File size: 26,282 Bytes
01d5a5d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
from datetime import datetime, timedelta
from enum import Enum
from typing import Any, Dict, List, Optional, Union
import json
import logging

import numpy as np


DEFAULT_EMBEDDING_DIM = 1536
DISTANCE_RATE = 0.8


class TimeType(str, Enum):
    RECENT = "recent"
    EARLIER = "earlier"


MIN_MEMORIES_N = {TimeType.RECENT: 3, TimeType.EARLIER: 10}

TIME_RANGE = {TimeType.RECENT: 60 * 60 * 24 * 1, TimeType.EARLIER: 60 * 60 * 24 * 7}


class MemoryType(str, Enum):
    TEXT = "TEXT"
    MARKDOWN = "MARKDOWN"
    PDF = "PDF"
    LINK = "LINK"


class AnalysisType(str, Enum):
    SUBJECT = "SUBJECT"
    OBJECT = "OBJECT"
    CHAT = "CHAT"


def datetime2timestamp(time_str: str) -> float:
    """Convert datetime string to timestamp.
    
    Args:
        time_str: String representation of datetime in TIME_FORMAT format.
        
    Returns:
        Timestamp in seconds.
        
    Raises:
        Exception: If time_str has invalid format.
    """
    try:
        timestamp = datetime.strptime(time_str, TIME_FORMAT).timestamp()
        return timestamp
    except Exception as e:
        logging.error(f"Invalid time format: {time_str}")
        raise e


OBJECT_NOTE_TYPE = [MemoryType.LINK]
SUBJECT_NOTE_TYPE = [
    MemoryType.TEXT,
    MemoryType.MARKDOWN,
    MemoryType.PDF,
]
TIME_FORMAT = "%Y-%m-%d %H:%M:%S"


TAG_TYPE = {
    TimeType.RECENT: {"time": "Today", "default": "Recent"},
    TimeType.EARLIER: {"time": "Earlier", "default": "Earlier"},
}


class Chunk:
    """Represents a chunk of document content with embedding information."""
    
    def __init__(
        self,
        id: int,
        document_id: int,
        content: str,
        embedding: Optional[Union[List[float], np.ndarray]] = None,
        tags: Optional[List[str]] = None,
        topic: Optional[str] = None,
    ):
        """Initialize a Chunk instance.
        
        Args:
            id: Unique identifier for the chunk.
            document_id: ID of the document this chunk belongs to.
            content: Text content of the chunk.
            embedding: Vector representation of the chunk content.
            tags: List of tags associated with the chunk.
            topic: Topic classification for the chunk.
        """
        self.id = id
        self.document_id = document_id
        self.content = content
        self.embedding = embedding.squeeze() if embedding is not None else None
        self.tags = tags
        self.topic = topic


class Note:
    """Represents a note with its content and metadata."""
    
    def __init__(
        self,
        noteId: int = None,
        content: str = "",
        createTime: str = "",
        memoryType: str = "",
        embedding: Optional[Union[List[float], np.ndarray]] = None,
        chunks: List[Chunk] = None,
        title: str = "",
        summary: str = "",
        insight: str = "",
        tags: List[str] = None,
        topic: str = None,
    ):
        """Initialize a Note instance.
        
        Args:
            noteId: Unique identifier for the note.
            content: Text content of the note.
            createTime: Creation timestamp in string format.
            memoryType: Type of the memory (TEXT, MARKDOWN, etc.).
            embedding: Vector representation of the note content.
            chunks: List of chunks the note is divided into.
            title: Title of the note.
            summary: Summary of the note content.
            insight: Insights extracted from the note.
            tags: List of tags associated with the note.
            topic: Topic classification for the note.
        """
        self.id = noteId
        self.content = content
        self.create_time = createTime
        self.memory_type = memoryType
        self.embedding = embedding.squeeze() if embedding is not None else None
        self.chunks = chunks or []
        self.title = title
        self.summary = summary
        self.insight = insight
        self.tags = tags
        self.topic = topic

    def __str__(self) -> str:
        """Return a string representation of the note.
        
