app.py

import inspect
import json
import logging
import os
from typing import List, Type

import gradio as gr
import spacy  # noqa
from dotenv import load_dotenv
from gradio import routes
from transformers import pipeline

load_dotenv()

TOKENS2INT_ERROR_INT = 32202

log = logging.getLogger()

ONES = [
    "zero", "one", "two", "three", "four", "five", "six", "seven", "eight",
    "nine", "ten", "eleven", "twelve", "thirteen", "fourteen", "fifteen",
    "sixteen", "seventeen", "eighteen", "nineteen",
]

# token_mapping = json.load(open('str_mapping.json'))
CHAR_MAPPING = {
    "-": " ",
    "_": " ",
}
CHAR_MAPPING.update((str(i), word) for i, word in enumerate([" " + s + " " for s in ONES]))

TOKEN_MAPPING = dict(enumerate([" " + s + " " for s in ONES]))

BQ_JSON = os.environ['BQ_JSON']


def tokenize(text):
    return text.split()


def detokenize(tokens):
    return ' '.join(tokens)


def replace_tokens(tokens, token_mapping=TOKEN_MAPPING):
    return [token_mapping.get(tok, tok) for tok in tokens]


def replace_chars(text, char_mapping=CHAR_MAPPING):
    return ''.join((char_mapping.get(c, c) for c in text))


def tokens2int(tokens, numwords={}):
    """ Convert an English str containing number words into an int
    >>> text2int("nine")
    9
    >>> text2int("forty two")
    42
    >>> text2int("1 2 three")
    123
    """
    if not numwords:

        tens = ["", "", "twenty", "thirty", "forty", "fifty", "sixty", "seventy", "eighty", "ninety"]

        scales = ["hundred", "thousand", "million", "billion", "trillion"]

        numwords["and"] = (1, 0)
        for idx, word in enumerate(ONES):
            numwords[word] = (1, idx)
        for idx, word in enumerate(tens):
            numwords[word] = (1, idx * 10)
        for idx, word in enumerate(scales):
            numwords[word] = (10 ** (idx * 3 or 2), 0)

    current = result = 0

    for word in tokens:
        if word not in numwords:
            raise Exception("Illegal word: " + word)

        scale, increment = numwords[word]
        current = current * scale + increment
        if scale > 100:
            result += current
            current = 0

    return str(result + current)


def text2int(text):
    return tokens2int(tokenize(replace_chars(text)))


def try_text2int(text):
    text = str(text)
    try:
        intstr = tokens2int(tokens2int(tokenize(replace_chars(text))))
    except Exception as e:
        log.error(str(e))
        log.error(f'User input: {text}')
        intstr = TOKENS2INT_ERROR_INT
    return str(intstr)


def try_text2int_preprocessed(text):
    text = str(text)
    try:
        tokens = replace_tokens(tokenize(replace_chars(str(text))))
    except Exception as e:
        log.error(str(e))
        tokens = text.split()
    try:
        intstr = tokens2int(tokens)
    except Exception as e:
        log.error(str(e))
        intstr = str(TOKENS2INT_ERROR_INT)
    return intstr


def get_types(cls_set: List[Type], component: str):
    docset = []
    types = []
    if component == "input":
        for cls in cls_set:
            doc = inspect.getdoc(cls)
            doc_lines = doc.split("\n")
            docset.append(doc_lines[1].split(":")[-1])
            types.append(doc_lines[1].split(")")[0].split("(")[-1])
    else:
        for cls in cls_set:
            doc = inspect.getdoc(cls)
            doc_lines = doc.split("\n")
            docset.append(doc_lines[-1].split(":")[-1])
            types.append(doc_lines[-1].split(")")[0].split("(")[-1])
    return docset, types


routes.get_types = get_types

sentiment = pipeline(task="sentiment-analysis", model="distilbert-base-uncased-finetuned-sst-2-english")


def get_sentiment(text):
    return sentiment(text)


with gr.Blocks() as html_block:
    gr.Markdown("# Rori - Mathbot")

