Blog to blogger from the command line.
I tried. Didn't work. googlecl
is just too rough and isn't easy to
script. For ex:
- No way to update a post
- Doesn't work with blog and post ids.
- and others...
- Provides a command line tool and create, update or delete posts on Blogger hosted blogs.
- Post content can be piped in - so you can use your favourite way to generate html markup
- Pandoc goodness - so that you can write your doc in any of the input formats that Pandoc supports
- More Pandoc goodness - supports pandoc filters so you can do nice things like
create diagrams with
mermaid-filter
- AsciiDoc support - Supports asciidoc as a source format as well using
asciidoctor
&asciidoctor-diagram
- You can also export your existing posts to your favourite lightweight markup format like markdown etc as individual files. Then edit them in a real editor, and publish them back! All pandoc output formats!
- Understands specially marked comments - so you can just hand it a file and it'll figure out what to do - great for posting from vim etc.
# Now live on PyPI
sudo pip install EasyBlogger
This installs EasyBlogger and its dependencies. It also installs the
easyblogger
script
Install pandoc If
you're on cygwin, you can just install the windows dl and put
pandoc.exe
somewhere on path
The tool needs to be granted access to manage the blog. Google APIs use OAuth2.
-
First, get your blog id. You can find the blog id by viewing the source. For ex. on my blog, I have the following lines near the top:
<link rel="alternate" type="application/atom+xml" title="Nifty Tidbits - Atom" href="http://blog.rraghur.in/feeds/posts/default" /> <link rel="alternate" type="application/rss+xml" title="Nifty Tidbits - RSS" href="http://blog.rraghur.in/feeds/posts/default?alt=rss" /> <link rel="service.post" type="application/atom+xml" title="Nifty Tidbits - Atom" href="http://www.blogger.com/feeds/7642453/posts/default" />
On the last link, the number
7642453
is the blogId -
Authorize
On Linux
# run through OAuth2 hoops... following needs to be run as root # First find your blog Id easyblogger --blogid <yourblogid> get # This will first open a browser and ask you to sign in and # approve access to your blog
This will open a browser. You may see a chrome warning that it can't be run as root - but you can ignore that. Once you authorize,
~/.easyblogger.credentials
is created with your OAuth2 tokenOn Windows
If your
PATH
variable has the python Scripts folder, then you should just be able to runeasyblogger --blogid <id> get
in a command window. If not, then open acmd
window and navigate to<python install folder>\Scripts
and runpython easyblogger --blogid <yourblogid> get
-
All set!
That's it - you're all set!
You will need to repeat the OAuth2 authorization process if you ever change the blog, or revoke permissions or if the auth token expires.
- Stick the following in your
~/.vimrc
func! s:systemwrapper(cmd)
echom a:cmd
let output=system(a:cmd)
return output
endfunction
func! BlogSave(file)
" to debug, replace with
" exec "!easyblogger file " . a:file
let output=s:systemwrapper("easyblogger file ". a:file)
echom output
endfunction
command! BlogSave call BlogSave(expand("%:p"))
- Start writing a post - create a markdown file (.md) with frontmatter in a comment
<!--
id:
title : title
tags : [any, comma, separated, labels]
format : markdown
published: true
filters: <path to your filter>
-->
Note that as of Easyblogger 3.0, the preferred frontmatter format is borrowed from Hugo.
The original frontmatter header in earlier versions is deprecated. However, if easyblogger
finds header using the older keys, then it will use them.
