#! /usr/bin/gawk -f
# Last edited on 2012-05-05 22:14:47 by stolfilocal
# Sampling functions for latn/ock
# The Dialogus of William of Ockam, III.1.2, mapped to lowercase, with abbrevs.
# To be included in wds-to-tlw

function smp_define_patterns(smp,sec)
{
  # No patterns needed 
  if (sec != "tot.1")
    { data_error(("invalid section \"" sec "\"")); }
}

function smp_is_good_word(smp,sec,type,wd)
{ 
  # Accept only lowercase alpha and "~" (abbrev marker).
  # The text uses neither hyphen nor apostrophe. 
  return (wd ~ /^[~a-z]+$/);
}

function smp_reclassify_word(smp,sec,cursec,curlin,type,wd)
{
  # Delete anything outside section {sec}:
  if (sec == "tot.1")
    { if (cursec !~ /^{3[.]1[.][0-9]+}{c[1-9][0-9]*}{tx/) { return "n"; } }
  else  
    { arg_error(("bad subsection \"" subsec "\"")); }
  
  # Delete titles:
  if (cursec ~ /{(tt|cn)}$/) { return "n"; }
  
  # Reject any subsection that is not prose text:
  if (cursec !~ /{tx[1-9][0-9]*[md]}$/) { return "x"; };
  
  # Discard punctuation other than parag breaks: 
  if ((type == "p") && (wd != "=")) { return "n"; }

  return type;
}

function smp_fix_word(smp,sec,type,wd)
{
  # Map upper case to lower case (just in case).
  wd = tolower(wd);
  return wd;
}