diff --git a/.gitignore b/.gitignore
index e788c0d..0378848 100644
--- a/.gitignore
+++ b/.gitignore
@@ -175,8 +175,21 @@ venv.bak/
 .dmypy.json
 dmypy.json
 
-# Pyre type checker
-.pyre/
+### Python Patch ###
+.venv/
+
+### Python.VirtualEnv Stack ###
+# Virtualenv
+# http://iamzed.com/2009/05/07/a-primer-on-virtualenv/
+[Bb]in
+[Ii]nclude
+[Ll]ib
+[Ll]ib64
+[Ll]ocal
+[Ss]cripts
+pyvenv.cfg
+pip-selfcheck.json
+
 
 # End of https://www.gitignore.io/api/python,emacs
 with_my
diff --git a/goodreads/__init__.py b/goodreads/__init__.py
new file mode 100755
index 0000000..accac9f
--- /dev/null
+++ b/goodreads/__init__.py
@@ -0,0 +1,159 @@
+#!/usr/bin/env python3
+from pathlib import Path
+from typing import List, Dict, NamedTuple, Iterator, Optional, Sequence
+from datetime import datetime
+import pytz
+
+from lxml import etree as ET # type: ignore
+
+BPATH = Path("/L/backups/goodreads")
+
+# TODO might be useful to keep track of updates?...
+# then I need some sort of system to store diffs in generic way...
+# althogh... coud use same mechanism as for filtering
+def get_last() -> Path:
+    return max(sorted(BPATH.glob('*.xmll')))
+
+_SP = '</review>'
+
+def get_reviews():
+    fname = get_last()
+    xmls = []
+    with open(fname, 'r') as fo:
+        data = fo.read()
+        for xx in data.split(_SP):
+            if len(xx.strip()) == 0:
+                break
+            xmls.append(ET.fromstring(xx + _SP))
+    return xmls
+
+class Book(NamedTuple):
+    bid: str
+    title: str
+    authors: Sequence[str]
+    shelves: Sequence[str]
+    date_added: datetime
+    date_started: Optional[datetime]
+    date_read: Optional[datetime]
+
+from kython import the
+
+
+def _parse_date(s: Optional[str]) -> Optional[datetime]:
+    if s is None:
+        return None
+    res = datetime.strptime(s, "%a %b %d %H:%M:%S %z %Y")
+    assert res.tzinfo is not None
+    return res
+
+
+def iter_books() -> Iterator[Book]:
+    for r in get_reviews():
+        # review_xml = the(review.childNodes)
+        # rdict = {n.tagName: n for n in review_xml.childNodes if isinstance(n, Element)}
+        # fuck xml...
+
+        be    = the(r.xpath('book'))
+        title = the(be.xpath('title/text()'))
+        authors = be.xpath('authors/author/name/text()')
+
+        bid     = the(r.xpath('id/text()'))
+        # isbn_element   = the(book_element.getElementsByTagName('isbn'))
+        # isbn13_element = the(book_element.getElementsByTagName('isbn13'))
+        date_added     = the(r.xpath('date_added/text()'))
+        sss = r.xpath('started_at/text()')
+        rrr = r.xpath('read_at/text()')
+        started_at     = None if len(sss) == 0 else the(sss)
+        read_at        = None if len(rrr) == 0 else the(rrr)
+
+        shelves = r.xpath('shelves/shelf/name/text()')
+
+        # if isbn_element.getAttribute('nil') != 'true':
+        #     book['isbn'] = isbn_element.firstChild.data
+        # else:
+        #     book['isbn'] = ''
+
+        # if isbn13_element.getAttribute('nil') != 'true':
+        #     book['isbn13'] = isbn13_element.firstChild.data
+        # else:
+        #     book['isbn13'] = ''
+
+        da = _parse_date(date_added)
+        assert da is not None
+        yield Book(
+            bid=bid,
+            title=title,
+            authors=authors,
+            shelves=shelves,
+            date_added=da,
+            date_started=_parse_date(started_at),
+            date_read=_parse_date(read_at),
+        )
+
+def get_books():
+    return list(iter_books())
+
+
+def test_books():
+    books = get_books()
+    assert len(books) > 10
+
+
+class Event(NamedTuple):
+    dt: datetime
+    summary: str
+    eid: str
+
+
+def get_events():
+    events = []
+    for b in get_books():
+        events.append(Event(
+            dt=b.date_added,
+            summary=f'Added book "{b.title}"', # TODO shelf?
+            eid=b.bid
+        ))
+        # TODO finished? other updates?
+    return sorted(events, key=lambda e: e.dt)
+
+
+def test():
+    assert len(get_events()) > 20
+
+
+def print_read_history():
+    def key(b):
+        read = b.date_read
+        if read is None:
+            return datetime.fromtimestamp(0, pytz.utc)
+        else:
+            return read
+
+    def fmtdt(dt):
+        if dt is None:
+            return dt
+        tz = pytz.timezone('Europe/London')
+        return dt.astimezone(tz)
+    for b in sorted(iter_books(), key=key):
+        print(f"""
+{b.title} by {', '.join(b.authors)}
+    started : {fmtdt(b.date_started)}
+    finished: {fmtdt(b.date_read)}
+        """)
+
+
+def main():
+    import argparse
+    p = argparse.ArgumentParser()
+    sp = p.add_argument('mode', nargs='?')
+    args = p.parse_args()
+
+    if args.mode == 'history':
+        print_read_history()
+    else:
+        assert args.mode is None
+        for b in iter_books():
+            print(b)
+
+if __name__ == '__main__':
+    main()