pydatamax 0.1.14__tar.gz → 0.1.15.post2__tar.gz

{pydatamax-0.1.14 → pydatamax-0.1.15.post2}/LICENSE

@@ -1,21 +1,21 @@
-MIT License
-
-Copyright (c) 2024 Hi-Dolphin
-
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
+MIT License
+
+Copyright (c) 2024 Hi-Dolphin
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

{pydatamax-0.1.14 → pydatamax-0.1.15.post2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pydatamax
-Version: 0.1.14
+Version: 0.1.15.post2
 Summary: A library for parsing and converting various file formats.
 Home-page: https://github.com/Hi-Dolphin/datamax
 Author: ccy
@@ -105,10 +105,15 @@ cleaned_data = dm.clean_data(method_list=["abnormal", "private", "filter"])
 
 # AI annotation
 qa_data = dm.get_pre_label(
-    api_key="your-api-key",
-    base_url="https://api.openai.com/v1",
-    model_name="gpt-3.5-turbo"
+    api_key="sk-xxx",
+    base_url="https://api.provider.com/v1",
+    model_name="model-name",
+    chunk_size=500,        # 文本块大小
+    chunk_overlap=100,     # 重叠长度
+    question_number=5,     # 每块生成问题数
+    max_workers=5          # 并发数
 )
+dm.save_label_data(res)
 ```
 
 ## 📖 Detailed Documentation
@@ -138,8 +143,54 @@ dm = DataMax(file_path="document.docx", to_markdown=True)
 # Image OCR
 dm = DataMax(file_path="image.jpg", use_ocr=True)
 ```
+### Batch Processing
+```python
+# Parse multiple files in batch
+dm = DataMax(
+    file_path=["file1.pdf", "file2.docx"],
+    use_mineru=True
+)
+data = dm.get_data()
+```
+
+### Cache parsed results
+```python
+# Cache parsed results to avoid repeated parsing
+dm = DataMax(
+    file_path=["file1.pdf", "file2.docx"],
+    ttl=3600  # Cache duration in seconds, default 3600s, 0 means no caching
+)
+data = dm.get_data()
+```
 
 ### Data Cleaning
+## Exception Handling
+
+- remove_abnormal_chars Remove abnormal characters from text
+- remove_html_tags Remove HTML tags
+- convert_newlines Convert \r to \n and merge multiple \n into single \n
+- single_space Convert multiple spaces (more than 2) to single space
+- tabs_to_spaces Convert tabs to 4 spaces
+- remove_invisible_chars Remove invisible ASCII characters
+- simplify_chinese Convert traditional Chinese to simplified Chinese
+
+## Text Filtering
+
+- filter_by_word_repetition Filter by word repetition rate
+- filter_by_char_count Filter by character count
+- filter_by_numeric_content Filter by numeric content ratio
+
+## Privacy Desensitization
+
+- replace_ip
+- replace_email
+- replace_customer_number Clean hotline numbers like 4008-123-123
+- replace_bank_id
+- replace_phone_number
+- replace_qq
+- replace_id_card
+
+
 
 ```python
 # Three cleaning modes
@@ -148,6 +199,67 @@ dm.clean_data(method_list=[
     "private",   # Privacy information masking
     "filter"     # Text filtering and normalization
 ])
+
+# Custom cleaning mode
+from datamax.utils.data_cleaner import TextFilter, PrivacyDesensitization, AbnormalCleaner
+dm = DataMax(
+    file_path=r"C:\Users\cykro\Desktop\HongKongDevMachine.txt"
+)
+parsed_data = dm.get_data().get('content')
+# 1. Text filtering
+tf = TextFilter(parsed_data=parsed_data)
+    # Word repetition filtering - default threshold is 0.6 (max 60% of characters can be repeated)
+tf_bool = tf.filter_by_word_repetition(threshold=0.6)
+if tf_bool:
+    print("Text passed word repetition filtering")
+else:
+    print("Text failed word repetition filtering")
+    
+# Character count filtering - default min_chars=30 (minimum 30 chars), max_chars=500000 (maximum 500000 chars)
+tf_bool = tf.filter_by_char_count(min_chars=30, max_chars=500000)
+if tf_bool:
+    print("Text passed character count filtering")
+else:
+    print("Text failed character count filtering")
+
+# Numeric content filtering - default threshold=0.6 (max 60% of characters can be digits)
+tf_bool = tf.filter_by_numeric_content(threshold=0.6)
+if tf_bool:
+    print("Text passed numeric ratio filtering")
+else:
+    print("Text failed numeric ratio filtering")
+
+# 2. Privacy desensitization
+pd = PrivacyDesensitization(parsed_data=parsed_data)
+res = pd.replace_ip(
+    token="MyIP"
+)
+print(res)
+
+# 3. Abnormal character cleaning
+ac = AbnormalCleaner(parsed_data=parsed_data)
+res = ac.remove_abnormal_chars()
+res = ac.remove_html_tags()
+res = ac.convert_newlines()
+res = ac.single_space()
+res = ac.tabs_to_spaces()
+res = ac.remove_invisible_chars()
+res = ac.simplify_chinese()
+print(res)
+```
+# Text Segmentation
+```python
+dm.split_data(
+    chunk_size=500,      # Chunk size
+    chunk_overlap=100,    # Overlap length
+    use_langchain=True   # Use LangChain for text segmentation
+)
+
+# When use_langchain is False, use custom segmentation method
+# Using 。！？ as separators, consecutive separators will be merged
+# chunk_size strictly limits the string length
+for chunk in parser.split_data(chunk_size=500, chunk_overlap=100, use_langchain=False).get("content"):
+    print(chunk)
 ```
 
