feat(user/transcribe_whisper.sh):Shuffle input files

[BK-2020-03.git] / unitproc / bkcsvjoin

# Date: 2020-02-21T21:23Z;

# Author: Steven Baltakatei Sandoval

# License: This bash script, 'bkcsvjoin.sh', is licensed under GPLv3 or
# later by Steven Baltakatei Sandoval:
#
#    'bkcsvjoin.sh' sorts and joins two CSV files based on first field
#    Copyright (C) 2020  Steven Baltakatei Sandoval (baltakatei.com)
#
#    This program is free software: you can redistribute it and/or modify
#    it under the terms of the GNU General Public License as published by
#    the Free Software Foundation, either version 3 of the License, or
#    any later version.
#
#    This program is distributed in the hope that it will be useful,
#    but WITHOUT ANY WARRANTY; without even the implied warranty of
#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#    GNU General Public License for more details.
#
#    A copy of the GNU General Public License may be found at
#    <https://www.gnu.org/licenses/>.

# Description: Sorts and joins two CSV files based upon the contents
# of the first field ("field1") in each line of CSV file. It assumes
# the first line is a header labelling field names. It outputs several
# files into a new directory in the working directory. Set theory
# characters in output file names indicate file 1 elements with "A"
# and file 2 elements with "B". Input assumes UNIX-style line
# endings. The output files use the following numbers and definitions:
#
#    1. A⋃B     The union of A and B. Contains all field1 elements from file 1 and 2.
#
#    2. A⋃(A⋂B) Contains all field1 elements of file 1 plus the matches between
#               file 1 and 2.
#
#    3. B⋃(B⋂A) Contains all field1 elements of file 2 plus the matches between
#               file 1 and 2.
#
#    4. A⋂B     Contains only field1 elements present in both file 1 and file 2.
#
#    5. A-B     Contains all field1 elements of file 1 except when a matching
#               element is present in file 2.
#
#    6. B-A     Contains all field1 elements of file 2 except when a matching
#               element is present in file 1.
#
#    7. A⊖B     Contains all field1 elements of file 1 and file 2 except for
#               elements which are present in both file 1 and 2.

# Usage: bkcsvjoin.sh  [ file1 ]  [ file2 ] > fileOut.csv

# Dependencies: GNU Coreutils 8.30-3 (bash, join, sort, head, tail,
# mkdir) (see end of file for license details). Note: only tested on
# Debian GNU/Linux 10.

# Version 0.0.1

echoerr() { echo "$@" 1>&2; } # Function for outputing text to stderr.

SCRIPT_DATE=$(date +%Y%m%dT%H%M%S%z)
SCRIPT_PATH=$(pwd)

# Create temporary working directory
TEMP_DIRPATH="$SCRIPT_PATH"/"$SCRIPT_DATE"..temp
mkdir "$TEMP_DIRPATH"

# Read file names as input arguments
FILE1_PATH="$1"
FILE2_PATH="$2"
FILE1_BASENAME=$(basename "$FILE1_PATH")
FILE2_BASENAME=$(basename "$FILE2_PATH")
#[[ -f "$FILE1_PATH" ]] && echoerr "DEBUG:$FILE1_PATH is file with basename $FILE1_BASENAME."
#[[ -f "$FILE2_PATH" ]] && echoerr "DEBUG:$FILE2_PATH is file with basename $FILE2_BASENAME."

# Warn and prompt if carriage return character is used.
if [[ $(grep -Uc $'\x0D' "$FILE1_PATH") -gt 0 ]] || [[ $(grep -Uc $'\x0D' "$FILE2_PATH") -gt 0 ]]; then
    echoerr "ERROR: Carriage Return character(s) detected. Windows text file?"
    read -p "Do you wish to attempt to convert DOS-style line endings to UNIX-style by removing all carriage returns? (probably okay) [y/n]" choice
    case "$choice" in
	y|Y ) echoerr "yes"
	      OPTION_DOS2UNIX="true"
	      ;;
	n|N ) echoerr "no"
	      OPTION_DOS2UNIX="false"
	      echoerr "WARNING:Output may contain mix of DOS-style (CRLF) and UNIX-style (LF) line endings!"
	      echoerr "  See https://stackoverflow.com/a/2613834 "
	      sleep 2
	      ;;
	* ) echo "Invalid selection."
	    echoerr "Exiting." && exit 1;; 
    esac
fi

