docler 2.1.1__tar.gz → 2.2.0__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.
Files changed (98) hide show
  1. {docler-2.1.1 → docler-2.2.0}/PKG-INFO +29 -42
  2. {docler-2.1.1 → docler-2.2.0}/README.md +24 -37
  3. docler-2.2.0/pyproject.toml +338 -0
  4. docler-2.1.1/pyproject.toml → docler-2.2.0/pyproject.toml.orig +34 -5
  5. {docler-2.1.1 → docler-2.2.0}/src/docler/__main__.py +136 -133
  6. {docler-2.1.1 → docler-2.2.0}/src/docler/annotators/__init__.py +1 -1
  7. {docler-2.1.1 → docler-2.2.0}/src/docler/annotators/base.py +36 -36
  8. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/ai_chunker/chunker.py +1 -1
  9. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/token_chunker/chunker.py +1 -1
  10. {docler-2.1.1 → docler-2.2.0}/src/docler/common_types.py +49 -49
  11. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/azure_provider/__init__.py +5 -5
  12. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/azure_provider/provider.py +1 -1
  13. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/datalab_provider/__init__.py +5 -5
  14. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/datalab_provider/provider.py +1 -1
  15. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/dir_converter.py +179 -179
  16. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/docling_provider/__init__.py +5 -5
  17. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/docling_provider/provider.py +2 -2
  18. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/docling_provider/utils.py +82 -82
  19. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/docling_remote_provider/provider.py +1 -1
  20. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/exceptions.py +9 -9
  21. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/llamaparse_provider/__init__.py +5 -5
  22. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/llamaparse_provider/provider.py +29 -8
  23. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/llm_provider/__init__.py +5 -5
  24. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/llm_provider/provider.py +2 -3
  25. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/marker_provider/__init__.py +5 -5
  26. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/marker_provider/provider.py +1 -1
  27. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/markitdown_provider/__init__.py +5 -5
  28. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/mistral_provider/__init__.py +5 -5
  29. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/mistral_provider/provider.py +3 -3
  30. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/mistral_provider/utils.py +1 -1
  31. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/upstage_provider/__init__.py +5 -5
  32. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/upstage_provider/provider.py +4 -0
  33. {docler-2.1.1 → docler-2.2.0}/src/docler/diffs.py +90 -90
  34. {docler-2.1.1 → docler-2.2.0}/src/docler/mime_types.py +174 -174
  35. {docler-2.1.1 → docler-2.2.0}/src/docler/process_runner.py +276 -276
  36. {docler-2.1.1 → docler-2.2.0}/src/docler/processors/base.py +35 -35
  37. {docler-2.1.1 → docler-2.2.0}/src/docler/provider.py +66 -66
  38. {docler-2.1.1 → docler-2.2.0}/src/docler_api/__init__.py +3 -3
  39. {docler-2.1.1 → docler-2.2.0}/src/docler_config/__init__.py +91 -91
  40. {docler-2.1.1 → docler-2.2.0}/src/docler_config/annotator_configs.py +117 -117
  41. {docler-2.1.1 → docler-2.2.0}/src/docler_config/chunker_configs.py +146 -146
  42. {docler-2.1.1 → docler-2.2.0}/src/docler_config/converter_configs.py +17 -8
  43. {docler-2.1.1 → docler-2.2.0}/src/docler_config/processor_configs.py +79 -79
  44. {docler-2.1.1 → docler-2.2.0}/src/docler_config/provider.py +44 -44
  45. {docler-2.1.1 → docler-2.2.0}/src/docler_streamlit/__init__.py +6 -6
  46. {docler-2.1.1 → docler-2.2.0}/src/docler_streamlit/ocr_app.py +56 -56
  47. {docler-2.1.1 → docler-2.2.0}/LICENSE +0 -0
  48. {docler-2.1.1 → docler-2.2.0}/src/docler/__init__.py +0 -0
  49. {docler-2.1.1 → docler-2.2.0}/src/docler/annotators/ai_document_annotator.py +0 -0
  50. {docler-2.1.1 → docler-2.2.0}/src/docler/annotators/ai_image_annotator.py +0 -0
  51. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/__init__.py +0 -0
  52. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/ai_chunker/__init__.py +0 -0
  53. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/ai_chunker/models.py +0 -0
  54. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/ai_chunker/utils.py +0 -0
  55. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/base.py +0 -0
  56. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/llamaindex_chunker/__init__.py +0 -0
  57. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/llamaindex_chunker/chunker.py +0 -0
  58. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/markdown_chunker/__init__.py +0 -0
  59. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/markdown_chunker/chunker.py +0 -0
