Downloading archived web pages from the Wayback Machine.
Internet-archive is a nice source for several OSINT-information. This tool is a work in progress to query and fetch archived web pages.
This tool allows you to download content from the Wayback Machine (archive.org). You can use it to download either the latest version or all versions of web page snapshots within a specified range.
- Install the package
pip install pywaybackup
- Run the tool
waybackup -h
- Clone the repository
git clone https://github.com/bitdruid/python-wayback-machine-downloader.git
- Install
pip install .
- in a virtual env or use
--break-system-package
- in a virtual env or use
- Linux recommended: On Windows machines, the path length is limited. This can only be overcome by editing the registry. Files that exceed the path length will not be downloaded.
- If you query an explicit file (e.g. a query-string
?query=this
orlogin.html
), the--explicit
-argument is recommended as a wildcard query may lead to an empty result. - The tool will inform you if your query has an immense amount of snapshots which could consume your system memory and lead to a crash. Consider splitting your query into smaller jobs by specifying a range e.g.
--start 2023 --end 2024
or--range 1
.
-h
,--help
: Show the help message and exit.-a
,--about
: Show information about the tool and exit.
-u
,--url
:
The URL of the web page to download. This argument is required.
-c
,--current
:
Download the latest version of each file snapshot. You will get a rebuild of the current website with all available files (but not any original state because new and old versions are mixed).-f
,--full
:
Download snapshots of all timestamps. You will get a folder per timestamp with the files available at that time.-s
,--save
:
Save a page to the Wayback Machine. (beta)
-
-l
,--list
:
Only print the snapshots available within the specified range. Does not download the snapshots. -
-e
,--explicit
:
Only download the explicit given URL. No wildcard subdomains or paths. Use e.g. to get root-only snapshots. This is recommended for explicit files likelogin.html
or?query=this
. -
-o
,--output
:
Defaults towaybackup_snapshots
in the current directory. The folder where downloaded files will be saved. -
Range Selection:
Specify the range in years or a specific timestamp either start, end, or both. If you specify therange
argument, thestart
andend
arguments will be ignored. Format for timestamps: YYYYMMDDhhmmss. You can only give a year or increase specificity by going through the timestamp starting on the left.
(year 2019, year+month 201901, year+month+day 20190101, year+month+day+hour 2019010112)-r
,--range
:
Specify the range in years for which to search and download snapshots.--start
:
Timestamp to start searching.--end
:
Timestamp to end searching.
-
--csv
<path>
:
Path defaults to output-dir. Saves a CSV file with the json-response for successfull downloads. If--list
is set, the CSV contains the CDX list of snapshots. If--current
or--full
is set, CSV contains downloaded files. Named aswaybackup_<sanitized_url>.csv
. -
--skip
<path>
:
Path defaults to output-dir. Checks for an existingwaybackup_<sanitized_url>.csv
for URLs to skip downloading. Useful for interrupted downloads. Files are checked by their root-domain, ensuring consistency across queries. This means that if you downloadhttp://example.com/subdir1/
and laterhttp://example.com
, the second query will skip the first path. -
--no-redirect
:
Disables following redirects of snapshots. Useful for preventing timestamp-folder mismatches caused by Archive.org redirects. -
--verbosity
<level>
:
Sets verbosity level. Options arejson
(prints JSON response) orprogress
(shows progress bar).
-
--log
<path>
:
Path defaults to output-dir. Saves a log file with the output of the tool. Named aswaybackup_<sanitized_url>.log
. -
--workers
<count>
:
Sets the number of simultaneous download workers. Default is 1, safe range is about 10. Be cautious as too many workers may lead to refused connections from the Wayback Machine. -
--retry
<attempts>
:
Specifies number of retry attempts for failed downloads. -
--delay
<seconds>
:
Specifies delay between download requests in seconds. Default is no delay (0). -
--limit
<count>
:
Limits the amount of snapshots to query from the CDX server. If an existing CDX file is injected (with--cdxinject
or--auto
), the limit will have no effect.
CDX Query Result Handling:
-
--cdxbackup
<path>
:
Path defaults to output-dir. Saves the result of CDX query as a file. Useful for later downloading snapshots and overcoming refused connections by CDX server due to too many queries. Named aswaybackup_<sanitized_url>.cdx
. -
--cdxinject
<filepath>
:
Injects a CDX query file to download snapshots. Ensure the query matches the previous--url
for correct folder structure.
Auto:
--auto
:
If set, csv, skip and cdxbackup/cdxinject are handled automatically. Keep the files and folders as they are. Otherwise they will not be recognized when restarting a download.
Download latest snapshot of all files:
waybackup -u http://example.com -c
Download latest snapshot of a specific file:
waybackup -u http://example.com/subdir/file.html -c
Download all snapshots sorted per timestamp with a specified range and do not follow redirects:
waybackup -u http://example.com -f -r 5 --no-redirect
Download all snapshots sorted per timestamp with a specified range and save to a specified folder with 3 workers:
waybackup -u http://example.com -f -r 5 -o /home/user/Downloads/snapshots --workers 3
Download all snapshots from 2020 to 12th of December 2022 with 4 workers, save a csv and show a progress bar:
waybackup -u http://example.com -f --start 2020 --end 20221212 --workers 4 --csv --verbosity progress
Download all snapshots and output a json response:
waybackup -u http://example.com -f --verbosity json
List available snapshots per timestamp without downloading and save a csv file to home folder:
waybackup -u http://example.com -f -l --csv /home/user/Downloads
The output path is currently structured as follows by an example for the query:
http://example.com/subdir1/subdir2/assets/
:
For the current version (-c
):
- The requested path will only include all files/folders starting from your query-path.
your/path/waybackup_snapshots/
└── the_root_of_your_query/ (example.com/)
└── subdir1/
└── subdir2/
└── assets/
├── image.jpg
├── style.css
...
For all versions (-f
):
- Will currently create a folder named as the root of your query. Inside this folder, you will find all timestamps and per timestamp the path you requested.
your/path/waybackup_snapshots/
└── the_root_of_your_query/ (example.com/)
├── yyyymmddhhmmss/
│ ├── subidr1/
│ │ └── subdir2/
│ │ └── assets/
│ │ ├── image.jpg
│ │ └── style.css
├── yyyymmddhhmmss/
│ ├── subdir1/
│ │ └── subdir2/
│ │ └── assets/
│ │ ├── image.jpg
│ │ └── style.css
...
For download queries:
[
{
"file": "/your/path/waybackup_snapshots/example.com/yyyymmddhhmmss/index.html",
"id": 1,
"redirect_timestamp": "yyyymmddhhmmss",
"redirect_url": "http://web.archive.org/web/yyyymmddhhmmssid_/http://example.com/",
"response": 200,
"timestamp": "yyyymmddhhmmss",
"url_archive": "http://web.archive.org/web/yyyymmddhhmmssid_/http://example.com/",
"url_origin": "http://example.com/"
},
...
]
For list queries:
[
{
"digest": "DIGESTOFSNAPSHOT",
"id": 1,
"mimetype": "text/html",
"status": "200",
"timestamp": "yyyymmddhhmmss",
"url": "http://example.com/"
},
...
]
The csv contains the json response in a table format.
Exceptions will be written into waybackup_error.log
(each run overwrites the file).
I'm always happy for some feature requests to improve the usability of this tool. Feel free to give suggestions and report issues. Project is still far from being perfect.