While there should be no reason to prefer the old format, if you need that for whatever
reason, you must specify --legacy-frontmatter
flag in the get
subcommand. For more
details, refer to the Frontmatter section
LEGACY FRONTMATTER FORMAT still supported but you're encouraged to use the new format
<!--
PostId:
Title : title
Labels : any, comma, separated, labels
Format : markdown
Published: true
filters: <path to your filter>
-->
- If you prefer using
asciidoc
, then use the following comment header:
+++
id:
title : title
tags : [any, comma, separated, labels]
format : asciidoc
published: true
+++
LEGACY FRONTMATTER FORMAT still supported but you're encouraged to use the new format
////
PostId:
Title : title
Labels : any, comma, separated, labels
Format : asciidoc
Published: true
////
Asciidoc does not require filters - it has a better system of plugins. Just
ensure that you have installed asciidoctor
and asciidoctor-diagram
gems
- When done, call
:BlogSave
and your blog will be published
-
Get a list of posts post Id, title and url are output by default.
# get a list of posts # param : Blog Id - look at your blog's atom pub url - its the number in the url. easyblogger --blogid 7642453 get 4424091495287409038,Moving from Wordpress.com to Blogger,http://blog.rraghur.in/2013/08/moving-from-wordpresscom-to-blogger.html ... ... # 10 rows shown
-
Filter by labels or search; specify
max
results to be returned.# get only the last 5 with tag 'vim' # you can specify multiple tags - separate them with commas easyblogger --blogid 7642453 get -l vim -c 5 # search for all posts with 'vim' easyblogger --blogid 7642453 get -q vim -c 5
-
Get a specific post by its id
# get a specific post by id easyblogger --blogid 7642453 get -p 3728334747597998671
-
Get a specific post by its url
# get a specific post by url easyblogger --blogid 7642453 get -u https://blog.rraghur.in/2015/06/js-development-2015.html
-
Control which fields are printed out.
# output field control easyblogger get -p 3728334747597998671 -f "id,title,url,labels" 3728334747597998671,Rewriting history with Git,http://blog.rraghur.in/2012/12/rewriting-history-with-git.html,[u'git', u'HOWTO', u'Tips']
-
Output in (lightweight) markup - very good for updates.
-
If its a single post, then its printed to console.
easyblogger --blogid 7642453 get -p 3728334747597998671 -d markdown
It also includes a header that will allow you to edit the file and update the post back with the file subcommand below
-
if more than one post, then each post is written to a file (name derived from the title)
easyblogger --blogid 7642453 get -l vim -d markdown
-
If you'd like to get a single post as a file, specific
-w
or--write-files
easyblogger --blogid 7642453 get -p 3728334747597998671 -d markdown --write-files
-
Supports all mark up formats supported by
pandoc
# Output formats: native, json, docx, odt, epub, epub3, fb2, html, html5, s5, slidy, slideous, dzslides, docbook, opendocument, latex, beamer, context, texinfo, man, markdown, markdown_strict, markdown_phpextra, markdown_github, markdown_mmd, plain, rst, mediawiki, textile, rtf, org, asciidoc
-
Specifying --blogid each time is just painful. You can set a default
blogId by creating a default args file ~/.easyblogger
as follows:
cat > ~/.easyblogger
--blogid
234567
And now you can type the command as:
easyblogger get
You can override the args from file by providing the argument on the command line
Note: Blogger API v3 does not support/expose API for creating posts as drafts. Please ask for this feature on Google's blogger dev group - I'll add that capability once/if it becomes available.
Blogs are created as drafts by default now. You can override this with the --publish
flag
which will post the blog directly (current behavior)
# create a post from stdin with title and labels
easyblogger post -t "Hello World" -l "python,hello" -c "Hello world!!!" --date 2018-01-01T10:00:00
Pipe out from any HTML generation mechanism
pandoc -f markdown - | easyblogger --blogid 6136696198547817747 post -t 'Hello from Pandoc' -f -
# Hello from Pandoc
this is a nice long post
3295765957555899963
Or just tell easyblogger to convert from markdown
with the --format
arg
# --format supports
# native,json,markdown,
# markdown_strict,markdown_phpextra,
# markdown_mmd,rst,mediawiki,
# docbook,textile,html,latex
easyblogger post -t 'Hello from Pandoc' --format markdown -c "##heading2"
2342323423423423423
Update works with a patch request - only specify what you need changed Blogger API does not allow the blog permalink to be updated - so in case you want to change that you'll need to delete and create another post (though you will probably lose comments etc - so only viable if you've just published it)
easyblogger update -t 'A new title' -l "new,labels" 3295765957555899963 --date 2018-01-01T10:00:00
You can also update the contents by passing in the --file
argument.