  60. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/markdown_chunker/utils.py +0 -0
  61. {docler-2.1.1 → docler-2.2.0}/src/docler/chunkers/token_chunker/__init__.py +0 -0
  62. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/__init__.py +0 -0
  63. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/aggregated_converter.py +0 -0
  64. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/azure_provider/utils.py +0 -0
  65. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/base.py +0 -0
  66. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/datalab_provider/utils.py +0 -0
  67. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/docling_remote_provider/__init__.py +0 -0
  68. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/docling_remote_provider/utils.py +0 -0
  69. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/llamaparse_provider/utils.py +0 -0
  70. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/markitdown_provider/provider.py +0 -0
  71. {docler-2.1.1 → docler-2.2.0}/src/docler/converters/registry.py +0 -0
  72. {docler-2.1.1 → docler-2.2.0}/src/docler/log.py +0 -0
  73. {docler-2.1.1 → docler-2.2.0}/src/docler/models.py +0 -0
  74. {docler-2.1.1 → docler-2.2.0}/src/docler/pdf_utils.py +0 -0
  75. {docler-2.1.1 → docler-2.2.0}/src/docler/processors/__init__.py +0 -0
  76. {docler-2.1.1 → docler-2.2.0}/src/docler/processors/ai_processor.py +0 -0
  77. {docler-2.1.1 → docler-2.2.0}/src/docler/py.typed +0 -0
  78. {docler-2.1.1 → docler-2.2.0}/src/docler/resources/pdf_sample.pdf +0 -0
  79. {docler-2.1.1 → docler-2.2.0}/src/docler/resources/samplepptx.pptx +0 -0
  80. {docler-2.1.1 → docler-2.2.0}/src/docler/resources/test.docx +0 -0
  81. {docler-2.1.1 → docler-2.2.0}/src/docler/resources/test.jpg +0 -0
  82. {docler-2.1.1 → docler-2.2.0}/src/docler/resources/test.pptx +0 -0
  83. {docler-2.1.1 → docler-2.2.0}/src/docler/resources/test.xls +0 -0
  84. {docler-2.1.1 → docler-2.2.0}/src/docler/resources/test.xlsx +0 -0
  85. {docler-2.1.1 → docler-2.2.0}/src/docler/utils.py +0 -0
  86. {docler-2.1.1 → docler-2.2.0}/src/docler_api/README.md +0 -0
  87. {docler-2.1.1 → docler-2.2.0}/src/docler_api/main.py +0 -0
  88. {docler-2.1.1 → docler-2.2.0}/src/docler_api/py.typed +0 -0
  89. {docler-2.1.1 → docler-2.2.0}/src/docler_api/routes.py +0 -0
  90. {docler-2.1.1 → docler-2.2.0}/src/docler_config/py.typed +0 -0
  91. {docler-2.1.1 → docler-2.2.0}/src/docler_mcp/__init__.py +0 -0
  92. {docler-2.1.1 → docler-2.2.0}/src/docler_mcp/py.typed +0 -0
  93. {docler-2.1.1 → docler-2.2.0}/src/docler_mcp/server.py +0 -0
  94. {docler-2.1.1 → docler-2.2.0}/src/docler_streamlit/chunk_app.py +0 -0
  95. {docler-2.1.1 → docler-2.2.0}/src/docler_streamlit/chunkers.py +0 -0
  96. {docler-2.1.1 → docler-2.2.0}/src/docler_streamlit/converters.py +0 -0
  97. {docler-2.1.1 → docler-2.2.0}/src/docler_streamlit/py.typed +0 -0
  98. {docler-2.1.1 → docler-2.2.0}/src/docler_streamlit/utils.py +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: docler
3
- Version: 2.1.1
3
+ Version: 2.2.0
4
4
  Summary: Abstractions & Tools for OCR / document processing
5
5
  Keywords:
6
6
  Author: Philipp Temminghoff
@@ -29,18 +29,18 @@ Requires-Dist: pypdf
29
29
  Requires-Dist: pytest-docker>=3.2.1
30
30
  Requires-Dist: schemez>=0.0.1
31
31
  Requires-Dist: upathtools>=0.4.3
32
- Requires-Dist: docler[marker,azure,diffs,docling,llama-parse,marker,markitdown,mcp,mistralai,openai,server] ; extra == 'all'
33
- Requires-Dist: agentpool[default] ; extra == 'all'
32
+ Requires-Dist: docler[azure,diffs,docling,llama-parse,markitdown,mcp,mistralai,openai,server] ; extra == 'all'
33
+ Requires-Dist: agentpool>=2.10.0 ; extra == 'all'
34
34
  Requires-Dist: azure-ai-documentintelligence ; extra == 'azure'
35
35
  Requires-Dist: diff-match-patch ; extra == 'diffs'
36
36
  Requires-Dist: docling[ocrmac,vlm,rapidocr] ; python_full_version < '3.14' and extra == 'docling'
37
37
  Requires-Dist: docler[mistralai,diffs,llama-parse,openai,azure] ; extra == 'light'
38
- Requires-Dist: agentpool[default] ; extra == 'light'
38
+ Requires-Dist: agentpool>=2.10.0 ; extra == 'light'
39
39
  Requires-Dist: llama-parse ; python_full_version < '3.14' and extra == 'llama-parse'
40
40
  Requires-Dist: marker-pdf ; python_full_version < '3.14' and extra == 'marker'
41
41
  Requires-Dist: markitdown[all] ; python_full_version < '3.14' and extra == 'markitdown'
42
42
  Requires-Dist: fastmcp ; extra == 'mcp'
43
- Requires-Dist: mistralai ; extra == 'mistralai'
43
+ Requires-Dist: mistralai>=3 ; extra == 'mistralai'
44
44
  Requires-Dist: olmocr ; python_full_version < '3.14' and extra == 'olmocr'
45
45
  Requires-Dist: accelerate ; python_full_version < '3.14' and extra == 'olmocr'
46
46
  Requires-Dist: openai ; extra == 'openai'
@@ -118,6 +118,7 @@ A unified Python library for document conversion and OCR that provides a consist
118
118
  import asyncio
119
119
  from docler import MistralConverter
120
120
 