        Returns:
            Formatted string with note metadata and content.
        """
        note_statement = "---\n"
        if self.id:
            note_statement += f"[ID]: {self.id}\n"
        if self.title:
            note_statement += f"[Title]: {self.title}\n"
        if self.create_time:
            note_statement += f"[Date]: {self.create_time}\n"
        if self.memory_type:
            note_statement += f"[Type]: {self.memory_type}\n"
        note_statement += "---\n\n"
        if self.summary:
            note_statement += f"----- Doc Summary -----\n{self.summary}\n\n"
        if self.insight:
            note_statement += f"----- Doc Insight -----\n{self.insight}\n\n"
        if not (self.insight or self.summary):
            note_statement += f"----- Doc Content -----\n{self.content[:4000]}\n\n"

        return note_statement

    def to_json(self) -> Dict[str, Any]:
        """Convert the note to a JSON-serializable dictionary.
        
        Returns:
            Dictionary representation of the note.
        """
        if hasattr(self, "processed"):
            return {
                "id": self.id,
                "insight": self.insight,
                "summary": self.summary,
                "memory_type": self.memory_type,
                "create_time": self.create_time,
                "title": self.title,
                "content": self.content,
                "processed": self.processed,
            }
        else:
            return {
                "id": self.id,
                "insight": self.insight,
                "summary": self.summary,
                "memory_type": self.memory_type,
                "create_time": self.create_time,
                "title": self.title,
                "content": self.content,
            }

    def to_str(self, analysis_type: AnalysisType = None) -> str:
        """Convert the note to a string based on analysis type.
        
        Args:
            analysis_type: Type of analysis to determine format.
            
        Returns:
            Formatted string representation of the note.
            
        Raises:
            ValueError: If memory_type or analysis_type is invalid.
        """
        if not analysis_type:
            if self.memory_type in SUBJECT_NOTE_TYPE:
                analysis_type = AnalysisType.SUBJECT
            elif self.memory_type in OBJECT_NOTE_TYPE:
                analysis_type = AnalysisType.OBJECT
            else:
                raise ValueError(f"Invalid memory type: {self.memory_type}")

        if analysis_type == AnalysisType.SUBJECT:
            return self.to_subject_str()
        elif analysis_type == AnalysisType.OBJECT:
            return self.to_object_str()
        else:
            raise ValueError(f"Invalid analysis type: {analysis_type}")

    def to_subject_str(self) -> str:
        """Convert the note to a string formatted as subject.
        
        Returns:
            Formatted string for subject analysis.
        """
        note_statement = "---\n"
        if self.id:
            note_statement += f"[ID]: {self.id}\n"
        if self.title:
            note_statement += f"[Title]: {self.title}\n"
        if self.create_time:
            note_statement += f"[Date]: {self.create_time}\n"
        if self.memory_type:
            note_statement += f"[Type]: {self.memory_type}\n"
        note_statement += "---\n\n"
        if self.summary:
            note_statement += f"----- Doc Summary -----\n{self.summary}\n\n"
        if self.insight:
            note_statement += f"----- Doc Insight -----\n{self.insight}\n\n"
        if not (self.insight or self.summary):
            note_statement += f"----- Doc Content -----\n{self.content[:4000]}\n\n"

        return note_statement

    def to_object_str(self) -> str:
        """Convert the note to a string formatted as object.
        
        Returns:
            Formatted string for object analysis.
        """
        note_statement = "---\n"
        if self.id:
            note_statement += f"[ID]: {self.id}\n"
        if self.title:
            note_statement += f"[Title]: {self.title}\n"
        if self.create_time:
            note_statement += f"[Read Time]: {self.create_time}\n"
        if self.memory_type:
            note_statement += f"[Meta Type]: {self.memory_type}\n"
        note_statement += "---\n\n"
        if self.summary:
            note_statement += f"----- Doc Summary -----\n{self.summary}\n\n"
        if not self.summary and self.insight:
            note_statement += f"----- Doc Insight -----\n{self.insight}\n\n"
        if not (self.insight or self.summary):
            note_statement += f"----- Doc Content -----\n{self.content[:4000]}\n\n"
        return note_statement


