Some handy short scripts I like, many of which I've written all by myself!
cd ~
git clone git://github.com/chbrown/scripts.git
echo 'export PATH=$HOME/scripts:$PATH' >> .bashrc.localPrerequisite for many of these scripts:
pip install requests coloramaCopyright © 2012–2013 Christopher Brown. MIT Licensed.
Except for following files, I am the sole author of all of these scripts.
autoreloadhtpasswd.pyotf2ttf2eot.shsoundcloudtransposelib/and everything in this directory.
Create 10 alphadecimal passwords of length 16 each.
alphadec --length 16 --count 10
Watch a directory for changed files and restart a process when a change is detected.
autoreload python main.py
From stevekrenzel/autoreload (no license).
Measured from the top left of the screen, left-click once.
click -x 600 -y 400
Beautify all source code in ./src/*.c and ./src/*.h, in-place.
Uses astyle with the following options:
astyle
--style=java
--
--unpad-paren
--delete-empty-lines
--add-brackets
--convert-tabs
--align-pointer=type
--lineend=linux
--suffix=none $file
Get astyle from your package manager, e.g.:
brew install astyle
A shortcut to this wget sequence:
wget -e robots=off --recursive --no-clobber --page-requisites --convert-links --restrict-file-names=windows $1
E.g.,
csv2html batch_results.csv > batch_results.csv.html
Uses Python's dialects to convert.
cat sb5b.csv | tr '\r' '\n' | csv2tsv > sb5b.tsv
Wait until a page doesn't scroll down anymore, and then output the DOM to STDOUT.
curlweb2.0 http://animalstalkinginallcaps.tumblr.com/
Open any GitHub pages which the current git repo has as remotes.
cd ~/scripts
gh # opens https://github.com/chbrown/scripts in your browser
One-liner for when you have some lengthy lines and you want to kill the wrap:
tr '\t' ' ' | cut -c -$(tput cols)
Wrapping around git ls-remote --tags git://... to get the good stuff:
git-remote-tags chbrown/amulet
Assumes github.
Until git 1.8.3 rolls around:
git-submodule-rm static/lib
Thanks goes to stackoverflow.
Script originally from a guy named "Eli Carter."
After alphadec'ing a password, say, vsw8lq4NuM0S, create an htpasswd file called nginx.htpasswd with that password for the user scriptuser. I think it adds to an existing file if nginx.htpasswd already exists.
htpasswd.py -c -b nginx.htpasswd scriptuser vsw8lq4NuM0S
Simply grep for a Mac LaunchAgent that matches the given argument, and start it. Easy way to have databases around but not always use them when not developing.
launch redis
> launchctl start homebrew.mxcl.redis
It found the LaunchAgent called homebrew.mxcl.redis and started it.
Basically undo what launch does.
launchrm mongo
Given a bunch of plain text files, concatenate them all, tranforming each into the format:
[File creation date in ISO format] Filename.txt
<File contents>
So, if you have a folder of iPad notes, say, simply do something like:
cd ~/Dropbox/ipad-notes
lsmcat Ideas/*.txt > Ideas.txt
rm -r Ideas/
cat Ideas.txt
Display your current MAC address:
mac.py --display
Generate a random valid MAC address:
mac.py --gen
Reset your current MAC address to a random new one:
sudo mac.py
MD5 hash the last command line argument.
md5py freshplum
Copy the MongoDB somewhere to somewhere else.
mongomigrate.rb asl:drags localhost:drags
MySQL helper for creating a table directly from a csv file.
mysqlshove somedb maps/coords.csv --table some_new_table -u root
The first column of the table should be unique integers, and it will be named pkid.
Print out the current datetime in some variation on ISO-8601:
| output | timezone | date | time | file | eol |
|---|---|---|---|---|---|
2013-07-08 |
UTC | True | False | False | False |
18:16:52 |
UTC | False | True | False | False |
2013-07-08T18:16:52 |
UTC | False | False | False | False |
2013-07-08T18:16:52\n |
UTC | False | False | False | True |
2013-07-08T18-16-52 |
UTC | False | False | True | False |
2013-07-08 |
None | True | False | False | False |
13:16:52 |
None | False | True | False | False |
2013-07-08T13:16:52 |
None | False | False | False | False |
2013-07-08T13:16:52\n |
None | False | False | False | True |
2013-07-08T13-16-52 |
None | False | False | True | False |
$ now --utc -l
2013-07-08T18:18:51
$Convert an OTF font file to EOT with fontforge
otf2ttf2eot.sh TimesNewRoman.otf
Like col, but auto-adjusts with more text. Slower, obviously.
cat ANEW2010All.txt | pcol -s \\t
gs | pdftk cat output helper.
pdfcat page1.pdf page2.pdf bothpages.pdf
Or
pdfcat page1.pdf page2.pdf > bothpages.pdf
Just pipes pdftotext (comes with LaTeX, I think) through wc -w:
pdfcount LSA.pdf
Given a list of PDFs, run pdfcrop on all of them with meager (but non-zero) margins, letting pdfcrop rename them as it likes.
pdfcropall Readings/2011-10-02/*.pdf
Watermark only the first page of a PDF with another PDF.
pdfstamp Chomsky_2012.pdf fair_use.pdf
Like mongomigrate, but shorter (and not ruby)!
pg_migrate dark:twitter_dev localhost:twitter_dev
Uses ssh -C and pg_dump but that's about it. You'll need to dropdb and createdb, etc., yourself.
