Archive Now (archivenow) currently is configured to push resources into four public web archives. You can easily add more archives by writing a new archive handler (e.g., myarchive_handler.py) and place it inside the folder "handlers".
Originally, archivenow was configured to push to 6 different public web archives. The two removed web archives are WebCite and archive.st. WebCite was removed from archivenow as they are no longer accepting archiving requests. Archive.st was removed from archivenow due to encountering a Captcha when attempting to push to the archive. In addition to removing those 2 archives, the method for pushing to archive.today and megalodon.jp from archivenow has been updated. In order to push to archive.today and megalodon.jp, Selenium is used.
As explained below, this library can be used through:
- Command Line Interface (CLI)
- A Web Service
- A Docker Container
- Python
The latest release of archivenow can be installed using pip:
$ pip install archivenow
The latest development version containing changes not yet released can be installed from source:
$ git clone git@github.com:oduwsdl/archivenow.git
$ cd archivenow
$ pip install -r requirements.txt
$ pip install ./
In order to push to archive.today and megalodon.jp, archivenow must use Selenium, which has already been added to the requirements.txt. However, Selenium additionally needs a driver to interface with the chosen browser. It is recommended to use Selenium and archivenow with Firefox and Firefox's corresponding GeckoDriver.
You can download the latest versions of Firefox and the GeckoDriver to use with archivenow.
After installing the driver, you can push to archive.today and megalodon.jp from archivenow.
Usage of sub-commands in archivenow can be accessed through providing the -h or --help flag, like any of the below.
$ archivenow -h
usage: archivenow.py [-h] [--mg] [--cc] [--cc_api_key [CC_API_KEY]]
[--is] [--ia] [--warc [WARC]] [-v] [--all]
[--server] [--host [HOST]] [--agent [AGENT]]
[--port [PORT]]
[URI]
positional arguments:
URI URI of a web resource
optional arguments:
-h, --help show this help message and exit
--mg Use Megalodon.jp
--cc Use The Perma.cc Archive
--cc_api_key [CC_API_KEY]
An API KEY is required by The Perma.cc Archive
--is Use The Archive.is
--ia Use The Internet Archive
--warc [WARC] Generate WARC file
-v, --version Report the version of archivenow
--all Use all possible archives
--server Run archiveNow as a Web Service
--host [HOST] A server address
--agent [AGENT] Use "wget" or "squidwarc" for WARC generation
--port [PORT] A port number to run a Web Service
To save the web page (www.foxnews.com) in the Internet Archive:
$ archivenow --ia www.foxnews.com
https://web.archive.org/web/20170209135625/http://www.foxnews.com
By default, the web page (e.g., www.foxnews.com) will be saved in the Internet Archive if no optional arguments are provided:
$ archivenow www.foxnews.com
https://web.archive.org/web/20170215164835/http://www.foxnews.com
To save the web page (www.foxnews.com) in the Internet Archive (archive.org) and Archive.is:
$ archivenow --ia --is www.foxnews.com
https://web.archive.org/web/20170209140345/http://www.foxnews.com
http://archive.is/fPVyc
To save the web page (https://nypost.com/) in all configured web archives. In addition to preserving the page in all configured archives, this command will also locally create a WARC file:
$ archivenow --all https://nypost.com/ --cc_api_key $Your-Perma-CC-API-Key
http://archive.is/dcnan
https://perma.cc/53CC-5ST8
https://web.archive.org/web/20181002081445/https://nypost.com/
https://megalodon.jp/2018-1002-1714-24/https://nypost.com:443/
https_nypost.com__96ec2300.warc
To download the web page (https://nypost.com/) and create a WARC file:
$ archivenow --warc=mypage --agent=wget https://nypost.com/
mypage.warc
You can run archivenow as a web service. You can specify the server address and/or the port number (e.g., --host localhost --port 12345)
$ archivenow --server
Running on http://0.0.0.0:12345/ (Press CTRL+C to quit)
To save the web page (www.foxnews.com) in The Internet Archive through the web service:
$ curl -i http://0.0.0.0:12345/ia/www.foxnews.com
HTTP/1.0 200 OK
Content-Type: application/json