 ### AI Annotation
@@ -225,4 +337,4 @@ This project is licensed under the [MIT License](LICENSE).
 
 ---
 
-⭐ If this project helps you, please give us a star! 
+⭐ If this project helps you, please give us a star!

{pydatamax-0.1.14 → pydatamax-0.1.15.post2}/README.md

@@ -44,10 +44,15 @@ cleaned_data = dm.clean_data(method_list=["abnormal", "private", "filter"])
 
 # AI annotation
 qa_data = dm.get_pre_label(
-    api_key="your-api-key",
-    base_url="https://api.openai.com/v1",
-    model_name="gpt-3.5-turbo"
+    api_key="sk-xxx",
+    base_url="https://api.provider.com/v1",
+    model_name="model-name",
+    chunk_size=500,        # 文本块大小
+    chunk_overlap=100,     # 重叠长度
+    question_number=5,     # 每块生成问题数
+    max_workers=5          # 并发数
 )
+dm.save_label_data(res)
 ```
 
 ## 📖 Detailed Documentation
@@ -77,8 +82,54 @@ dm = DataMax(file_path="document.docx", to_markdown=True)
 # Image OCR
 dm = DataMax(file_path="image.jpg", use_ocr=True)
 ```
+### Batch Processing
+```python
+# Parse multiple files in batch
+dm = DataMax(
+    file_path=["file1.pdf", "file2.docx"],
+    use_mineru=True
+)
+data = dm.get_data()
+```
+
+### Cache parsed results
+```python
+# Cache parsed results to avoid repeated parsing
+dm = DataMax(
+    file_path=["file1.pdf", "file2.docx"],
+    ttl=3600  # Cache duration in seconds, default 3600s, 0 means no caching
+)
+data = dm.get_data()
+```
 