121
+
121
122
  async def main():
122
123
  # Use the aggregated converter for automatic provider selection
123
124
  converter = MistralConverter()
@@ -130,6 +131,7 @@ async def main():
130
131
  print(f"Images: {len(result.images)} extracted")
131
132
  print(f"Pages: {result.page_count}")
132
133
 
134
+
133
135
  asyncio.run(main())
134
136
  ```
135
137
 
@@ -142,11 +144,7 @@ asyncio.run(main())
142
144
  ```python
143
145
  from docler import AzureConverter
144
146
 
145
- converter = AzureConverter(
146
- endpoint="your-endpoint",
147
- api_key="your-key",
148
- model="prebuilt-layout"
149
- )
147
+ converter = AzureConverter(endpoint="your-endpoint", api_key="your-key", model="prebuilt-layout")
150
148
  ```
151
149
 
152
150
  #### Mistral OCR
@@ -154,10 +152,7 @@ converter = AzureConverter(
154
152
  ```python
155
153
  from docler import MistralConverter
156
154
 
157
- converter = MistralConverter(
158
- api_key="your-key",
159
- languages=["en", "fr", "de"]
160
- )
155
+ converter = MistralConverter(api_key="your-key", languages=["en", "fr", "de"])
161
156
  ```
162
157
 
163
158
  #### LlamaParse
@@ -165,10 +160,7 @@ converter = MistralConverter(
165
160
  ```python
166
161
  from docler import LlamaParseConverter
167
162
 
168
- converter = LlamaParseConverter(
169
- api_key="your-key",
170
- adaptive_long_table=True
171
- )
163
+ converter = LlamaParseConverter(api_key="your-key", adaptive_long_table=True)
172
164
  ```
173
165
 
174
166
  #### Upstage Document AI
@@ -176,10 +168,7 @@ converter = LlamaParseConverter(
176
168
  ```python
177
169
  from docler import UpstageConverter
178
170
 
