data-science-document-ai 1.40.4__py3-none-any.whl → 1.42.0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (47) hide show
  1. {data_science_document_ai-1.40.4.dist-info → data_science_document_ai-1.42.0.dist-info}/METADATA +1 -1
  2. data_science_document_ai-1.42.0.dist-info/RECORD +57 -0
  3. src/docai.py +14 -5
  4. src/excel_processing.py +14 -4
  5. src/io.py +26 -1
  6. src/llm.py +10 -3
  7. src/pdf_processing.py +46 -10
  8. src/postprocessing/common.py +34 -5
  9. src/prompts/library/bookingConfirmation/evergreen/placeholders.json +17 -17
  10. src/prompts/library/bookingConfirmation/evergreen/prompt.txt +1 -0
  11. src/prompts/library/bookingConfirmation/hapag-lloyd/placeholders.json +18 -18
  12. src/prompts/library/bookingConfirmation/hapag-lloyd/prompt.txt +1 -1
  13. src/prompts/library/bookingConfirmation/maersk/placeholders.json +17 -17
  14. src/prompts/library/bookingConfirmation/maersk/prompt.txt +1 -1
  15. src/prompts/library/bookingConfirmation/msc/placeholders.json +17 -17
  16. src/prompts/library/bookingConfirmation/msc/prompt.txt +1 -1
  17. src/prompts/library/bookingConfirmation/oocl/placeholders.json +17 -17
  18. src/prompts/library/bookingConfirmation/oocl/prompt.txt +3 -1
  19. src/prompts/library/bookingConfirmation/other/placeholders.json +17 -17
  20. src/prompts/library/bookingConfirmation/other/prompt.txt +1 -1
  21. src/prompts/library/bookingConfirmation/yangming/placeholders.json +17 -17
  22. src/prompts/library/bookingConfirmation/yangming/prompt.txt +1 -1
  23. src/prompts/library/bundeskasse/other/placeholders.json +19 -19
  24. src/prompts/library/bundeskasse/other/prompt.txt +1 -1
  25. src/prompts/library/commercialInvoice/other/prompt.txt +2 -1
  26. src/prompts/library/customsAssessment/other/prompt.txt +1 -1
  27. src/prompts/library/customsInvoice/other/placeholders.json +19 -19
  28. src/prompts/library/customsInvoice/other/prompt.txt +1 -1
  29. src/prompts/library/deliveryOrder/other/placeholders.json +15 -17
  30. src/prompts/library/deliveryOrder/other/prompt.txt +1 -1
  31. src/prompts/library/draftMbl/hapag-lloyd/prompt.txt +2 -1
  32. src/prompts/library/draftMbl/maersk/prompt.txt +2 -0
  33. src/prompts/library/draftMbl/other/prompt.txt +1 -1
  34. src/prompts/library/finalMbL/hapag-lloyd/prompt.txt +1 -1
  35. src/prompts/library/finalMbL/maersk/prompt.txt +2 -0
  36. src/prompts/library/finalMbL/other/prompt.txt +1 -1
  37. src/prompts/library/packingList/other/prompt.txt +1 -1
  38. src/prompts/library/partnerInvoice/other/placeholders.json +12 -60
  39. src/prompts/library/partnerInvoice/other/prompt.txt +1 -1
  40. src/prompts/library/shippingInstruction/other/prompt.txt +1 -0
  41. src/prompts/prompt_library.py +4 -0
  42. src/setup.py +5 -1
  43. src/utils.py +64 -4
  44. data_science_document_ai-1.40.4.dist-info/RECORD +0 -59
  45. src/prompts/library/customsAssessment/other/placeholders.json +0 -19
  46. src/prompts/library/finalMbL/other/placeholders.json +0 -80
  47. {data_science_document_ai-1.40.4.dist-info → data_science_document_ai-1.42.0.dist-info}/WHEEL +0 -0
@@ -16,7 +16,8 @@ containers:
16
16
  vessel: The name of the vessel.
17
17
 
