Question

我希望能够基于输入搜索的结果返回填充的破折号表。到目前为止，我已经尝试了2种方法-在回调输出中返回整个DashTable，并在回调中分别返回列和数据。两种选择都不适用于我。我已经包含了每个选项的相关代码以及每个选项所产生的错误消息：

分别返回数据和列：

        @app.callback(
        [Output('table', 'data'),
        Output('table', 'columns')],
        [Input("button", "n_clicks")], state=[State('url', 'value')])
    def update_table(n_click:int, url):
        if n_click>1:
            summary, table = summarizer(url)
            columns=[{"name": i, "id": i, "deletable": True, "selectable": True} for i in table.columns]
            table = table.to_dict('records')
            return table, columns
        else: 
            return [], []

app.layout包含以下行

     html.Div(dt.DataTable(id='table'))

由此导致的错误消息是：

对象不能作为React子对象

第二种方法是通过回调传递整个DataTable并仅使用html.Div在这样的布局中显示它

@app.callback(
    Output('table', 'children'),
    [Input("button", "n_clicks")], state=[State('url', 'value')])
def update_table(n_click:int, url):
    if n_click>1:
        summary, table = summarizer(url)
        columns=[{"name": i, "id": i, "deletable": True, "selectable": True} for i in table.columns]
        table = table.to_dict('records')
        return dt.DataTable(data=table, columns=columns)
    else: 
        return []

html.Div(id='table')

相应的错误是

[Objects are not valid as a React child][2]

这个错误令我感到困惑，因为它似乎与列定义有关，但是我无法传递数组，并且文档要求提供字典。

完整代码示例：

import dash
import dash_core_components as dcc
import dash_html_components as html
import dash_bootstrap_components as dbc
import dash_table as dt
from dash.dependencies import Input, Output, State
import sd_material_ui
from newspaper import Article
import gensim
from gensim.summarization import summarize
from dash.exceptions import PreventUpdate
from newspaper import fulltext
import requests
import pandas as pd
import yake
import nltk
from newsapi import NewsApiClient


leftSources = ["cnn", "buzzfeed", "the-washington-post", "bbc-news", "vice-news", "newsweek", "techcrunch", "reuters", "politico", "newsweek", "msnbc"]
rightSources = ["fox-news", "national-review", "new-york-magazine", "breitbart-news", "business-insider", "the-wall-street-journal", "bloomberg", "the-washington-times", "the-hill", "the-american-conservative"]

# importing CSS
external_stylesheets = ['https://codepen.io/chriddyp/pen/bWLwgP.css']

# similarArticleURL
getSimilarArticlesURL = "https://us-central1-secure-site-266302.cloudfunctions.net/getSimilarArticles?keywords="
getKeywordsURL = "https://us-central1-secure-site-266302.cloudfunctions.net/getKeyword?text="
getArticleTextURL = "https://us-central1-secure-site-266302.cloudfunctions.net/getArticleText?url="

allData = pd.DataFrame()


# instantiating dash application
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
server = app.server # the flask app

# helper functions
def generate_table(dataframe, max_rows=10):
    return html.Table([
        html.Thead(
            html.Tr([html.Th(col) for col in dataframe.columns])
        ),
        html.Tbody([
            html.Tr([
                html.Td(dataframe.iloc[i][col]) for col in dataframe.columns
            ]) for i in range(min(len(dataframe), max_rows))
        ])
    ])

app.layout = html.Div([

    html.Div(html.H3("Brief.Me"), style={'font-weight':'bold','background-color':'darkorange', 'color':'white','text-align':'center'}),

    html.Br(),
    html.Br(),

    dbc.Row([

        dbc.Col(dbc.Input(id='url', type='url', size=30, placeholder="Type or copy/paste an URL"), width={'size':6, 'order':1, 'offset':3}),
        dbc.Col(dbc.Button("Summarize", id='button', n_clicks=1, color="primary", className="mr-1"), width={'order':2})

        ]),

    html.Br(),

    # dbc.Row([

    #     dbc.Col(dcc.Loading(html.Div(html.Div(id="summary"), style={'font-weight':'bold'})), width={'size':6, 'offset':3})

    # ]),

    html.Div(id='table')

    ],
)

def fetch_similar_articles(keyword):

    tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')

    newsapi = NewsApiClient(api_key='ce7482cbd40f4d90a8eea404e7702db6')

    top_headlines = newsapi.get_top_headlines(q=keyword,
                                              sources='bbc-news,the-wall-street-journal,the-washington-post,fox-news,bloomberg, vice-news, politico, reuters, the-hill',
                                              language='en')

    return top_headlines["articles"]

def fetch_article_text(url):
  try:
    article = Article(url)
    article.download()
    article.parse()
    return article.text
  except:
    return None

def summarizer(url):
    global allData

    leftSummaries, rightSummaries = {}, {}
    text = fetch_article_text(url)
    main_summary = summarize(text)

    keywords = extract_keywords(text)
    urls = []

    rightData, leftData, allData = get_articles_content(keywords)
    rightDf, leftDf = pd.DataFrame(rightData), pd.DataFrame(leftData)
    allSources = pd.concat([rightDf, leftDf], axis=1)

    return main_summary, allData

def get_articles_content(keywords):
    '''
    This function will return a row of the dataframe where there is a title, source, url and summary. 
    '''
    allResults, leftRows, rightRows = [], [], []

    for keyword in keywords:

        articleList = fetch_similar_articles(keyword)

        for elem in articleList:
            source = elem['source']
            url = elem['url']
            title = elem['title']
            text = fetch_article_text(url)

            if text is not None and len(text) > 1:
                summary = summarize(text)

                allResults.append({'title': title, 'url': url,'source': source, 'summary': summary})
                if source in leftSources:
                    leftRows.append(pd.DataFrame({'title': title, 'url': url,'source': source, 'summary': summary}))
                elif source in rightSources:
                    rightRows.append(pd.DataFrame({'title': title, 'url': url, 'source': source, 'summary': summary}))

    allResults = pd.DataFrame(allResults)
    return leftRows, rightRows, allResults

def extract_keywords_yake(text, phrase_length, num_keywords):

  custom_kw_extractor = yake.KeywordExtractor(n=phrase_length, top=num_keywords)
  keywords = custom_kw_extractor.extract_keywords(text)

  return keywords

def extract_keywords(text):
    '''
    Returns a list of keywords given the article text. 
    '''
    global getKeywordsURL

    getKeywordsURL += text
    keywordRes = extract_keywords_yake(text, 2, 5)
    keywords = []

    for pair in keywordRes:
        keywords.append(pair[1])

    return keywords


@app.callback( # Output('summary', 'children')
    Output('table', 'children'),
    [Input("button", "n_clicks")], state=[State('url', 'value')])
def update_table(n_click:int, url):
    if n_click>1:
        summary, table = summarizer(url)
        columns=[{"name": i, "id": i, "deletable": True, "selectable": True} for i in table.columns]
        table = table.to_dict('records')
        return dt.DataTable(data=table, columns=columns)
    else: 
        return [], []

if __name__ == '__main__':
    app.run_server(debug=True, host='0.0.0.0', port=8080)

通过回调显示破折号数据表

0 个答案: