PDFminer 从每页 headers 获取字体大小（迭代）

Question

我对 python 和 PDFminer 很陌生，这对我来说有点复杂，我想要实现的是从 pdf 文件或幻灯片中提取每一页的标题。

我的方法是获取文本行的列表和每页的字体大小，然后我会选择最大的数字作为幻灯片标题，通常以较大的字体大小书写。

这是我目前所做的：

假设我想从此 pdf 文件中获取第 #8 页的标题。 File sample

这是第 8 页内容的样子：

这是获取所有页面每行字体大小的代码：

from pdfminer.high_level import extract_pages
from pdfminer.layout import LTTextContainer, LTChar,LTLine,LAParams
import os
path=r'cov.pdf'

Extract_Data=[]

for page_layout in extract_pages(path):
    for element in page_layout:
        if isinstance(element, LTTextContainer):
            for text_line in element:
                for character in text_line:
                    if isinstance(character, LTChar):
                        Font_size=character.size
            Extract_Data.append([Font_size,(element.get_text())])

生成的列表Extract_Data适用于pdf文档的所有页面。我的问题是如何为文档的每一页（迭代）获取此列表？

仅第 8 页的预期输出，依此类推/然后如果我想选择页面标题，它将是字体大小值最高的项目（行）：

[[32.039999999999964, 'Pandemic declaration \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0,
  '•  On March 11, 2020, the World Health Organization \n(WHO) characterized COVID-19 as a pandemic. \n \n•  It has caused severe illness and death. It features \n \nsustained person-to-person spread worldwide. \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, '•  It poses an especially high risk for the elderly (60 or \n \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0,
  'older), people with preexisting health conditions such \nas high blood pressure, heart disease, lung disease, \n  \ndiabetes, autoimmune disorders, and certain workers. \n \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [24.0, ' \n'],
 [14.04, '8 \n']]

Answer 1

完全公开，我是 pdfminer.six 的维护者之一。

执行此操作的 pythonic 方式如下。

import os

from pdfminer.high_level import extract_pages
from pdfminer.layout import LTTextContainer, LTChar


def get_font_sizes(paragraph: LTTextContainer):
    """Get the font sizes for every LTChar element in this LTTextContainer"""
    return [
        char.size
        for line in paragraph
        for char in line
        if isinstance(char, LTChar)
    ]


def list_sized_paragraphs(page):
    """List all the paragraphs and their maximum font size on this page"""
    return [
        (max(get_font_sizes(paragraph)), paragraph.get_text())
        for paragraph in page
        if isinstance(paragraph, LTTextContainer)
    ]


file_path = '~/Downloads/covid_19_training_tool_v3_01.05.2021_508.pdf'
for page in extract_pages(os.path.expanduser(file_path)):
    _, text = max(list_sized_paragraphs(page))
    print('---')
    print(text.strip())

第 8 页打印：

Pandemic declaration

注意：这不适用于所有页面，因为有时警告或注释的字体比 header 大。

PDFminer 从每页 headers 获取字体大小（迭代）

PDFminer get font size from headers per each page (iteration)

python

python-3.x

pdf

pdfminer