Call Sequencer

Call Sequencer is a Python package for sequencing function calls as blocks.

Installation

You can install the package via pip:

pip install call-sequencer

Usage

Using Chains of Functions

from call_sequencer import CallSequencer
# Example Usage
@CallSequencer.simple
def sanitize(param) -> str:...

@CallSequencer.simple
def retrive_data(param)-> str:...

@CallSequencer.with_args
def query_llm(param, *args, **kwargs)-> str:...

@CallSequencer.simple
def extract_code(param)-> str:...

@CallSequencer.simple
def execute_code(param)-> str:...


code_evaluator = (
    CallSequencer.start("Hello")  # Initial input block
    >> sanitize
    >> retrive_data
    >> query_llm("extra_param")
    >> extract_code
    >> execute_code
    >> query_llm("another_param")
)

# Run the chain
result = code_evaluator()
print("Final Result:", result)

Predefined Functions

Explanation of Markdown Parsing Functions

1. extract_title

   • Extracts the title from a Markdown string, identified by a single # at the beginning of a line. Returns None if no title is found.

2. extract_code_blocks

   • Extracts all code blocks from a Markdown string delimited by triple backticks (```). Returns a list of code block strings.

3. extract_tables

   • Extracts tables from a Markdown string formatted with pipes (|). Returns a list of table row strings.

4. extract_lists

   • Extracts both ordered (numbered) and unordered (bulleted) lists from a Markdown string. Returns a list of list item strings.

5. extract_first_code_block

   • Extracts the content of the first code block encountered in a Markdown string. Returns None if no code block is found.

6. extract_headings

   • Extracts all headings from a Markdown string, identified by lines starting with multiple # characters. Returns a list of heading strings.

7. extract_bold_text

   • Extracts text formatted in bold (**text**) from a Markdown string. Returns a list of bold text strings.

8. extract_italic_text

   • Extracts text formatted in italics (*text*) from a Markdown string. Returns a list of italic text strings.

9. extract_links

   • Extracts hyperlinks in the format [text](url) from a Markdown string. Returns a list of tuples where each tuple contains the text and URL.

10. extract_all_text

    • Strips out all Markdown formatting (bold, italic, links) and returns plain text from a Markdown string. Returns a plain string of text content.

These functions are designed to parse and extract specific elements from Markdown strings, useful for processing and analyzing text generated by large language models or similar applications.

Explanation of Code Runner Functions

1. run_python_code

   • Runs a given Python code string and returns the result. Accepts additional global variables through **globals_dict.

2. run_shell_command

   • Executes a shell command and returns the output as a dictionary containing stdout, stderr, and returncode.

3. evaluate_expression

   • Evaluates a Python expression and returns the result.

4. install_package

   • Installs a Python package using pip and returns the output as a dictionary containing stdout, stderr, and returncode.

5. uninstall_package

   • Uninstalls a Python package using pip and returns the output as a dictionary containing stdout, stderr, and returncode.

6. list_installed_packages

   • Lists all installed Python packages using pip and returns the output as a dictionary containing stdout, stderr, and returncode.

7. get_package_info

   • Retrieves information about a Python package using pip show and returns the output as a dictionary containing stdout, stderr, and returncode.

8. write_file

   • Writes a string content to a specified file. If the operation is successful, returns None; otherwise, returns an error message.

9. read_file

   • Reads the content of a specified file and returns it as a string. If the operation fails, returns an error message.

10. append_to_file

    • Appends string content to a specified file. If the operation is successful, returns None; otherwise, returns an error message.

License

This project is licensed under the MIT License - see the LICENSE file for details.