class Memory:
    def __init__(self, memoryId: int, embedding: List[float] = None):
        self.memory_id = memoryId

        if embedding is not None:
            self.embedding = np.array(embedding).squeeze()
        else:
            self.embedding = None

    def to_json(self):
        return {"memoryId": self.memory_id}


class Cluster:
    def __init__(
        self,
        clusterId: int,
        memoryList: List[Optional[Union[Dict, Memory]]] = [],
        centerEmbedding: List[float] = None,
        is_new=False,
    ):
        self.cluster_id = clusterId
        memory_list = [
            memory if isinstance(memory, Memory) else Memory(**memory)
            for memory in memoryList
        ]
        self.memory_list = memory_list
        self.is_new = is_new
        self.size = len(memory_list)
        self.cluster_center = (
            np.array(centerEmbedding)
            if centerEmbedding
            else np.zeros(DEFAULT_EMBEDDING_DIM)
        )
        self.merge_list = []

    def add_memory(self, memory: Memory):
        self.memory_list.append(memory)
        self.size += 1
        self.get_cluster_center()

    def extend_memory_list(self, memory_list: List[Memory]):
        self.memory_list.extend(memory_list)
        self.size += len(memory_list)
        self.get_cluster_center()

    def get_cluster_center(self):
        if not self.memory_list:
            self.cluster_center = np.zeros(DEFAULT_EMBEDDING_DIM)
        else:
            self.cluster_center = np.mean(
                [memory.embedding for memory in self.memory_list], axis=0
            )

    def prune_outliers_from_cluster(self):
        if not self.memory_list:
            self.get_cluster_center()
        memory_list = sorted(
            self.memory_list,
            key=lambda x: np.linalg.norm(x.embedding - self.cluster_center),
        )
        memory_list = memory_list[: max(int(self.size * DISTANCE_RATE), 1)]
        self.memory_list = memory_list
        self.size = len(memory_list)
        self.get_cluster_center()

    def to_json(self):
        return {
            "clusterId": self.cluster_id if not self.is_new else None,
            "memoryList": [memory.to_json() for memory in self.memory_list],
            "centerEmbedding": self.cluster_center.tolist(),
            "mergeList": self.merge_list,
        }


class ShadeTimeline:
    def __init__(
        self,
        refMemoryId: int = None,
        createTime: str = "",
        descSecondView: str = "",
        descThirdView: str = "",
        is_new: bool = False,
    ):
        self.create_time = createTime
        self.ref_memory_id = refMemoryId
        self.desc_second_view = descSecondView
        self.desc_third_view = descThirdView
        self.is_new = is_new

    @classmethod
    def from_raw_format(cls, raw_format: Dict[str, Any]):
        return cls(
            refMemoryId=raw_format.get("refMemoryId", None),
            createTime=raw_format.get("createTime", ""),
            descSecondView="",
            descThirdView=raw_format.get("description", ""),
            is_new=True,
        )

    def add_second_view(self, description):
        self.desc_second_view = description

    def to_json(self):
        return {
            "createTime": self.create_time,
            "refMemoryId": self.ref_memory_id,
            "descThirdView": self.desc_third_view,
            "descSecondView": self.desc_second_view,
        }


class ConfidenceLevel(str, Enum):
    VERY_LOW = "VERY LOW"
    LOW = "LOW"
    MEDIUM = "MEDIUM"
    HIGH = "HIGH"
    VERY_HIGH = "VERY HIGH"


