PgsFile 0.2.2__py3-none-any.whl → 0.2.3__py3-none-any.whl

PgsFile/PgsFile.py

@@ -700,6 +700,21 @@ def find_txt_files_with_keyword(root_folder, keyword, case_sensitive=None):
                     matches.append(os.path.join(root, filename))
     return matches    
 
+import fnmatch
+def find_user_files_in_upper_folder(directory, user_file_name):
+    # Get the direct upper folder path
+    upper_folder = os.path.dirname(os.path.abspath(directory))
+
+    # List to store matching file paths
+    matching_files = []
+
+    # Walk through the upper folder
+    for root, dirs, files in os.walk(upper_folder):
+        for filename in fnmatch.filter(files, f'{user_file_name}.user'):
+            matching_files.append(os.path.join(root, filename))
+
+    return matching_files
+
 # Standard sentence tokenizer.
 def sent_tokenize(text, lang=None):
     import pysbd
@@ -782,6 +797,7 @@ def cs1(text):
         sentences=sentences
     return sentences
 
+                
 def word_tokenize(text, pos_tagged=False):
     '''
     Parameters
@@ -816,10 +832,22 @@ def word_tokenize(text, pos_tagged=False):
             try:
                 from nlpir import tools
                 tools.update_license()
+                print("\n\nThe user file is ready. Please restart your kernel and run the Python script!")
             except Exception as err2:
-                print("You need a VPN to try this service!", err2)
+                print("\n*****SOLUTION WARNING! \nYOU MAY NEED A VPN TO TRY THIS SERVICE!*****\n\n", err2)
         else:
-            print(err)
+            try:
+                if "Can not open" in str(err):
+                    user_folder=get_library_location("PgsFile")+"/PgsFile/Corpora/Stopwords"
+                    destination_folder=get_library_location("nlpir-python")+"/nlpir/Data"
+                    source_file=find_user_files_in_upper_folder(user_folder, "NLPIR")[0]
+                    copy_file(source_file, destination_folder)  
+                    print("The user file is ready. Please restart your kernel and run the Python script!")
+                else:
+                    print(err)
+            except Exception as rer:
+                print(rer)
+
     return words  
 
 def pad_sequence(
@@ -1547,6 +1575,29 @@ def move_file(source_file, destination_folder, new_file_name=None):
     shutil.move(source_file, destination_file)
 
     print(f"File moved from {source_file} to {destination_file}")
+
+def copy_file(source_file, destination_folder, new_file_name=None):
+    """
+    Copy a file to another folder.
+
+    Parameters:
+    source_file (str): The path to the source file.
+    destination_folder (str): The path to the destination folder.
+    new_file_name (str, optional): The new name for the file in the destination folder. Defaults to None.
+    """
+    # Ensure the destination folder exists
+    if not os.path.exists(destination_folder):
+        os.makedirs(destination_folder)
+
+    # Construct the destination file path
+    if new_file_name:
+        destination_file = os.path.join(destination_folder, new_file_name)
+    else:
+        destination_file = os.path.join(destination_folder, os.path.basename(source_file))
+
+    # Copy the file to the destination folder
+    shutil.copy2(source_file, destination_file)
+
     
 def check_empty_cells(file_path):
     """

PgsFile/__init__.py

@@ -22,7 +22,7 @@ from .PgsFile import FilePath, FileName, DirList
 from .PgsFile import get_subfolder_path
 from .PgsFile import makedirec, makefile
 from .PgsFile import source_path, next_folder_names, get_directory_tree_with_meta, find_txt_files_with_keyword
-from .PgsFile import remove_empty_folders, remove_empty_txts, remove_empty_lines, remove_empty_last_line, move_file
+from .PgsFile import remove_empty_folders, remove_empty_txts, remove_empty_lines, remove_empty_last_line, move_file, copy_file
 from .PgsFile import concatenate_excel_files
 
 # 6. Data cleaning

{PgsFile-0.2.2.dist-info → PgsFile-0.2.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: PgsFile
-Version: 0.2.2
+Version: 0.2.3
 Summary: This module aims to simplify Python package management, script execution, file handling, web scraping, multimedia download, data cleaning, NLP tasks like Chinese word tokenization and POS tagging, and word list generation for literary students, making it more accessible and convenient to use.
 Home-page: https://mp.weixin.qq.com/s/12-KVLfaPszoZkCxuRd-nQ?token=1589547443&lang=zh_CN
 Author: Pan Guisheng

