- Python 3 only.
- Running in docker container is strongly recommended.
- Install tornado server (to run server).
- Insert your Bitly OAuth access token in config file. Visit https://bitly.com/a/oauth_apps to get your access token.
- Insert your Bing search api key in config file. Visit https://www.microsoft.com/cognitive-services/en-us/bing-web-search-api to get your api key.
- (optional) Change server ip/port number in config.
- All other packages that generate error "No module named" does exist, must be installed.
It is recommended to use Docker to install and run this application.
To download the Docker image:
$ docker image pull oduwsdl/carbondate
Once you have downloaded the image, run it as a server or a local command line program (in one-off mode).
To run it as a server:
$ docker container run --rm -it -p 8888:8888 oduwsdl/carbondate ./main.py -s
Then access it at http://localhost:8888/.
To run it in one-off mode:
$ docker container run --rm -it oduwsdl/carbondate ./main.py -l {URI-R}
To run it as a server:
$ ./main.py -s
Then access it at http://localhost:8888/.
To run it in one-off mode:
$ ./main.py -l {URI-R}
Carbon Date provides the option of passing in environment variables for both the Bing and Bitly services.
For Bitly, the environment key is CD_Bitly_token
. For Bing, the environment key is CD_Bing_key
.
Environment variables can be passed into docker using the -e
or --env
arguments before executing the Carbon Date application like so:
$ docker run -e "CD_Bitly_token=foo" -e "CD_Bing_key=bar" -it --rm oduwsdl/carbondate ./main.py -l http://www.cs.odu.edu/
CarbonDate provides the option of preventing searching for specified modules.
For example, if a user wants to disable backlinks and google modules the user can add the -e
argument after a URI-R is specified like so:
$ ./main.py -l "https://www.cs.odu.edu/" -e cdGetBacklinks cdGetGoogle
A complete list of all the modules a user can disable is as follows:
cdGetPubdate
cdGetArchives
cdGetBing
cdGetBitly
cdGetTwitter
cdGetBacklinks
cdGetGoogle
cdGetLastModified
Name your module's main script as cdGet<ModuleName>.py
. If your module relies on extra code in a different folder you may bring this directory into the modules directory, the CarbonDate library will ignore subfolders while loading modules. If the extra code is not in a subfolder, after copying it to the ./modules
folder, add the file names into the config file under the SystemUtility
field.
And ensure the entry function is named
get<Module name>(url,outputArray, indexOfOutputArray,verbose=False,**kwargs)
or customize your own entry function name by assigning a string value to 'entry' variable in the beginning of your script.
For example if your module name is Service, then the script should be named cdGetService, and the interface function should be named:
getService (url,outputArray, indexOfOutputArray,verbose=False,**kwargs)
Copy your scripts to the folder ./modules, then the system will automatically detect and load it.
The data returned from your module can be a string of date in the following format: '1995-01-01T12:00:00'
or a dictionary with earliest
as a mandatory key for a source you define along with any optional keys your module provides.
{
"foo": {
"earliest": "2017-07-04T15:10:24",
"optional-fields": "extra stuff"
}
}
If your module gathers multiple sources you can also return these sources in a dictionary but each source must have a key named earliest
. For example your method can return a dictionary as follows:
{
"foo": {
"earliest": "2017-07-04T15:10:24"
},
"bar": {
"earliest": "2015-02-01T06:05:30",
"extra": "stuff"
}
}
Put your result date in to outputArray[indexOfOutputArray] for result comparison,
outputArray[outputArrayIndex] = time
and put the result and other data you want to show in the "displayArray" like:
kwargs['displayArray'][outputArrayIndex] = dictionary_or_datestring_variable
Where the variable outputArray, indexOfOutputArray and displayArray are past in by the system.