7 Commits

18 changed files with 407 additions and 15358 deletions

1
.gitignore vendored
View File

@@ -1,3 +1,4 @@
/.DS_Store
/.project
/.pydevproject
bananaSPLIT/build

Binary file not shown.

File diff suppressed because it is too large Load Diff

View File

@@ -1,210 +0,0 @@
('D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\dist\\bananaSPLIT.exe',
True,
False,
False,
'banana.ico',
None,
False,
False,
'<?xml version="1.0" encoding="UTF-8" standalone="yes"?><assembly manifestVersion="1.0" xmlns="urn:schemas-microsoft-com:asm.v1"><assemblyIdentity name="bananaSPLIT.exe" processorArchitecture="amd64" type="win32" version="1.0.0.0"/><dependency><dependentAssembly><assemblyIdentity language="*" name="Microsoft.Windows.Common-Controls" processorArchitecture="*" publicKeyToken="6595b64144ccf1df" type="win32" version="6.0.0.0"/><compatibility xmlns="urn:schemas-microsoft-com:compatibility.v1"/></dependentAssembly></dependency><compatibility xmlns="urn:schemas-microsoft-com:compatibility.v1"><application><supportedOS Id="{e2011457-1546-43c5-a5fe-008deee3d3f0}"/><supportedOS Id="{35138b9a-5d96-4fbd-8e2d-a2440225f93a}"/><supportedOS Id="{4a2f28e3-53b9-4441-ba9c-d69d4a4a6e38}"/><supportedOS Id="{1f676c76-80e1-4239-95bb-83d0f6d0da78}"/><supportedOS Id="{8e0f7a12-bfb3-4fe8-b9a5-48fd50a15a9a}"/></application></compatibility></assembly>',
True,
'bananaSPLIT.pkg',
[('PYZ-00.pyz',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\PYZ-00.pyz',
'PYZ'),
('struct',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\localpycos\\struct.pyo',
'PYMODULE'),
('pyimod01_os_path',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyimod01_os_path.pyc',
'PYMODULE'),
('pyimod02_archive',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyimod02_archive.pyc',
'PYMODULE'),
('pyimod03_importers',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyimod03_importers.pyc',
'PYMODULE'),
('pyiboot01_bootstrap',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyiboot01_bootstrap.py',
'PYSOURCE'),
('main',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\main.py',
'PYSOURCE'),
('api-ms-win-crt-math-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-math-l1-1-0.dll',
'BINARY'),
('VCRUNTIME140.dll',
'c:\\program files\\python37\\VCRUNTIME140.dll',
'BINARY'),
('api-ms-win-crt-locale-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-locale-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-heap-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-heap-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-stdio-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-stdio-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-runtime-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-runtime-l1-1-0.dll',
'BINARY'),
('python37.dll', 'c:\\program files\\python37\\python37.dll', 'BINARY'),
('ucrtbase.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\ucrtbase.dll',
'BINARY'),
('api-ms-win-crt-convert-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-convert-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-string-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-string-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-environment-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-environment-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-conio-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-conio-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-filesystem-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-filesystem-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-time-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-time-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-process-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-process-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-file-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-file-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-sysinfo-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-sysinfo-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-file-l2-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-file-l2-1-0.dll',
'BINARY'),
('api-ms-win-core-namedpipe-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-namedpipe-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-interlocked-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-interlocked-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-processthreads-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-processthreads-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-console-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-console-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-heap-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-heap-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-file-l1-2-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-file-l1-2-0.dll',
'BINARY'),
('api-ms-win-core-memory-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-memory-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-synch-l1-2-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-synch-l1-2-0.dll',
'BINARY'),
('api-ms-win-core-errorhandling-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-errorhandling-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-libraryloader-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-libraryloader-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-processthreads-l1-1-1.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-processthreads-l1-1-1.dll',
'BINARY'),
('api-ms-win-core-util-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-util-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-localization-l1-2-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-localization-l1-2-0.dll',
'BINARY'),
('api-ms-win-core-handle-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-handle-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-string-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-string-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-processenvironment-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-processenvironment-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-synch-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-synch-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-rtlsupport-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-rtlsupport-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-profile-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-profile-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-debug-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-debug-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-timezone-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-timezone-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-datetime-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-datetime-l1-1-0.dll',
'BINARY'),
('_ssl', 'c:\\program files\\python37\\DLLs\\_ssl.pyd', 'EXTENSION'),
('unicodedata',
'c:\\program files\\python37\\DLLs\\unicodedata.pyd',
'EXTENSION'),
('pyexpat', 'c:\\program files\\python37\\DLLs\\pyexpat.pyd', 'EXTENSION'),
('_hashlib', 'c:\\program files\\python37\\DLLs\\_hashlib.pyd', 'EXTENSION'),
('select', 'c:\\program files\\python37\\DLLs\\select.pyd', 'EXTENSION'),
('_bz2', 'c:\\program files\\python37\\DLLs\\_bz2.pyd', 'EXTENSION'),
('_lzma', 'c:\\program files\\python37\\DLLs\\_lzma.pyd', 'EXTENSION'),
('_socket', 'c:\\program files\\python37\\DLLs\\_socket.pyd', 'EXTENSION'),
('_ctypes', 'c:\\program files\\python37\\DLLs\\_ctypes.pyd', 'EXTENSION'),
('_contextvars',
'c:\\program files\\python37\\DLLs\\_contextvars.pyd',
'EXTENSION'),
('_decimal', 'c:\\program files\\python37\\DLLs\\_decimal.pyd', 'EXTENSION'),
('libssl-1_1-x64.dll',
'c:\\program files\\python37\\DLLs\\libssl-1_1-x64.dll',
'BINARY'),
('libcrypto-1_1-x64.dll',
'c:\\program files\\python37\\DLLs\\libcrypto-1_1-x64.dll',
'BINARY'),
('api-ms-win-crt-utility-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-utility-l1-1-0.dll',
'BINARY'),
('text_unidecode\\data.bin',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\text_unidecode\\data.bin',
'DATA'),
('base_library.zip',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\base_library.zip',
'DATA'),
('bananaSPLIT.exe.manifest',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\bananaSPLIT.exe.exe.manifest',
'BINARY'),
('pyi-windows-manifest-filename bananaSPLIT.exe.manifest', '', 'OPTION')],
[],
False,
False,
1572446705,
[('run.exe',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\bootloader\\Windows-64bit\\run.exe',
'EXECUTABLE')])

