diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..2f0f6a3
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,3 @@
+*-raw.txt
+scan
+pts
\ No newline at end of file
diff --git a/build-pts.js b/build-pts.js
new file mode 100644
index 0000000..af858d2
--- /dev/null
+++ b/build-pts.js
@@ -0,0 +1,67 @@
+const fs=require("fs");
+const folder="../pts-dhammakaya/data/";
+const {createbuilder}=require("dengine");
+
+const booknames=[
+	'mv','cv','vb1'	,'vb2','pvr'
+	,'dn1','dn2','dn3'//6~8
+	,'mn1','mn2','mn3'//9~11
+	,'sn1','sn2','sn3','sn4','sn5'//12~16
+	,'an1','an2','an3','an4','an5'//17~21
+	,'kp','dhp','ud','iti','snp'  // 26
+	,'vv','pv', 'thag'            //27,28,29
+	,'ja1','ja2','ja3','ja4','ja5','ja6'  //30~35
+	,'mnd','cnd','ps1','ps2'              //36,37,38,39
+	,'ap'  ,'bv','cp',                  //40,41,42
+	,'ds','vb','dt','pp','kv'         //43,44,45,46,47
+	,'ya1','ya2'                       // 48,49
+	,'pt1','pt2','pt3','pt4','pt5','pt6'//50,51,52,53,54,55
+	
+]
+const LASTBOOK='pt1';//pt is not complete in dhammakaya database 
+const rawtext=[];
+
+const dofile=(fn,builder)=>{
+	console.log("processing",fn);
+	let content=fs.readFileSync(fn,"utf8").split(/\r?\n/);
+	let prevbk=0,page=0,line=0;
+	const MAXLINE=content.length;
+	for (var i=0;i<MAXLINE;i++) {
+		const arr=content[i].split("\t");
+		let bkseq=parseInt(arr[0])-1;
+		if (bkseq>52)bkseq=52;
+		const bk=booknames[bkseq],page=parseInt(arr[1]);
+
+		if (arr.length<3) continue;
+		const lines=arr[2].split("\\n");
+		if (lines[lines.length-1]=="") lines.pop();
+
+		if (prevbk&&prevbk!==bk) {
+			builder.addbook(prevbk);
+		}
+		for (var j=0;j<lines.length;j++){
+			builder.addline(lines[j]);
+			if (outputrawtext) {
+				rawtext.push(bk+":"+page+"."+(j+1)+"\t"+lines[j]);
+			}
+		}
+		builder.addpage(page);
+		prevbk=bk;
+	}
+	builder.addbook(LASTBOOK);
+}
+
+const outputrawtext=true;
+
+const build=(tsv,name)=>{
+	const out=[],footnote=[];
+	const builder=createbuilder({name});
+	dofile(folder+tsv,builder);
+	builder.done();
+
+	if (outputrawtext) {
+		fs.writeFileSync("pts-raw.txt",rawtext.join("\n"),"utf8");
+	}
+}
+
+build('palipg1.tsv','pts');
\ No newline at end of file
diff --git a/gen-pts.js b/gen-pts.js
new file mode 100644
index 0000000..e69de29
diff --git a/pts-errata.js b/pts-errata.js
new file mode 100644
index 0000000..6b67248
--- /dev/null
+++ b/pts-errata.js
@@ -0,0 +1,31 @@
+word: Kāma
+— -ādhikaraṇa having its cause in desire mn1_p85 ; sn1_p74 ;
+should be mn1_p86
+
+"word": "iddhi", nested ref
+<span class='ref'>Vin.209</span>, <span class='ref'>Vin.240</span>; <span class='ref'>Vin.<span class='ref'>DN.iii.</span>iii.67</span>;
+
+
+	
+	
+			"kalya-rūpa",
+	move from nati natta
+
+
+
+       "word": "māsati",
+        "text": "<dl id='māsati'><dt><dfn>Māsati, Māsana, Māsin</dfn></dt><dd><p> touch, touching, etc. in sense of eating or taking in. So is probably to be read for āsati etc. in the foll. passages, where <i class='term'>m</i> precedes this <i class='term'><a href='/define/ā'>ā</a></i> in all cases Otherwise we have to refer them to a root <i class='term'>ās = as</i> (to eat) and consider the <i class='term'>m</i> as partly euphonic.