personal memory agent
Merge branch 'hopper-a5od7hcg-import-dry-run'
+181
-18
+103
tests/test_importer.py
+103
tests/test_importer.py
···
413
413
seg_key, seg_dir, seg_files = segments[0]
414
414
assert seg_key == "120001_300" # Deconflicted key
415
415
assert seg_dir == day_dir / "import.audio" / "120001_300"
416
+
417
+
418
+
def test_importer_dry_run_text(tmp_path, monkeypatch, capsys):
419
+
"""Test --dry-run for text import prints plan without writing files."""
420
+
mod = importlib.import_module("think.importers.cli")
421
+
422
+
txt = tmp_path / "sample.txt"
423
+
txt.write_text("hello\nworld\n")
424
+
425
+
monkeypatch.setenv("JOURNAL_PATH", str(tmp_path))
426
+
monkeypatch.setattr(
427
+
"sys.argv",
428
+
["sol import", str(txt), "--timestamp", "20240101_120000", "--dry-run"],
429
+
)
430
+
431
+
mod.main()
432
+
433
+
captured = capsys.readouterr()
434
+
assert "File:" in captured.out
435
+
assert "Size:" in captured.out
436
+
assert "Timestamp:" in captured.out
437
+
assert "Source:" in captured.out
438
+
assert "Stream:" in captured.out
439
+
assert "Target day:" in captured.out
440
+
assert "Content:" in captured.out
441
+
assert "characters" in captured.out
442
+
assert "lines" in captured.out
443
+
assert "import.text" in captured.out
444
+
assert "20240101" in captured.out
445
+
assert "12 characters" in captured.out
446
+
assert "2 lines" in captured.out
447
+
448
+
assert not (tmp_path / "imports").exists()
449
+
assert not (tmp_path / "20240101").exists()
450
+
451
+
452
+
def test_importer_dry_run_audio(tmp_path, monkeypatch, capsys):
453
+
"""Test --dry-run for audio import prints plan without writing files."""
454
+
mod = importlib.import_module("think.importers.cli")
455
+
456
+
mp3 = tmp_path / "sample.mp3"
457
+
mp3.write_bytes(b"fake audio")
458
+
459
+
monkeypatch.setenv("JOURNAL_PATH", str(tmp_path))
460
+
monkeypatch.setattr(mod, "_get_audio_duration", lambda p: 420.0)
461
+
callosum_cls = MagicMock()
462
+
monkeypatch.setattr(mod, "CallosumConnection", callosum_cls)
463
+
monkeypatch.setattr(
464
+
"sys.argv",
465
+
["sol import", str(mp3), "--timestamp", "20240101_120000", "--dry-run"],
466
+
)
467
+
468
+
mod.main()
469
+
470
+
captured = capsys.readouterr()
471
+
assert "File:" in captured.out
472
+
assert "Size:" in captured.out
473
+
assert "Timestamp:" in captured.out
474
+
assert "Source:" in captured.out
475
+
assert "Stream:" in captured.out
476
+
assert "Target day:" in captured.out
477
+
assert "Duration:" in captured.out
478
+
assert "Segments:" in captured.out
479
+
assert "Keys:" in captured.out
480
+
assert "import.audio" in captured.out
481
+
assert "20240101" in captured.out
482
+
assert "7.0 minutes" in captured.out
483
+
assert "2 (5-minute chunks)" in captured.out
484
+
assert "120000_300" in captured.out
485
+
assert "120500_300" in captured.out
486
+
487
+
assert not (tmp_path / "imports").exists()
488
+
assert not (tmp_path / "20240101").exists()
489
+
assert callosum_cls.call_count == 0
490
+
491
+
492
+
def test_importer_dry_run_auto(tmp_path, monkeypatch, capsys):
493
+
"""Test --dry-run with --auto detects timestamp and prints summary."""