View File

@@ -1,202 +0,0 @@
('D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\PKG-00.pkg',
{'BINARY': 1,
'DATA': 1,
'EXECUTABLE': 1,
'EXTENSION': 1,
'PYMODULE': 1,
'PYSOURCE': 1,
'PYZ': 0},
[('PYZ-00.pyz',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\PYZ-00.pyz',
'PYZ'),
('struct',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\localpycos\\struct.pyo',
'PYMODULE'),
('pyimod01_os_path',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyimod01_os_path.pyc',
'PYMODULE'),
('pyimod02_archive',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyimod02_archive.pyc',
'PYMODULE'),
('pyimod03_importers',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyimod03_importers.pyc',
'PYMODULE'),
('pyiboot01_bootstrap',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\PyInstaller\\loader\\pyiboot01_bootstrap.py',
'PYSOURCE'),
('main',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\main.py',
'PYSOURCE'),
('api-ms-win-crt-math-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-math-l1-1-0.dll',
'BINARY'),
('VCRUNTIME140.dll',
'c:\\program files\\python37\\VCRUNTIME140.dll',
'BINARY'),
('api-ms-win-crt-locale-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-locale-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-heap-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-heap-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-stdio-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-stdio-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-runtime-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-runtime-l1-1-0.dll',
'BINARY'),
('python37.dll', 'c:\\program files\\python37\\python37.dll', 'BINARY'),
('ucrtbase.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\ucrtbase.dll',
'BINARY'),
('api-ms-win-crt-convert-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-convert-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-string-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-string-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-environment-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-environment-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-conio-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-conio-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-filesystem-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-filesystem-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-time-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-time-l1-1-0.dll',
'BINARY'),
('api-ms-win-crt-process-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-process-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-file-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-file-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-sysinfo-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-sysinfo-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-file-l2-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-file-l2-1-0.dll',
'BINARY'),
('api-ms-win-core-namedpipe-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-namedpipe-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-interlocked-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-interlocked-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-processthreads-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-processthreads-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-console-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-console-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-heap-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-heap-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-file-l1-2-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-file-l1-2-0.dll',
'BINARY'),
('api-ms-win-core-memory-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-memory-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-synch-l1-2-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-synch-l1-2-0.dll',
'BINARY'),
('api-ms-win-core-errorhandling-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-errorhandling-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-libraryloader-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-libraryloader-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-processthreads-l1-1-1.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-processthreads-l1-1-1.dll',
'BINARY'),
('api-ms-win-core-util-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-util-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-localization-l1-2-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-localization-l1-2-0.dll',
'BINARY'),
('api-ms-win-core-handle-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-handle-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-string-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-string-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-processenvironment-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-processenvironment-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-synch-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-synch-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-rtlsupport-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-rtlsupport-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-profile-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-profile-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-debug-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-core-debug-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-timezone-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-timezone-l1-1-0.dll',
'BINARY'),
('api-ms-win-core-datetime-l1-1-0.dll',
'C:/Program '
'Files/Java/jre1.8.0_231/bin\\api-ms-win-core-datetime-l1-1-0.dll',
'BINARY'),
('_ssl', 'c:\\program files\\python37\\DLLs\\_ssl.pyd', 'EXTENSION'),
('unicodedata',
'c:\\program files\\python37\\DLLs\\unicodedata.pyd',
'EXTENSION'),
('pyexpat', 'c:\\program files\\python37\\DLLs\\pyexpat.pyd', 'EXTENSION'),
('_hashlib', 'c:\\program files\\python37\\DLLs\\_hashlib.pyd', 'EXTENSION'),
('select', 'c:\\program files\\python37\\DLLs\\select.pyd', 'EXTENSION'),
('_bz2', 'c:\\program files\\python37\\DLLs\\_bz2.pyd', 'EXTENSION'),
('_lzma', 'c:\\program files\\python37\\DLLs\\_lzma.pyd', 'EXTENSION'),
('_socket', 'c:\\program files\\python37\\DLLs\\_socket.pyd', 'EXTENSION'),
('_ctypes', 'c:\\program files\\python37\\DLLs\\_ctypes.pyd', 'EXTENSION'),
('_contextvars',
'c:\\program files\\python37\\DLLs\\_contextvars.pyd',
'EXTENSION'),
('_decimal', 'c:\\program files\\python37\\DLLs\\_decimal.pyd', 'EXTENSION'),
('libssl-1_1-x64.dll',
'c:\\program files\\python37\\DLLs\\libssl-1_1-x64.dll',
'BINARY'),
('libcrypto-1_1-x64.dll',
'c:\\program files\\python37\\DLLs\\libcrypto-1_1-x64.dll',
'BINARY'),
('api-ms-win-crt-utility-l1-1-0.dll',
'C:/Program Files/Java/jre1.8.0_231/bin\\api-ms-win-crt-utility-l1-1-0.dll',
'BINARY'),
('text_unidecode\\data.bin',
'C:\\Users\\Emanuele '
'Trabattoni\\AppData\\Roaming\\Python\\Python37\\site-packages\\text_unidecode\\data.bin',
'DATA'),
('base_library.zip',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\base_library.zip',
'DATA'),
('bananaSPLIT.exe.manifest',
'D:\\Emanuele\\Documenti\\workspace\\bananaSPLIT\\bananaSPLIT\\build\\bananaSPLIT.exe\\bananaSPLIT.exe.exe.manifest',
'BINARY'),
('pyi-windows-manifest-filename bananaSPLIT.exe.manifest', '', 'OPTION')],
False,
False,
False)

File diff suppressed because it is too large Load Diff

View File

@@ -1,19 +0,0 @@
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<assembly manifestVersion="1.0" xmlns="urn:schemas-microsoft-com:asm.v1">
<assemblyIdentity name="bananaSPLIT.exe" processorArchitecture="amd64" type="win32" version="1.0.0.0"/>
<dependency>
<dependentAssembly>
<assemblyIdentity language="*" name="Microsoft.Windows.Common-Controls" processorArchitecture="*" publicKeyToken="6595b64144ccf1df" type="win32" version="6.0.0.0"/>
<compatibility xmlns="urn:schemas-microsoft-com:compatibility.v1"/>
</dependentAssembly>
</dependency>
<compatibility xmlns="urn:schemas-microsoft-com:compatibility.v1">
<application>
<supportedOS Id="{e2011457-1546-43c5-a5fe-008deee3d3f0}"/>
<supportedOS Id="{35138b9a-5d96-4fbd-8e2d-a2440225f93a}"/>
<supportedOS Id="{4a2f28e3-53b9-4441-ba9c-d69d4a4a6e38}"/>
<supportedOS Id="{1f676c76-80e1-4239-95bb-83d0f6d0da78}"/>
<supportedOS Id="{8e0f7a12-bfb3-4fe8-b9a5-48fd50a15a9a}"/>
</application>
</compatibility>
</assembly>

View File

@@ -1,33 +0,0 @@
This file lists modules PyInstaller was not able to find. This does not
necessarily mean this module is required for running you program. Python and
Python 3rd-party packages include a lot of conditional or optional module. For
example the module 'ntpath' only exists on Windows, whereas the module
'posixpath' only exists on Posix systems.
Types if import:
* top-level: imported at the top-level - look at these first
* conditional: imported within an if-statement
* delayed: imported from within a function
* optional: imported within a try-except-statement
IMPORTANT: Do NOT post this list to the issue-tracker. Use it as a basis for
yourself tracking down the missing module. Thanks!
missing module named resource - imported by posix (top-level), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named posix - imported by os (conditional, optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named _posixsubprocess - imported by subprocess (conditional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named readline - imported by cmd (delayed, conditional, optional), code (delayed, conditional, optional), pdb (delayed, optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
excluded module named _frozen_importlib - imported by importlib (optional), importlib.abc (optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named _frozen_importlib_external - imported by importlib._bootstrap (delayed), importlib (optional), importlib.abc (optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named _winreg - imported by platform (delayed, optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named _scproxy - imported by urllib.request (conditional)
missing module named java - imported by platform (delayed), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named 'java.lang' - imported by platform (delayed, optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level), xml.sax._exceptions (conditional)
missing module named vms_lib - imported by platform (delayed, conditional, optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named termios - imported by tty (top-level), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level), getpass (optional)
missing module named grp - imported by shutil (optional), tarfile (optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named org - imported by pickle (optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level)
missing module named pwd - imported by posixpath (delayed, conditional), shutil (optional), tarfile (optional), http.server (delayed, optional), webbrowser (delayed), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level), netrc (delayed, conditional), getpass (delayed)
missing module named 'org.python' - imported by copy (optional), D:\Emanuele\Documenti\workspace\bananaSPLIT\bananaSPLIT\main.py (top-level), xml.sax (delayed, conditional)
missing module named htmlentitydefs - imported by slugify.slugify (optional)

File diff suppressed because it is too large Load Diff

Binary file not shown.