18
18
 
19
- Your task is to extract the text value of the following entities:
19
+ Your task is to extract the text value of the following entities and page numbers starting from 0 where the value was found in the document:
20
+
20
21
 
21
22
  Keywords for datapoints:
22
23
  - blNumber: Bill of Lading number, bill of landing no., swb-no., b/l no.
@@ -1,4 +1,6 @@
1
1
  Extract the following information from the sea waybill document.
2
+ Your task is to extract the text value of the following entities and page numbers starting from 0 where the value was found in the document:
3
+
2
4
 
3
5
  **blNumber:** Find the value labeled as "B/L No.".
4
6
  **voyage:** Get the "Voyage No." value.
@@ -16,7 +16,7 @@ containers:
16
16
  vessel: The name of the vessel.
17
17
 
18
18
 
19
- Your task is to extract the text value of the following entities:
19
+ Your task is to extract the text value of the following entities and page numbers starting from 0 where the value was found in the document:
20
20
 
21
21
  Keywords for datapoints:
22
22
  - blNumber: Bill of Lading number, bill of landing no., swb-no., b/l no.
@@ -16,7 +16,7 @@ containers:
16
16
  vessel: The name of the vessel.
17
17
 
18
18
 
19
- Your task is to extract the text value of the following entities:
19
+ Your task is to extract the text value of the following entities and page numbers starting from 0 where the value was found in the document:
20
20
 
21
21
  Keywords for datapoints:
22
22
  - blNumber: Bill of Lading number, bill of landing no., swb-no., b/l no.
@@ -1,4 +1,6 @@
1
1
  Extract the following information from the sea waybill document.
2
+ Your task is to extract the text value of the following entities and page numbers starting from 0 where the value was found in the document:
3
+
2
4
 
3
5
  **blNumber:** Find the value labeled as "B/L No.".
4
6
  **voyage:** Get the "Voyage No." value.
@@ -16,7 +16,7 @@ containers:
16
16
  vessel: The name of the vessel.
17
17
 
18
18
 
19
- Your task is to extract the text value of the following entities:
19
+ Your task is to extract the text value of the following entities and page numbers starting from 0 where the value was found in the document:
20
20
 
21
21
  Keywords for datapoints:
22
22
  - blNumber: Bill of Lading number, bill of landing no., swb-no., b/l no.
@@ -2,7 +2,7 @@ Task: You are a document entity extraction specialist. Given a document, your ta
2
2
 
3
3
  Extract all the data points from the given document.
4
4
  Each data point is part of a master field called "skuData". There may be multiple sku entries in a document.
5
- Your goal is to extract all instances.
5
+ Your task is to extract the text value of the following entities and their page numbers starting from 0 where the value was found in the document:
6
6
 
