tame/build-aux/csvm-expand

#!/usr/bin/awk -f
#
# Expands a "magic" CSV file into a normal CSV
#
#   Copyright (C) 2014-2023 Ryan Specialty, LLC.
#
#   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
#   (at your option) 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.
#
#   You should have received a copy of the GNU General Public License
#   along with this program.  If not, see <http://www.gnu.org/licenses/>.
#
# "Magic" CSVs simply exist to make life easier: they permit comments, blank
# lines, variables, sub-delimiter expansion, and any number of ranges per line.
# Ranges will be expanded in every combination, making rate tables highly
# maintainable.
#
# Variables are also supported when defined using :var=val. Variables may
# expand into ranges, 'cause they're awesome. Multiple variables may be
# delimited by semi-colons, as may multiple values.
#
# For example:
#   :foo=1--3
#   $foo;7;9--10:$foo, 5--10,1/1/2017
#
# Would generate:
#   1, 5, 1483246800
#   1, 6, 1483246800
#   ...
#   5, 10, 1483246800
#   2, 5, 1483246800
#   ...
#   9, 5, 14832468005
#   ...
#   1, 5, 1483246800
#   1, 6, 1483246800
#   ...
##

BEGIN {
  date_cmd = "stdbuf -o0 date -f- +%s"
}

END {
  close( date_cmd )
}


# Parse a date string into a Unix timestamp (memoized)
#
# This spawns a single process for date and reads from standard in.  Even
# then, though, date parsing is very slow for many thousands of rows, so the
# output is also cached in `date_cache'.
function parse_date( i,   src )
{
  src = $i

  if ( date_cache[ src ] )
  {
    $i = date_cache[ src ]
    return
  }

  print $i |& date_cmd
  date_cmd |& getline $i

  date_cache[ src ] = $i;
}


# Expand variable with its value, if any
function expand_vars( s,   value )
{
  # attempt to parse variable (may expand into a range)
  if ( match( s, /^\$([a-zA-Z_-]+)$/, m ) )
  {
    value = vars[ m[1] ];

    if ( value == "" )
    {
      print "error: unknown variable reference: `$" m[1] "'" > "/dev/stderr"
      exit 1
    }

    return value
  }

  return s
}


# Expand line
function parseline( i, m, j, me, orig )
{
  if ( i > NF )
  {
    print
    return
  }

  orig = $i

  # expand variables before any processing so that expansions
  # can include any type of formatting
  $i = expand_vars( $i )

  if ( match( $i, /^([0-9]+\/){2}[0-9]+$/, m ) )
  {
    parse_date( i );
  }

  # check first for delimiters
  if ( match( $i, /^([^;]+);(.*)$/, m ) )
  {
    # give it a shot with the first value
    $i = m[1]
    parseline( i )

    # strip off the first value and process with following value(s)
    $i = m[2]
    parseline( i )

    # we've delegated; we're done
    $i = orig
    return
  }

  # parse range
  if ( match( $i, /^([^-]+)--([^-]+)$/, m ) )
  {
    j  = expand_vars( m[1] )
    me = expand_vars( m[2] )

    if ( !match( j, /^[0-9]+$/ ) || !match( me, /^[0-9]+$/ ) )
    {
      print "error: invalid range: `" $i "'" > "/dev/stderr"
      exit 1
    }

    do
    {
      $i = j
      parseline( i + 1 )
    } while ( j++ < me )
  }
  else
  {
    parseline( i + 1 );
  }

  # restore to original value
  $i = orig
}


BEGIN {
  # we're parsing CSVs
  FS = "[[:space:]]*,[[:space:]]*"
  OFS = ","

  has_directives = 0
  directives     = "!(NODIRECTIVES)"
}


# skip all lines that begin with `#', which denotes a comment, or are empty
/^#|^$/ { next; }

# directives are echoed back and are intended for processing by
# the parent csvm2csv script
/^!/ && output_started {
  print "error: directive must appear before header: `" $0 "'" > "/dev/stderr"
  exit 1
}
/^!/ && has_directives {
  print "error: all directives must be on one line: `" $0 "'" > "/dev/stderr"
  exit 1
}
/^!/ {
  has_directives = 1
  directives     = $0

  next
}

# lines that begin with a colon are variable definitions
/^:/ {
  if ( !match( $0, /^:([a-zA-Z_-]+)=(.*?)$/, m ) )
  {
      print "error: invalid variable definition: `" $0 "'" > "/dev/stderr"
      exit 1
  }

  vars[ m[1] ] = m[2]
  next
}

# Always begin output with a line for directives, even if there are
# none.  This makes subsequent processing much easier, since we won't have
# to conditionally ignore the top line.
!output_started {
  print directives

  output_started = 1
}

# lines that need any sort of processing (ranges, dates, etc)
/--|;|\$[a-zA-Z_-]|\// { parseline( 1 ); next; }

# all other lines are normal; simply output them verbatim
{
  # this assignment will ensure that awk processes the output, ensuring that
  # extra spaces between commas are stripped
  $1=$1
  print
}