File size: 4,383 Bytes
d660b02
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import hashlib
from abc import ABC, abstractmethod
from typing import Generic, TypeVar
from uuid import UUID

from llm_engineering.domain.chunks import ArticleChunk, Chunk, PostChunk, RepositoryChunk
from llm_engineering.domain.cleaned_documents import (
    CleanedArticleDocument,
    CleanedDocument,
    CleanedPostDocument,
    CleanedRepositoryDocument,
)

from .operations import chunk_article, chunk_text

CleanedDocumentT = TypeVar("CleanedDocumentT", bound=CleanedDocument)
ChunkT = TypeVar("ChunkT", bound=Chunk)


class ChunkingDataHandler(ABC, Generic[CleanedDocumentT, ChunkT]):
    """

    Abstract class for all Chunking data handlers.

    All data transformations logic for the chunking step is done here

    """

    @property
    def metadata(self) -> dict:
        return {
            "chunk_size": 500,
            "chunk_overlap": 50,
        }

    @abstractmethod
    def chunk(self, data_model: CleanedDocumentT) -> list[ChunkT]:
        pass


class PostChunkingHandler(ChunkingDataHandler):
    @property
    def metadata(self) -> dict:
        return {
            "chunk_size": 250,
            "chunk_overlap": 25,
        }

    def chunk(self, data_model: CleanedPostDocument) -> list[PostChunk]:
        data_models_list = []

        cleaned_content = data_model.content
        chunks = chunk_text(
            cleaned_content, chunk_size=self.metadata["chunk_size"], chunk_overlap=self.metadata["chunk_overlap"]
        )

        for chunk in chunks:
            chunk_id = hashlib.md5(chunk.encode()).hexdigest()
            model = PostChunk(
                id=UUID(chunk_id, version=4),
                content=chunk,
                platform=data_model.platform,
                document_id=data_model.id,
                author_id=data_model.author_id,
                author_full_name=data_model.author_full_name,
                image=data_model.image if data_model.image else None,
                metadata=self.metadata,
            )
            data_models_list.append(model)

        return data_models_list


class ArticleChunkingHandler(ChunkingDataHandler):
    @property
    def metadata(self) -> dict:
        return {
            "min_length": 1000,
            "max_length": 2000,
        }

    def chunk(self, data_model: CleanedArticleDocument) -> list[ArticleChunk]:
        data_models_list = []

        cleaned_content = data_model.content
        chunks = chunk_article(
            cleaned_content, min_length=self.metadata["min_length"], max_length=self.metadata["max_length"]
        )

        for chunk in chunks:
            chunk_id = hashlib.md5(chunk.encode()).hexdigest()
            model = ArticleChunk(
                id=UUID(chunk_id, version=4),
                content=chunk,
                platform=data_model.platform,
                link=data_model.link,
                document_id=data_model.id,
                author_id=data_model.author_id,
                author_full_name=data_model.author_full_name,
                metadata=self.metadata,
            )
            data_models_list.append(model)

        return data_models_list


class RepositoryChunkingHandler(ChunkingDataHandler):
    @property
    def metadata(self) -> dict:
        return {
            "chunk_size": 1500,
            "chunk_overlap": 100,
        }

    def chunk(self, data_model: CleanedRepositoryDocument) -> list[RepositoryChunk]:
        data_models_list = []

        cleaned_content = data_model.content
        chunks = chunk_text(
            cleaned_content, chunk_size=self.metadata["chunk_size"], chunk_overlap=self.metadata["chunk_overlap"]
        )

        for chunk in chunks:
            chunk_id = hashlib.md5(chunk.encode()).hexdigest()
            model = RepositoryChunk(
                id=UUID(chunk_id, version=4),
                content=chunk,
                platform=data_model.platform,
                name=data_model.name,
                link=data_model.link,
                document_id=data_model.id,
                author_id=data_model.author_id,
                author_full_name=data_model.author_full_name,
                metadata=self.metadata,
            )
            data_models_list.append(model)

        return data_models_list