7
7
  Instructions:
8
8
  - Populate fields as defined in the response schema.
@@ -110,66 +110,18 @@
110
110
  "containerSize"
111
111
  ]
112
112
  },
113
- "mblNumber": {"type": "STRING", "nullable": true,
114
- "description": "Bill of Lading number (B/L NO.), a document issued by the carrier."
115
- },
116
- "partnerReference": {"type": "STRING", "nullable": true,
117
- "description": "A partnerReference can be a shipment ID. It starts with 'S' followed by 6 or 7 digits (e.g., 'S1234567')."
118
- },
119
- "paymentTerm": {"type": "STRING", "nullable": true,
120
- "description": "The payment term indicates the conditions under which the payment should be made. E.g., 'In 10 TAGEN', '14 TAGEN', '14 days', etc."},
121
- "portOfDischarge": {"type": "STRING", "nullable": true,
122
- "description": "The port where the goods are discharged from the vessel. This is the destination port for the shipment."},
123
- "portOfLoading": {"type": "STRING", "nullable": true,
124
- "description": "The origin port where the goods are loaded onto the vessel. Find information like 'Ladehafen' or 'Port of Loading' in the invoice."},
125
- "recipientAddress": {"type": "STRING", "nullable": true,
126
- "description": "Majority of the times, it is 'Forto Logistics SE & Co KG' Address depends on the entity."},
127
- "recipientName": {"type": "STRING", "nullable": true,
128
- "description": "The name of the recipient who is responsible for making the payment. This is often the 'Forto Logistics SE & Co KG' entity or partner."},
129
- "serviceDate": {"type": "STRING", "nullable": true,
130
- "description": "The date when the service was provided. If Service date is not available in the invoice, Estimated Time of Arrival (ETA) can be used."},
131
- "vatId": {"type": "STRING", "nullable": true,
132
- "description": "The VAT ID of the vendor. This is used for tax purposes and to identify the vendor in financial transactions."},
133
- "vendorAddress": {"type": "STRING", "nullable": true,
134
- "description": "The address of the vendor to whom the payment should be made."},
135
- "vendorName": {"type": "STRING", "nullable": true,
136
- "description": "The name of the vendor to whom the payment should be made. Extract the main vendor details incase the invoice contains 'As Agent For'."},
137
- "agentName": {
138
- "type": "STRING",
139
- "nullable": true,
140
- "description": "The name of the agent or intermediary involved in the transaction, if applicable."},
141
- "agentKeyWord": {
142
- "type": "STRING",
143
- "nullable": true,
144
- "description": "A keyword or phrase that indicates the presence of an agent or intermediary in the transaction, such as 'As Agent For', 'Acting Agent', 'Issuing agent', 'Contact Agent', or similar words."},
145
- "paymentInformation": {
146
- "type": "OBJECT",
147
- "properties": {
148
- "paidAmount": {
149
- "type": "STRING",
150
- "nullable": true,
151
- "description": "The amount that has been paid so far. You can identify this in the invoice by looking for terms like 'Vorschuss'."
152
- },
153
- "remainingAmountToPay": {
154
- "type": "STRING",
155
- "nullable": true,
156
- "description": "The amount that is still due for payment (e.g., 'Bitte zahlen', 'Zu zahlen' only). This can be negative & ensure the negative sign is captured if applicable."
157
- },
158
- "currency": {
159
- "type": "STRING",
160
- "nullable": true,
161
- "description": "Currency code associated with the paidAmount and remainngAmountToPay"
162
- },
163
- "sentence": {"type": "STRING", "nullable": true,
164
- "description": "A sentence that indicates the payment status, such as 'Vorschuss', 'Vorauszahlung', 'Paid', 'Partially Paid', or 'Unpaid'. This is used to summarize the payment status of the invoice."}
165
- }
166
-
167
- },
168
- "reverseChargeSentence": {
169
- "type": "STRING",
170
- "nullable": true,
171
- "description": "A sentence which indicate that the reverse charge applies. Mostly found as VAT/Tax Clause."
172
- }
113
+ "mblNumber": {"type": "STRING", "nullable": true},
114
+ "partnerReference": {"type": "STRING", "nullable": true},
115
+ "paymentTerm": {"type": "STRING", "nullable": true},
116
+ "portOfDischarge": {"type": "STRING", "nullable": true},
117
+ "portOfLoading": {"type": "STRING", "nullable": true},
118
+ "recipientAddress": {"type": "STRING", "nullable": true},
119
+ "recipientName": {"type": "STRING", "nullable": true},
120
+ "serviceDate": {"type": "STRING", "nullable": true},
121
+ "vatId": {"type": "STRING", "nullable": true},
122
+ "vendorAddress": {"type": "STRING", "nullable": true},
123
+ "vendorName": {"type": "STRING", "nullable": true},
124
+ "reverseChargeSentence": {"type": "STRING", "nullable": true}
173
125
  },
174
126
  "required": [
175
127
  "bankAccount",
@@ -1,6 +1,6 @@
1
1
  <PERSONA> You are an efficient document entity data extraction specialist working for a Freight Forwarding company. <PERSONA>
2
2
 
3
- <TASK>Your task is to extract data from invoice documents as per the given response schema structure.<TASK>
3
+ <TASK>Your task is to extract data and page numbers starting from 0 from invoice documents as per the given response schema structure.<TASK>
4
4
 