 ### Data Cleaning
+## Exception Handling
+
+- remove_abnormal_chars Remove abnormal characters from text
+- remove_html_tags Remove HTML tags
+- convert_newlines Convert \r to \n and merge multiple \n into single \n
+- single_space Convert multiple spaces (more than 2) to single space
+- tabs_to_spaces Convert tabs to 4 spaces
+- remove_invisible_chars Remove invisible ASCII characters
+- simplify_chinese Convert traditional Chinese to simplified Chinese
+
+## Text Filtering
+
+- filter_by_word_repetition Filter by word repetition rate
+- filter_by_char_count Filter by character count
+- filter_by_numeric_content Filter by numeric content ratio
+
+## Privacy Desensitization
+
+- replace_ip
+- replace_email
+- replace_customer_number Clean hotline numbers like 4008-123-123
+- replace_bank_id
+- replace_phone_number
+- replace_qq
+- replace_id_card
+
+
 
 ```python
 # Three cleaning modes
@@ -87,6 +138,67 @@ dm.clean_data(method_list=[
     "private",   # Privacy information masking
     "filter"     # Text filtering and normalization
 ])
+
+# Custom cleaning mode
+from datamax.utils.data_cleaner import TextFilter, PrivacyDesensitization, AbnormalCleaner
+dm = DataMax(
+    file_path=r"C:\Users\cykro\Desktop\HongKongDevMachine.txt"
+)
+parsed_data = dm.get_data().get('content')
+# 1. Text filtering
+tf = TextFilter(parsed_data=parsed_data)
+    # Word repetition filtering - default threshold is 0.6 (max 60% of characters can be repeated)
+tf_bool = tf.filter_by_word_repetition(threshold=0.6)
+if tf_bool:
+    print("Text passed word repetition filtering")
+else:
+    print("Text failed word repetition filtering")
+    
+# Character count filtering - default min_chars=30 (minimum 30 chars), max_chars=500000 (maximum 500000 chars)
+tf_bool = tf.filter_by_char_count(min_chars=30, max_chars=500000)
+if tf_bool:
+    print("Text passed character count filtering")
+else:
+    print("Text failed character count filtering")
+
+# Numeric content filtering - default threshold=0.6 (max 60% of characters can be digits)
+tf_bool = tf.filter_by_numeric_content(threshold=0.6)
+if tf_bool:
+    print("Text passed numeric ratio filtering")
+else:
+    print("Text failed numeric ratio filtering")
+
+# 2. Privacy desensitization
+pd = PrivacyDesensitization(parsed_data=parsed_data)
+res = pd.replace_ip(
+    token="MyIP"
+)
+print(res)
+
+# 3. Abnormal character cleaning
+ac = AbnormalCleaner(parsed_data=parsed_data)
+res = ac.remove_abnormal_chars()
+res = ac.remove_html_tags()
+res = ac.convert_newlines()
+res = ac.single_space()
+res = ac.tabs_to_spaces()
+res = ac.remove_invisible_chars()
+res = ac.simplify_chinese()
+print(res)
+```
+# Text Segmentation
+```python
+dm.split_data(
+    chunk_size=500,      # Chunk size
+    chunk_overlap=100,    # Overlap length
+    use_langchain=True   # Use LangChain for text segmentation
+)
+
+# When use_langchain is False, use custom segmentation method
+# Using 。！？ as separators, consecutive separators will be merged
+# chunk_size strictly limits the string length
+for chunk in parser.split_data(chunk_size=500, chunk_overlap=100, use_langchain=False).get("content"):
+    print(chunk)
 ```
 
 ### AI Annotation
@@ -164,4 +276,4 @@ This project is licensed under the [MIT License](LICENSE).
 
 ---
 
-⭐ If this project helps you, please give us a star! 
+⭐ If this project helps you, please give us a star!

