apehex commited on
Commit
63e2f1c
1 Parent(s): ac7d850

Enforce schema during conversion from binary to hex

Browse files
Files changed (2) hide show
  1. README.md +1 -1
  2. hexify.py +2 -2
README.md CHANGED
@@ -58,7 +58,7 @@ In binary form, the raw bytes are stored in most fields:
58
  ```python
59
  datasets.Features({
60
  "chain_id": datasets.features.Value(dtype='uint64'),
61
- "block_number": datasets.features.Value(dtype='uint32'),
62
  'block_hash': datasets.features.Value(dtype='large_binary'),
63
  'transaction_hash': datasets.features.Value(dtype='large_binary'),
64
  'deployer_address': datasets.features.Value(dtype='large_binary'),
 
58
  ```python
59
  datasets.Features({
60
  "chain_id": datasets.features.Value(dtype='uint64'),
61
+ "block_number": datasets.features.Value(dtype='uint64'),
62
  'block_hash': datasets.features.Value(dtype='large_binary'),
63
  'transaction_hash': datasets.features.Value(dtype='large_binary'),
64
  'deployer_address': datasets.features.Value(dtype='large_binary'),
hexify.py CHANGED
@@ -16,7 +16,7 @@ OUTPUT_DATASET_PATH = 'hex/ethereum/train/'
16
 
17
  INPUT_SCHEMA = pa.schema(fields=[
18
  pl.field('chain_id', pa.uint64()),
19
- pl.field('block_number', pa.uint32()),
20
  pl.field('block_hash', pa.large_binary()),
21
  pl.field('transaction_hash', pa.large_binary()),
22
  pl.field('deployer_address', pa.large_binary()),
@@ -84,7 +84,7 @@ def process_dataset(dataset: pq.ParquetDataset, output: str=OUTPUT_DATASET_PATH,
84
  for __f in dataset.fragments:
85
  __path = os.path.join(output, os.path.basename(__f.path))
86
  # encode each record
87
- __table = pl.Table.from_pylist(process_table(table=__f.to_table()))
88
  # save
89
  pq.write_table(table=__table, where=__path)
90
  # log
 
16
 
17
  INPUT_SCHEMA = pa.schema(fields=[
18
  pl.field('chain_id', pa.uint64()),
19
+ pl.field('block_number', pa.uint64()),
20
  pl.field('block_hash', pa.large_binary()),
21
  pl.field('transaction_hash', pa.large_binary()),
22
  pl.field('deployer_address', pa.large_binary()),
 
84
  for __f in dataset.fragments:
85
  __path = os.path.join(output, os.path.basename(__f.path))
86
  # encode each record
87
+ __table = pl.Table.from_pylist(process_table(table=__f.to_table()), schema=schema)
88
  # save
89
  pq.write_table(table=__table, where=__path)
90
  # log