    with gr.Tab("Text to integer"):
        inputs_text2int = [
            gr.Text(placeholder="Type a number as text or a sentence", label="Text to process",
                    value="forty two"),
        ]

        outputs_text2int = gr.Textbox(label="Output integer")

        button_text2int = gr.Button("text2int")

        button_text2int.click(
            fn=try_text2int,
            inputs=inputs_text2int,
            outputs=outputs_text2int,
            api_name="text2int",
        )

        examples_text2int = [
            "one thousand forty seven",
            "one hundred",
        ]

        gr.Examples(examples=examples_text2int, inputs=inputs_text2int)

        gr.Markdown(r"""

        ## API
        ```python
        import requests

        requests.post(
            url="https://tangibleai-mathtext.hf.space/run/text2int", json={"data": ["one hundred forty five"]}
        ).json()
        ```

        Or using `curl`:

        ```bash
        curl -X POST https://tangibleai-mathtext.hf.space/run/text2int -H 'Content-Type: application/json' -d '{"data": ["one hundred forty five"]}'
        ```
        {bq_json}""" + f"{json.loads(BQ_JSON)['type']}")

    with gr.Tab("Text to integer preprocessed"):
        inputs_text2int_preprocessed = [
            gr.Text(placeholder="Type a number as text or a sentence", label="Text to process",
                    value="forty two"),
        ]

        outputs_text2int_preprocessed = gr.Textbox(label="Output integer")

        button_text2int = gr.Button("text2int preprocessed")

        button_text2int.click(
            fn=try_text2int_preprocessed,
            inputs=inputs_text2int_preprocessed,
            outputs=outputs_text2int_preprocessed,
            api_name="text2int-preprocessed",
        )

        examples_text2int_preprocessed = [
            "one thousand forty seven",
            "one hundred",
        ]

        gr.Examples(examples=examples_text2int_preprocessed, inputs=inputs_text2int_preprocessed)

        gr.Markdown(r"""

        ## API
        ```python
        import requests

        requests.post(
            url="https://tangibleai-mathtext.hf.space/run/text2int-preprocessed", json={"data": ["one hundred forty five"]}
        ).json()
        ```

        Or using `curl`:

        ```bash
        curl -X POST https://tangibleai-mathtext.hf.space/run/text2int-preprocessed -H 'Content-Type: application/json' -d '{"data": ["one hundred forty five"]}'
        ```
        {bq_json}""" + f"{json.loads(BQ_JSON)['type']}")

    with gr.Tab("Sentiment Analysis"):
        inputs_sentiment = [
            gr.Text(placeholder="Type a number as text or a sentence", label="Text to process",
                    value="I really like it!"),
        ]

        outputs_sentiment = gr.Textbox(label="Sentiment result")

        button_sentiment = gr.Button("sentiment analysis")

        button_sentiment.click(
            get_sentiment,
            inputs=inputs_sentiment,
            outputs=outputs_sentiment,
            api_name="sentiment-analysis"
        )

        examples_sentiment = [
            ["Totally agree!"],
            ["Sorry, I can not accept this!"],
        ]

        gr.Examples(examples=examples_sentiment, inputs=inputs_sentiment)

        gr.Markdown(r"""
    
        ## API
        ```python
        import requests
        
        requests.post(
            url="https://tangibleai-mathtext.hf.space/run/sentiment-analysis", json={"data": ["You are right!"]}
        ).json()
        ```
        
        Or using `curl`:
        
        ```bash
        curl -X POST https://tangibleai-mathtext.hf.space/run/sentiment-analysis -H 'Content-Type: application/json' -d '{"data": ["You are right!"]}'
        ```
        {bq_json}""" + f"{json.loads(BQ_JSON)['type']}")

# interface = gr.Interface(lambda x: x, inputs=["text"], outputs=["text"])
# html_block.input_components = interface.input_components
# html_block.output_components = interface.output_components
# html_block.examples = None

html_block.predict_durations = []

html_block.launch()