{PgsFile-0.2.2.dist-info → PgsFile-0.2.3.dist-info}/RECORD

@@ -1,5 +1,5 @@
-PgsFile/PgsFile.py,sha256=BQNZBrBYgyB_4TVxD0CJ6cMpiaaDzL9b7c7kcYtxmwQ,83682
-PgsFile/__init__.py,sha256=-Vy1SIh-BYopiEan-EjBtwqZsNteNrOqkws7hUj1d2w,2378
+PgsFile/PgsFile.py,sha256=MpXQK6MLMBh1JMAcBw5sRiRof--x4OyARcCsWwn7Z4A,85828
+PgsFile/__init__.py,sha256=E4VfPu1BxCBcZ5WXi5E6faPaNt_Shpvgh9LvBlg7eA0,2389
 PgsFile/Corpora/Idioms/English_Idioms_8774.txt,sha256=qlsP0yI_XGECBRiPZuLkGZpdasc77sWSKexANu7v8_M,175905
 PgsFile/Corpora/Monolingual/Chinese/People's Daily 20130605/Raw/00000000.txt,sha256=SLGGSMSb7Ff1RoBstsTW3yX2wNZpqEUchFNpcI-mrR4,1513
 PgsFile/Corpora/Monolingual/Chinese/People's Daily 20130605/Raw/00000001.txt,sha256=imOa6UoCOIZoPXT4_HNHgCUJtd4FTIdk2FZNHNBgJyg,3372
@@ -2569,6 +2569,7 @@ PgsFile/Corpora/Parallel/TED_EC_2017-2020/YvetteAlberdingkThijm_2017X公民影
 PgsFile/Corpora/Parallel/TED_EC_2017-2020/ZacharyRWood_2018为什么持有不同意见的人值得被聆听..txt,sha256=4SFYMhlFSHP2aEVvNS1CBeogq0D2lPTE5VhFsZjlZnM,9546
 PgsFile/Corpora/Parallel/TED_EC_2017-2020/ZeynepTufekci_2017G为了让人们点击广告_我们正在建造一个反乌托邦..txt,sha256=S3BSXKsNAX0ugVqBPhmJyaRF8MYAHapDMR12DoBYZgc,32353
 PgsFile/Corpora/Parallel/Xi's Speech_CE_2021/Speech at a Ceremony Marking the Centenary of the CPC.txt,sha256=3suCjs2LF2_Endg2i_hc3GX1N8lTBORlqpMWEKsXFeM,54282
+PgsFile/Corpora/Stopwords/NLPIR.user,sha256=DykLJdr8_cVHrdCnDJES1O5dgmnYqfaSO1_dtAVKYJk,3356
 PgsFile/Corpora/Stopwords/arabic.txt,sha256=yL9id0vdNF20WEvM0buRnRt1ByEeRGJuGDiY3jE7tlQ,1287
 PgsFile/Corpora/Stopwords/bulgarian.txt,sha256=eiIwYk1TU8YcYYPbMPjUzZSZlgd7gl5o7d0LIthzqHQ,2409
 PgsFile/Corpora/Stopwords/catalan.txt,sha256=8OyAOBHfWsEvKuLEphCfdiWhuxyVg1sOWV5gi2DJLwY,699
@@ -2618,8 +2619,8 @@ PgsFile/models/slovene.pickle,sha256=faxlAhKzeHs5mWwBvSCEEVST5vbsOQurYfdnUlsIuOo
 PgsFile/models/spanish.pickle,sha256=Jx3GAnxKrgVvcqm_q1ZFz2fhmL9PlyiVhE5A9ZiczcM,597831
 PgsFile/models/swedish.pickle,sha256=QNUOva1sqodxXy4wCxIX7JLELeIFpUPMSlaQO9LJrPo,1034496
 PgsFile/models/turkish.pickle,sha256=065H12UB0CdpiAnRLnUpLJw5KRBIhUM0KAL5Xbl2XMw,1225013
-PgsFile-0.2.2.dist-info/LICENSE,sha256=cE5c-QToSkG1KTUsU8drQXz1vG0EbJWuU4ybHTRb5SE,1138
-PgsFile-0.2.2.dist-info/METADATA,sha256=1fm2uh-uYKgDe26DvUGCmj2LbMcjwDum113nbmW-MIA,5070
-PgsFile-0.2.2.dist-info/WHEEL,sha256=eOLhNAGa2EW3wWl_TU484h7q1UNgy0JXjjoqKoxAAQc,92
-PgsFile-0.2.2.dist-info/top_level.txt,sha256=028hCfwhF3UpfD6X0rwtWpXI1RKSTeZ1ALwagWaSmX8,8
-PgsFile-0.2.2.dist-info/RECORD,,
+PgsFile-0.2.3.dist-info/LICENSE,sha256=cE5c-QToSkG1KTUsU8drQXz1vG0EbJWuU4ybHTRb5SE,1138
+PgsFile-0.2.3.dist-info/METADATA,sha256=a9KMN6LpC2raZYhWwrFhWCXKl7nWneiXT7KtvA74ruY,5070
+PgsFile-0.2.3.dist-info/WHEEL,sha256=eOLhNAGa2EW3wWl_TU484h7q1UNgy0JXjjoqKoxAAQc,92
+PgsFile-0.2.3.dist-info/top_level.txt,sha256=028hCfwhF3UpfD6X0rwtWpXI1RKSTeZ1ALwagWaSmX8,8
+PgsFile-0.2.3.dist-info/RECORD,,