179
- converter = UpstageConverter(
180
- api_key="your-key",
181
- chart_recognition=True
182
- )
171
+ converter = UpstageConverter(api_key="your-key", chart_recognition=True)
183
172
  ```
184
173
 
185
174
  #### DataLab
@@ -189,7 +178,7 @@ from docler import DataLabConverter
189
178
 
190
179
  converter = DataLabConverter(
191
180
  api_key="your-key",
192
- use_llm=False # Enable for higher accuracy
181
+ use_llm=False, # Enable for higher accuracy
193
182
  )
194
183
  ```
195
184
 
@@ -203,7 +192,7 @@ from docler import MarkerConverter
203
192
  converter = MarkerConverter(
204
193
  dpi=192,
205
194
  use_llm=True, # Requires local LLM setup
206
- llm_provider="ollama"
195
+ llm_provider="ollama",
207
196
  )
208
197
  ```
209
198
 
@@ -212,10 +201,7 @@ converter = MarkerConverter(
212
201
  ```python
213
202
  from docler import DoclingConverter
214
203
 
215
- converter = DoclingConverter(
216
- ocr_engine="easy_ocr",
217
- image_scale=2.0
218
- )
204
+ converter = DoclingConverter(ocr_engine="rapid_ocr", image_scale=2.0)
219
205
  ```
220
206
 
221
207
  #### Docling Remote
@@ -223,10 +209,7 @@ converter = DoclingConverter(
223
209
  ```python
224
210
  from docler import DoclingRemoteConverter
225
211
 
226
- converter = DoclingRemoteConverter(
227
- endpoint="http://localhost:5001",
228
- pdf_backend="dlparse_v4"
229
- )
212
+ converter = DoclingRemoteConverter(endpoint="http://localhost:5001", pdf_backend="dlparse_v4")
230
213
  ```
231
214
 
232
215
  #### MarkItDown (Microsoft)
@@ -246,7 +229,7 @@ from docler import LLMConverter
246
229
 
247
230
  converter = LLMConverter(
248
231
  model="gpt-4o", # or claude-3-5-sonnet, etc.
249
- system_prompt="Extract text preserving formatting..."
232
+ system_prompt="Extract text preserving formatting...",
250
233
  )
251
234
  ```
252
235
 
@@ -315,13 +298,13 @@ All converters return a standardized `Document` object with:
315
298
 
316
299
  ```python
317
300
  class Document:
318
- content: str # Extracted text in markdown format
319
- images: list[Image] # Extracted images with metadata
320
- title: str # Document title
321
- source_path: str # Original file path
322
- mime_type: str # File MIME type
323
- metadata: dict # Provider-specific metadata
324
- page_count: int # Number of pages processed
301
+ content: str # Extracted text in markdown format
302
+ images: list[Image] # Extracted images with metadata
303
+ title: str # Document title
304
+ source_path: str # Original file path
305
+ mime_type: str # File MIME type
306
+ metadata: dict # Provider-specific metadata
307
+ page_count: int # Number of pages processed
325
308
  ```
326
309
 
327
310
  The markdown content includes standardized metadata for page breaks and structure:
@@ -344,11 +327,15 @@ pip install docler
344
327
 
345
328
  # With specific provider dependencies
346
329
  pip install docler[azure] # Azure Document Intelligence
347
- pip install docler[mistral] # Mistral OCR
330
+ pip install docler[mistralai] # Mistral OCR
348
331
  pip install docler[marker] # Marker PDF processing
349
- pip install docler[all] # All providers
332
+ pip install docler[all] # All providers except Marker
350
333
  ```
351
334
 
335
+ Marker cannot be installed together with `docler[all]`/`docler[light]`: marker-pdf
336
+ requires `openai<3`, while the AI features (agentpool) require `openai>=3.31`.
337
+ Install `docler[marker]` in a separate environment.
338
+
352
339
  ## Environment Variables
353
340
 
354
341
  Configure API keys via environment variables:
@@ -44,6 +44,7 @@ A unified Python library for document conversion and OCR that provides a consist
44
44
  import asyncio
45
45
  from docler import MistralConverter
46
46
 
47
+
47
48
  async def main():
48
49
  # Use the aggregated converter for automatic provider selection
49
50
  converter = MistralConverter()
@@ -56,6 +57,7 @@ async def main():
56
57
  print(f"Images: {len(result.images)} extracted")
57
58
  print(f"Pages: {result.page_count}")
58
59
 
60
+
59
61
  asyncio.run(main())
60
62
  ```
61
63
 
@@ -68,11 +70,7 @@ asyncio.run(main())
68
70
  ```python
69
71
  from docler import AzureConverter
70
72
 
71
- converter = AzureConverter(
72
- endpoint="your-endpoint",
73
- api_key="your-key",
74
- model="prebuilt-layout"
75
- )
73
+ converter = AzureConverter(endpoint="your-endpoint", api_key="your-key", model="prebuilt-layout")
76
74
  ```
77
75
 
78
76
  #### Mistral OCR
@@ -80,10 +78,7 @@ converter = AzureConverter(
80
78
  ```python
81
79
  from docler import MistralConverter
82
80
 
83
- converter = MistralConverter(
84
- api_key="your-key",
85
- languages=["en", "fr", "de"]
86
- )
81
+ converter = MistralConverter(api_key="your-key", languages=["en", "fr", "de"])
87
82
  ```
88
83
 
89
84
  #### LlamaParse
@@ -91,10 +86,7 @@ converter = MistralConverter(
91
86
  ```python
92
87
  from docler import LlamaParseConverter
93
88
 
94
- converter = LlamaParseConverter(
95
- api_key="your-key",
96
- adaptive_long_table=True
97
- )
89
+ converter = LlamaParseConverter(api_key="your-key", adaptive_long_table=True)
98
90
  ```
99
91
 
100
92
  #### Upstage Document AI
@@ -102,10 +94,7 @@ converter = LlamaParseConverter(
102
94
  ```python
103
95
  from docler import UpstageConverter
104
96
 
105
- converter = UpstageConverter(
106
- api_key="your-key",
107
- chart_recognition=True
108
- )
97
+ converter = UpstageConverter(api_key="your-key", chart_recognition=True)
109
98
  ```
110
99
 
111
100
  #### DataLab
@@ -115,7 +104,7 @@ from docler import DataLabConverter
115
104
 
116
105
  converter = DataLabConverter(
117
106
  api_key="your-key",
118
- use_llm=False # Enable for higher accuracy
107
+ use_llm=False, # Enable for higher accuracy
119
108
  )
120
109
  ```
121
110
 
@@ -129,7 +118,7 @@ from docler import MarkerConverter
129
118
  converter = MarkerConverter(
130
119
  dpi=192,
131
120
  use_llm=True, # Requires local LLM setup
132
- llm_provider="ollama"
121
+ llm_provider="ollama",
133
122
  )
134
123
  ```
135
124
 
@@ -138,10 +127,7 @@ converter = MarkerConverter(
138
127
  ```python
139
128
  from docler import DoclingConverter
140
129
 
141
- converter = DoclingConverter(
142
- ocr_engine="easy_ocr",
143
- image_scale=2.0
144
- )
130
+ converter = DoclingConverter(ocr_engine="rapid_ocr", image_scale=2.0)
145
131
  ```
146
132
 
147
133
  #### Docling Remote
@@ -149,10 +135,7 @@ converter = DoclingConverter(
149
135
  ```python
150
136
  from docler import DoclingRemoteConverter
151
137
 
152
- converter = DoclingRemoteConverter(
153
- endpoint="http://localhost:5001",
154
- pdf_backend="dlparse_v4"
155
- )
138
+ converter = DoclingRemoteConverter(endpoint="http://localhost:5001", pdf_backend="dlparse_v4")
156
139
  ```
157
140
 
158
141
  #### MarkItDown (Microsoft)
@@ -172,7 +155,7 @@ from docler import LLMConverter
172
155
 