5
5
  <CONTEXT>
6
6
  The Freight Forwarding company receives invoices from Carrier (Shipping Lines) partners and Customs Brokers. These include Partner Invoices (COGS Invoices) and COGS Customs Invoices.
@@ -1,4 +1,5 @@
1
1
  Task: Extract data from the provided shipping instruction PDF document and populate the following dictionary based on the given schema.
2
+ Your task is to extract the text value of the following entities and page numbers starting from 0 where the value was found in the document:
2
3
 
3
4
  ### Instructions:
4
5
  1. Extract all data points from the shipping instruction document.
@@ -4,6 +4,8 @@ import os
4
4
  from pathlib import Path
5
5
  from typing import Dict
6
6
 
7
+ from src.utils import transform_schema_strings
8
+
7
9
 
8
10
  class PromptLibrary:
9
11
  """
@@ -41,6 +43,8 @@ class PromptLibrary:
41
43
  if file == "placeholders.json":
42
44
  with open(path_to_library / prompt_type / prompt_subtype / file) as f:
43
45
  placeholders = json.load(f)
46
+ if prompt_type not in ["postprocessing", "preprocessing"]:
47
+ placeholders = transform_schema_strings(placeholders)
44
48
  self.library[prompt_type][prompt_subtype][
45
49
  "placeholders"
46
50
  ] = placeholders
src/setup.py CHANGED
@@ -18,7 +18,7 @@ from src.constants import project_parameters
18
18
  from src.constants_sandbox import project_parameters_sandbox
19
19
 
20
20
  # Parent repos are imported without .
21
- from src.io import download_dir_from_bucket, get_storage_client, logger
21
+ from src.io import download_dir_from_bucket, get_bq_client, get_storage_client, logger
22
22
  from src.llm import LlmClient
23
23
 
24
24
 
@@ -118,6 +118,10 @@ def setup_params(args=None):
118
118
 
119
119
  params = setup_docai_client_and_path(params)
120
120
 
121
+ # Set up BigQuery client for logging
122
+ bq_client, _ = get_bq_client(params)
123
+ params["bq_client"] = bq_client
124
+
121
125
  # Set up Vertex AI for text embeddings
122
126
  setup_vertexai(params)
123
127
 
src/utils.py CHANGED
@@ -14,7 +14,7 @@ import requests
14
14
  from google.cloud import documentai_v1beta3 as docu_ai_beta
15
15
  from PyPDF2 import PdfReader, PdfWriter
16
16
 
17
- from src.io import get_bq_client, get_storage_client, logger
17
+ from src.io import get_storage_client, logger
18
18
 
19
19
 
20
20
  def bq_logs(data_to_insert, params):
@@ -24,8 +24,8 @@ def bq_logs(data_to_insert, params):
24
24
  data_to_insert (list): The data to insert into BigQuery.
25
25
  params (dict): The parameters dictionary.
26
26
  """
27
- # Get the BigQuery client
28
- bq_client, config = get_bq_client(params)
27
+ # Use the pre-initialized BigQuery client
28
+ bq_client = params["bq_client"]
29
29
  # Get the table string
30
30
  table_string = f"{params['g_ai_project_name']}.{params['g_ai_gbq_db_schema']}.{params['g_ai_gbq_db_table_out']}"
31
31
 
@@ -139,7 +139,12 @@ def store_json_in_gcs(
139
139
  bucket = storage_client.bucket(params.get("doc_ai_bucket_name"))
140
140
  full_object_name = folder_path + document_id
141
141
  blob = bucket.blob(full_object_name)
142
- blob.upload_from_string(json_data, content_type="application/json")
142
+
143
+ # Convert dict to JSON string if needed
144
+ json_string = (
145
+ json.dumps(json_data) if isinstance(json_data, dict) else json_data
146
+ )
147
+ blob.upload_from_string(json_string, content_type="application/json")
143
148
 
