Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Browse files

starting to work on doc files

  • Loading branch information...
commit 8e2a48102d3d7998fce5d604df0e6df869da89bb 1 parent 51aac7d
@mrgaaron authored
View
72 doc/overview.txt
@@ -0,0 +1,72 @@
+This library is intended to be used with the LinkedIn API. It is not
+an application in and of itself, so you will need to obtain your own
+API key to use the library properly.
+
+This overview will go over the library in the following sections:
+
+I. Design Philosophy
+II. Library Structure
+III. Additional Information
+
+I. Design Philosophy
+
+The library is highly object-oriented in nature. Pretty much everything
+you can do will end up being passed through a series of objects. The
+client interface is an object, all parsers are objects, and even XML nodes
+from API returns are objects.
+
+Most importantly, all returns from the LinkedIn API come back as objects,
+not XML string literals. This includes API errors caused by bad URL
+formats, bad access tokens, etc. As of right now, the original XML
+used to construct the objects will not be available directly for you
+to inspect, although it can easily be accessed by calling
+etree.tostring() from the lxml library on the return object's xml
+attribute.
+
+In keeping with that and the Python design philosophy, most interactions
+with the API are done with standard Python objects and data types. To use
+the search API, you pass a dict with your search criteria as key-value pairs.
+To use field selectors, you simply pass a list of strings. To specify timestamps,
+you can either pass the UTC string/integer or a standard Python datetime object.
+Lowlevel type conversions and URL formatting should be done for you.
+
+If it should work, it probably does. If it doesn't, let me know or contribute
+code to make it work :).
+
+II. Library Structure
+
+ liclient/
+ __init__.py - contains the LinkedInAPI class itself and is
+ the entry point for the library
+ analysis/
+ __init__.py
+ nlp.py - contains utilities for NLP analysis
+ of profiles (requires NLTK)
+ oauth2/
+ __init__.py - 3rd party module for dealing with
+ oauth, MIT license is included in
+ the module itself
+ parsers/
+ __init__.py
+ lixml.py - contains the parsers for XML returns
+ helpers.py - helper functions for data conversion
+ mappers.py - classes for mapping XML returns
+
+III. Additional Information
+
+I work at a small recruiting firm and wrote the library to help us do
+LinkedIn integrations with our website and CRM system. Some of the design
+choices probably do reflect that a little bit (like the update_content
+attribute that appears on each network update - it's not a part of the
+update itself but is a generated string based on the type of update).
+
+Any questions or comments can be emailed to me (my address is in the
+readme). Issues can be raised and fixes contributed at the code's
+Github repo.
+
+Thanks for giving this library a look.
+
+Aaron
+
+
+
View
67 doc/overview.txt~
@@ -0,0 +1,67 @@
+This library is intended to be used with the LinkedIn API. It is not
+an application in and of itself, so you will need to obtain your own
+API key to use the library properly.
+
+This overview will go over the library in the following sections:
+
+I. Design Philosophy
+II. Library Structure
+III. Additional Information
+
+I. Design Philosophy
+
+The library is highly object-oriented in nature. Pretty much everything
+you can do will end up being passed through a series of objects. The
+client interface is an object, all parsers are objects, and even XML nodes
+from API returns are objects.
+
+Most importantly, all returns from the LinkedIn API come back as objects,
+not XML string literals. This includes API errors caused by bad URL
+formats, bad access tokens, etc. As of right now, the original XML
+used to construct the objects will not be available directly for you
+to inspect, although it can easily be accessed by calling
+etree.tostring() from the lxml library on the return object's xml
+attribute.
+
+In keeping with that and the Python design philosophy, most interactions
+with the API are done with standard Python objects and data types. To use
+the search API, you pass a dict with your search criteria as key-value pairs.
+To use field selectors, you simply pass a list of strings. To specify timestamps,
+you can either pass the UTC string/integer or a standard Python datetime object.
+Lowlevel type conversions and URL formatting should be done for you.
+
+If it should work, it probably does. If it doesn't, let me know or contribute
+code to make it work :).
+
+II. Library Structure
+
+ liclient/
+ __init__.py - contains the LinkedInAPI class itself and is
+ the entry point for the library
+ analysis/
+ __init__.py
+ nlp.py - contains utilities for NLP analysis
+ of profiles (requires NLTK)
+ oauth2/
+ __init__.py - 3rd party module for dealing with
+ oauth, MIT license is included in
+ the module itself
+ parsers/
+ __init__.py
+ lixml.py - contains the parsers for XML returns
+ helpers.py - helper functions for data conversion
+ mappers.py - classes for mapping XML returns
+
+III. Additional Information
+
+I work at a small recruiting firm and wrote the library to help us do
+LinkedIn integrations with our website and CRM system. Some of the design
+choices probably do reflect that a little bit (like the update_content
+attribute that appears on each network update - it's not a part of the
+update itself but is a generated string based on the type of update).
+
+Any questions or comments can be emailed to me (my address is in the
+readme).
+
+
+
View
0  doc/tutorial.txt
No changes.
Please sign in to comment.
Something went wrong with that request. Please try again.