CONFIDENCE_LEVELS_INT = {
    ConfidenceLevel.VERY_LOW: 1,
    ConfidenceLevel.LOW: 2,
    ConfidenceLevel.MEDIUM: 3,
    ConfidenceLevel.HIGH: 4,
    ConfidenceLevel.VERY_HIGH: 5,
}


class ShadeInfo:
    def __init__(
        self,
        id: int = None,
        name: str = "",
        aspect: str = "",
        icon: str = "",
        descThirdView: str = "",
        contentThirdView: str = "",
        descSecondView: str = "",
        contentSecondView: str = "",
        timelines: List[Dict[str, Any]] = [],
        confidenceLevel: str = None,
    ):
        self.id = id
        self.name = name
        self.aspect = aspect
        self.icon = icon
        self.desc_second_view = descSecondView
        self.desc_third_view = descThirdView
        self.content_third_view = contentThirdView
        self.content_second_view = contentSecondView
        if confidenceLevel:
            self.confidence_level = ConfidenceLevel(confidenceLevel)
        else:
            self.confidence_level = None

        self.timelines = [ShadeTimeline(**timeline) for timeline in timelines]

    def imporve_shade_info(
        self,
        improveDesc: str,
        improveContent: str,
        improveTimelines: List[Dict[str, Any]],
    ):
        self.desc_third_view = improveDesc
        self.content_third_view = improveContent
        self.timelines.extend(
            [ShadeTimeline.from_raw_format(timeline) for timeline in improveTimelines]
        )

    def add_second_view(
        self,
        domainDesc: str,
        domainContent: str,
        domainTimeline: List[Dict[str, Any]],
        *args,
        **kwargs,
    ):
        self.desc_second_view = domainDesc
        self.content_second_view = domainContent
        timelime_dict = {
            timelime.ref_memory_id: timelime for timelime in self.timelines
        }
        for timeline in domainTimeline:
            ref_memory_id = timeline.get("refMemoryId", None)
            if not (ref_memory_id and ref_memory_id in timelime_dict):
                logging.error(
                    f"Timeline with refMemoryId {ref_memory_id} already exists, skipping"
                )
                continue
            timelime_dict[ref_memory_id].add_second_view(
                timeline.get("description", "")
            )

    def _preview_(self, second_view: bool = False):
        if second_view:
            return f"- **{self.name}**: {self.desc_second_view}"
        return f"- **{self.name}**: {self.desc_third_view}"

    def to_str(self):
        shade_statement = f"---\n**[Name]**: {self.name}\n**[Aspect]**: {self.aspect}\n**[Icon]**: {self.icon}\n"
        shade_statement += f"**[Description]**: \n{self.desc_third_view}\n\n**[Content]**: \n{self.content_third_view}\n"
        shade_statement += "---\n\n[Timelines]:\n"
        for timeline in self.timelines:
            shade_statement += f"- {timeline.create_time}, {timeline.desc_third_view}, {timeline.ref_memory_id}\n"
        return shade_statement

    def to_json(self):
        return {
            "id": self.id,
            "name": self.name,
            "aspect": self.aspect,
            "icon": self.icon,
            "descSecondView": self.desc_second_view,
            "descThirdView": self.desc_third_view,
            "contentThirdView": self.content_third_view,
            "contentSecondView": self.content_second_view,
            "confidenceLevel": self.confidence_level if self.confidence_level else None,
            "timelines": [timeline.to_json() for timeline in self.timelines],
        }


class AttributeInfo:
    def __init__(
        self,
        id: int = None,
        name: str = "",
        description: str = "",
        confidenceLevel: Optional[Union[str, ConfidenceLevel]] = None,
    ):
        self.id = id
        self.name = name
        self.description = description
        if confidenceLevel and isinstance(confidenceLevel, str):
            self.confidence_level = ConfidenceLevel(confidenceLevel)
        elif isinstance(confidenceLevel, ConfidenceLevel):
            self.confidence_level = confidenceLevel
        else:
            self.confidence_level = None

    def to_str(self):
        # - **[Attribute Name]**: (Attribute Description), Confidence level: [LOW/MEDIUM/HIGH]
        return f"- **{self.name}**: {self.description}, Confidence level: {self.confidence_level.value}"

    def to_json(self):
        return {
            "id": self.id,
            "name": self.name,
            "description": self.description,
            "confidenceLevel": self.confidence_level.value
            if self.confidence_level
            else None,
        }