{pydatamax-0.1.14 → pydatamax-0.1.15.post2}/datamax/__init__.py

@@ -1 +1 @@
-from .parser import DataMax
+from .parser import DataMax

{pydatamax-0.1.14 → pydatamax-0.1.15.post2}/datamax/loader/core.py

@@ -1,119 +1,119 @@
-import os
-from typing import List
-from datamax.loader.minio_handler import MinIOClient
-from datamax.loader.oss_handler import OssClient
-
-
-class DataLoader:
-    def __init__(self, endpoint: str = None, secret_key: str = None, access_key: str = None,
-                 bucket_name: str = None, source: str = None):
-        if source and source == 'Oss':
-            self.oss = OssClient(
-                oss_endpoint=endpoint,
-                oss_access_key_secret=secret_key,
-                oss_access_key_id=access_key,
-                oss_bucket_name=bucket_name
-            )
-        elif source and source == 'MinIO':
-            self.mi = MinIOClient(
-                endpoint=endpoint,
-                secret_key=secret_key,
-                access_key=access_key,
-                bucket_name=bucket_name
-            )
-        self.download_path = str('./download_file')
-        self.source = source
-        self.bucket_name = bucket_name
-
-    @staticmethod
-    def load_from_file(local_file_path) -> List[str]:
-        if os.path.isfile(local_file_path):
-            if os.path.exists(local_file_path):
-                if os.access(local_file_path, os.R_OK):
-                    return [local_file_path]
-                else:
-                    return []
-            else:
-                return []
-        elif os.path.isdir(local_file_path):
-            access_path = []
-            for root, dirs, files in os.walk(local_file_path):
-                for file in files:
-                    file_path = os.path.join(root, file)
-                    if os.path.exists(file_path):
-                        if os.access(file_path, os.R_OK):
-                            access_path.append(file_path)
-                        else:
-                            continue
-                    else:
-                        continue
-            return access_path
-        else:
-            return []
-
-    def load_from_oss_source(self, oss_path: str) -> List[str]:
-        if not os.path.exists(self.download_path):
-            os.makedirs(self.download_path)
-
-        self.download(oss_path=oss_path)
-
-        file_list = []
-        for root, dirs, files in os.walk(self.download_path):
-            for file in files:
-                file_path = os.path.join(self.download_path, file)
-                file_list.append(file_path)
-
-        success_file_list = []
-        for file_path in file_list:
-            if self.load_from_file(file_path):
-                success_file_list.append(file_path)
-
-        return success_file_list
-
-    def download(self, oss_path: str):
-        if self.source == 'MinIO':
-            file_list = self.mi.list_objects(bucket_name=self.bucket_name, prefix=oss_path)
-            for path in file_list:
-                self.mi.download_file(bucket_name=self.bucket_name, object_name=path,
-                                      file_path=f'{self.download_path}/{path.split("/")[-1]}')
-        elif self.source == "Oss":
-            keys = self.oss.get_objects_in_folders(prefix=oss_path)
-            for path in keys:
-                self.oss.get_object_to_file(object_name=path,
-                                            file_path=f'{self.download_path}/{path.split("/")[-1]}')
-
-    def upload(self, local_file_path: str, save_prefix: str):
-        if self.source == 'MinIO':
-            if os.path.isdir(local_file_path):
-                for root, dirs, files in os.walk(local_file_path):
-                    for file in files:
-                        file_path = os.path.join(root, file)
-                        self.mi.upload_file(bucket_name=self.bucket_name, object_name=save_prefix + f'{file}',
-                                            file_path=file_path)
-            elif os.path.isfile(local_file_path):
-                self.mi.upload_file(bucket_name=self.bucket_name,
-                                    object_name=save_prefix + os.path.basename(local_file_path),
-                                    file_path=local_file_path)
-            else:
-                pass
-
-        elif self.source == "Oss":
-            if os.path.isdir(local_file_path):
-                self.oss.put_object_from_folder(object_folder_name=save_prefix, local_folder_path=local_file_path)
-            elif os.path.isfile(local_file_path):
-                self.oss.put_object_from_file(object_name=save_prefix + os.path.basename(local_file_path),
-                                              file_path=local_file_path)
-        else:
-            pass
-
-    def share(self, oss_path: str,
-              expires: int = None,
-              aliyun_oss_url_prefix: str = None,
-              csnt_url_prefix: str = None):
-        if self.source == 'MinIO':
-            return self.mi.get_object_tmp_link(bucket_name=self.bucket_name, object_name=oss_path, expires=expires)
-        elif self.source == "Oss":
-            return self.oss.get_oss_url(object_name=oss_path,
-                                        url_expires_time=expires,