Content-Length: 95
Server: Werkzeug/0.11.15 Python/2.7.10
Date: Tue, 02 Oct 2018 08:20:18 GMT
{
"results": [
"https://web.archive.org/web/20181002082007/http://www.foxnews.com"
]
}
To save the web page (www.foxnews.com) in all configured archives though the web service:
$ curl -i http://0.0.0.0:12345/all/www.foxnews.com
HTTP/1.0 200 OK
Content-Type: application/json
Content-Length: 385
Server: Werkzeug/0.11.15 Python/2.7.10
Date: Tue, 02 Oct 2018 08:23:53 GMT
{
"results": [
"Error (The Perma.cc Archive): An API Key is required ",
"http://archive.is/ukads",
"https://web.archive.org/web/20181002082007/http://www.foxnews.com",
"Error (Megalodon.jp): We can not obtain this page because the time limit has been reached or for technical ... ",
"http://www.webcitation.org/72rbKsX8B"
]
}
Because an API Key is required by Perma.cc, the HTTP request should be as follows:
$ curl -i http://127.0.0.1:12345/all/https://nypost.com/?cc_api_key=$Your-Perma-CC-API-Key
Or use only Perma.cc:
$ curl -i http://127.0.0.1:12345/cc/https://nypost.com/?cc_api_key=$Your-Perma-CC-API-Key
$ docker image pull oduwsdl/archivenow
Different ways to run archivenow
$ docker container run -it --rm oduwsdl/archivenow -h
Accessible at 127.0.0.1:12345:
$ docker container run -p 12345:12345 -it --rm oduwsdl/archivenow --server --host 0.0.0.0
Accessible at 127.0.0.1:22222:
$ docker container run -p 22222:11111 -it --rm oduwsdl/archivenow --server --port 11111 --host 0.0.0.0
To save the web page (http://www.cnn.com) in The Internet Archive
$ docker container run -it --rm oduwsdl/archivenow --ia http://www.cnn.com
>>> from archivenow import archivenow
To save the web page (www.foxnews.com) in all configured archives:
>>> archivenow.push("www.foxnews.com","all")
['https://web.archive.org/web/20170209145930/http://www.foxnews.com','http://archive.is/oAjuM','http://www.webcitation.org/6o9LcQoVV','Error (The Perma.cc Archive): An API KEY is required]
To save the web page (www.foxnews.com) in The Perma.cc:
>>> archivenow.push("www.foxnews.com","cc",{"cc_api_key":"$YOUR-Perma-cc-API-KEY"})
['https://perma.cc/8YYC-C7RM']
To start the server from Python do the following. The server/port number can be passed (e.g., start(port=1111, host='localhost')):
>>> archivenow.start()
2017-02-09 15:02:37
Running on http://127.0.0.1:12345
(Press CTRL+C to quit)
Additional archives may be added by creating a handler file in the "handlers" directory.
For example, if I want to add a new archive named "My Archive", I would create a file "ma_handler.py" and store it in the folder "handlers". The "ma" will be the archive identifier, so to push a web page (e.g., www.cnn.com) to this archive through the Python code, I should write:
archivenow.push("www.cnn.com","ma")
In the file "ma_handler.py", the name of the class must be "MA_handler". This class must have at least one function called "push" which has one argument. See the existing handler files for examples on how to organized a newly configured archive handler.
Removing an archive can be done by one of the following options:
- Removing the archive handler file from the folder "handlers"
- Renaming the archive handler file to other name that does not end with "_handler.py"
- Setting the variable "enabled" to "False" inside the handler file
The Internet Archive (IA) sets a time gap of at least two minutes between creating different copies of the "same" resource.
For example, if you send a request to IA to capture (www.cnn.com) at 10:00pm, IA will create a new copy (C) of this URI. IA will then return C for all requests to the archive for this URI received until 10:02pm. Using this same submission procedure for Archive.is requires a time gap of five minutes.
@INPROCEEDINGS{archivenow-jcdl2018,
AUTHOR = {Mohamed Aturban and
Mat Kelly and
Sawood Alam and
John A. Berlin and
Michael L. Nelson and
Michele C. Weigle},
TITLE = {{ArchiveNow}: Simplified, Extensible, Multi-Archive Preservation},
BOOKTITLE = {Proceedings of the 18th {ACM/IEEE-CS} Joint Conference on Digital Libraries},
SERIES = {{JCDL} '18},
PAGES = {321--322},
MONTH = {June},
YEAR = {2018},
ADDRESS = {Fort Worth, Texas, USA},
URL = {https://doi.org/10.1145/3197026.3203880},
DOI = {10.1145/3197026.3203880}
}