Piping it in works too - use --file -
; like so
pandoc -f markdown - | easyblogger update -t 'Hello from Pandoc' --file - 3295765957555899963
# This is h1
## this is h2
Some para text
[EOF]
I wrote easyblogger
script primarily so I can blog from Vim. If your
file has properly formatted comments, then EasyBlogger
will figure out
what to do (insert or update) based on the metadata.
So, you can author a post like so:
cat MyBlogPost.md
<!--
Title: This is your title
PostId:
Labels: a,b,c
format: markdown
published: false
PublishDate: 2018-01-01T10:00:00
filters: <path to your installed filter>
-->
# This is my content
The example above uses legacy frontmatter format. You're encouraged to use the new format which allows for additional metadata.
And post it to your blog like so:
easyblogger file MyBlogPost.md
And easyblogger
will update your post doc back with the postId
of
the generated post. Now, if you edit the doc and publish again with the
same command, your post will be updated.
To delete posts, you need to specify the post id
easyblogger delete 234546720561632959
As you've seen, easyblogger relies on a comment header with specific keys for
metadata about the post as well as to drive the behavior of the program.
When EasyBlogger
started, I had come up with my own set of (minimal) keys.
Somewhere in the 2.x days, I built support for the frontmatter format as defined
in Hugo project(along with some specific keys used for Blogger) - this is especially
useful if you want to migrate off Blogger to Hugo.
The header format can be either TOML or YAML. The new frontmatter keys are the default both for input and output.
When writing to output files with get
, easyblogger will write the header in
- Document format - asciidoc: Header in TOML enclosed by
+++
- Legacy header keys - only if the command line specifies the
--legacy-frontmatter
flag
- Header enclosed with
+++
- parse as TOML - Header encosed with HTML comment or
////
- parse as YAML - If doc is TOML, then default format is supposed to be 'asciidoc' if not specified.
- If doc is YAML, then default format is supposed to be 'markdown' if not specified.
If any of the legacy frontmatter keys (Title
, PostId
etc) are present, then the legacy keys
are expected. Otherwise the new style Hugo compliant headers are expected.
- New style (Hugo)
+++ title = "Proxy PAC file for work" id = "293493242234" tags = [ "Rants", "Tips", "Utilities",] aliases = [ "http://niftybytes.blogspot.com/2018/04/proxy-pac-file-for-work_30.html",] publishdate = 2018-04-30T12:42:00+05:30 draft = false date = 2018-04-30T12:42:00+05:30 lastmod = 2018-04-30T12:47:37+05:30 +++
- Old style (Easyblogger)
<!-- Labels: [Rants, Tips, Utilities] PostId: '8010087245053438499' Published: true PublishDate: 2018-04-30T12:42:00 Title: Proxy PAC file for work -->
Feel free to use the EasyBlogger class in your own tool/utility whatever else. Just remember:
- Use your own API client id (see below)
- Include an attribution and a link to EasyBlogger - not mandatory -but just be nice:)
If you're using EasyBlogger class in your tool/utility, please then
register for API access at Google's API
console. Create a client id and
secret key at the API access page and request for Blogger API access.
Once you have API access authorized, you're good to get started. Just
create the EasyBlogger
constructor with your client id and secret
If you're integrating by shelling out, then stick the API key and client
id in the command line with --clientid
and --secret
args. You can
also stick them in the ~/.easyblogger
file to avoid specifying them
each time
- Clone the repo
- Start a virtualenv -
virtualenv .dev
- Activate it -
.dev\Scripts\activate
- Install dependencies -
pip install -r requirements.txt
pip install -e .
- Exit out of any virtualenvs
- Run
tox