File size: 7,154 Bytes
6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 6866239 6d55408 |
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 |
import os
import hashlib
import io
import pandas as pd
from PyPDF2 import PdfReader
from docx import Document
from langchain_huggingface import HuggingFaceEmbeddings
from pymilvus import connections, FieldSchema, CollectionSchema, DataType, Collection
import json
class FileHandler:
def __init__(self,api_token,logger):
self.logger = logger
self.logger.info("Initializing FileHandler...")
# Initialize the embedding model using Hugging Face
self.embeddings = HuggingFaceEmbeddings(
model_name="sentence-transformers/all-MiniLM-L6-v2",
model_kwargs={"token": api_token},
)
def handle_file_upload(self, file, document_name, document_description):
try:
content = file.read()
file_hash = hashlib.md5(content).hexdigest()
collection_name = f"collection_{file_hash}"
# Check if the collection exists
if connections._fetch_handler().has_collection(collection_name):
self.logger.info(f"Collection '{collection_name}' already exists.")
return {"message": "File already processed."}
# Process file based on type
if file.name.endswith(".pdf"):
texts, metadatas = self.load_and_split_pdf(file)
elif file.name.endswith(".docx"):
texts, metadatas = self.load_and_split_docx(file)
elif file.name.endswith(".txt"):
texts, metadatas = self.load_and_split_txt(content)
elif file.name.endswith(".xlsx"):
texts, metadatas = self.load_and_split_table(content)
elif file.name.endswith(".csv"):
texts, metadatas = self.load_and_split_csv(content)
else:
self.logger.info("Unsupported file format.")
raise ValueError("Unsupported file format.")
if not texts:
return {"message": "No text extracted from the file. Check the file content."}
# self._store_vectors(collection_name, texts, metadatas)
filename = file.name
filelen = len(content)
self._store_vectors(collection_name, texts, metadatas, document_name, document_description,filename,filelen)
self.logger.info(f"File processed successfully. Collection name: {collection_name}")
return {"message": "File processed successfully."}
except Exception as e:
self.logger.error(f"Error processing file: {str(e)}")
return {"message": f"Error processing file: {str(e)}"}
def _store_vectors(self, collection_name, texts, metadatas, document_name, document_description,file_name,file_len):
fields = [
FieldSchema(name="pk", dtype=DataType.INT64, is_primary=True, auto_id=True),
FieldSchema(name="embedding", dtype=DataType.FLOAT_VECTOR, dim=384),
FieldSchema(name="file_name_hash", dtype=DataType.INT64), # Hash of file name
FieldSchema(name="document_name_hash", dtype=DataType.INT64), # Hash of document name
FieldSchema(name="document_description_hash", dtype=DataType.INT64), # Hash of document description
FieldSchema(name="file_meta_hash", dtype=DataType.INT64),
FieldSchema(name="file_size", dtype=DataType.INT64),
]
schema = CollectionSchema(fields, description="Document embeddings with metadata")
collection = Collection(name=collection_name, schema=schema)
# Generate embeddings
embeddings = [self.embeddings.embed_query(text) for text in texts]
# Convert metadata to hashed values
file_name_hash = int(hashlib.md5(file_name.encode('utf-8')).hexdigest(), 16) % (10 ** 12)
document_name_hash = int(hashlib.md5((document_name or "Unknown Document").encode('utf-8')).hexdigest(), 16) % (
10 ** 12)
document_description_hash = int(
hashlib.md5((document_description or "No Description Provided").encode('utf-8')).hexdigest(), 16) % (
10 ** 12)
# Convert metadata list to JSON string and hash it
metadata_string = json.dumps(metadatas, ensure_ascii=False)
file_meta_hash = int(hashlib.md5(metadata_string.encode('utf-8')).hexdigest(), 16) % (10 ** 12)
# Prepare data for insertion
data = [
embeddings,
[file_name_hash] * len(embeddings),
[document_name_hash] * len(embeddings),
[document_description_hash] * len(embeddings),
[file_meta_hash] * len(embeddings),
[file_len or 0] * len(embeddings),
]
# Insert data into collection
collection.insert(data)
collection.load()
def load_and_split_pdf(self, file):
reader = PdfReader(file)
texts = []
metadatas = []
for page_num, page in enumerate(reader.pages):
text = page.extract_text()
if text:
texts.append(text)
metadatas.append({"page_number": page_num + 1})
return texts, metadatas
def load_and_split_docx(self, file):
doc = Document(file)
texts = []
metadatas = []
for para_num, paragraph in enumerate(doc.paragraphs):
if paragraph.text:
texts.append(paragraph.text)
metadatas.append({"paragraph_number": para_num + 1})
return texts, metadatas
def load_and_split_txt(self, content):
text = content.decode("utf-8")
lines = text.split('\n')
texts = [line for line in lines if line.strip()]
metadatas = [{}] * len(texts)
return texts, metadatas
def load_and_split_table(self, content):
excel_data = pd.read_excel(io.BytesIO(content), sheet_name=None)
texts = []
metadatas = []
for sheet_name, df in excel_data.items():
df = df.dropna(how='all', axis=0).dropna(how='all', axis=1)
df = df.fillna('N/A')
for _, row in df.iterrows():
row_dict = row.to_dict()
# Combine key-value pairs into a string
row_text = ', '.join([f"{key}: {value}" for key, value in row_dict.items()])
texts.append(row_text)
metadatas.append({"sheet_name": sheet_name})
return texts, metadatas
def load_and_split_csv(self, content):
csv_data = pd.read_csv(io.StringIO(content.decode('utf-8')))
texts = []
metadatas = []
csv_data = csv_data.dropna(how='all', axis=0).dropna(how='all', axis=1)
csv_data = csv_data.fillna('N/A')
for _, row in csv_data.iterrows():
row_dict = row.to_dict()
row_text = ', '.join([f"{key}: {value}" for key, value in row_dict.items()])
texts.append(row_text)
metadatas.append({"row_index": _})
return texts, metadatas
|