####################################################
# UNIX - THE BASIC COMMANDS
# -- created by judith degen on 10/17/2009
####################################################

************************************************
* PART I: navigating the directory structure
************************************************

# Log onto slate:
# ssh USERNAME@slate.hlp.rochester.edu
ssh jdegen@slate.hlp.rochester.edu

# Change your password!
passwd

# Show the contents of your current directory. If you just logged on, this will be your home directory.
ls

# For more information about file size/type/permissions
ls -l

# Show the contents of the directory /corpora/TDTlite.
ls /p/hlp/tools/TDTlite


# Move to the directory /corpora/TDTlite
cd /corpora/TDTlite

# Show the contents of the directory /corpora/TDTlite/sample_project
ls /p/hlp/tools/TDTlite/sample_project

# Move back to your home directory.
cd

# Move one directory up.
cd ..

# Check where you are.
pwd

# Figure out how to use a command.
# man COMMANDNAME
man cp

# Copy the sample_project directory to your home directory with cp and rename it with mv.
cp -r /p/hlp/tools/TDTlite/sample_project .
mv sample_project myproject

# Create a directory.
mkdir mydir

# Remove the directory you just created from your home directory. BE VERY CAREFUL WITH THE rm COMMAND!
rm -r mydir

# To copy file myfile.txt from your home directory on the server to your current directory on your computer:
scp jdegen@slate.hlp.rochester.edu:./myfile.txt .

# To copy the directory myproject from your home directory on the server to your current directory on your computer:
scp -r jdegen@slate.hlp.rochester.edu:./myproject .

************************************************
* PART II: the basics of tgrep2
************************************************

# Run tgrep2
tgrep2 "ADJP"

# Run tgrep2 on the Wall Street Journal. Search for all VPs
tgrep2 -c /corpora/TGrep2able/wsj_mrg.t2c.gz "ADJP"

# The same, but print only terminals
tgrep2 -c /corpora/TGrep2able/wsj_mrg.t2c.gz -t "ADJP"

# Print the entire sentence
tgrep2 -c /corpora/TGrep2able/wsj_mrg.t2c.gz -tw "ADJP"

# Save the output to a file
tgrep2 -c /corpora/TGrep2able/wsj_mrg.t2c.gz -tw "ADJP" > adjp.txt

# View the contents of adjp.txt
less adjp.txt

# Search inside adjp.txt for "awesome":
/awesome

# Count the lines in adjp.txt
wc -l adjp.txt

# Output the match ID in front of the match itself. \t is a special character that inserts a tab, similarly \n inserts a newline
tgrep2 -c /corpora/TGrep2able/wsj_mrg.t2c.gz -m '%xm\t%tm\n' "ADJP"

# Always use the -af options, they make sure all your matches are found if for example there are multiple matches within one sentence

# Two ADJP that are sisters, print first one, tab, second one
tgrep2 -c /corpora/TGrep2able/wsj_mrg.t2c.gz -m '%t=a1= \t %t=a2=\n' "ADJP=a1 $ ADJP=a2"

# Create a MACRO file
vi

# In the vi, there are a number of commands you can use:
# :w FILENAME - save file as FILENAME
# :q - quit
# :wq - save and quit
# 0 - move to start of line
# $ - move to end of line
# 1G - move to first line
# G - move to last line
# i - insert text before cursor, until <Esc> is hit
# a - insert text after cursor, until <Esc> is hit
# r - replace single character under cursor
# R - replace characters until <Esc> is hit
# x - delete character under cursor
# dd - delete entire current line
# yy - copy the current line
# p - paste the line(s) in the buffer into the text after the current line

# Create a macro @AA that contains the ADJP pattern from above:
i
@ AA	ADJP=a1 $ ADJP=a2;
<Esc>
:w MACRO.ptn
:q

************************************************
* PART III: regular expressions
************************************************

# Regular expressions in tgrep2 belong between //
# Probably the most useful one will be /^NODENAME_START/ - which finds all nodes that begin with NODENAME_START
tgrep2 -c /corpora/TGrep2able/wsj_mrg.t2c.gz "/^ADJP/"

# Special characters:
# ^ - start of string
# $ - end of string
# . - any character
# * - any node
# | - any of the strings separated by |

# /^AD/ matches "ADJP", "ADJP-PRD", "ADVP", "ADVP-LOC", "ADVP-MNR"...
# /VP$/ matches "ADVP", "VP"
# /AD.P/ matches "ADVP", "ADJP", "WHADVP"
# /ADVP|ADJP/ matches "ADVP", "ADJP"