You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
* Implement parameter parsing as in this [ServiceX story](ssl-hep/ServiceX#395)
Individual CommitsL
* Parse a `files` count thing
* Get rid of post-3.6 dependencies
* Search for the parsed DID, not the original parameterized one!
* Fix up flake8 errors
* Test for parsed parameter removal
* Update docs and comments
* Convert so everything is multi-path
* Make sure multi-path works ok
* Fix flake8
* Obey all vs avail get modes
* Get logic of using get and files correct
Copy file name to clipboardExpand all lines: README.md
+10Lines changed: 10 additions & 0 deletions
Original file line number
Diff line number
Diff line change
@@ -128,3 +128,13 @@ In the end, all DID finders for ServiceX will run under Kubernetes. ServiceX com
128
128
Note the parameter `request-id`: this marks the log messages with the request id that triggered this DID request. This will enable the system to track all log messages across all containers connected with this particular request id - making debugging a lot easier.
129
129
130
130
The `start_did_finder` will configure the python root logger properly to dump messages with a request ID in them.
131
+
132
+
## URI Format
133
+
134
+
All the incoming DID's are expected to be URI's without the schema. As such, there are two parameters that are currently parsed by the library. The rest are let through and routed to the callback:
135
+
136
+
*`files` - Number of files to report back to ServiceX. All files from the dataset are found, and then sorted in order. The first n files are then
137
+
sent back. Default is all files.
138
+
*`get` - If the value is `all` (the default) then all files in the dataset must be returned. If the value is `available`, then only files that are accessible need be returned.
139
+
140
+
As am example, if the following URI is given to ServiceX, "rucio://dataset_name?files=20&get=available", then the first 20 available files of the dataset will be processed by the rest of servicex.
0 commit comments