View File

@@ -37,8 +37,7 @@
"DOCUMENT-TYPE:",
"CHARTS:"
]
},
"settings":
},
"settings": {
"encoding": "utf-8",
"monthPosition": 0,

View File

@@ -15,26 +15,24 @@ from pprint import pprint
from glob import glob
from copy import deepcopy
from slugify import slugify
####### VAR GLOBALI #######
####### FUNZIONI GLOBALI #######
####### FUNZIONI GLOBALI #######
def printTitle():
print(". . . , \n| | _ | _. _ ._ _ _ -+- _ * \n|/\|(/,|(_.(_)[ | )(/, | (_) *")
print(". . . , \n| | _ | _. _ ._ _ _ -+- _ * \n|/\|(/,|(_.(_)[ | )(/, | (_) *")
print(Fore.LIGHTYELLOW_EX,\
' _ ___________ _ _____ _____ \n\
| | / ___| ___ \ | |_ _|_ _| \n\
| |__ __ _ _ __ __ _ _ __ __ _\ `--.| |_/ / | | | | | \n\
| \'_ \ / _` | \'_ \ / _` | \'_ \ / _` |`--. \ __/| | | | | | \n\
| |_) | (_| | | | | (_| | | | | (_| /\__/ / | | |_____| |_ | | \n\
|_.__/ \__,_|_| |_|\__,_|_| |_|\__,_\____/\_| \_____/\___/ \_/')
print ('\t\t\t\t\t\t\tVersione 0.4a\n\n', Style.RESET_ALL)
print('Iniziamo!!')
pass
print(Fore.LIGHTYELLOW_EX,' _ ___________ _ _____ _____ \n\
| | / ___| ___ \ | |_ _|_ _| \n\
| |__ __ _ _ __ __ _ _ __ __ _\ `--.| |_/ / | | | | | \n\
| \'_ \ / _` | \'_ \ / _` | \'_ \ / _` |`--. \ __/| | | | | | \n\
| |_) | (_| | | | | (_| | | | | (_| /\__/ / | | |_____| |_ | | \n\
|_.__/ \__,_|_| |_|\__,_|_| |_|\__,_\____/\_| \_____/\___/ \_/')
print ('\t\t\t\t\t\t\tVersione 0.4b\n\n', Style.RESET_ALL)
print('Iniziamo!!')
pass
#############################################################
####################### MAIN ################################
#############################################################
@@ -44,82 +42,86 @@ status={'first':0, 'head':1, 'body':2, 'tail':3}
s=status['first']
try:
conffiles = glob('*.json')
if len(conffiles) < 1:
raise
fileValid = False
while not fileValid:
print ('Seleziona un file di configurazione per la lingua:')
for cf in enumerate(conffiles):
print('[{0}] - {1}'.format(cf[0],cf[1]))
try:
fn = int(input('Scrivi il numero del file e premi Invio: '))
if (fn > len(conffiles)-1 or fn < 0):
raise
fileValid = True
except:
print('Scusa, non ho capito bene, ricominciamo..\n')
fileValid = False
conffiles = glob('*.json')
if len(conffiles) < 1:
raise
fileValid = False
while not fileValid:
print ('Seleziona un file di configurazione per la lingua:')
for cf in enumerate(conffiles):
print('[{0}] - {1}'.format(cf[0],cf[1]))
try:
fn = int(input('Scrivi il numero del file e premi Invio: '))
if (fn > len(conffiles)-1 or fn < 0):
raise
fileValid = True
except:
print('Scusa, non ho capito bene, ricominciamo..\n')
fileValid = False
except:
print(Fore.LIGHTRED_EX, 'OOPS!! File di configurazione non selezionato o non presente..', Style.RESET_ALL)
input()
sys.exit()
print(Fore.LIGHTRED_EX, 'OOPS!! File di configurazione non selezionato o non presente..', Style.RESET_ALL)
input()
sys.exit()
try:
print('\nApro il file di configurazione [{}]...'.format(conffiles[fn]))
fp = open(conffiles[fn],'r')
cfg = json.load(fp)
fp.close
try:
print('Carico i parametri...')
docParams=cfg['docStruct']
basePath=cfg['INworkPath']+"{0}.{1}"
settings=cfg['settings']
delChars=settings['delChars'].split('|')
docSep=re.compile(docParams['docSep'])
except:
print(Fore.LIGHTRED_EX,'OOPS! Qualcosa e\' andato storto, non riesco a caricare la configurazione, controlla la sintassi! :)', Style.RESET_ALL)
input()
sys.exit()
print('\nApro il file di configurazione [{}]...'.format(conffiles[fn]))
fp = open(conffiles[fn],'r')
cfg = json.load(fp)
fp.close
try:
print('Carico i parametri...')
docParams=cfg['docStruct']
basePath=cfg['INworkPath']+"{0}.{1}"
settings=cfg['settings']
delChars=settings['delChars'].split('|')
docSep=re.compile(docParams['docSep'])
except:
print(Fore.LIGHTRED_EX,'OOPS! Qualcosa e\' andato storto, non riesco a caricare la configurazione, controlla la sintassi! :)', Style.RESET_ALL)
input()
sys.exit()
except IOError as e:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non trovo il file di configurazione: {}'.format(e), Style.RESET_ALL)
input()
sys.exit()
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non trovo il file di configurazione: {}'.format(e), Style.RESET_ALL)
input()
sys.exit()
files=[]
try:
print('Ottengo la lista dei file da separare...')
if settings['loadTXT']:
files+=glob(basePath.format('*','txt'))
if settings['loadDOCX']:
print(Fore.LIGHTRED_EX, 'OOPS! Scusa ma non posso accontentarti, per ora non so leggere i file DOCX.. :(', Style.RESET_ALL)
input()
sys.exit()
#files+=glob(basePath.format('*','docx'))
if len(files)<=0:
raise
pprint(files)
while True:
r=input('\nVuoi davvero bananaSPLITTARE questi documenti? [y/n]:')
r.strip()
if r=='Y' or r =='y':
break
elif r=='n' or r=='N':
print('OK! Nessun problema, ci vediamo dopo :)')
input()
sys.exit()
else:
print('Non ho capito la risposta, sii un po\' piu\' specifico... [y/n]\n')
pass
if settings['removeOldFiles']:
print("Rimuovo i vecchi file dalla cartella di destinazione..")
