The script that was originally included here is now superseded by:
https://github.com/KBNLresearch/detectStorageMediaType
See also my blog post:
Identification of physical storage media and devices with Python and the Windows API
#!/bin/bash | |
# Location of Kakadu binaries | |
kduPath=/Applications/kakadu | |
# Add Kakadu path to LD_LIBRARY_PATH | |
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$kduPath | |
# Create TIFF from existing JP2 | |
/Applications/kakadu/kdu_expand -i aware.jp2 -o aware.tif |
{ | |
"tweet" : { | |
"edit_info" : { | |
"initial" : { | |
"editTweetIds" : [ | |
"1588159317974319106" | |
], | |
"editableUntil" : "2022-11-03T13:51:02.000Z", | |
"editsRemaining" : "5", | |
"isEditEligible" : false |
<?xml version='1.0' encoding='UTF-8'?> | |
<jpylyzer xmlns="http://openpreservation.org/ns/jpylyzer/v2/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://openpreservation.org/ns/jpylyzer/v2/ http://jpylyzer.openpreservation.org/jpylyzer-v-2-0.xsd"> | |
<toolInfo> | |
<toolName>jpylyzer</toolName> | |
<toolVersion>2.0.0</toolVersion> | |
</toolInfo> | |
<file> | |
<fileInfo> | |
<fileName>6e3Gf8Mu</fileName> | |
<filePath>/home/johan/test/6e3Gf8Mu</filePath> |
The script that was originally included here is now superseded by:
https://github.com/KBNLresearch/detectStorageMediaType
See also my blog post:
Identification of physical storage media and devices with Python and the Windows API
<?xml version="1.0" ?> | |
<isolyzer xmlns="http://kb.nl/ns/isolyzer/v1/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://kb.nl/ns/isolyzer/v1/ https://raw.githubusercontent.com/KBNLresearch/isolyzer/xsd/xsd/isolyzer-v-1-0.xsd"> | |
<toolInfo> | |
<toolName>cli.py</toolName> | |
<toolVersion>1.4.0a2</toolVersion> | |
</toolInfo> | |
<image> | |
<fileInfo> | |
<fileName>iso9660.iso</fileName> | |
<filePath>/home/johan/isolyzer/testFiles/iso9660.iso</filePath> |
<?xml version="1.0" ?> | |
<isolyzer> | |
<toolInfo> | |
<toolName>cli.py</toolName> | |
<toolVersion>1.4.0a1</toolVersion> | |
</toolInfo> | |
<image> | |
<fileInfo> | |
<fileName>BOOKSHELF.iso01.iso</fileName> | |
<filePath>/home/johan/kb/iso-identification/HSF/BOOKSHELF.iso01.iso</filePath> |
import sys | |
import os | |
import glob | |
import platform | |
import codecs | |
import argparse | |
# Create parser | |
parser = argparse.ArgumentParser( | |
description="Test CLI input with wildcards, multiple platforms") |
#! /usr/bin/env python3 | |
# | |
""" | |
Save web pages to Wayback Machine. Argument urlsIn can either be | |
a text file with URLs (each line contains one URL), or a single | |
URL. In the first (input file) case it will simply save each URL. | |
In the latter case (input URL) it will extract all links from the URL, and | |
save those as well as the root URL (useful for saving a page with all | |
of its direct references). The optional --extensions argument can be used | |
to limit this to one or more specific file extensions. E.g. the following |
#! /usr/bin/env python3 | |
from warcio.capture_http import capture_http | |
import requests | |
def main(): | |
# Existing warc.gz file (created with wget, then compressed using warcio's | |
# 'recompress' command) | |
with capture_http("ziklies.home.xs4all.nl.warc.gz"): | |
for indexOnder in range(1, 8): | |
for indexMidden in range(1, 8): |
Gebruik omSipCreator voor de tests op kopieën van batches, en NIET op de originele opslaglocaties! Dit is vooral omdat omSipCreator in 'prune' modus (opschoonfunctie) batches wijzigt en daarbij data verwijdert!!
Ik neem in de voorbeelden hieronder even aan dat Python onder de volgende folder geïnstalleerd is:
C:\Python37\