Tofugu Pronunciation Examples

Added some examples showing off the tool using the pitch patterns in the
Tofugu Pronunciation article.

tofugu_files/convert_tofugu_list.sh

@@ -0,0 +1,71 @@
+#!/bin/bash
+
+# First argument is input file
+if [[ $# -ne 1 ]]; then
+	echo """\
+$0:
+A tool to convert copy/paste source files from the Tofugu site's pitch
+tables to more pretty graphics for humans.
+
+Incorrect usage:
+	$0 <input_filename.list>
+	
+Format example:
+	LH 	名 	な
+	LHH 	水 	みず
+	LHHH 	会社 	かいしゃ
+	LHHHH 	大学 	だいがく
+	LHHHHH 	中国語 	ちゅうごくご
+	<pitch>\\t<unique name>\\t<phonetics>
+	
+	Input is proccessed with bash's \`read\` so spacing is not critical."""
+	
+	exit
+fi
+
+SRC="$1"
+MODIFY_SET=(ゃ ゅ ょ ぁ ぃ ぅ ぇ ぉ) # these characters don't count towards a mora length
+MARU='○'
+
+IND=0
+PFX="1-"
+while read INPUT; do
+	IND=$(($IND+1))
+	#echo "$INPUT"
+	ARGS=($INPUT)
+	CODE=${ARGS[0]}
+	NAME=${ARGS[1]}
+	KANA=${ARGS[2]}
+	
+	# Compute core word length
+	KANA_TMP=$KANA
+	for MiniMora in ${MODIFY_SET[@]}; do
+		KANA_TMP=$(sed "s/$MiniMora//g" <<< $KANA_TMP)
+	done
+	MORA_LENGTH=${#KANA_TMP}
+	# first find the length of the word, count the LH stuff
+	if [[ ${CODE:0:1} == 'H' ]]; then # 頭高
+		DOWN_STEP=1
+	else
+		CODE_TMP="${CODE:1}"
+		CODE_TMP=${CODE_TMP%H*}
+		DOWN_STEP=$((${#CODE_TMP}+2)) # downstep location
+		if [[ $DOWN_STEP -gt $MORA_LENGTH ]]; then
+			DOWN_STEP=0
+		fi
+	fi
+
+	# Lad when tofugu assumes more mora in their code (or a particle)
+	DELTA=$(($MORA_LENGTH-${#CODE}))
+	if [[ $DELTA -eq -1 ]]; then
+		KANA="$KANA$MARU"
+	elif [[ $DELTA -ne 0 ]]; then
+		echo "WARNING: Length error in line $IND: '$INPUT'" >&2
+	fi
+	
+	#echo $DOWN_STEP/$MORA_LENGTH
+	echo ${PFX}${IND}-$NAME:$KANA:$DOWN_STEP/$MORA_LENGTH
+done < "$SRC"
+
+
+