for x in glob(cfg['OUTworkPath']+'*.txt'):
os.remove(x)
print('Ottengo la lista dei file da separare...')
if settings['loadTXT']:
files+=glob(basePath.format('*','txt'))
if settings['loadDOCX']:
print(Fore.LIGHTRED_EX, 'OOPS! Scusa ma non posso accontentarti, per ora non so leggere i file DOCX.. :(', Style.RESET_ALL)
input()
sys.exit()
#files+=glob(basePath.format('*','docx'))
if len(files)<=0:
raise
pprint(files)
while True:
r=input('\nVuoi davvero bananaSPLITTARE questi documenti? [y/n]:')
r.strip()
if r=='Y' or r =='y':
break
elif r=='n' or r=='N':
print('OK! Nessun problema, ci vediamo dopo :)')
input()
sys.exit()
else:
print('Non ho capito la risposta, sii un po\' piu\' specifico... [y/n]\n')
pass
except:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco a trovare nessun file da leggere :(', Style.RESET_ALL)
input()
sys.exit()
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco a trovare nessun file da leggere :(', Style.RESET_ALL)
input()
sys.exit()
try:
if settings['removeOldFiles']:
print("Rimuovo i vecchi file dalla cartella di destinazione..")
for x in glob(cfg['OUTworkPath']+'*.txt'):
os.remove(x)
except:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco a rimuovere i vecchi file :(', Style.RESET_ALL)
lastTime=time.time()
fileCounter = 1
fileBodyCounter = 1
@@ -128,201 +130,201 @@ totSkipped=0
# per ogni file nella lista
for f in files:
s=status['first']
prevLine=''
doc={'title':'',
'date':'',
'content':[]
}
fileContent=''
tempContent=list()
fileBaseName=os.path.split(f)[1].split('.')[0]
fileBodyList=list()
fileContent = list()
try:
print()
print('-'*50)
print('Apro il file: {}'.format(fileBaseName))
fp = open(f,mode='r', encoding=settings['encoding'])
fileContent=fp.readlines() #leggi le linee del file
fp.close()
except IOError as e:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco ad aprire il file: {}'.format(fileBaseName), Style.RESET_ALL)
continue
print('Elimino le righe vuote...')
for ll in fileContent:
for c in delChars:
ll=ll.replace(c,'')
if ll not in ['\n','\r']:
tempContent.append(ll)
fileContent=deepcopy(tempContent)
tempContent=list()
print('Individuo il contenuto...')
docNumber = 0
docSkipped = 0
bodyCounter = 0
duplicateNumber = 0
docDate = {}
prevLine = ''
newsPaperName = ''
titleBegin = False
s=status['first']
prevLine=''
doc={'title':'',
'date':'',
'content':[]
}
fileContent=''
tempContent=list()
fileBaseName=os.path.split(f)[1].split('.')[0]
fileBodyList=list()
fileContent = list()
try:
print()
print('-'*50)
print('Apro il file: {}'.format(fileBaseName))
fp = open(f,mode='r', encoding=settings['encoding'])
fileContent=fp.readlines() #leggi le linee del file
fp.close()
except IOError as e:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco ad aprire il file: {}'.format(fileBaseName), Style.RESET_ALL)
continue
print('Elimino le righe vuote...')
for ll in fileContent:
for c in delChars:
ll=ll.replace(c,'')
if ll not in ['\n','\r']:
tempContent.append(ll)
fileContent=deepcopy(tempContent)
tempContent=list()
print('Individuo il contenuto...')
docNumber = 0
docSkipped = 0
bodyCounter = 0
duplicateNumber = 0
docDate = {}
prevLine = ''
newsPaperName = ''
titleBegin = False
for l in fileContent: #per ogni linea del file
lineWords=l.lstrip().split(' ') #dividi la riga in parole
if s==status['first']:
try:
#prendo il numero di documento per vedere se ci sono buchi
try:
nn = parse.parse("{current:d} Of {total} Documents",l.strip().capitalize()).named
if nn["current"]-docNumber==1:
pass
else:
if settings["showSkipped"]:
print(Fore.LIGHTRED_EX,"OOPS!! Il conto dei documenti non torna! LexisNexis \
ne ha saltato qualcuno!\nPrecedente:{0}-Attuale:{1}".format(docNumber,nn["current"]), Style.RESET_ALL)
docSkipped+=1
docNumber = nn["current"]
except:
pass
if (lineWords[settings['monthPosition']]).capitalize() in docParams['dateWords']:
try:
docDate=parse.parse(docParams['dateFormat'],l).named
docDate['month']=docDate['month'].lstrip().rstrip().capitalize()
docDate['month']=docParams['dateWords'].index(docDate['month'])+1
title = ''
titleBegin=True
if settings['getNewsPaperName']:
try:
if prevLine.split(' ')[0].strip().isalpha():
newsPaperName = prevLine.strip()
else:
newsPaperName = settings['nameNotFoundStr']
except:
print(Fore.LIGHTRED_EX, "OOPS! E' successo qualcosa mentre stavo cercando il nome della pubblicazione,\
controlla i file di uscita! \n\t[{}]".format(prevLine.strip()), Style.RESET_ALL)
else:
newsPaperName = settings['nameNotFoundStr']
except:
print(Fore.LIGHTRED_EX, "OOPS! Ho trovato una riga ambigua.. potrebbe essere una data ma non so: \n\t[{}]". format(l.strip('\r\n')),
Style.RESET_ALL)
elif lineWords[0] in docParams['headWords']:
s=status['head']
newDoc=deepcopy(doc)
newDoc['title']=title
newDoc['date']=docDate
newDoc['newsPaperName'] = newsPaperName
titleBegin=False
else:
if titleBegin:
title += l.strip().capitalize()