# Sort files on field1 and write to temporary directory. See [1], and [2].
FILE1_SORTED_PATH="$TEMP_DIRPATH"/"$FILE1_BASENAME"_SORTED.csv
FILE2_SORTED_PATH="$TEMP_DIRPATH"/"$FILE2_BASENAME"_SORTED.csv
( head -n1 $FILE1_PATH && tail -n+2 $FILE1_PATH | sort -t, -k1,1 | uniq ) > $FILE1_SORTED_PATH
( head -n1 $FILE2_PATH && tail -n+2 $FILE2_PATH | sort -t, -k1,1 | uniq ) > $FILE2_SORTED_PATH

# 1. A⋃B. Join files into Union of (file 1) + (file 2) (i.e. throwing nothing out). See [3], [6].
FILES_JOINED_COMPLETE="$( join -t, -a1 -a2 -o auto --header "$FILE1_SORTED_PATH" "$FILE2_SORTED_PATH" )"

# 2. A⋃(A⋂B). Join files, file 1 uniques only (omit unmatched field1 entries from file 2). See [6].
FILES_JOINED_F1UO="$( join -t, -a1 -o auto --header "$FILE1_SORTED_PATH" "$FILE2_SORTED_PATH" )"

# 3. B⋃(A⋂B). Join files, file 2 uniques only (omit unmatched field1 entries from file 1). See [6].
FILES_JOINED_F2UO="$( join -t, -a2 -o auto --header "$FILE1_SORTED_PATH" "$FILE2_SORTED_PATH" )"

# 4. A⋂B. Join files, only entries present in both file 1 and file 2 ( file 1 ⋂ file 2 ). See [6].
FILES_JOINED_INTERSECTION="$( join -t, -o auto --header "$FILE1_SORTED_PATH" "$FILE2_SORTED_PATH" )"

# 5. A-B. Join files into Relative Complement of (file 1) - (file 2) field1 entries. See See [5], [6].
FILES_JOINED_DIFFERENCE_AB="$( join -t, -v1 -o auto --header "$FILE1_SORTED_PATH" "$FILE2_SORTED_PATH" )"

# 6. B-A. Join files into Relative Complement of (file 2) - (file 1) field1 entries. See [5], [6].
FILES_JOINED_DIFFERENCE_BA="$( join -t, -v2 -o auto --header "$FILE1_SORTED_PATH" "$FILE2_SORTED_PATH" )"

# 7. A⊖B. Join files into Symmetric Difference of (file 2) ⊖ (file 1) field1 entries. See [5], [6].
FILES_JOINED_DIFFERENCE_SYM="$( join -t, -v1 -v2 -o auto --header "$FILE1_SORTED_PATH" "$FILE2_SORTED_PATH" )"

# Attempt to restore to UNIX-style line endings by removing all carriage return (CR) characters OPTION_DOS2UNIX set to true. See [7].
if [[ $OPTION_DOS2UNIX == "true" ]]; then
    FILES_JOINED_COMPLETE="$(echo "$FILES_JOINED_COMPLETE" | tr -d '\r')"
    FILES_JOINED_F1UO="$(echo "$FILES_JOINED_F1UO" | tr -d '\r')"
    FILES_JOINED_F2UO="$(echo "$FILES_JOINED_F2UO" | tr -d '\r')"
    FILES_JOINED_INTERSECTION="$(echo "$FILES_JOINED_INTERSECTION" | tr -d '\r')"
    FILES_JOINED_DIFFERENCE_AB="$(echo "$FILES_JOINED_DIFFERENCE_AB" | tr -d '\r')"
    FILES_JOINED_DIFFERENCE_BA="$(echo "$FILES_JOINED_DIFFERENCE_BA" | tr -d '\r')"
    FILES_JOINED_DIFFERENCE_SYM="$(echo "$FILES_JOINED_DIFFERENCE_SYM" | tr -d '\r')"
fi

