be52af70
 tangwang
first commit
|
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
|
import time
class BulkIndexer:
"""Bulk indexer for Elasticsearch with batching and error handling."""
def __init__(
self,
es_client: ESClient,
index_name: str,
batch_size: int = 500,
max_retries: int = 3
):
"""
Initialize bulk indexer.
Args:
es_client: Elasticsearch client
index_name: Target index name
batch_size: Number of documents per batch
max_retries: Maximum retry attempts for failed batches
"""
self.es_client = es_client
self.index_name = index_name
self.batch_size = batch_size
self.max_retries = max_retries
def index_documents(
self,
documents: List[Dict[str, Any]],
id_field: str = "skuId",
show_progress: bool = True
) -> Dict[str, Any]:
"""
Index documents in bulk.
Args:
documents: List of documents to index
id_field: Field to use as document ID
show_progress: Whether to print progress
Returns:
Dictionary with indexing statistics
"""
total_docs = len(documents)
success_count = 0
failed_count = 0
errors = []
print(f"[BulkIndexer] Starting bulk indexing of {total_docs} documents...")
start_time = time.time()
# Process in batches
for i in range(0, total_docs, self.batch_size):
batch = documents[i:i + self.batch_size]
batch_num = (i // self.batch_size) + 1
total_batches = (total_docs + self.batch_size - 1) // self.batch_size
if show_progress:
print(f"[BulkIndexer] Processing batch {batch_num}/{total_batches} "
f"({len(batch)} documents)...")
# Prepare actions for bulk API
actions = []
for doc in batch:
action = {
'_index': self.index_name,
'_source': doc
}
# Use specified field as document ID if present
if id_field and id_field in doc:
action['_id'] = doc[id_field]
actions.append(action)
# Try to index batch with retries
batch_success = False
for attempt in range(self.max_retries):
try:
success, failed = bulk(
self.es_client.client,
actions,
raise_on_error=False,
raise_on_exception=False
)
success_count += success
if failed:
failed_count += len(failed)
errors.extend(failed)
batch_success = True
break
except BulkIndexError as e:
if attempt < self.max_retries - 1:
print(f"[BulkIndexer] Batch {batch_num} failed, retrying... "
f"(attempt {attempt + 1}/{self.max_retries})")
time.sleep(1)
else:
print(f"[BulkIndexer] Batch {batch_num} failed after "
f"{self.max_retries} attempts")
failed_count += len(batch)
errors.append({
'batch': batch_num,
'error': str(e)
})
except Exception as e:
print(f"[BulkIndexer] Unexpected error in batch {batch_num}: {e}")
failed_count += len(batch)
errors.append({
'batch': batch_num,
'error': str(e)
})
break
elapsed_time = time.time() - start_time
# Refresh index to make documents searchable
self.es_client.refresh(self.index_name)
results = {
'total': total_docs,
'success': success_count,
'failed': failed_count,
'elapsed_time': elapsed_time,
'docs_per_second': total_docs / elapsed_time if elapsed_time > 0 else 0,
'errors': errors[:10] # Keep only first 10 errors
}
print(f"[BulkIndexer] Indexing complete!")
print(f" - Total: {total_docs}")
print(f" - Success: {success_count}")
print(f" - Failed: {failed_count}")
print(f" - Time: {elapsed_time:.2f}s")
print(f" - Speed: {results['docs_per_second']:.2f} docs/s")
return results
def delete_by_query(self, query: Dict[str, Any]) -> int:
"""
Delete documents matching a query.
Args:
query: ES query DSL
Returns:
Number of documents deleted
"""
try:
response = self.es_client.client.delete_by_query(
index=self.index_name,
body={"query": query}
)
deleted = response.get('deleted', 0)
print(f"[BulkIndexer] Deleted {deleted} documents")
return deleted
except Exception as e:
print(f"[BulkIndexer] Delete by query failed: {e}")
return 0
def update_by_query(self, query: Dict[str, Any], script: Dict[str, Any]) -> int:
"""
Update documents matching a query.
Args:
query: ES query DSL
script: Update script
Returns:
Number of documents updated
"""
try:
response = self.es_client.client.update_by_query(
index=self.index_name,
body={
"query": query,
"script": script
}
)
updated = response.get('updated', 0)
print(f"[BulkIndexer] Updated {updated} documents")
return updated
except Exception as e:
print(f"[BulkIndexer] Update by query failed: {e}")
return 0
class IndexingPipeline:
"""Complete indexing pipeline from source data to ES."""
def __init__(
self,
config,
es_client: ESClient,
data_transformer,
recreate_index: bool = False
):
"""
Initialize indexing pipeline.
Args:
|
|
be52af70
tangwang
first commit
|
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
|
es_client: Elasticsearch client
data_transformer: Data transformer instance
recreate_index: Whether to recreate index if exists
"""
self.config = config
self.es_client = es_client
self.transformer = data_transformer
self.recreate_index = recreate_index
def run(self, df, batch_size: int = 100) -> Dict[str, Any]:
"""
Run complete indexing pipeline.
Args:
df: Source dataframe
batch_size: Batch size for processing
Returns:
Indexing statistics
"""
|
|
be52af70
tangwang
first commit
|
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
|
# Generate and create index
mapping_gen = MappingGenerator(self.config)
mapping = mapping_gen.generate_mapping()
index_name = self.config.es_index_name
if self.recreate_index:
if self.es_client.index_exists(index_name):
print(f"[IndexingPipeline] Deleting existing index: {index_name}")
self.es_client.delete_index(index_name)
if not self.es_client.index_exists(index_name):
print(f"[IndexingPipeline] Creating index: {index_name}")
self.es_client.create_index(index_name, mapping)
else:
print(f"[IndexingPipeline] Using existing index: {index_name}")
# Transform data
print(f"[IndexingPipeline] Transforming {len(df)} documents...")
documents = self.transformer.transform_batch(df, batch_size=batch_size)
print(f"[IndexingPipeline] Transformed {len(documents)} documents")
# Bulk index
indexer = BulkIndexer(self.es_client, index_name, batch_size=500)
results = indexer.index_documents(documents, id_field="skuId")
return results
|