class Bio:
    def __init__(
        self,
        contentThirdView: str = "",
        content: str = "",
        summaryThirdView: str = "",
        summary: str = "",
        attributeList: List[Dict[str, Any]] = [],
        shadesList: List[Dict[str, Any]] = [],
    ):
        self.content_third_view = contentThirdView
        self.content_second_view = content
        self.summary_third_view = summaryThirdView
        self.summary_second_view = summary
        self.attribute_list = sorted(
            [AttributeInfo(**attribute) for attribute in attributeList],
            key=lambda x: CONFIDENCE_LEVELS_INT[x.confidence_level],
            reverse=True,
        )
        self.shades_list = sorted(
            [ShadeInfo(**shade) for shade in shadesList],
            key=lambda x: len(x.timelines),
            reverse=True,
        )

    def to_str(self) -> str:
        global_bio_statement = ""
        if self.is_raw_bio():
            global_bio_statement += (
                f"**[Origin Analysis]**\n{self.summary_third_view}\n"
            )
        # global_bio_statement += f"**[Identity Attributes]**\n"
        # global_bio_statement += '\n'.join([attribute.to_str() for attribute in self.attribute_list])

        global_bio_statement += f"\n**[Current Shades]**\n"
        for shade in self.shades_list:
            global_bio_statement += shade.to_str()
            global_bio_statement += "\n==============\n"
        return global_bio_statement

    def complete_content(self, second_view: bool = False) -> str:
        interests_preference_field = (
            "\n### User's Interests and Preferences ###\n"
            + "\n".join([shade._preview_(second_view) for shade in self.shades_list])
        )
        if not second_view:
            conclusion_field = "\n### Conclusion ###\n" + self.summary_third_view
        else:
            conclusion_field = "\n### Conclusion ###\n" + self.summary_second_view
        return f"""## Comprehensive Analysis Report ##
{interests_preference_field}
{conclusion_field}"""

    def is_raw_bio(self) -> bool:
        if not self.content_third_view and not self.summary_third_view:
            return True
        return False

    def to_json(self) -> Dict[str, Any]:
        return {
            "contentThirdView": self.content_third_view,
            "content": self.content_second_view,
            "summaryThirdView": self.summary_third_view,
            "summary": self.summary_second_view,
            "shadesList": [shade.to_json() for shade in self.shades_list],
        }


class ShadeMergeInfo:
    def __init__(
        self,
        id: int = None,
        name: str = "",
        aspect: str = "",
        icon: str = "",
        desc_third_view: str = "",
        content_third_view: str = "",
        desc_second_view: str = "",
        content_second_view: str = "",
        cluster_info: Optional[Dict[str, Any]] = None,
    ):
        self.id = id
        self.name = name
        self.aspect = aspect
        self.icon = icon
        self.desc_second_view = desc_second_view
        self.desc_third_view = desc_third_view
        self.content_third_view = content_third_view
        self.content_second_view = content_second_view
        self.cluster_info = cluster_info

    def improve_shade_info(self, improveDesc: str, improveContent: str):
        self.desc_third_view = improveDesc
        self.content_third_view = improveContent

    def add_second_view(self, domainDesc: str, domainContent: str):
        self.desc_second_view = domainDesc
        self.content_second_view = domainContent

    def _preview_(self, second_view: bool = False):
        if second_view:
            return f"- **{self.name}**: {self.desc_second_view}"
        return f"- **{self.name}**: {self.desc_third_view}"

    def to_str(self):
        shade_statement = f"---\n**[Name]**: {self.name}\n**[Aspect]**: {self.aspect}\n**[Icon]**: {self.icon}\n"
        shade_statement += f"**[Description]**: \n{self.desc_third_view}\n\n**[Content]**: \n{self.content_third_view}\n"
        shade_statement += "---\n\n"
        if self.cluster_info:
            shade_statement += (
                f"**[Cluster Info]**: \n{json.dumps(self.cluster_info, indent=2)}\n"
            )
        return shade_statement

    def to_json(self):
        return {
            "id": self.id,
            "name": self.name,
            "aspect": self.aspect,
            "icon": self.icon,
            "descSecondView": self.desc_second_view,
            "descThirdView": self.desc_third_view,
            "contentThirdView": self.content_third_view,
            "contentSecondView": self.content_second_view,
            "clusterInfo": self.cluster_info,
        }


