quantum-spec-parser 2025.12.21131621

A new package designed to extract and structure key technical specifications from quantum computing research and development texts. The package takes unstructured text input describing quantum process

quantum-spec-parser

quantum-spec-parser is a lightweight Python package that extracts and structures key technical specifications from unstructured quantum computing texts. It parses descriptions of quantum processors and returns a standardized list containing essential details such as qubit count, material composition, and fidelity ranges.

The package relies on the llmatch‑messages library for robust pattern matching and uses ChatLLM7 (from langchain_llm7) as the default language model. You can also supply any LangChain‑compatible LLM of your choice.

---

📦 Installation

pip install quantum_spec_parser

---

🚀 Quick Start

from quantum_spec_parser import quantum_spec_parser

# Example unstructured description of a quantum processor
user_input = """
The new Q-42 chip features 56 superconducting transmon qubits fabricated on a silicon substrate.
Gate fidelity lies between 99.2% and 99.7%, and the coherence time averages 120 µs.
"""

# Simple call – uses the default ChatLLM7 internally
specs = quantum_spec_parser(user_input)

print(specs)
# → ['56 qubits', 'silicon substrate', 'fidelity 99.2%‑99.7%', 'coherence time 120 µs']

---

📚 Detailed Usage

Function Signature

def quantum_spec_parser(
    user_input: str,
    api_key: Optional[str] = None,
    llm: Optional[BaseChatModel] = None,
) -> List[str]:

Parameter	Type	Description
user_input	str	The raw text containing quantum processor specifications.
api_key	Optional[str]	Your LLM7 API key. If omitted, the function reads LLM7_API_KEY from the environment or falls back to a placeholder.
llm	Optional[BaseChatModel]	A LangChain LLM instance to use instead of the default ChatLLM7. Any model that implements BaseChatModel works.

If llm is not provided, the function automatically creates a ChatLLM7 instance using the supplied (or env‑provided) API key.

---

Using a Custom LLM

You can replace the default ChatLLM7 with any LangChain‑compatible model, such as OpenAI, Anthropic, or Google Generative AI.

OpenAI

from langchain_openai import ChatOpenAI
from quantum_spec_parser import quantum_spec_parser

llm = ChatOpenAI(model="gpt-4o-mini")
response = quantum_spec_parser(user_input, llm=llm)

Anthropic

from langchain_anthropic import ChatAnthropic
from quantum_spec_parser import quantum_spec_parser

llm = ChatAnthropic(model="claude-3-haiku-20240307")
response = quantum_spec_parser(user_input, llm=llm)

Google Generative AI

from langchain_google_genai import ChatGoogleGenerativeAI
from quantum_spec_parser import quantum_spec_parser

llm = ChatGoogleGenerativeAI(model="gemini-1.5-flash")
response = quantum_spec_parser(user_input, llm=llm)

---

Supplying Your Own LLM7 API Key

The free tier of LLM7 usually suffices, but for higher rate limits you can provide a personal key:

response = quantum_spec_parser(user_input, api_key="your-llm7-api-key")

Or set it globally via the environment:

export LLM7_API_KEY="your-llm7-api-key"

You can obtain a free key by registering at https://token.llm7.io/.

---

🛠️ Under the Hood

1. Prompt construction – The package builds system and human prompts defined in prompts.py.
2. Pattern matching – A regular expression (pattern) is compiled and passed to llmatch from llmatch_messages.
3. LLM call – The selected LLM generates a response.
4. Extraction – llmatch validates the response against the regex and returns the extracted data as a list of strings.

If the LLM call fails or the response does not match the pattern, a RuntimeError is raised with the underlying error message.

---

🐞 Issues & Contributions

If you encounter bugs or have feature requests, please open an issue:

https://github....

Pull requests are welcome! Feel free to fork the repository and submit your improvements.

---

✍️ Author

Eugene Evstafev – [e-mail]
GitHub: chigwell

---

📜 License

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