Add 'setup.py' for install

This update allows users to import xsearch in an individual conda environment. 

* add setup.py
* update README.md to document alternate installation method using `setup.py`

Co-authored-by: Stephen Po-Chedley <[email protected]>

Author: lee1043

README.md

@@ -22,7 +22,9 @@ If the xsearch version is incremented you will receive a warning on import:
 
 `xsearch` will create a file in your home directory to support version checking.
 
-Alternatively, you can download the contents of this repository and import the package (or the `search.py` file). In this case, we suggest you watch the repository (in GitHub) so that you know when there are version changes.
+Alternatively, you can download the contents of this repository and import the package (or the `search.py` file). In this case, we suggest you watch the repository (in GitHub) so that you know when there are version changes. You can either add `export PYTHONPATH='/PATH/TO/YOUR/XSEARCH/'` in your `~/.bashrc` or use the following commend to make xsearch importable:
+
+    python setup.py install
 
 `xsearch` is built around metadata logic documented in the [CMIP6 Global Attributes, DRS, Filenames, Directory Structure, and CV’s document](https://goo.gl/v1drZl). These logic underpin the filename and directory creation, written by [CMOR](https://cmor.llnl.gov/) and other CMIP-writing libraries. The [CMIP6 CMOR Tables](https://github.com/PCMDI/cmip6-cmor-tables/tree/master/Tables) (e.g. Amon - designate A=atmospheric realm, and mon=monthly frequency), similar logic underpinned the [CMIP5](https://github.com/PCMDI/cmip5-cmor-tables/tree/master/Tables) and [CMIP3](https://github.com/PCMDI/cmip3-cmor-tables/tree/master/Tables) phases.
 
@@ -156,4 +158,4 @@ This isn't typically a problem if you specify an exact experiment, variable, and
 
 ### Acknowledgements
 
-Much of the de-duplication logic and ideas from this are from the [durolib](https://github.com/durack1/durolib) library, created by Paul Durack [@durack1](https://github.com/durack1) to search for CMIP data. durolib handled [CDAT](https://cdat.llnl.gov/) xml files, which could be used to read in multi-file datasets. Logic to produce the xml files was refactored in [xagg](https://github.com/pochedls/xagg). xagg xmls are being phased out in favor the current approach: rapidly searchable json files, which allow the users to locate datasets and read them in with xarray based tools. Stephen Po-Chedley [@pochedls](https://github.com/pochedls/) produced the `xsearch` initial version.
+Much of the de-duplication logic and ideas from this are from the [durolib](https://github.com/durack1/durolib) library, created by Paul Durack [@durack1](https://github.com/durack1) to search for CMIP data. durolib handled [CDAT](https://cdat.llnl.gov/) xml files, which could be used to read in multi-file datasets. Logic to produce the xml files was refactored in [xagg](https://github.com/pochedls/xagg). xagg xmls are being phased out in favor the current approach: rapidly searchable json files, which allow the users to locate datasets and read them in with xarray based tools. Stephen Po-Chedley [@pochedls](https://github.com/pochedls/) produced the `xsearch` initial version.

setup.py

@@ -0,0 +1,15 @@
+from setuptools import find_packages, setup
+
+version = '0.0.3'
+
+packages = find_packages(include=['xsearch'])
+
+setup(
+    name="xsearch",
+    version=version,
+    author="@pochedls",
+    description="xsearch search utility",
+    url="https://github.com/pochedls/xsearch",
+    packages=packages,
+)
+

xsearch/__init__.py

@@ -12,7 +12,7 @@
 import os
 
 # version
-__version__ = "0.0.4"
+__version__ = "0.0.5"
 
 # since this software is installed centrally and may be updated
 # this section of code stores the xsearch version in a hidden