144
149
  logger.info(
145
150
  f"JSON object stored successfully in gs://{params.get('doc_ai_bucket_name')}/{full_object_name}" # noqa
@@ -298,6 +303,9 @@ def generate_schema_structure(params, input_doc_type):
298
303
  "type": "string",
299
304
  }
300
305
 
306
+ # update schema to extract value-page_number pairs
307
+ response_schema = transform_schema_strings(response_schema)
308
+
301
309
  return response_schema
302
310
 
303
311
 
@@ -412,3 +420,55 @@ def get_tms_mappings(
412
420
  )
413
421
 
414
422
  return formatted_values
423
+
424
+
425
+ def transform_schema_strings(schema):
426
+ """
427
+ Recursively transforms a schema dictionary, replacing all "type": "STRING"
428
+ definitions with a new object containing "value" and "page_number" fields.
429
+ It preserves 'nullable' and 'description' fields by moving them to the
430
+ new 'value' property.
431
+
432
+ Args:
433
+ schema (dict): The input schema dictionary.
434
+
435
+ Returns:
436
+ dict: The transformed schema dictionary.
437
+ """
438
+ # Base case: if the current schema definition is for a string
439
+ if isinstance(schema, dict) and schema.get("type").upper() == "STRING":
440
+ new_schema = {
441
+ "type": "OBJECT",
442
+ "properties": {
443
+ "value": {"type": "STRING"},
444
+ "page_number": {
445
+ "type": "STRING",
446
+ "description": "Number of a page where the value was found in the document starting from 0.",
447
+ },
448
+ },
449
+ "required": [],
450
+ }
451
+
452
+ # Preserve original properties like nullable and description on the new 'value' key
453
+ if "nullable" in schema:
454
+ new_schema["properties"]["value"]["nullable"] = schema["nullable"]
455
+ if "description" in schema:
456
+ new_schema["properties"]["value"]["description"] = schema["description"]
457
+
458
+ return new_schema
459
+
460
+ # Recursive case: if the schema is a dictionary
461
+ elif isinstance(schema, dict) and schema.get("type").upper() == "OBJECT":
462
+ transformed_schema = schema.copy()
463
+ for key, value in schema.get("properties").items():
464
+ transformed_schema["properties"][key] = transform_schema_strings(value)
465
+ return transformed_schema
466
+
467
+ # Recursive case: if the schema is a list
468
+ elif isinstance(schema, dict) and schema.get("type").upper() == "ARRAY":
469
+ schema["items"] = transform_schema_strings(schema["items"])
470
+ return schema
471
+
472
+ # Base case: for non-dict/list values (e.g., None, bool, str)
473
+ else:
474
+ return schema
@@ -1,59 +0,0 @@
1
- src/constants.py,sha256=TF_UblovdXZnKIb1lnyJwUqQncJCbzBVihoelI6foSU,3579
2
- src/constants_sandbox.py,sha256=Iu6HdjCoNSmOX0AwoL9qUQkhq_ZnIN5U9e-Q2UfNuGc,547
3
