docketalarm_api 1.1.3

Wrapper to interact with DocketAlarm API

DocketAlarmAPI

API Wrapper for DocketAlarm.

Table of Contents

1. Installation and usage
2. Get Login Token
3. Search
4. Scroll
5. Get Docket
6. Get Document
7. Ask Docket
8. Case Matcher
9. Smart Search
10. Attorney Billing Rates
11. Get Complaint Summary
12. Get Cause of Action
13. Area of Law
14. Forum and Venues

usage

Use pip install docketalarm_api to install this library.
To use the DocketAlarmClient you'll need your DocketAlarm user and password, and if desired to interact with OpenAI powered endpoints, an OpenAI API Key.
Import and initialize the client as follows:

from docketalarm_api import DocketAlarmClient

da_client = DocketAlarmClient(<your-docketalarm-user>, <your-docketalarm-password>, <your-openai-api-key>, <your-anthropic-api-key>)

Without AI keys

You can initialize the client without an OpenAI API key or an Anthropic API key, but you will not be able to interact with AI endpoints.

from docketalarm_api import DocketAlarmClient

da_client = DocketAlarmClient(<your-docketalarm-user>, <your-docketalarm-password>)

With OpenAI key only

You can initialize the client with only an OpenAI API key as follows:

from docketalarm_api import DocketAlarmClient, GPT_4O_MINI

da_client = DocketAlarmClient(<your-docketalarm-user>, <your-docketalarm-password>, <your-openai-api-key>)

In this case you will be able to interact with AI endpoints directly without having to explicitly specify a model (GPT_4O is used as default) but you can import some models as in the example.

With Anthropic key only

You can also initialize the client with only an Anthropic API key as follows:

from docketalarm_api import DocketAlarmClient, CLAUDE_3_5_SONNET

da_client = DocketAlarmClient(<your-docketalarm-user>, <your-docketalarm-password>, anthropic_api_key=<your-anthropic-api-key>)

IMPORTANT

When using both keys or only an Anthropic API key you will have to explicitly provide the claude_model param/keyword argument to be used on the call, if it's desired to use Claude.

An AI method will attempt to use the OpenAI key by default

You can import different models from the "docketalarm_api" module:

GPT_3_5_TURBO, GPT_4_TURBO, GPT_4O_MINI, GPT_4O, CLAUDE_3_HAIKU, CLAUDE_3_SONNET, CLAUDE_3_5_SONNET

General Endpoints

These are methods available directly to your DocketAlarm user through a DocketAlarmClient instance.

get_login_token

Get authentication token from DocketAlarm.

Example:

    login_token = da_client.get_login_token()

search

Perform a search on DocketAlarm API's search endpoint.

Parameter	Description
query	The DocketAlarm query to use.
order	The order to get the results by.
limit	The search limit, must be 50 or less.
offset	Offset for the search, useful for pagination.
login_token	Will be created if not provided.
return	Dictionary with JSON response.

Example:

    response = da_client.search("is:docket AND is:state", "random", limit=10)
    search_results = response["search_results"]

search_scroll

Perform a scroll search on DocketAlarm API.

Parameter	Description
query	The DocketAlarm query to use.
order	The order to get the results by.
limit	The search limit, must be 50 or less.
offset	Offset for the search, useful for pagination.
login_token	Will be created if not provided.
scroll_parallel	Number of parallel threads, or bins, to divide the search results into for scanning
scroll_index	Individual thread, or bin, of the scroll_parallel threads indexed from 0 to scroll_parallel-1
scroll	String produced for each thread when first setting scroll_index and scroll_parallel outputted as another field with key “scroll.” Pass this value into all subsequent calls
return	Dictionary with JSON response.

Example:

    # SET UP
    response_index_0 = da_client.search_scroll(query="is:docket from:-7days court:(California)",
                                               order="-date_filed", scroll_parallel=2, scroll_index=0)
    
    response_index_1 = da_client.search_scroll(query="is:docket from:-7days court:(California)",
                                               order="-date_filed", scroll_parallel=2, scroll_index=1)
                                        
    search_results = response_index_0["search_results"] + response_index_1["search_results"]
    
    # SCROLLING
    scrolling_index_0 = da_client.search_scroll(query="is:docket from:-7days court:(California)",
                                         order="-date_filed", scroll=response_index_0["scroll"])
    
    search_results.extend(scrolling_index_0.get("search_results", []))

    scrolling_index_1 = da_client.search_scroll(query="is:docket from:-7days court:(California)",
                                         order="-date_filed", scroll=response_index_1["scroll"])
    
    
    search_results.extend(scrolling_index_1.get("search_results", []))
    
    # Could loop until no more results in any of the responses.
    # Make sure to always use the previous call's scroll value as a scroll parameter

get_docket

Interact with getdocket endpoint, fetching a docket by court and docket number.