except IndexError:
print (Fore.LIGHTRED_EX, 'OOPS! Errore inaspettato, contatta il tuo sviluppatore di fiducia!', Style.RESET_ALL)
elif s==status['head']:
if lineWords[0] not in docParams['headWords']: #se la prima parola non e' tra quelle di inizio
tempContent.append(l) # vuol dire che ho trovato l'articolo
s=status['body']
elif s==status['body']:
if not lineWords[0] in docParams['tailWords']: #se la prima parola non e' tra quelle di fine
if settings['delLF']:
tempContent.append(l.strip('\n')) #allora sto leggendo l'articolo
else:
tempContent.append(l)
else:
s=status['tail']
anomaly = False
if docSep.match(l) is not None: #controlla se ci sono articoli che non hanno le parole chiave finali
print(Fore.YELLOW, "HEY! Ho individuato una separatore valido prima che si chiusesse l'articolo precedente, controlla i tuoi file in uscita!\n\
L'errore dovrebbe essere intorno all'articolo {} ma non sono sicuro! \n\t\t[{}]".format(docNumber,
l.strip()), Style.RESET_ALL)
s=status['tail']
anomaly = True
elif s==status['tail']: #cerco il separatore di articoli e aggiungo quello letto alla lista
if docSep.match(l) is not None or anomaly:
s=status['first']
anomaly = False
if settings['delWordBreak']:
tempContent=[ll.replace('-\n', '') for ll in tempContent]
newDoc['content']=deepcopy(''.join(tempContent))
fileBodyList.append(deepcopy(newDoc))
tempContent=list()
bodyCounter +=1
else:
pass
prevLine=l
pass
print ('Nel file ho trovato {0} articoli..'.format(bodyCounter))
if docSkipped > 0:
print (Fore.YELLOW, 'Attentione, LexisNexis ne ha saltati {} !!!'.format(docSkipped), Style.RESET_ALL)
print()
if bodyCounter >= 1:
if settings['removeDuplicates']:
print ('Controllo se ci sono dei duplicati..')
titleList=[]
duplicateList=[]
duplicateNumber=0
for idx, ff in enumerate(fileBodyList):
if ff['title'] not in titleList:
titleList.append(ff['title'])
ff['duplicate']=False
fileBodyList[idx]=ff
pass
else:
if ff['title'] not in duplicateList:
duplicateList.append(ff['title'])
if settings['showRemovedDuplicates']:
print ('Duplicato: {}'.format(ff['title'].strip()))
ff['duplicate'] = True
fileBodyList[idx]=ff
duplicateNumber+=1
print ('Ho rimosso {} duplicati di {} articoli..\n'. format(duplicateNumber, len(duplicateList)))
else:
for idx, ff in enumerate(fileBodyList):
ff['duplicate']=False
fileBodyList[idx]=ff
print('Salto il controllo dei duplicati..')
pass
if settings['saveSeparateFiles']:
print ('Salvo gli articoli in file separati...')
print ('Persorso: {0}'.format(cfg['OUTworkPath'].format('nomeFile')))
for ff in fileBodyList:
try:
if ff['duplicate'] == False:
fileName=cfg['OUTnameFormat'].format(title=slugify(ff['title'][:settings['maxTitleLen']]),\
filename=slugify(fileBaseName),\
docnum=fileBodyCounter,\
papername=ff['newsPaperName'].strip(),\
**ff['date'])
out=open(cfg['OUTworkPath']+'{0}'.format(fileName),'wb')
if settings['includeTitle']:
ff['content'] = ff['title']+os.linesep+ff['content']
out.write(ff['content'].encode(settings['encoding']))
out.close()
fileBodyCounter+=1
except IOError as e:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco a scrivere il file: {}'.format(e), Style.RESET_ALL)
continue
if settings['saveBodyFile']:
print('Salvo gli articoli in un singolo file vicino agli originali...')
print ('Persorso: {0}'.format(cfg['OUTworkPath'].format('nomeFile')))
try:
fileName=slugify(fileBaseName)
fileName='BODYFILE_{0}_{1}.txt'.format(fileCounter,fileName[:settings['maxTitleLen']])
fileContent = os.linesep.join([cc['content'] for cc in fileBodyList])
out=open(cfg['OUTworkPath']+'{0}'.format(fileName),'wb')
out.write(fileContent.encode(settings['encoding']))
out.close()
except IOError as e:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco a scrivere il file: {}'.format(e), Style.RESET_ALL)
continue
fileCounter+=1
totSkipped+=docSkipped
totFound+=bodyCounter-duplicateNumber
else:
print('[{0}] non contiene articoli, \n controlla meglio le parole chiave! SGRUNT'.format(fileBaseName))
print('-'*50)
for l in fileContent: #per ogni linea del file
lineWords=l.lstrip().split(' ') #dividi la riga in parole
if s==status['first']:
try:
#prendo il numero di documento per vedere se ci sono buchi
try:
nn = parse.parse("{current:d} Of {total} Documents",l.strip().capitalize()).named
if nn["current"]-docNumber==1:
pass
else:
if settings["showSkipped"]:
print(Fore.LIGHTRED_EX,"OOPS!! Il conto dei documenti non torna! LexisNexis \
ne ha saltato qualcuno!\nPrecedente:{0}-Attuale:{1}".format(docNumber,nn["current"]), Style.RESET_ALL)
docSkipped+=1
docNumber = nn["current"]
except:
pass
if (lineWords[settings['monthPosition']]).capitalize() in docParams['dateWords']:
try:
docDate=parse.parse(docParams['dateFormat'],l).named
docDate['month']=docDate['month'].lstrip().rstrip().capitalize()
docDate['month']=docParams['dateWords'].index(docDate['month'])+1
title = ''
titleBegin=True