- src/docai.py,sha256=AepGdF3ZuSGkujLpewX393FgOBMy-e4sEudiGKho5EA,5280
4
- src/docai_processor_config.yaml,sha256=qOMmCIORpLQ_D-ytvejXxFvER0e0uGYuzPVdZBGv4Pc,2105
5
- src/excel_processing.py,sha256=ZUlZ5zgWObmQfAWHoSrEEITKwr-xXxuOiPC3qDnGjtQ,2459
6
- src/io.py,sha256=IXz4wWqiHa9mnHNgtrC6X9M2lItYp9eu6rHCThUIh5c,3585
7
- src/llm.py,sha256=aEK3rL8XvY7CakvkOJQmcHpEKwZRd8PPrLrzHiO-GFk,7827
8
- src/log_setup.py,sha256=RhHnpXqcl-ii4EJzRt47CF2R-Q3YPF68tepg_Kg7tkw,2895
9
- src/pdf_processing.py,sha256=S_eTsgaDIIr3SCrEmaQZyc7TDJlRI0GCuP0P9EGF1Xc,15385
10
- src/postprocessing/common.py,sha256=W4L455j7IvTRZDWiBizoj9KC_UGUflkL_hEkk5P0h0k,20391
11
- src/postprocessing/postprocess_booking_confirmation.py,sha256=nK32eDiBNbauyQz0oCa9eraysku8aqzrcoRFoWVumDU,4827
12
- src/postprocessing/postprocess_commercial_invoice.py,sha256=3I8ijluTZcOs_sMnFZxfkAPle0UFQ239EMuvZfDZVPg,1028
13
- src/postprocessing/postprocess_partner_invoice.py,sha256=cM4te4qjOI_bXyrF8Zhb6X7eNf5aMKoRaPCFfqFv-98,11538
14
- src/prompts/library/bookingConfirmation/evergreen/placeholders.json,sha256=Re2wBgZoaJ5yImUUAwZOZxFcKXHxi83TCZwTuqd2v2k,1405
15
- src/prompts/library/bookingConfirmation/evergreen/prompt.txt,sha256=qlBMFDHy-gwr2PVeuHrfMEg_8Ibdym243DnaCgINa7g,2614
16
- src/prompts/library/bookingConfirmation/hapag-lloyd/placeholders.json,sha256=Re2wBgZoaJ5yImUUAwZOZxFcKXHxi83TCZwTuqd2v2k,1405
17
- src/prompts/library/bookingConfirmation/hapag-lloyd/prompt.txt,sha256=sg11U3lIhhS36BsimX7IOzR7Pez_9gScdNmJna2pPuw,3355
18
- src/prompts/library/bookingConfirmation/maersk/placeholders.json,sha256=PKWXySGAls6A8tujbSjokYp4ldc3c0DmSP2ITKYiUF8,1405
19
- src/prompts/library/bookingConfirmation/maersk/prompt.txt,sha256=-00tzWzXtQnXX3EPtaCBM39leCoLa4FB52_t7Z3eoQk,3148
20
- src/prompts/library/bookingConfirmation/msc/placeholders.json,sha256=Re2wBgZoaJ5yImUUAwZOZxFcKXHxi83TCZwTuqd2v2k,1405
21
- src/prompts/library/bookingConfirmation/msc/prompt.txt,sha256=9wdbLofnp5s1acD19jCmQuw__HMcVq1yr4vIJNJlKVM,4630
22
- src/prompts/library/bookingConfirmation/oocl/placeholders.json,sha256=NnXjMiEsTCzTDWs2WY7BIMo2p4_98-DL3v1r7x-FL3A,1403
23
- src/prompts/library/bookingConfirmation/oocl/prompt.txt,sha256=aGowVvOgl4w6TjX5O2RtD4QOiWC1JnXiWgg0t0chThU,2060
24
- src/prompts/library/bookingConfirmation/other/placeholders.json,sha256=Re2wBgZoaJ5yImUUAwZOZxFcKXHxi83TCZwTuqd2v2k,1405
25
- src/prompts/library/bookingConfirmation/other/prompt.txt,sha256=XOrq5Ns0nl8lDI9VvoOEbIMbOQdv8mcM8HqP8-eIjc4,3304
26
- src/prompts/library/bookingConfirmation/yangming/placeholders.json,sha256=Re2wBgZoaJ5yImUUAwZOZxFcKXHxi83TCZwTuqd2v2k,1405
27
- src/prompts/library/bookingConfirmation/yangming/prompt.txt,sha256=BSFy-6zDlAmOH2uZjsp-zZkR_Uy3RS5sGtdv9wysiSI,3151
