Skip to content

jobh/komle

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

komle

komle, a python library for WITSML, uses PyXB to marshal/unmarshal xml files according to the schemas.

Some of the features are:

  • WITSML data model bindings for schema v1.4.1.1 and v2.0
    • Note just one version can be used in the same runtime, due to namespace collision
  • WITSML to dict, for use in a pandas dataframe or json
  • Unit converter based on witsmlUnitDict
  • Soap client to request data from a witsml server, according to the webservice description
  • Validation that xml files conforms to the WITSML schema
  • Support for the generated write schemas, to be used for WMLS_AddToStore
    • Note that write_bindings can not be imported at the same time as read_bindings. See below for details.

Install

pip3 install git+ssh://git@github.com/kle043/komle.git

Or if the repo is cloned

pip3 install -U .

Getting started

from komle.bindings.v1411.read import witsml
from komle import utils as ku
import pandas as pd # Not part of komle setup

with open('log.xml', 'r') as log_file:
    # logs is a regular python object according to the witsml schema
    logs = witsml.CreateFromDocument(log_file.read())

# Print the witsml documentation for logs
print(logs._element().documentation())

# Print the schema location for logCurveInfo, nice to have for reference
print(logs.log[0].logCurveInfo[0]._element().xsdLocation().locationBase)

print([l.name for l in logs.log])

# Convert logdata to a dict
log = logs.log[0]

data_dict = ku.logdata_dict(log)

# Create a dataframe, if you have installed pandas
df_data = pd.DataFrame(data_dict)

# Do the same for the plural logCurveInfo element
df_curve = pd.DataFrame(ku.plural_dict(log.logCurveInfo))

witsml.CreateFromDocument works on any witsml object, like trajectorys, mudLogs, tubulars etc, and returns a python representation according to the schema. Nodes are converted to there corresponding python types and accessed like any other python object, the exception is leaf nodes with attributes where one must call value() since primitive types in python does not have custom attributes. For example mdTop.value() where mdTop also has the attribute mdTop.uom, also see examples/hello_witsml.py.

Usage of different schemas

The difference between the schemas is described here. In summary,

  • Read Schemas: [...] a copy of the normative files except that all choices, elements and attributes are optional. [...] these schema files must represent the XMLout response from the WITSML WMLS_GetFromStore method.
  • Write Schemas: [...] a copy of the normative files except that some unique identifier attributes have had their optionality changed. [...] these schema files must represent the XMLin input to the WITSML WMLS_AddToStore method.
  • Update Schemas (not currently supported): [...] a copy of the normative files with all elements and attributes optional except that all unique identifier attributes and uom attributes are mandatory. [...] these schema files must represent the XMLin input to the WITSML WMLS_UpdateInStore method.
  • Delete Schemas (not currently supported): [...] a copy of the normative files with all elements and attributes optional except for all object uids and parentage-pointers which are mandatory. [...] these schema files must represent the QueryIn input to the WITSML WMLS_DeleteFromStore method.

As a practical matter, any program needing to work on both read and write (and update/delete) should only import the read bindings, since they have the least restrictions. The read bindings will be valid also for write/update/delete, as long as the mandatory elements and attributes are present. For validation of the write schema, a separate test program which only imports the write bindings must be used. (Note that the multiprocessing module can not easily be used for this purpose, since the child processes will inherit the parent process' imports.)

About

A python library for WITSML

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 100.0%