Simply do a pip install -U <package> on all your installed packages.
pip-update-all
Pretty print JSON (using Python's json.dumps, sorting keys and indenting by two spaces):
cat Twitter/Texas_geolocated.json | ppjson
Pretty print xml, using Python's BeautifulSoup.
ppxml timeml/APW19980322.0749.tml
Not like xml gets pretty. It just gets prettier.
Prepend the UTF-8 byte order marker (BOM) to the input. Uses fileinput, so STDIN and all files supplied as command line arguments will be streamed to STDOUT after the BOM byte sequence, \xef\xbb\xbf.
<flat.txt prependbom > flat.utf8
Run pdflatex on the nearest .tex file, potentially specified by command line argument,
without interaction, and then open in TeXShop.
ptx Brown_FinalReport.tex
Or
ptx Brown_FinalReport
Or
$ ls *.tex
Brown_FinalReport.tex Figures.tex Zscores.tex
$ ptx
I use this in Sublime Text 2 in my LaTeX build system.
Do a combo redis KEYS "prefix:*" command and then DEL whatever it finds.
redis-del domains:*
Needs to be generalized. Currently copies everything from a hard-coded remote host, "dark", (over automatically opened SSH tunnel) to localhost, using the hard-coded in "forex:*" key wildcard.
redismigrate
But you get the idea.
Open up a webpage in PhantomJS and render it to an image file in the current directory. The filename is based on the given URL.
renderpage http://irinawerning.com/bttf2/back-to-the-future-2-2011/
It outputs the filename it used, among other things.
Little helper for the awesome sshfs tool that OS X Fuse provides. It'll make the given directory as needed, and die quietly if the connection already exists.
sf tacc: /Volumes/tacc
Python rendition of Luka Pusic's soundcloud.sh downloader. His wasn't unescaping higher order Unicode character escapes. Well, the Mac OS X filesystem totally supports UTF-8, and this script does too.
soundcloud https://soundcloud.com/oskmusic
Give it the directories of your git repositories and it'll run git status on them all and colorize them and tell you which ones need committing and pushing.
statuses ~/github/*/
Basic duration timer.
stopwatch
> Press \n to end.
↪
> 1.3509s
Just cat current user's id_rsa.pub into remote .ssh directory on the given server, creating .ssh and authorized_keys2 if necessary:
ssh-copy-rsa dark
You may need to fix the permissions on the remote's new ~/.ssh, i.e.:
700 .ssh
644 .ssh/authorized_keys2
Uh, what? So it looks through lines of json and prints just the "text" attribute. I guess if you want to look as just Tweet text. Something like npm install json and then cat feed.json | json -C text would do the same thing.
List all active ssh tunnels, looking for something like ssh ... -L ....
Wrapper around tar cz, basically:
tar -czf $1.tgz $1
mv $1 $1.tmp
But with some existing / missing checks.
Use like:
tgz penn-treebank-rel3/
Super-complicated Awk script for transposing like in Excel.
transpose alice-results.dat | pcol -g 2
Like ptx, but with a latex && dvipdf pipeline instead of pdflatex.
You'll probably only find this useful if you're still using qtree with arrows,
in which case you really ought to check out TikZ.
tx so_many_trees
Crawl vimeo for the most popular movies under a given channel, saving to ~/Movies/vimeo:
vimeo-crawler --channels staffpicks
The destination directory is created with os.makedirs if it doesn't exist. See --help for more options.
Requires requests, youtube-dl, and redis:
pip install requests
brew install youtube-dl
brew install redis
launch redis
Open all files in the current folder as images, split them into half, left and right, (like opening a normal book), and save as the original filenames suffixed with -left or -right, JPEGs with quality 95.
cd ~/Pictures/scans
vsplitimg *.jpg
Requires PIL: brew install PIL or pip install -U PIL if you're feeling optimistic
Using pyPdf, open up a PDF, cropping the left and right sides right down the middle, into left and right. Create a new PDF with these crops.
vsplitpdf reconstruction.pdf
A quick pip install pyPdf may be required.
wget | tar -xz helper. Get from url and unpackage based on extension (handles gz and bz2).
wgetar http://ftp.gnu.org/gnu/wget/wget-1.5.3.tar.gz
- Extract all strings that look something like domain names from the supplied files (uses Python's
fileinput) - Fetch all the whois records using the command line
whois - Cache raw whois results in Redis (using the bucket prefix
whois:) becausewhoisservers are finicky. If some whois request times out or the program crashes, simply restart the script and it will resume where it left off. - Extract the expiration date from the whois records and list them all, ordered from soonest expiration to most distant.
whois-domains ~/domains.yamlUsing Python's yaml and json modules, read in yaml and output json. Useful because Node.js's YAML support used to suck.
cat simple_spec.yaml | yaml2json > simple_spec.json
Print out the most common words in a plain text document.
cd ~/corpora/heliohost
<Fre_Newspapers.txt zipf