Parameter	Description
docket	The docket number obtained from search.
court	The court of the docket obtained from search.
timeout	Timeout for the GET request.
client_matter	The client matter for the API call.
normalize	Normalize option for getdocket endpoint.
cached	Defaults to True, gets cached version of the docket.
login_token	If not provided it's auto-generated.
check_is_pacer	Include a boolean stating if the case is from a PACER court
add_documents_by_entries	Include a list of all documents per entry on the response
return	Dictionary with JSON response.

Example:

    docket_data = da_client.get_docket(docket="JC205-106", court="Texas State, Grayson County, Justice of the Peace 2")

get_document_binary

Fetches the binary content of a pdf stored in DocketAlarm or directly from the court.

Parameter	Description
doc_url	URL for the document.
login_token	Token for OpenAI authentication.
client_matter	The matter or client for the API call.
cached	Boolean stating if desired to use cached version of the document or not.
return	Document binary (bytes).

Example:

    document_bytes = da_client.get_document_binary(
        'https://www.docketalarm.com/cases/Arkansas_State_Faulkner_County_Circuit_Court/23DR-98-821/OCSE_V_JOHN_TAYLOR/docs/28082014_OTHER_0.pdf'
    )

AI Endpoints

These methods require the use of an AI API key, supplied when initializing the instance. This AI API key can be an OpenAI API key or an Anthropic API key.

Remember that when it's desired to interact with an AI endpoint using Claude, the claude_model param has to be provided

ask_docket

Interact with DocketAlarm's ask_docket OpenAI endpoint.

Parameter	Description
docket	The docket number as extracted from search.
court	The court of the docket as extracted from search.
question	The question to ask to the docket data.
output_format	The output format of the desired response in natural language.
target	The target for ask_docket, "docket", "documents" or "both".
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
cached	Gets cached version of the docket on the interaction, defaults to False.
show_relevant	Gets relevant data used by ask_docket.
login_token	If not provided, it is autogenerated.
timeout	The timeout for the request.
return	Dictionary with JSON response.

Examples:

Using OpenAI

    response = da_client.ask_docket(docket='JC205-106', court='Texas State, Grayson County, Justice of the Peace 2',
                                    question='is the case pre or post discovery stages?',
                                    output_format='Enum["pre" or "post" or "unknown"]',
                                    target="docket", cached = True)

    openai_answer = response["from_dockets"]["openai_answer"]

Using Claude

    from docketalarm_api import CLAUDE_3_5_SONNET
    
    
    response = da_client.ask_docket(docket='JC205-106', court='Texas State, Grayson County, Justice of the Peace 2',
                                    question='is the case pre or post discovery stages?',
                                    output_format='Enum["pre" or "post" or "unknown"]',
                                    target="docket", cached=True, claude_model=CLAUDE_3_5_SONNET)

    claude_answer = response["from_dockets"]["claude_answer"]

case_matcher

Match a case from any input provided using DocketAlarm's OpenAI powered case_matcher endpoint.

Parameter	Description
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
timeout	The timeout for the request.
kwargs	Provide any argument and it will be used as inputs in case matcher.
return	Dictionary with result from case matcher and AI costs incurred.

Examples:

Using OpenAI

    from docketalarm_api import GPT_4_TURBO
    
    
    response = da_client.case_matcher(openai_model=GPT_4_TURBO,
                                      description="A PACER case involving Cloud Systems HoldCo in California")
    case_link = response["result"]["Link"]

Using Claude

    from docketalarm_api import CLAUDE_3_5_SONNET
    
    
    response = da_client.case_matcher(claude_model=CLAUDE_3_5_SONNET,
                                      description="A PACER case involving Cloud Systems HoldCo in California")
    case_link = response["result"]["Link"]

smart_search

Return a query for DocketAlarm search based on instructions in natural language

Parameter	Description
instructions	Instructions to build a query by.
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
login_token	If not provided, it will be auto-generated.
timeout	The timeout for the request.
return	Dictionary with query and AI costs incurred.

Examples:

Using OpenAI:

    response = da_client.smart_search(
        instructions="Cases involving Ford Motor in New York, that span from December 2014 to june 2019"
    )
    query = response["query"]

Using Claude:

    from docketalarm_api import CLAUDE_3_5_SONNET


    response = da_client.smart_search(
        instructions="Cases involving Ford Motor in New York, that span from December 2014 to june 2019",
        claude_model=CLAUDE_3_5_SONNET
    )
    query = response["query"]

attorney_billing_rates

Extract attorney billing rates by name and state.

Parameter	Description
attorney_name	The name of the attorney for which billing rates are to be extracted.
state	The state of the attorney.
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
login_token	Auto-generated by default.
client_matter	Empty by default.
timeout	The timeout for the request.
return	Dictionary with result and AI costs incurred.

