HowdenParser 2.0.1__tar.gz → 2.0.2__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.
- {howdenparser-2.0.1 → howdenparser-2.0.2}/HowdenParser/parameter/llamaparser.py +0 -4
- {howdenparser-2.0.1 → howdenparser-2.0.2}/HowdenParser/parser.py +8 -4
- {howdenparser-2.0.1 → howdenparser-2.0.2}/PKG-INFO +1 -1
- {howdenparser-2.0.1 → howdenparser-2.0.2}/pyproject.toml +1 -1
- {howdenparser-2.0.1 → howdenparser-2.0.2}/HowdenParser/__init__.py +0 -0
- {howdenparser-2.0.1 → howdenparser-2.0.2}/HowdenParser/parameter/__init__.py +0 -0
- {howdenparser-2.0.1 → howdenparser-2.0.2}/HowdenParser/parameter/huggingface.py +0 -0
- {howdenparser-2.0.1 → howdenparser-2.0.2}/HowdenParser/parameter/mistralocr.py +0 -0
- {howdenparser-2.0.1 → howdenparser-2.0.2}/README.md +0 -0
|
@@ -2,12 +2,8 @@
|
|
|
2
2
|
from typing import Literal
|
|
3
3
|
from pydantic import BaseModel
|
|
4
4
|
|
|
5
|
-
class Hest(BaseModel):
|
|
6
|
-
provider_and_model: str = "huggingface:trocr-large-printed"
|
|
7
|
-
|
|
8
5
|
|
|
9
6
|
class Parameter(BaseModel):
|
|
10
|
-
model1: Hest = Hest()
|
|
11
7
|
provider_and_model: str = "llamaparser:"
|
|
12
8
|
result_type: Literal["md"] = "md"
|
|
13
9
|
mode: bool = False
|
|
@@ -144,22 +144,26 @@ class LangChainParser(BaseParser, name="langchain"):
|
|
|
144
144
|
|
|
145
145
|
|
|
146
146
|
class LlamaParser(BaseParser, name="llamaparser"):
|
|
147
|
-
def __init__(self, result_type: str, mode: bool) -> None:
|
|
147
|
+
def __init__(self, result_type: str, mode: bool, provider_and_model) -> None:
|
|
148
148
|
logging.info("Initializing LlamaParser...")
|
|
149
149
|
|
|
150
|
+
self.result_type = result_type
|
|
151
|
+
self.mode = mode
|
|
152
|
+
self.provider_and_model = provider_and_model
|
|
153
|
+
|
|
150
154
|
from llama_parse import LlamaParse, ResultType
|
|
151
155
|
|
|
152
156
|
if result_type.lower() in ("md", "markdown"):
|
|
153
|
-
|
|
157
|
+
result_type = ResultType.MD
|
|
154
158
|
|
|
155
159
|
api_key = os.getenv("LLAMA-PARSER-API-TOKEN")
|
|
156
160
|
if not api_key:
|
|
157
161
|
raise EnvironmentError("Missing LLAMA-PARSER-API-TOKEN in .env file.")
|
|
158
162
|
|
|
159
|
-
self.
|
|
163
|
+
self._parser = LlamaParse(api_key=api_key, result_type=result_type, premium_mode=self.mode)
|
|
160
164
|
|
|
161
165
|
def parse(self, file_path: Path) -> str:
|
|
162
|
-
documents = self.
|
|
166
|
+
documents = self._parser.load_data(str(file_path))
|
|
163
167
|
return "\n".join(doc.text for doc in documents)
|
|
164
168
|
|
|
165
169
|
|
|
@@ -14,7 +14,7 @@ build-backend = "poetry.core.masonry.api"
|
|
|
14
14
|
|
|
15
15
|
[tool.poetry]
|
|
16
16
|
name = "HowdenParser"
|
|
17
|
-
version = "2.0.
|
|
17
|
+
version = "2.0.2"
|
|
18
18
|
description = "A simple configuration manager with Pydantic and JSON export."
|
|
19
19
|
authors = [ "JesperThoftIllemannJ <jesper.jaeger@howdendanmark.dk>",]
|
|
20
20
|
readme = "README.md"
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|