173
156
  converter = LLMConverter(
174
157
  model="gpt-4o", # or claude-3-5-sonnet, etc.
175
- system_prompt="Extract text preserving formatting..."
158
+ system_prompt="Extract text preserving formatting...",
176
159
  )
177
160
  ```
178
161
 
@@ -241,13 +224,13 @@ All converters return a standardized `Document` object with:
241
224
 
242
225
  ```python
243
226
  class Document:
244
- content: str # Extracted text in markdown format
245
- images: list[Image] # Extracted images with metadata
246
- title: str # Document title
247
- source_path: str # Original file path
248
- mime_type: str # File MIME type
249
- metadata: dict # Provider-specific metadata
250
- page_count: int # Number of pages processed
227
+ content: str # Extracted text in markdown format
228
+ images: list[Image] # Extracted images with metadata
229
+ title: str # Document title
230
+ source_path: str # Original file path
231
+ mime_type: str # File MIME type
232
+ metadata: dict # Provider-specific metadata
233
+ page_count: int # Number of pages processed
251
234
  ```
252
235
 
253
236
  The markdown content includes standardized metadata for page breaks and structure:
@@ -270,11 +253,15 @@ pip install docler
270
253
 
271
254
  # With specific provider dependencies
272
255
  pip install docler[azure] # Azure Document Intelligence
273
- pip install docler[mistral] # Mistral OCR
256
+ pip install docler[mistralai] # Mistral OCR
274
257
  pip install docler[marker] # Marker PDF processing
275
- pip install docler[all] # All providers
258
+ pip install docler[all] # All providers except Marker
276
259
  ```
277
260
 
261
+ Marker cannot be installed together with `docler[all]`/`docler[light]`: marker-pdf
262
+ requires `openai<3`, while the AI features (agentpool) require `openai>=3.31`.
263
+ Install `docler[marker]` in a separate environment.
264
+
278
265
  ## Environment Variables
279
266
 
