trustgraph-ocr 2.3.19__tar.gz → 2.4.0__tar.gz
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.
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/PKG-INFO +2 -2
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/pyproject.toml +1 -1
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph/decoding/ocr/pdf_decoder.py +3 -5
- trustgraph_ocr-2.4.0/trustgraph/ocr_version.py +1 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph_ocr.egg-info/PKG-INFO +2 -2
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph_ocr.egg-info/requires.txt +1 -1
- trustgraph_ocr-2.3.19/trustgraph/ocr_version.py +0 -1
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/README.md +0 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/setup.cfg +0 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph/decoding/ocr/__init__.py +0 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph/decoding/ocr/__main__.py +0 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph_ocr.egg-info/SOURCES.txt +0 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph_ocr.egg-info/dependency_links.txt +0 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph_ocr.egg-info/entry_points.txt +0 -0
- {trustgraph_ocr-2.3.19 → trustgraph_ocr-2.4.0}/trustgraph_ocr.egg-info/top_level.txt +0 -0
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.4
|
|
2
2
|
Name: trustgraph-ocr
|
|
3
|
-
Version: 2.
|
|
3
|
+
Version: 2.4.0
|
|
4
4
|
Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
|
|
5
5
|
Author-email: "trustgraph.ai" <security@trustgraph.ai>
|
|
6
6
|
Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph
|
|
@@ -8,7 +8,7 @@ Classifier: Programming Language :: Python :: 3
|
|
|
8
8
|
Classifier: Operating System :: OS Independent
|
|
9
9
|
Requires-Python: >=3.8
|
|
10
10
|
Description-Content-Type: text/markdown
|
|
11
|
-
Requires-Dist: trustgraph-base<2.
|
|
11
|
+
Requires-Dist: trustgraph-base<2.5,>=2.4
|
|
12
12
|
Requires-Dist: pulsar-client
|
|
13
13
|
Requires-Dist: prometheus-client
|
|
14
14
|
Requires-Dist: boto3
|
|
@@ -10,7 +10,7 @@ description = "TrustGraph provides a means to run a pipeline of flexible AI proc
|
|
|
10
10
|
readme = "README.md"
|
|
11
11
|
requires-python = ">=3.8"
|
|
12
12
|
dependencies = [
|
|
13
|
-
"trustgraph-base>=2.
|
|
13
|
+
"trustgraph-base>=2.4,<2.5",
|
|
14
14
|
"pulsar-client",
|
|
15
15
|
"prometheus-client",
|
|
16
16
|
"boto3",
|
|
@@ -91,7 +91,7 @@ class Processor(FlowProcessor):
|
|
|
91
91
|
if v.document_id:
|
|
92
92
|
doc_meta = await self.librarian.fetch_document_metadata(
|
|
93
93
|
document_id=v.document_id,
|
|
94
|
-
|
|
94
|
+
workspace=flow.workspace,
|
|
95
95
|
)
|
|
96
96
|
if doc_meta and doc_meta.kind and doc_meta.kind != "application/pdf":
|
|
97
97
|
logger.error(
|
|
@@ -106,7 +106,7 @@ class Processor(FlowProcessor):
|
|
|
106
106
|
logger.info(f"Fetching document {v.document_id} from librarian...")
|
|
107
107
|
content = await self.librarian.fetch_document_content(
|
|
108
108
|
document_id=v.document_id,
|
|
109
|
-
|
|
109
|
+
workspace=flow.workspace,
|
|
110
110
|
)
|
|
111
111
|
if isinstance(content, str):
|
|
112
112
|
content = content.encode('utf-8')
|
|
@@ -141,7 +141,7 @@ class Processor(FlowProcessor):
|
|
|
141
141
|
await self.librarian.save_child_document(
|
|
142
142
|
doc_id=page_doc_id,
|
|
143
143
|
parent_id=source_doc_id,
|
|
144
|
-
|
|
144
|
+
workspace=flow.workspace,
|
|
145
145
|
content=page_content,
|
|
146
146
|
document_type="page",
|
|
147
147
|
title=f"Page {page_num}",
|
|
@@ -163,7 +163,6 @@ class Processor(FlowProcessor):
|
|
|
163
163
|
metadata=Metadata(
|
|
164
164
|
id=pg_uri,
|
|
165
165
|
root=v.metadata.root,
|
|
166
|
-
user=v.metadata.user,
|
|
167
166
|
collection=v.metadata.collection,
|
|
168
167
|
),
|
|
169
168
|
triples=set_graph(prov_triples, GRAPH_SOURCE),
|
|
@@ -175,7 +174,6 @@ class Processor(FlowProcessor):
|
|
|
175
174
|
metadata=Metadata(
|
|
176
175
|
id=pg_uri,
|
|
177
176
|
root=v.metadata.root,
|
|
178
|
-
user=v.metadata.user,
|
|
179
177
|
collection=v.metadata.collection,
|
|
180
178
|
),
|
|
181
179
|
document_id=page_doc_id,
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
__version__ = "2.4.0"
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.4
|
|
2
2
|
Name: trustgraph-ocr
|
|
3
|
-
Version: 2.
|
|
3
|
+
Version: 2.4.0
|
|
4
4
|
Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
|
|
5
5
|
Author-email: "trustgraph.ai" <security@trustgraph.ai>
|
|
6
6
|
Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph
|
|
@@ -8,7 +8,7 @@ Classifier: Programming Language :: Python :: 3
|
|
|
8
8
|
Classifier: Operating System :: OS Independent
|
|
9
9
|
Requires-Python: >=3.8
|
|
10
10
|
Description-Content-Type: text/markdown
|
|
11
|
-
Requires-Dist: trustgraph-base<2.
|
|
11
|
+
Requires-Dist: trustgraph-base<2.5,>=2.4
|
|
12
12
|
Requires-Dist: pulsar-client
|
|
13
13
|
Requires-Dist: prometheus-client
|
|
14
14
|
Requires-Dist: boto3
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
__version__ = "2.3.19"
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|