# Output CSV files to temp directory with carriage returns ('\r') removed.
echo "$FILES_JOINED_COMPLETE" > "$TEMP_DIRPATH"/"1.A⋃B..UNION.csv" && echoerr "1.A⋃B..UNION.csv written to $TEMP_DIRPATH"
echo "$FILES_JOINED_F1UO" > "$TEMP_DIRPATH"/"2.A⋃(A⋂B)..FILE-1-UNIQUES-ONLY.csv" && echoerr "2.A⋃(A⋂B)..FILE-1-UNIQUES-ONLY.csv written to $TEMP_DIRPATH"
echo "$FILES_JOINED_F2UO" > "$TEMP_DIRPATH"/"3.B⋃(A⋂B)..FILE-2-UNIQUES-ONLY.csv" && echoerr "3.B⋃(A⋂B)..FILE-2-UNIQUES-ONLY.csv written to $TEMP_DIRPATH"
echo "$FILES_JOINED_INTERSECTION" > "$TEMP_DIRPATH"/"4.A⋂B..INTERSECTION.csv" && echoerr "4.A⋂B..INTERSECTION.csv written to $TEMP_DIRPATH"
echo "$FILES_JOINED_DIFFERENCE_AB" > "$TEMP_DIRPATH"/"5.A-B..DIFFERENCE_A_MINUS_B.csv" && echoerr "5.A-B..DIFFERENCE_A_MINUS_B.csv written to $TEMP_DIRPATH"
echo "$FILES_JOINED_DIFFERENCE_BA" > "$TEMP_DIRPATH"/"6.B-A..DIFFERENCE_B_MINUS_A.csv" && echoerr "6.B-A..DIFFERENCE_B_MINUS_A.csv written to $TEMP_DIRPATH"
echo "$FILES_JOINED_DIFFERENCE_SYM" > "$TEMP_DIRPATH"/"7.A⊖B..DIFFERENCE_SYM.csv" && echoerr "7.A⊖B..DIFFERENCE_SYM.csv written to $TEMP_DIRPATH"

# ==References==
# [1]: How to sort a csv file while retaining the first line. https://stackoverflow.com/a/14562674
# [2]: How to sort a csv file by sorting on a single field. https://stackoverflow.com/a/44744800
# [3]: How to join multiple sorted csv files into one. https://stackoverflow.com/a/27606199
# [4]: How to join a csv file retaining the first line, sorting by certain fields, and not omitting nonmatching fields. https://superuser.com/a/1527095
# [5]: How to determine symmetric difference between two sorted files. https://www.gnu.org/software/coreutils/manual/html_node/Set-operations.html
# [6]: Set theory symbols https://www.rapidtables.com/math/symbols/Set_Symbols.html
# [7]: How to convert DOS-style to UNIX-style line endings. https://stackoverflow.com/a/2613834

# ==Dependencies==

    # GNU bash, version 5.0.3(1)-release (x86_64-pc-linux-gnu)
    # Copyright (C) 2019 Free Software Foundation, Inc.
    # License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>

    # This is free software; you are free to change and redistribute it.
    # There is NO WARRANTY, to the extent permitted by law.


    # join (GNU coreutils) 8.30
    # Copyright (C) 2018 Free Software Foundation, Inc.
    # License GPLv3+: GNU GPL version 3 or later <https://gnu.org/licenses/gpl.html>.
    # This is free software: you are free to change and redistribute it.
    # There is NO WARRANTY, to the extent permitted by law.

    # Written by Mike Haertel.


    # sort (GNU coreutils) 8.30
    # Copyright (C) 2018 Free Software Foundation, Inc.
    # License GPLv3+: GNU GPL version 3 or later <https://gnu.org/licenses/gpl.html>.
    # This is free software: you are free to change and redistribute it.
    # There is NO WARRANTY, to the extent permitted by law.

    # Written by Mike Haertel and Paul Eggert.


    # head (GNU coreutils) 8.30
    # Copyright (C) 2018 Free Software Foundation, Inc.
    # License GPLv3+: GNU GPL version 3 or later <https://gnu.org/licenses/gpl.html>.
    # This is free software: you are free to change and redistribute it.
    # There is NO WARRANTY, to the extent permitted by law.

    # Written by David MacKenzie and Jim Meyering.


    # tail (GNU coreutils) 8.30
    # Copyright (C) 2018 Free Software Foundation, Inc.
    # License GPLv3+: GNU GPL version 3 or later <https://gnu.org/licenses/gpl.html>.
    # This is free software: you are free to change and redistribute it.
    # There is NO WARRANTY, to the extent permitted by law.

    # Written by Paul Rubin, David MacKenzie, Ian Lance Taylor,
    # and Jim Meyering.


    # mkdir (GNU coreutils) 8.30
    # Copyright (C) 2018 Free Software Foundation, Inc.
    # License GPLv3+: GNU GPL version 3 or later <https://gnu.org/licenses/gpl.html>.
    # This is free software: you are free to change and redistribute it.
    # There is NO WARRANTY, to the extent permitted by law.

    # Written by David MacKenzie.

# ==Metadata==
# - Created in task 20200220_90eb0c7d