From 975f9dd110ab7fcd35b4b74c2686660f14c1b311 Mon Sep 17 00:00:00 2001
From: Dima Gerasimov <karlicoss@gmail.com>
Date: Fri, 31 Jul 2020 16:46:07 +0100
Subject: [PATCH] rescuetime: get rid of kython, use cachew

---
 my/rescuetime.py | 66 +++++++++++++++++++++++-------------------------
 1 file changed, 32 insertions(+), 34 deletions(-)

diff --git a/my/rescuetime.py b/my/rescuetime.py
index 5bf136c..4b46e2e 100644
--- a/my/rescuetime.py
+++ b/my/rescuetime.py
@@ -4,13 +4,13 @@ Rescuetime (activity tracking) data
 
 from pathlib import Path
 from datetime import datetime, timedelta
-from typing import NamedTuple, Dict, List, Set, Optional
+from typing import Sequence, Iterable
 
-from .common import get_files, LazyLogger
-from .error import Res, split_errors
+from .core import get_files, LazyLogger
+from .core.common import mcachew
+from .core.error import Res, split_errors
 
-# TODO get rid of it
-from kython import group_by_cmp # type: ignore
+import more_itertools
 
 from my.config import rescuetime as config
 
@@ -18,51 +18,49 @@ from my.config import rescuetime as config
 log = LazyLogger(__package__, level='info')
 
 
-def inputs():
-    return get_files(config.export_path, '*.json')
+def inputs() -> Sequence[Path]:
+    return get_files(config.export_path)
 
 
-import my.config.repos.rescuexport.model as rescuexport
-Model = rescuexport.Model
+import my.config.repos.rescuexport.dal as dal
+DAL = dal.DAL
+Entry = dal.Entry
 
 
-# TODO cache?
-def get_model(last=0) -> Model:
-    return Model(inputs()[-last:])
-
-
-def _without_errors():
-    model = get_model()
-    it = model.iter_entries()
+# todo needs to be cumulative cache
+@mcachew
+def entries(files=inputs()) -> Iterable[Entry]:
+    dal = DAL(files)
+    it = dal.iter_entries()
     vit, eit = split_errors(it, ET=Exception)
-    # TODO FIXME handle eit somehow?
+    # todo handle errors, I guess initially I didn't because it's unclear how to easily group?
     yield from vit
 
 
+def groups(gap=timedelta(hours=3)):
+    vit = entries()
+    from more_itertools import split_when
+    yield from split_when(vit, lambda a, b: (b.dt - a.dt) > gap)
 
-def get_groups(gap=timedelta(hours=3)):
-    vit = _without_errors()
-    lit = list(vit) # TODO get rid of it...
-    return group_by_cmp(lit, lambda a, b: (b.dt - a.dt) <= gap, dist=1)
+
+def stats():
+    from .core import stat
+    return {
+        **stat(groups),
+        **stat(entries),
+    }
+
+
+# todo not sure if I want to keep these here? vvv
 
 
 def print_groups():
-    for gr in get_groups():
+    for gr in groups():
          print(f"{gr[0].dt}--{gr[-1].dt}")
     # TODO merged db?
     # TODO ok, it summarises my sleep intervals pretty well. I guess should adjust it for the fact I don't sleep during the day, and it would be ok!
 
 
-def check_backed_up(hours=24):
-    vit = _without_errors()
-    # TODO use some itertools stuff to get a window only?
-    last = list(vit)[-1]
-    latest_dt = last.dt
-
-    assert (datetime.now() - latest_dt) < timedelta(hours=hours)
-    # TODO move this to backup checker??
-
-
 def fill_influxdb():
     from influxdb import InfluxDBClient # type: ignore
     client = InfluxDBClient()
@@ -70,7 +68,7 @@ def fill_influxdb():
     db = 'test'
     client.drop_database(db)
     client.create_database(db)
-    vit = _without_errors()
+    vit = entries()
     jsons = [{
         "measurement": 'phone',
         "tags": {},