280
267
  Configure API keys via environment variables:
@@ -0,0 +1,338 @@
1
+ [project]
2
+ name = "docler"
3
+ version = "2.2.0"
4
+ description = "Abstractions & Tools for OCR / document processing"
5
+ readme = "README.md"
6
+ requires-python = ">=3.13"
7
+ license = "MIT"
8
+ license-files = ["LICENSE"]
9
+ keywords = []
10
+ classifiers = [
11
+ "Development Status :: 4 - Beta",
12
+ "Framework :: Pydantic",
13
+ "Framework :: Pydantic :: 2",
14
+ "Intended Audience :: Developers",
15
+ "Operating System :: OS Independent",
16
+ "Programming Language :: Python :: 3",
17
+ "Programming Language :: Python :: 3 :: Only",
18
+ "Programming Language :: Python :: 3.12",
19
+ "Programming Language :: Python :: 3.13",
20
+ "Programming Language :: Python :: 3.14",
21
+ "Topic :: Documentation",
22
+ "Topic :: Software Development",
23
+ "Topic :: Utilities",
24
+ "Typing :: Typed",
25
+ ]
26
+ dependencies = [
27
+ "anyenv>=0.4.1",
28
+ "mkdown>=0.12.1",
29
+ "pydantic",
30
+ "pydantic-settings>=2.8.1",
31
+ "pypdf",
32
+ "pytest-docker>=3.2.1",
33
+ "schemez>=0.0.1",
34
+ "upathtools>=0.4.3",
35
+ ]
36
+
37
+ [[project.authors]]
38
+ name = "Philipp Temminghoff"
39
+ email = "philipptemminghoff@googlemail.com"
40
+
41
+ [project.urls]
42
+ "Code coverage" = "https://app.codecov.io/gh/phil65/docler"
43
+ Discussions = "https://github.com/phil65/docler/discussions"
44
+ Documentation = "https://phil65.github.io/docler/"
45
+ Issues = "https://github.com/phil65/docler/issues"
46
+ Source = "https://github.com/phil65/docler"
47
+
48
+ [project.scripts]
49
+ docler = "docler.__main__:cli"
50
+
51
+ [project.optional-dependencies]
52
+ all = [
53
+ "docler[azure,diffs,docling,llama-parse,markitdown,mcp,mistralai,openai,server]",
54
+ "agentpool>=2.10.0",
55
+ ]
56
+ azure = ["azure-ai-documentintelligence"]
57
+ diffs = ["diff-match-patch"]
58
+ docling = ["docling[ocrmac,vlm,rapidocr]; python_version < '3.14'"]
59
+ light = [
60
+ "docler[mistralai,diffs,llama-parse,openai,azure]",
61
+ "agentpool>=2.10.0",
62
+ ]
63
+ llama-parse = ["llama-parse; python_version < '3.14'"]
64
+ marker = ["marker-pdf; python_version < '3.14'"]
65
+ markitdown = ["markitdown[all]; python_version < '3.14'"]
66
+ mcp = ["fastmcp"]
67
+ mistralai = ["mistralai>=3"]
68
+ olmocr = [
69
+ "olmocr; python_version < '3.14'",
70
+ "accelerate; python_version < '3.14'",
71
+ ]
72
+ openai = ["openai"]
73
+ server = [
74
+ "uvicorn",
75
+ "fastapi",
76
+ ]
77
+ streamlit = [
78
+ "streamlit; python_version < '3.14'",
79
+ "streambricks; python_version < '3.14'",
80
+ ]
81
+ token-chunking = ["tokonomics"]
82
+
83
+ [dependency-groups]
84
+ dev = [
85
+ "devtools",
86
+ "pyreadline3",
87
+ "pytest",
88
+ "pytest-asyncio",
89
+ "pytest-cov",
90
+ "syrupy>=4.9.1",
91
+ ]
92
+ benchmark = ["pyinstrument"]
93
+ docs = [
94
+ "mkdocs-material",
95
+ "mkdocs-mknodes",
96
+ ]
97
+ lint = [
98
+ "mypy; python_version >= '3.14'",
99
+ "mypy[faster-cache]; python_version < '3.14'",
100
+ "ruff",
101
+ ]
102
+
103
+ [build-system]
104
+ requires = ["uv_build>=0.8.22"]
105
+ build-backend = "uv_build"
106
+
107
+ [tool.coverage.report]
108
+ exclude_lines = [
109
+ "pragma: no cover",
110
+ "if TYPE_CHECKING:",
111
+ "@overload",
112
+ "except ImportError",
113
+ 'if __name__ == "__main__":',
114
+ ]
115
+
116
+ [tool.mknodes]
117
+ allowed-commit-types = [
118
+ "fix",
119
+ "feat",
120
+ "refactor",
121
+ "docs",
122
+ "test",
123
+ "build",
124
+ "chore",
125
+ ]
126
+ docstring-style = "google"
127
+
128
+ [tool.mypy]
129
+ python_version = "3.13"
130
+ disable_error_code = ["import"]
131
+ pretty = true
132
+ strict = true
133
+ check_untyped_defs = true
134
+ exclude = [
135
+ "venv/",
136
+ ".venv/",
137
+ "tests/",
138
+ ]
139
+ plugins = ["pydantic.mypy"]
140
+
141
+ [tool.pyright]
142
+ venvPath = "."