-                                        aliyun_oss_url_prefix=aliyun_oss_url_prefix,
+import os
+from typing import List
+from datamax.loader.minio_handler import MinIOClient
+from datamax.loader.oss_handler import OssClient
+
+
+class DataLoader:
+    def __init__(self, endpoint: str = None, secret_key: str = None, access_key: str = None,
+                 bucket_name: str = None, source: str = None):
+        if source and source == 'Oss':
+            self.oss = OssClient(
+                oss_endpoint=endpoint,
+                oss_access_key_secret=secret_key,
+                oss_access_key_id=access_key,
+                oss_bucket_name=bucket_name
+            )
+        elif source and source == 'MinIO':
+            self.mi = MinIOClient(
+                endpoint=endpoint,
+                secret_key=secret_key,
+                access_key=access_key,
+                bucket_name=bucket_name
+            )
+        self.download_path = str('./download_file')
+        self.source = source
+        self.bucket_name = bucket_name
+
+    @staticmethod
+    def load_from_file(local_file_path) -> List[str]:
+        if os.path.isfile(local_file_path):
+            if os.path.exists(local_file_path):
+                if os.access(local_file_path, os.R_OK):
+                    return [local_file_path]
+                else:
+                    return []
+            else:
+                return []
+        elif os.path.isdir(local_file_path):
+            access_path = []
+            for root, dirs, files in os.walk(local_file_path):
+                for file in files:
+                    file_path = os.path.join(root, file)
+                    if os.path.exists(file_path):
+                        if os.access(file_path, os.R_OK):
+                            access_path.append(file_path)
+                        else:
+                            continue
+                    else:
+                        continue
+            return access_path
+        else:
+            return []
+
+    def load_from_oss_source(self, oss_path: str) -> List[str]:
+        if not os.path.exists(self.download_path):
+            os.makedirs(self.download_path)
+
+        self.download(oss_path=oss_path)
+
+        file_list = []
+        for root, dirs, files in os.walk(self.download_path):
+            for file in files:
+                file_path = os.path.join(self.download_path, file)
+                file_list.append(file_path)
+
+        success_file_list = []
+        for file_path in file_list:
+            if self.load_from_file(file_path):
+                success_file_list.append(file_path)
+
+        return success_file_list
+
+    def download(self, oss_path: str):
+        if self.source == 'MinIO':
+            file_list = self.mi.list_objects(bucket_name=self.bucket_name, prefix=oss_path)
+            for path in file_list:
+                self.mi.download_file(bucket_name=self.bucket_name, object_name=path,
+                                      file_path=f'{self.download_path}/{path.split("/")[-1]}')
+        elif self.source == "Oss":
+            keys = self.oss.get_objects_in_folders(prefix=oss_path)
+            for path in keys:
+                self.oss.get_object_to_file(object_name=path,
+                                            file_path=f'{self.download_path}/{path.split("/")[-1]}')
+
+    def upload(self, local_file_path: str, save_prefix: str):
+        if self.source == 'MinIO':
+            if os.path.isdir(local_file_path):
+                for root, dirs, files in os.walk(local_file_path):
+                    for file in files:
+                        file_path = os.path.join(root, file)
+                        self.mi.upload_file(bucket_name=self.bucket_name, object_name=save_prefix + f'{file}',
+                                            file_path=file_path)
+            elif os.path.isfile(local_file_path):
+                self.mi.upload_file(bucket_name=self.bucket_name,
+                                    object_name=save_prefix + os.path.basename(local_file_path),
+                                    file_path=local_file_path)
+            else:
+                pass
+
+        elif self.source == "Oss":
+            if os.path.isdir(local_file_path):
+                self.oss.put_object_from_folder(object_folder_name=save_prefix, local_folder_path=local_file_path)
+            elif os.path.isfile(local_file_path):
+                self.oss.put_object_from_file(object_name=save_prefix + os.path.basename(local_file_path),
+                                              file_path=local_file_path)
+        else:
+            pass
+
+    def share(self, oss_path: str,
+              expires: int = None,
+              aliyun_oss_url_prefix: str = None,
+              csnt_url_prefix: str = None):
+        if self.source == 'MinIO':
+            return self.mi.get_object_tmp_link(bucket_name=self.bucket_name, object_name=oss_path, expires=expires)
+        elif self.source == "Oss":
+            return self.oss.get_oss_url(object_name=oss_path,
+                                        url_expires_time=expires,
+                                        aliyun_oss_url_prefix=aliyun_oss_url_prefix,
                                         csnt_url_prefix=csnt_url_prefix)