494
+
mod = importlib.import_module("think.importers.cli")
495
+
496
+
txt = tmp_path / "notes.txt"
497
+
txt.write_text("meeting notes")
498
+
499
+
monkeypatch.setenv("JOURNAL_PATH", str(tmp_path))
500
+
monkeypatch.setattr(
501
+
mod, "detect_created", lambda p, **kw: {"day": "20240315", "time": "140000"}
502
+
)
503
+
monkeypatch.setattr(
504
+
"sys.argv",
505
+
["sol import", str(txt), "--auto", "--dry-run"],
506
+
)
507
+
508
+
mod.main()
509
+
510
+
captured = capsys.readouterr()
511
+
assert "Detected timestamp: 20240315_140000" in captured.out
512
+
assert "auto-importing" in captured.out
513
+
assert "import.text" in captured.out
514
+
assert "Target day: 20240315" in captured.out
515
+
assert "Content:" in captured.out
516
+
517
+
assert not (tmp_path / "imports").exists()
518
+
assert not (tmp_path / "20240315").exists()
+78
-18
think/importers/cli.py
+78
-18
think/importers/cli.py
···
14
14
15
15
from think.callosum import CallosumConnection
16
16
from think.detect_created import detect_created
17
-
from think.importers.audio import prepare_audio_segments
17
+
from think.importers.audio import _get_audio_duration, prepare_audio_segments
18
18
from think.importers.shared import (
19
19
_get_relative_path,
20
20
_is_in_imports,
21
21
_setup_import,
22
22
)
23
-
from think.importers.text import process_transcript
23
+
from think.importers.text import _read_transcript, process_transcript
24
24
from think.importers.utils import save_import_segments
25
25
from think.streams import stream_name, update_stream, write_segment_stream
26
26
from think.utils import day_path, get_journal, get_rev, segment_key, setup_cli
···
195
195
help="Auto-accept detected timestamp and proceed with import",
196
196
)
197
197
parser.add_argument(
198
+
"--dry-run",
199
+
action="store_true",
200
+
help="Show what would be imported without writing to the journal",
201
+
)
202
+
parser.add_argument(
198
203
"--backends",
199
204
action="store_true",
200
205
help="List syncable importer backends",
···
267
272
if not TIME_RE.fullmatch(args.timestamp):
268
273
raise SystemExit("timestamp must be in YYYYMMDD_HHMMSS format")
269
274
275
+
base_dt = dt.datetime.strptime(args.timestamp, "%Y%m%d_%H%M%S")
276
+
day = base_dt.strftime("%Y%m%d")
277
+
278
+
# Derive stream identity for this import
279
+
if args.source:
280
+
import_source = args.source
281
+
else:
282
+
# Auto-detect from file extension
283
+
_ext = os.path.splitext(args.media)[1].lower()
284
+
if _ext == ".m4a":
285
+
import_source = "apple"
286
+
elif _ext in {".txt", ".md", ".pdf"}:
287
+
import_source = "text"
288
+
else:
289
+
import_source = "audio"
290
+
stream = stream_name(import_source=import_source)
291
+
292
+
if args.dry_run:
293
+
from think.importers.plaud import format_size
294
+
295
+
# Print summary without writing anything
296
+
file_size = os.path.getsize(args.media)
297
+
display = _format_timestamp_display(args.timestamp)
298
+
299
+
print()
300
+
print(f" File: {args.media}")
301
+
print(f" Size: {format_size(file_size)}")
302
+
print(f" Timestamp: {args.timestamp} ({display})")
303
+
print(f" Source: {import_source}")
304
+
print(f" Stream: {stream}")
305
+
print(f" Target day: {day}")
306
+
307
+
ext = os.path.splitext(args.media)[1].lower()
308
+
if ext in {".txt", ".md", ".pdf"}:
309
+
text = _read_transcript(args.media)
310
+
chars = len(text)
311
+
lines = text.count("\n") + (1 if text and not text.endswith("\n") else 0)
312
+
print()
313
+
print(f" Content: {chars:,} characters, {lines:,} lines")
314
+
else:
315
+
duration = _get_audio_duration(args.media)
316
+
if duration is not None:
317
+
segment_duration = 300
318
+
num_segments = int(
319
+
(duration + segment_duration - 1) // segment_duration
320
+
)
321
+
if num_segments == 0:
322
+
num_segments = 1
323
+
324
+
keys = []
325
+
for i in range(num_segments):
326
+
ts = base_dt + dt.timedelta(minutes=i * 5)
327
+
keys.append(f"{ts.strftime('%H%M%S')}_{segment_duration}")
328
+
329
+
if duration < 60:
330
+
dur_str = f"{duration:.0f} seconds"
331
+
elif duration < 3600:
332
+
dur_str = f"{duration / 60:.1f} minutes"
333
+
else:
334
+
dur_str = f"{duration / 3600:.1f} hours"
335
+
336
+
print()
337
+
print(f" Duration: {dur_str}")
338
+
print(f" Segments: {num_segments} (5-minute chunks)")
339
+
print(f" Keys: {', '.join(keys)}")
340
+
else:
341
+
print()
342
+
print(" Duration: unknown (ffprobe failed)")
343
+
print()
344
+
return
345
+
270
346
# Check if file needs setup (not already in imports/)
271
347
needs_setup = not _is_in_imports(args.media)
272
348
···
282
358
)
283
359
print("Starting import...")
284
360
285
-
base_dt = dt.datetime.strptime(args.timestamp, "%Y%m%d_%H%M%S")
286
-
day = base_dt.strftime("%Y%m%d")
287
361
logger.info(f"Using provided timestamp: {args.timestamp}")
288
362
day_dir = str(day_path(day))
289
-
290
-
# Derive stream identity for this import
291
-
if args.source:
292
-
import_source = args.source
293
-
else:
294
-
# Auto-detect from file extension
295
-
_ext = os.path.splitext(args.media)[1].lower()
296
-
if _ext == ".m4a":
297
-
import_source = "apple"
298
-
elif _ext in {".txt", ".md", ".pdf"}:
299
-
import_source = "text"
300
-
else:
301
-
import_source = "audio"
302
-
stream = stream_name(import_source=import_source)
303
363
304
364
# Initialize importer tract state
305
365
_import_id = args.timestamp