143
+ venv = ".venv"
144
+ pythonVersion = "3.13"
145
+ pythonPlatform = "All"
146
+ typeCheckingMode = "basic"
147
+ deprecateTypingAliases = true
148
+ reportMissingTypeStubs = false
149
+ reportUnusedCallResult = false
150
+ reportUnknownVariableType = false
151
+ reportAny = false
152
+ reportImplicitOverride = false
153
+ reportUnusedFunction = false
154
+ reportImplicitStringConcatenation = false
155
+ reportIgnoreCommentWithoutRule = false
156
+ reportUnannotatedClassAttribute = false
157
+ reportSelfClsParameterName = false
158
+ reportPrivateImportUsage = false
159
+
160
+ [tool.pytest]
161
+ log_cli = true
162
+ log_date_format = "%Y-%m-%d %H:%M:%S"
163
+ log_format = "%(asctime)s %(levelname)s %(message)s"
164
+ minversion = "9.0"
165
+ testpaths = ["tests/"]
166
+ asyncio_mode = "auto"
167
+
168
+ [tool.ruff]
169
+ line-length = 100
170
+ extend-exclude = ["docs"]
171
+ target-version = "py313"
172
+
173
+ [tool.ruff.lint]
174
+ future-annotations = true
175
+ select = [
176
+ "A",
177
+ "B",
178
+ "BLE",
179
+ "C",
180
+ "C4",
181
+ "COM",
182
+ "D",
183
+ "E",
184
+ "EM",
185
+ "EXE",
186
+ "F",
187
+ "FA",
188
+ "FLY",
189
+ "G",
190
+ "I",
191
+ "ICN",
192
+ "INP",
193
+ "INT",
194
+ "ISC",
195
+ "N",
196
+ "PERF",
197
+ "PIE",
198
+ "PLE",
199
+ "PLC",
200
+ "PLR",
201
+ "PT",
202
+ "PTH",
203
+ "PYI",
204
+ "Q",
205
+ "RET",
206
+ "RSE",
207
+ "RUF",
208
+ "SIM",
209
+ "SLF",
210
+ "SLOT",
211
+ "T10",
212
+ "TC",
213
+ "TID",
214
+ "TRY",
215
+ "UP",
216
+ "W",
217
+ "YTT",
218
+ ]
219
+ ignore = [
220
+ "C408",
221
+ "B905",
222
+ "C901",
223
+ "COM812",
224
+ "D100",
225
+ "D101",
226
+ "D102",
227
+ "D103",
228
+ "D104",
229
+ "D105",
230
+ "D106",
231
+ "D107",
232
+ "D203",
233
+ "D204",
234
+ "D213",
235
+ "D215",
236
+ "D400",
237
+ "D401",
238
+ "D404",
239
+ "D406",
240
+ "D407",
241
+ "D408",
242
+ "D409",
243
+ "D413",
244
+ "EM101",
245
+ "EM102",
246
+ "ISC001",
247
+ "PLC0415",
248
+ "PLR0912",
249
+ "PLR0913",
250
+ "PLR0915",
251
+ "PLR0917",
252
+ "SLF001",
253
+ "TRY003",
254
+ "TC006",
255
+ ]
256
+
257
+ [tool.ruff.lint.flake8-quotes]
258
+ docstring-quotes = "double"
259
+
260
+ [tool.ruff.lint.flake8-type-checking]
261
+ runtime-evaluated-base-classes = [
262
+ "pydantic.BaseModel",
263
+ "sqlalchemy.orm.DeclarativeBase",
264
+ "schemez.Schema",
265
+ "sqlmodel.SQLModel",
266
+ ]
267
+ runtime-evaluated-decorators = [
268
+ "pydantic.validate_call",
269
+ "fastapi.FastAPI.get",
270
+ "fastapi.FastAPI.post",
271
+ "fastmcp.FastMCP.tool",
272
+ "typer.Typer.command",
273
+ ]
274
+
275
+ [tool.ruff.lint.isort]
276
+ lines-after-imports = 2
277
+ force-sort-within-sections = true
278
+ combine-as-imports = true
279
+
280
+ [tool.ruff.lint.mccabe]
281
+ max-complexity = 15
282
+
283
+ [tool.ruff.lint.per-file-ignores]
284
+ "__init__.py" = [
285
+ "E402",
286
+ "I001",
287
+ ]
288
+ "scripts/*" = ["INP001"]
289
+ "*tests/*" = ["D100"]
290
+
291
+ [tool.ruff.format]
292
+ preview = true
293
+
294
+ [tool.ty.environment]
295
+ python-version = "3.13"
296
+ python-platform = "all"
297
+
298
+ [tool.uv]
299
+ default-groups = [
300
+ "dev",
301
+ "lint",
302
+ "docs",
303
+ ]
304
+ conflicts = [
305
+ [
306
+ { extra = "marker" },
307
+ { extra = "all" },
308
+ ],
309
+ [
310
+ { extra = "marker" },
311
+ { extra = "light" },
312
+ ],
313
+ ]
314
+
315
+ [tool.uv.build-backend]
316
+ module-name = [
317
+ "docler",
318
+ "docler_config",
319
+ "docler_streamlit",
320
+ "docler_api",
321
+ "docler_mcp",
322
+ ]
323
+ wheel-exclude = [
324
+ ".mypy_cache/**",
325
+ "__pycache__/**",
326
+ "*.pyc",
327
+ "*.pyo",
328
+ ".pytest_cache/**",
329
+ ".ruff_cache/**",
330
+ ]
331
+ source-exclude = [
332
+ ".mypy_cache/**",
333
+ "__pycache__/**",
334
+ "*.pyc",
335
+ "*.pyo",
336
+ ".pytest_cache/**",
337
+ ".ruff_cache/**",
338
+ ]