class ShadeMergeResponse:
    def __init__(self, result: Any, success: bool):
        self.success: bool = success
        self.message: str = ""
        self.merge_shade_list: Optional[List[Dict[str, Any]]] = None

        if not success:
            self.message = result if isinstance(result, str) else "Error occurred"
            logging.error(self.message)
        else:
            self.message = "Success"
            self.merge_shade_list = result.get("mergeShadeList")

    def to_json(self) -> dict:
        return {
            "success": self.success,
            "message": self.message,
            "mergeShadeList": self.merge_shade_list,
        }


class Todo:
    def __init__(
        self,
        todoId: int = 0,
        content: str = "",
        deadlineTime: str = "",
        createTime: str = "",
        status: str = "Done",
    ) -> None:
        self.todo_id = todoId
        self.content = content
        self.deadline_time = deadlineTime
        self.create_time = createTime
        self.status = status

    def __str__(self):
        todo_statement = "---\n"
        todo_statement += f"[Action] User have a Plan\n"
        if self.content:
            todo_statement += f"[Content]: {self.content}\n"
        if self.create_time:
            todo_statement += f"[Create Time]: {self.create_time}\n"
        if self.deadline_time:
            todo_statement += f"[Deadline Time]: {self.deadline_time}\n"
        if self.status:
            todo_statement += f"[Status]: {self.status}\n"
        return todo_statement


class Chat:
    def __init__(
        self,
        sessionId: str = "",
        summary: str = "",
        title: str = "",
        createTime: str = "",
    ) -> None:
        self.session_id = sessionId
        self.summary = summary
        self.title = title
        self.create_time = createTime

    def __str__(self):
        chat_statement = "---\n"
        chat_statement += f"[Action] User had a chat\n"
        if self.create_time:
            chat_statement += f"[Create Time]: {self.create_time}\n"
        if self.title:
            chat_statement += f"[Title]: {self.title}\n"
        if self.summary:
            chat_statement += f"{self.summary}\n"
        return chat_statement


class UserInfo:
    def __init__(
        self, cur_time: str, notes: List[Note], todos: List[Todo], chats: List[Chat]
    ):
        self.notes = notes
        self.todos = todos
        self.chats = chats
        self.cur_time = cur_time
        self.recent_tag = {k: v["default"] for k, v in TAG_TYPE.items()}
        self.memories = sorted(
            notes + todos + chats,
            key=lambda x: datetime2timestamp(x.create_time),
            reverse=True,
        )
        self.recent_memories = self.get_range_memories(TimeType.RECENT)
        self.earlier_memories = self.get_range_memories(TimeType.EARLIER)[
            len(self.recent_memories) :
        ]

    def __str__(self):
        user_memories_statement = "### {recent_type} Memory ###\n".format(
            recent_type=self.recent_tag[TimeType.RECENT]
        )
        user_memories_statement += "".join(
            [str(memory) for memory in self.recent_memories]
        )
        user_memories_statement += "\n\n### Earlier Memory ###\n"
        user_memories_statement += "".join(
            [str(memory) for memory in self.earlier_memories]
        )
        return user_memories_statement

    def get_range_memories(self, time_type: TimeType) -> List[Union[Note, Todo, Chat]]:
        if len(self.memories) < MIN_MEMORIES_N[time_type]:
            return self.memories
        recent_memories = []
        cur_datetime = datetime.fromtimestamp(datetime2timestamp(self.cur_time))
        end_datetime = cur_datetime + timedelta(days=1)
        end_timestamp = end_datetime.replace(hour=0, minute=0, second=0).timestamp()
        for memory in self.memories:
            if (
                end_timestamp - datetime2timestamp(memory.create_time)
                < TIME_RANGE[time_type]
            ):
                recent_memories.append(memory)
            else:
                break
        if len(recent_memories) >= MIN_MEMORIES_N[time_type]:
            self.recent_tag[time_type] = TAG_TYPE[time_type].get(
                "time", TAG_TYPE[time_type].get("default")
            )
            return recent_memories
        return self.memories[: MIN_MEMORIES_N[time_type]]