Because Blogofile(at least the 0.7.1 version I'm using) regenerates every file and directory every time it makes it hard to update just the changed files on Amazon's S3 (Simple Storage System) can scale incredibly high

So I wrote a shell script that updates only the changed files and also pings GooglePing so that Google and other services come along and read your update.

Over time, the difference between what the old page 2 on S3 has and what it should be will build. There's a simple answer, which is to just upload all of the page files.

Every 10 blog posts or so, you could just update all the category and archive files as well. Maybe that's somthing to automate as well

!/bin/bash

jdeibele [ at ] gmail . com

What this does:

builds your blog with blogofile, then builds a list of key

files that have changed. Blogofile is a python program

that uses Mako to generate static pages.

Blogofile: https://www.blogofile.com

For me, these files are:

the post itself

the first page of the categories it appears in

feeds for the categories (RSS and atom)

the archive page (year/month)

the first page of /page

feed for the blog (RSS and atom)

the main page (index.html) for the blog

BLOGURL="www.siriusventures.com" BLOGNAME="Sirius Stuff" BLOGHOME=$HOMEDIR/siriusventures

cd $BLOGHOME blogofile build rmdir *

blogofile builds extra directories for these but they're empty

cd $BLOGHOME/_posts file=ls -t * | head -1 permalink=grep "^permalink:" $file | cut -f3 -d: | cut -f4 -d"/" archive=grep "^date:" $file | cut -f2 -d: | cut -f1,2 -d/ | sed 's/ //g'

"slugify" the category names

categories=grep "^categories:" $file | tr [:upper:] [:lower:] | cut -f2 -d: | sed s'/^ //' | sed s'/ /-/' echo $categories cd $BLOGHOME/_site IFS=',' echo index.html > /tmp/postit for category in $categories do echo category/$category/index.html >> /tmp/postit echo category/$category/1/index.html >> /tmp/postit echo category/$category/feed/index.xml >> /tmp/postit echo category/$category/feed/atom/index.xml >> /tmp/postit done echo archive/$archive/1/index.html>> /tmp/postit echo feed/index.xml >> /tmp/postit echo feed/atom/index.xml >> /tmp/postit echo page/1/index.html>> /tmp/postit

s3cmd can be installed with homebrew - brew install s3cmd

s3cmd needs -recursive to upload a new directory

s3cmd put --recursive $BLOGHOME/_site/$permalink s3://$BLOGURL

while read file do s3cmd put $BLOGHOME/_site/$file s3://$BLOGURL/$file echo $BLOGHOME/_site/$file done