papers_please

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import subprocess
import sys
from pathlib import Path
import pickle
from paperqa import Settings, Docs

local_llm_config = {
    "model_list": [
        {
            "model_name": "ollama/llama3.1",
            "litellm_params": {
                "model": "ollama/llama3.1",
            },
        },
    ]
}
local_emb_config = {
    "model_list": [
        {
            "model_name": "ollama/mxbai-embed-large",
            "litellm_params": {
                "model": "ollama/mxbai-embed-large",
            },
        }
    ]
}

settings = Settings(
    llm="ollama/llama3.1",
    llm_config=local_llm_config,
    summary_llm="ollama/llama3.1",
    summary_llm_config=local_llm_config,
    embedding="ollama/mxbai-embed-large",
    embedding_config=local_emb_config,
)

def find_main_tex_file(folder_path: Path):
    """
    Find the main LaTeX file in the given folder.

    This function searches for a .tex file that is likely to be the main file
    of a LaTeX project. It first checks for common names like 'main.tex',
    then looks for files containing '\\documentclass', and finally returns
    the first .tex file if no other criteria are met.

    Args:
        folder_path (Path): The path to the folder to search in.

    Returns:
        Path: The path to the main .tex file, or None if no .tex files are found.
    """
    tex_files = list(folder_path.glob('**/*.tex'))
    if not tex_files:
        return None

    # Check for common main file names
    common_names = ['main.tex', 'paper.tex', 'article.tex']
    for name in common_names:
        if name in tex_files:
            return name

    # If no common name found, look for \documentclass
    for file in tex_files:
        with open(file, 'r', encoding='utf-8') as f:
            content = f.read()
            if '\\documentclass' in content:
                return file
    # If still not found, return the first .tex file
    return tex_files[0]

def run_latexpand(input_file, output_file):
    """
    Run the latexpand command on the input file and write the result to the output file.

    This function uses the latexpand tool to expand a LaTeX file, including all its
    inputs and packages, into a single file. The expanded content is then written
    to the specified output file.

    Args:
        input_file (str or Path): The path to the input LaTeX file.
        output_file (str or Path): The path where the expanded LaTeX content will be written.

    Raises:
        subprocess.CalledProcessError: If latexpand encounters an error during execution.
        FileNotFoundError: If the latexpand command is not found in the system PATH.
    """
    try:
        result = subprocess.run(['latexpand', input_file], 
                                capture_output=True, text=True, check=True)
        with open(output_file, 'w', encoding='utf-8') as output_file_handle:
            output_file_handle.write(result.stdout)
        print(f"Expanded LaTeX written to {output_file}")
    except subprocess.CalledProcessError as e:
        print(f"Error running latexpand: {e}")
    except FileNotFoundError:
        print("latexpand not found. Please make sure it's installed and in your PATH.")


cache_path = Path("pqa_index.pkl")

if cache_path.exists():
    with open(cache_path, "rb") as f:
        docs = pickle.load(f)
else:
    docs = Docs()
    for root, dirs, files in Path(".").walk():
        for dir_name in dirs:
            if dir_name.startswith("arXiv-"):
                dir_path = root / dir_name
                concat_main = dir_path / ".main.tex"
                try:
                    # Step 1: Find the main entry TeX file
                    main_file = find_main_tex_file(dir_path)
                    if not main_file:
                        raise ValueError("No main TeX file found.")
                    # Step 2 & 3: Run latexpand and write output
                    run_latexpand(main_file, dir_path / ".main.tex")
                except (ValueError, subprocess.CalledProcessError,
                        FileNotFoundError) as preprocess_error:
                    print(f"Failed to pre-process {dir_name}: {preprocess_error}")
                    continue
                print(f"adding {dir_path} (latex source)")
                try:
                    docs.add(concat_main, settings=settings, disable_check=True)
                except (IOError, OSError, ValueError) as add_error:
                    print(f"Failed to add {dir_path}: {add_error}")
                    continue
                dirs.remove(dir_name)
                break
        else:
            for file_name in files:
                if file_name.lower().endswith((".pdf", ".txt", ".md", ".tex")):
                    file_path = root / file_name
                    print(f"adding {file_path}")
                    docs.add(file_path, settings=settings, disable_check=True)

    with open(cache_path, "wb") as f:
        pickle.dump(docs, f)


if __name__ == "__main__":
    if len(sys.argv) > 1:
        QUERY = " ".join(sys.argv[1:])
        answer = docs.query(QUERY, settings=settings)
        print(answer)
    else:
        print("Please provide a query as a command-line argument.")
        print("Usage: python script_name.py 'Your query here'")