28
- src/prompts/library/bundeskasse/other/placeholders.json,sha256=vc_m1yZP__ZMsdvnIvkgKmwF1q_ZMnKuWeIRNWour3w,4054
29
- src/prompts/library/bundeskasse/other/prompt.txt,sha256=GrrLH8lSuTcWmKUPUgBob0-bS_a61jTFSf09xxcE-nU,2890
30
- src/prompts/library/commercialInvoice/other/prompt.txt,sha256=fYUF7btc48Uqv4mJH5BgJdY4JVwj9I1xKX_HRBIrN7M,2590
31
- src/prompts/library/customsAssessment/other/placeholders.json,sha256=5nSGsMbpfKrpKoYImcTto_RlOvPCHyld2RlwU0Zbbqw,361
32
- src/prompts/library/customsAssessment/other/prompt.txt,sha256=wgJ8PYM0PKXiIKSljhFXEFBQ23GRs2E2DE9lVwHDvBU,2116
33
- src/prompts/library/customsInvoice/other/placeholders.json,sha256=vDji41DWA1ViKgMGwSBt0vu3wBCACrJhprKb8vG8DwY,12153
34
- src/prompts/library/customsInvoice/other/prompt.txt,sha256=P9t3Ga0KITSM1vmQWiX3UJ0pBPmndfJj_rZrvrf7utA,9593
35
- src/prompts/library/deliveryOrder/other/placeholders.json,sha256=6b_6OVsxT7bjFnV_v0OZkGEy-GN5K4AjL0ATzuoLdOU,1286
36
- src/prompts/library/deliveryOrder/other/prompt.txt,sha256=MVSS5AhkiWT17G9X4xk_AgKzYElagvWjLPCMr_ZhmOs,2393
37
- src/prompts/library/draftMbl/hapag-lloyd/prompt.txt,sha256=0k1xLW4zWaenCSNQJxXMXenIwI-eYmGgpxnAAcM3HOg,2251
38
- src/prompts/library/draftMbl/maersk/prompt.txt,sha256=GxaIYlksORvD2uAbodRx_9JFJXD4XbDaVFYtpN9uzxc,2050
39
- src/prompts/library/draftMbl/other/placeholders.json,sha256=wIN06_NWsESDyNEDfOLPi3F2Vq-XPa4O3U32A32s-_Q,1736
40
- src/prompts/library/draftMbl/other/prompt.txt,sha256=gqbPm1joXKDUss0wU6vMc-269sx-fYWh90gWuNKOBQc,2166
41
- src/prompts/library/finalMbL/hapag-lloyd/prompt.txt,sha256=0k1xLW4zWaenCSNQJxXMXenIwI-eYmGgpxnAAcM3HOg,2251
42
- src/prompts/library/finalMbL/maersk/prompt.txt,sha256=GxaIYlksORvD2uAbodRx_9JFJXD4XbDaVFYtpN9uzxc,2050
43
- src/prompts/library/finalMbL/other/placeholders.json,sha256=K_yJYhQo2DnZV_Rg6xXjo6sHkSGB-SMO4IQnY47V43w,1735
44
- src/prompts/library/finalMbL/other/prompt.txt,sha256=gqbPm1joXKDUss0wU6vMc-269sx-fYWh90gWuNKOBQc,2166
45
- src/prompts/library/packingList/other/prompt.txt,sha256=Qw16n7_48GGFYWz2vRepNowZCX1UPXKetEZ1UqFXPdY,2764
46
- src/prompts/library/partnerInvoice/other/placeholders.json,sha256=y_S_iAr-hHFzrPJgquZEtR66ETMvdWp1rtwCLrLPQUU,12139
47
- src/prompts/library/partnerInvoice/other/prompt.txt,sha256=hHzI5Vq3BABBWPMsv72nlahfYjgq08OSpflDDhV0NX8,9360
48
- src/prompts/library/postprocessing/port_code/placeholders.json,sha256=2TiXf3zSzrglOMPtDOlCntIa5RSvyZQAKG2-IgrCY5A,22
49
- src/prompts/library/postprocessing/port_code/prompt_port_code.txt,sha256=--1wunSqEr2ox958lEhjO-0JFBfOLzA3qfKYIzG_Iok,884
50
- src/prompts/library/preprocessing/carrier/placeholders.json,sha256=1UmrQNqBEsjLIpOO-a39Az6bQ_g1lxDGlwqZFU3IEt0,408
51
