexport
or merge in whatever direction desired - Support auto sync with Google Drive: write a script that does a full
import
andexport
or merge in whatever direction desired