Examples

Using OpenAI:

    response = da_client.attorney_billing_rates(attorney_name="ashley marshal",
                                                state="connecticut")
    attorney_data = response["result"]

Using Claude:

    from docketalarm_api import CLAUDE_3_5_SONNET
    
    
    response = da_client.attorney_billing_rates(attorney_name="ashley marshal",
                                                state="connecticut",
                                                claude_model=CLAUDE_3_5_SONNET)
    attorney_data = response["result"]

get_complaint_summary

Get a summary of the legal complaint in the docket.

Parameter	Description
docket	Docket number.
court	The court of the docket.
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
login_token	Auto-generated by default.
cached	Bool stating if desired to use cached version of the docket.
short	Extract a short complaint summary.
timeout	The timeout for the request.
return	Dictionary with result and AI costs incurred.

Examples:

Using OpenAI:

    from docketalarm_api import GPT_4O
    
    
    response = da_client.get_complaint_summary(docket="JC205-106", court="Texas State, Grayson County, Justice of the Peace 2",
                                            short=True, openai_model=GPT_4O, cached=True)
    openai_answer = response["openai_answer"]

Using Claude:

    from docketalarm_api import CLAUDE_3_5_SONNET
    
    
    response = da_client.get_complaint_summary(docket="JC205-106", court="Texas State, Grayson County, Justice of the Peace 2",
                                               short=True, claude_model=CLAUDE_3_5_SONNET, cached=True)
    claude_answer = response["claude_answer"]

get_cause_of_action

Get the causes of action from a legal complaint.

Parameter	Description
docket	Docket number.
court	The court of the docket.
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
login_token	Auto-generated by default.
cached	Bool stating if desired to use cached version of the docket.
timeout	The timeout for the request.
return	Dictionary with result and AI costs incurred.

Examples:

Using OpenAI:

    from docketalarm_api import GPT_4O
    
    
    response = da_client.get_cause_of_action(docket="JC205-106", court="Texas State, Grayson County, Justice of the Peace 2",
                                            openai_model=GPT_4O, cached=True)
    openai_answer = response["openai_answer"]

Using Claude:

    from docketalarm_api import CLAUDE_3_5_SONNET
    
    
    response = da_client.get_cause_of_action(docket="JC205-106", court="Texas State, Grayson County, Justice of the Peace 2",
                                            claude_model=CLAUDE_3_5_SONNET, cached=True)
    claude_answer = response["claude_answer"]

entity_normalizer

Get a DocketAlarm query for the entity normalized

Parameter	Description
entity	The entity to normalize.
include_corp_group	Boolean stating if desired to include corporation group matches.
search_limit	The internal search limit when optimizing. Must be between 10 and 50.
login_token	If not provided, it is autogenerated.
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
timeout	The timeout for the request.
return	Dictionary with result and AI costs incurred.

Examples:

Using OpenAI:

    response = da_client.entity_normalizer(entity="Apple", search_limit=20,
                                           include_corp_group=True)
    query = response["query"]

Using Claude:

    from docketalarm_api import CLAUDE_3_5_SONNET
    
    
    response = da_client.entity_normalizer(entity="Apple", search_limit=20,
                                           include_corp_group=True,
                                           claude_model=CLAUDE_3_5_SONNET)
    query = response["query"]

area_of_law

Get the area of law SALI tag from a provided case type or NOS code.

Parameter	Description
case_type	The case type or NOS code from the docket
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
timeout	The timeout for the request.
login_token	If not provided, it is autogenerated.
return	Dictionary with result and AI costs incurred.

Examples:

Using OpenAI:

    response = da_client.area_of_law("Design - D23/262000")
    tag = response["tag"]

Using Claude:

    from docketalarm_api import CLAUDE_3_HAIKU
    
    
    response = da_client.area_of_law("Design - D23/262000", claude_model=CLAUDE_3_HAIKU)
    tag = response["tag"]

forum_and_venues

Get the forum and venues SALI tag from a provided court.

Parameter	Description
court	The court from the docket
openai_model	Model to be used on OpenAI interactions, defaults to gpt-4o.
claude_model	Model to be used on Claude interactions, defaults to claude-3.5-sonnet
timeout	The timeout for the request.
login_token	If not provided, it is autogenerated.
return	Dictionary with result and AI costs incurred.

Examples:

Using OpenAI:

    response = da_client.forum_and_venues("U.S. Patent Application")
    tag = response["tag"]

Using Claude:

    from docketalarm_api import CLAUDE_3_HAIKU
    
    
    response = da_client.forum_and_venues("U.S. Patent Application", claude_model=CLAUDE_3_HAIKU)
    tag = response["tag"]