<span class='square'>■</span> dumapakkāni <i class='term'>-māsita</i> <span class='ref'>Ja.ii.446</span> (C. reads māsita & explains by asita, dhāta); visa-māsita <span class='ref'>Mil.302</span> (T. reads visamāsita) having taken in poison; visa- <i class='term'>māsan-</i> ûpatāpa (id.) <span class='ref'>Vism.166</span>; tiṇa <i class='term'>-māsin</i> eating grass <span class='ref'>Ja.vi.354</span> (= tiṇakhādaka C.)<br><span class='square'>■</span> A similar case where Sandhi-m-has led to a wrong partition of syllables and has thus been lost through syncope may be P. eḷaka<sup>1</sup>, as comp<sup>d</sup> with Sk. methi (cp. Prk. meḍhi), pillar, post.</p><p class='eti'>fr. <i class='term'>mṛṣ</i>, for massati etc.; see <i class='term'><a href='/define/masati'>masati</a></i></p></dd></dl>"
+           
+//report to suttacentral/sc-data/issues#74
+
+
+ "word": "jimha",
+ 	 '(cp. vanka, opp. uju |','(cp. vanka, opp. uju )'
+ 	 
+
+        "word": "māsati",
+        "text": "<dl id='māsati'><dt><dfn>Māsati, Māsana, Māsin</dfn></dt><dd><p><span class='square'>■</span> dumapakkāni <i class='term'>-māsita</i> <span class='ref'>Ja.ii.446</span> (C. reads māsita & explains by asita, dhāta); visa-māsita <span class='ref'>Mil.302</span> (T. reads visamāsita) having taken in poison; visa- <i class='term'>māsan-</i> ûpatāpa (id.) <span class='ref'>Vism.166</span>; tiṇa <i class='term'>-māsin</i> eating grass <span class='ref'>Ja.vi.354</span> (= tiṇakhādaka C.)<br><span class='square'>■</span> A similar case where Sandhi-m-has led to a wrong partition of syllables and has thus been lost through syncope may be P. eḷaka<sup>1</sup>, as comp<sup>d</sup> with Sk. methi (cp. Prk. meḍhi), pillar, post.</p><p class='eti'>fr. <i class='term'>mṛṣ</i>, for massati etc.; see <i class='term'><a href='/define/masati'>masati</a></i>| touch, touching, etc. in sense of eating or taking in. So is probably to be read for āsati etc. in the foll. passages, where <i class='term'>m</i> precedes this <i class='term'><a href='/define/ā'>ā</a></i> in all cases Otherwise we have to refer them to a root <i class='term'>ās = as</i> (to eat) and consider the <i class='term'>m</i> as partly euphonic.</p></dd></dl>"
+        	
+        	
+        "word": "māsati",
+        "text": "<dl id='māsati'><dt><dfn>Māsati, Māsana, Māsin</dfn></dt><dd><p>touch, touching, etc. in sense of eating or taking in. So is probably to be read for āsati etc. in the foll. passages, where <i class='term'>m</i> precedes this <i class='term'><a href='/define/ā'>ā</a></i> in all cases Otherwise we have to refer them to a root <i class='term'>ās = as</i> (to eat) and consider the <i class='term'>m</i> as partly euphonic.</p><span class='square'>■</span> dumapakkāni <i class='term'>-māsita</i> <span class='ref'>Ja.ii.446</span> (C. reads māsita & explains by asita, dhāta); visa-māsita <span class='ref'>Mil.302</span> (T. reads visamāsita) having taken in poison; visa- <i class='term'>māsan-</i> ûpatāpa (id.) <span class='ref'>Vism.166</span>; tiṇa <i class='term'>-māsin</i> eating grass <span class='ref'>Ja.vi.354</span> (= tiṇakhādaka C.)<br><span class='square'>■</span> A similar case where Sandhi-m-has led to a wrong partition of syllables and has thus been lost through syncope may be P. eḷaka<sup>1</sup>, as comp<sup>d</sup> with Sk. methi (cp. Prk. meḍhi), pillar, post.</p><p class='eti'>fr. <i class='term'>mṛṣ</i>, for massati etc.; see <i class='term'><a href='/define/masati'>masati</a></i></p> </dd></dl>"
\ No newline at end of file