if settings['getNewsPaperName']:
try:
if prevLine.split(' ')[0].strip().isalpha():
newsPaperName = prevLine.strip()
else:
newsPaperName = settings['nameNotFoundStr']
except:
print(Fore.LIGHTRED_EX, "OOPS! E' successo qualcosa mentre stavo cercando il nome della pubblicazione,\
controlla i file di uscita! \n\t[{}]".format(prevLine.strip()), Style.RESET_ALL)
else:
newsPaperName = settings['nameNotFoundStr']
except:
print(Fore.LIGHTRED_EX, "OOPS! Ho trovato una riga ambigua.. potrebbe essere una data ma non so: \n\t[{}]". format(l.strip('\r\n')),
Style.RESET_ALL)
elif lineWords[0] in docParams['headWords']:
s=status['head']
newDoc=deepcopy(doc)
newDoc['title']=title
newDoc['date']=docDate
newDoc['newsPaperName'] = newsPaperName
titleBegin=False
else:
if titleBegin:
title += l.strip().capitalize()
except IndexError:
print (Fore.LIGHTRED_EX, 'OOPS! Errore inaspettato, contatta il tuo sviluppatore di fiducia!', Style.RESET_ALL)
elif s==status['head']:
if lineWords[0] not in docParams['headWords']: #se la prima parola non e' tra quelle di inizio
tempContent.append(l) # vuol dire che ho trovato l'articolo
s=status['body']
elif s==status['body']:
if not lineWords[0] in docParams['tailWords']: #se la prima parola non e' tra quelle di fine
if settings['delLF']:
tempContent.append(l.strip('\n')) #allora sto leggendo l'articolo
else:
tempContent.append(l)
else:
s=status['tail']
anomaly = False
if docSep.match(l) is not None: #controlla se ci sono articoli che non hanno le parole chiave finali
print(Fore.YELLOW, "HEY! Ho individuato una separatore valido prima che si chiusesse l'articolo precedente, controlla i tuoi file in uscita!\n\
L'errore dovrebbe essere intorno all'articolo {} ma non sono sicuro! \n\t\t[{}]".format(docNumber,
l.strip()), Style.RESET_ALL)
s=status['tail']
anomaly = True
elif s==status['tail']: #cerco il separatore di articoli e aggiungo quello letto alla lista
if docSep.match(l) is not None or anomaly:
s=status['first']
anomaly = False
if settings['delWordBreak']:
tempContent=[ll.replace('-\n', '') for ll in tempContent]
newDoc['content']=deepcopy(''.join(tempContent))
fileBodyList.append(deepcopy(newDoc))
tempContent=list()
bodyCounter +=1
else:
pass
prevLine=l
pass
print ('Nel file ho trovato {0} articoli..'.format(bodyCounter))
if docSkipped > 0:
print (Fore.YELLOW, 'Attentione, LexisNexis ne ha saltati {} !!!'.format(docSkipped), Style.RESET_ALL)
print()
if bodyCounter >= 1:
if settings['removeDuplicates']:
print ('Controllo se ci sono dei duplicati..')
titleList=[]
duplicateList=[]
duplicateNumber=0
for idx, ff in enumerate(fileBodyList):
if ff['title'] not in titleList:
titleList.append(ff['title'])
ff['duplicate']=False
fileBodyList[idx]=ff
pass
else:
if ff['title'] not in duplicateList:
duplicateList.append(ff['title'])
if settings['showRemovedDuplicates']:
print ('Duplicato: {}'.format(ff['title'].strip()))
ff['duplicate'] = True
fileBodyList[idx]=ff
duplicateNumber+=1
print ('Ho rimosso {} duplicati di {} articoli..\n'. format(duplicateNumber, len(duplicateList)))
else:
for idx, ff in enumerate(fileBodyList):
ff['duplicate']=False
fileBodyList[idx]=ff
print('Salto il controllo dei duplicati..')
pass
if settings['saveSeparateFiles']:
print ('Salvo gli articoli in file separati...')
print ('Persorso: {0}'.format(cfg['OUTworkPath'].format('nomeFile')))
for ff in fileBodyList:
try:
if ff['duplicate'] == False:
fileName=cfg['OUTnameFormat'].format(title=slugify(ff['title'][:settings['maxTitleLen']]),\
filename=slugify(fileBaseName),\
docnum=fileBodyCounter,\
papername=ff['newsPaperName'].strip(),\
**ff['date'])
out=open(cfg['OUTworkPath']+'{0}'.format(fileName),'wb')
if settings['includeTitle']:
ff['content'] = ff['title']+os.linesep+ff['content']
out.write(ff['content'].encode(settings['encoding']))
out.close()
fileBodyCounter+=1
except IOError as e:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco a scrivere il file: {}'.format(e), Style.RESET_ALL)
continue
if settings['saveBodyFile']:
print('Salvo gli articoli in un singolo file vicino agli originali...')
print ('Persorso: {0}'.format(cfg['OUTworkPath'].format('nomeFile')))
try:
fileName=slugify(fileBaseName)
fileName='BODYFILE_{0}_{1}.txt'.format(fileCounter,fileName[:settings['maxTitleLen']])
fileContent = os.linesep.join([cc['content'] for cc in fileBodyList])
out=open(cfg['OUTworkPath']+'{0}'.format(fileName),'wb')
out.write(fileContent.encode(settings['encoding']))
out.close()
except IOError as e:
print(Fore.LIGHTRED_EX, 'OOPS! Qualcosa e\' andato storto, non riesco a scrivere il file: {}'.format(e), Style.RESET_ALL)
continue
fileCounter+=1
totSkipped+=docSkipped
totFound+=bodyCounter-duplicateNumber
else:
print('[{0}] non contiene articoli, \n controlla meglio le parole chiave! SGRUNT'.format(fileBaseName))
print('-'*50)
print()
print (Fore.LIGHTYELLOW_EX, 'bananaSPLIT ha concluso con successo in {0:1.3f} secondi, \n\

