import pandas as pd

sheet_id = "1xDGQTL8QXFKK2t_FFZhswfuZc_BwznsWKHeOkFRRXYY"
sheet_name = "Token-Level_Annotation"

import pandas as pd

def df_to_conllu(df):
    conllu_sentences = []

    for sent_id, group in df.groupby("SENT_ID"):
        row0 = group.iloc[0]

        text = row0["TEXT"]
        text_en = row0["TEXT_EN"]
        genre = row0["GENRE"]
        source = row0["SOURCE"]
        annotator = row0["ANNOTATOR"]
        status = row0["STATUS"]

        lines = [
            f"# sent_id = {sent_id}",
            f"# text = {text}",
            f"# text_en = {text_en}",
            f"# genre = {genre}",
            f"# source = {source}",
            f"# annotator = {annotator}",
            f"# status = {status}",
        ]

        for _, row in group.iterrows():
            fields = [
                str(row["ID"]),
                str(row["FORM"]),
                str(row["LEMMA"]),
                str(row["UPOS"]),
                str(row["XPOS"]),
                str(row["FEATS"]),
                str(row["HEAD"]),
                str(row["DEPREL"]),
                str(row["DEPS"]),
                str(row["MISC"]),
            ]
            lines.append("\t".join(fields))

        conllu_sentences.append("\n".join(lines))

    return "\n\n".join(conllu_sentences) + "\n"


def convert(sheet_id, sheet_name):
    url = f"https://docs.google.com/spreadsheets/d/{sheet_id}/gviz/tq?tqx=out:csv&sheet={sheet_name}"
    df = pd.read_csv(url)
    cols = ", ".join(df.columns.tolist())
    conllu = df_to_conllu(df)
    return f"Columns: {cols}\n\n{conllu}"


# -------------------------
# RUN WITHOUT UI
# -------------------------

sheet_id = "1xDGQTL8QXFKK2t_FFZhswfuZc_BwznsWKHeOkFRRXYY"
sheet_name = "Token-Level_Annotation"

result = convert(sheet_id, sheet_name)
print(result)

import gradio as gr

def noop():
    return "Running."

demo = gr.Interface(fn=noop, inputs=None, outputs="text")
demo.launch()