Get a webpage and all the pages it links to for local viewing with wget

wget -r -l1 -p -np -k -nH -P 'outdir' http://www.aaa/bbb.html

#download bbb.html
#-r: recursivelly
#-l1: up to level 1 links, that is, download only the pages imediately linked to by www.aaa/bbb.html
#   (-l1, could be -l2, -l3, ... for nth level links)
#-p: saving all necessary files to view the page locally (images, css)
#-np: not downloading parent directories pages
#-k: converting links in htmls that were downloaded into local links for offline viewing
#-nH: not creating a directory 'www.aaa' as would be default
#-P 'outdir': saving output inside a directory 'outdir', and creating it if it doesn't exist

Synctex forward search example

1. make a multipage tex document with INPUT_PATH

   INPUT_PATH="/path/to/tex/file/asdf.tex"

2. compile with:

   pdflatex -synctex=1 "$INPUT_PATH"

   this should generate “asdf.pdf” and the synctex file ( because of -synctex=1)

3. run synctex as

   synctex view -i 77:1:/path/to/tex/file/./asdf.tex -o asdf.pdf

   77:1 means you are on line 77 column 1 of the tex file, and you want to get the position on the pdf. Note the ugly “/./” between the parent directory and the basename. It must be there!

4. sample output

   This is SyncTeX command line utility, version 1.2
   SyncTeX result begin
   Output:report.pdf
   Page:2
   x:261.757507
   y:492.330658
   h:255.095062
   v:495.319450
   W:100.061378
   H:13.902562
   before:
   offset:0
   middle:
   after:
   SyncTeX result end

   amongst other things, this is telling you that this position on the tex file corresponds to Page 2 of your pdf.

5. Now all you have to do is to tell your pdf viewer to open that page. In Okular this would be:

   okular -p 2 asdf.pdf

Reasons to learn the bash command line

Bash suffers from a few disadvantages

• Linux only
• quite complicated variable/parameters passing

which in my opinion make python a better choice for anything that is not a one liner.

On the other hand, learning bash and the basic Linux commands can be very useful for a programmer

• interface. the bash command line contains interface that is widely recognized and that has survived the test of time, meaning that you will be able to understand others programs and create programs that are more easily understandable. For example, if you name a method ‘find’, which does more or less the same as the Linux command, chances are that others will immediately understand and remember this, which is a crucial step to coping with complexity of programs.
• raw power. some standard bash utilities are so amazingly powerful for the amount of code that you have to write, that it is just worth learning it.

remove duplicates by content with fdupes

sudo apt-get install fdupes
fdupes -rd dirA dirB dirC

Will look for duplicates

• under relative paths dirA, dirB and dirC
• recursively (-r)
• ask print duplicate paths one by one and ask which you want to keep (can also type all to keep all) and delete the others (-d)

cd inside a script does not change the directory of current terminal

echo 'cd ..' > test && chmod +x test

./test

does not nothing because ./test is executed is a subshell

. test

works, because the ‘.’ operator does does the same thing as copying the file test and pasting inside the current shell

Bash pdf command

Linux command line is such a powerhouse it already comes with a pdf command! pdf –help to learn of course

Tee bash command

Useful, simple AND visually appealing. Beautiful. See Wiki link and by examples link.