64
bananaconfEN.json Normal file
View File

@@ -0,0 +1,64 @@
{
"INworkPath": "D:\\Test\\",
"OUTworkPath": "D:\\Test\\Separati\\",
"OUTnameFormat":"TEST_{docnum}_{year:04d}{month:02d}{day:02d}_{title}.txt",
"docStruct": {
"docSep": "\\s*Copyright [(0-9)]+",
"dateFormat":"{month} {day:d}, {year:d}{}",
"dateWords": [
"January",
"February",
"March",
"April",
"May",
"June",
"July",
"August",
"September",
"October",
"November",
"December"
],
"headWords": [
"BYLINE:",
"SECTION:",
"LENGTH:",
"DATELINE:",
"HIGHLIGHT:",
"Email:"
],
"tailWords": [
"Newstex ID" ,
"NOTES",
"LANGUAGE:",
"GRAPHIC:",
"TYPE:",
"URL:",
"LOAD-DATE:",
"PUBLICATION-TYPE:",
"DOCUMENT-TYPE:",
"CHARTS:",
"JOURNAL-CODE:"
]
},
"settings": {
"encoding": "utf-8",
"monthPosition": 0,
"getNewsPaperName": true,
"nameNotFoundStr" : "--ND--",
"includeTitle" : true,
"removeDuplicates": true,
"showSkipped": false,
"showRemovedDuplicates": true,
"maxTitleLen": 32,
"loadTXT": true,
"loadDOCX": false,
"removeOldFiles":true,
"saveSeparateFiles": true,
"saveBodyFile": true,
"saveBodyNumber":true,
"delLF": false,
"delWordBreak": true,
"delChars": "'|@|#"
}
}

61
bananaconfITA.json Normal file
View File

@@ -0,0 +1,61 @@
{
"INworkPath": "C:\\Test\\",
"OUTworkPath": "C:\\Test\\Separati\\",
"OUTnameFormat":"TEST_{docnum}_{year:04d}{month:02d}{day:02d}_{papername}_{title}.txt",
"docStruct": {
"docSep": "Copyright [(0-9)]+",
"dateFormat":"{day:d} {month} {year:d} {}",
"dateWords": [
"Gennaio",
"Febbraio",
"Marzo",
"Aprile",
"Maggio",
"Giugno",
"Luglio",
"Agosto",
"Settembre",
"Ottobre",
"Novembre",
"Dicembre"
],
"headWords": [
"BYLINE:",
"SECTION:",
"LENGTH:",
"DATELINE:",
"HIGHLIGHT:",
"Email:"
],
"tailWords": [
"LANGUAGE:",
"GRAPHIC:",
"TYPE:",
"URL:",
"LOAD-DATE:",
"PUBLICATION-TYPE:",
"DOCUMENT-TYPE:",
"CHARTS:"
]
},
"settings": {
"encoding": "utf-8",
"monthPosition": 0,
"getNewsPaperName": true,
"nameNotFoundStr" : "--ND--",
"includeTitle" : true,
"removeDuplicates": true,
"showSkipped": false,
"showRemovedDuplicates": true,
"maxTitleLen": 32,
"loadTXT": true,
"loadDOCX": false,
"removeOldFiles":true,
"saveSeparateFiles": true,
"saveBodyFile": true,
"saveBodyNumber":true,
"delLF": false,
"delWordBreak": true,
"delChars": "'|@|#"
}
}