- src/prompts/library/preprocessing/carrier/prompt.txt,sha256=NLvRZQCZ6aWC1yTr7Q93jK5z7Vi_b4HBaiFYYnIsO-w,134
52
- src/prompts/library/shippingInstruction/other/prompt.txt,sha256=fyC24ig4FyRNnLuQM69s4ZVajsK-LHIl2dvaaEXr-6Q,1327
53
- src/prompts/prompt_library.py,sha256=VJWHeXN-s501C2GiidIIvQQuZdU6T1R27hE2dKBiI40,2555
54
- src/setup.py,sha256=kPSZosrICfaGZeDaajr40Ha7Ok4XK4fo_uq35Omiwr0,7128
55
- src/tms.py,sha256=UXbIo1QE--hIX6NZi5Qyp2R_CP338syrY9pCTPrfgnE,1741
56
- src/utils.py,sha256=-1Yq_5ExZlFQRUPRsQHiBD3TthNSiPVPp46Dvdb9Kf0,13830
57
- data_science_document_ai-1.40.4.dist-info/METADATA,sha256=zM542Z9wdq9B2SaEmjoAEzj20BYlbL9LxwjZvMXb22w,2153
58
- data_science_document_ai-1.40.4.dist-info/WHEEL,sha256=zp0Cn7JsFoX2ATtOhtaFYIiE2rmFAD4OcMhtUki8W3U,88
59
- data_science_document_ai-1.40.4.dist-info/RECORD,,
@@ -1,19 +0,0 @@
1
- {
2
- "SCHEMA_PLACEHOLDER": {
3
- "consignee": "",
4
- "shipper":"",
5
- "countryOfOrigin": "",
6
- "MRN": "",
7
- "totalValueOfGoods": "",
8
- "containers": [
9
- {
10
- "containerNumber": "",
11
- "grossWeight": "",
12
- "goodsDescription": "",
13
- "nettWeight": "",
14
- "packagingNumber": "",
15
- "packagingType": ""
16
- }
17
- ]
18
- }
19
- }
@@ -1,80 +0,0 @@
1
- {
2
- "type": "OBJECT",
3
- "properties": {
4
- "blNumber": {
5
- "type": "string",
6
- "nullable": true,
7
- "description": ""
8
- },
9
- "voyage": {
10
- "type": "string",
11
- "nullable": true,
12
- "description": ""
13
- },
14
- "portOfLoading": {
15
- "type": "string",
16
- "nullable": true,
17
- "description": ""
18
- },
19
- "portOfDischarge": {
20
- "type": "string",
21
- "nullable": true,
22
- "description": ""
23
- },
24
- "bookingNumber": {
25
- "type": "string",
26
- "nullable": true,
27
- "description": ""
28
- },
29
- "containers": {
30
- "type": "ARRAY",
31
- "items": {
32
- "type": "OBJECT",
33
- "properties": {
34
- "containerType": {
35
- "type": "string",
36
- "nullable": true,
37
- "description": ""
38
- },
39
- "grossWeight": {
40
- "type": "string",
41
- "nullable": true,
42
- "description": ""
43
- },
44
- "measurements": {
45
- "type": "string",
46
- "nullable": true,
47
- "description": ""
48
- },
49
- "packageQuantity": {
50
- "type": "string",
51
- "nullable": true,
52
- "description": ""
53
- },
54
- "packageType": {
55
- "type": "string",
56
- "nullable": true,
57
- "description": ""
58
- },
59
- "containerNumber": {
60
- "type": "string",
61
- "nullable": true,
62
- "description": ""
63
- },
64
- "sealNumber": {
65
- "type": "string",
66
- "nullable": true,
67
- "description": ""
68
- }
69
- },
70
- "required": []
71
- }
72
- },
73
- "vessel": {
74
- "type": "string",
75
- "nullable": true,
76
- "description": ""
77
- }
78
- },
79
- "required": []
80
- }