pyhtmd 0.1.2

A Python HTML to Markdown parser

一款Python版本的HTML转markdown解析器，不使用任何第三方工具

install

pip install pyhtmd

usage

from pyhtmd import Pyhtmd
html="<code> Hello, world ! by Pyhtmd. </code>"
md= Pyhtmd(html)
content=md.markdown()
print(content) # `Hello, world ! by Pyhtmd.`

API

Pyhtmd(html, language="", img=True )

• language：类型 string （js、python、java等）
• img:{Boolean}，默认True，可以不需要img渲染

from pyhtmd import Pyhtmd
html="<pre><code>import time\n print(time.time()) </code><pre>"
md= Pyhtmd(html,language="python")
content=md.markdown()
print(content) # `Hello, world ! by Pyhtmd.`

todo 开发中

• 递归太多被打断了

demo1

Given a tensor <code translate="no" dir="ltr">t</code>, this operation returns a tensor of the same type andshape as <code translate="no" dir="ltr">t</code> with its values clipped to <code translate="no" dir="ltr">clip_value_min</code> and <code translate="no" dir="ltr">clip_value_max</code>.Any values less than  <code translate="no" dir="ltr">clip_value_min</code> are set to <code translate="no" dir="ltr">clip_value_min</code>. Any valuesgreater than <code translate="no" dir="ltr">clip_value_max</code> are  set to <code translate="no" dir="ltr">clip_value_max</code>. 

---

Given a tensor t, this operation returns a tensor of the same type andshape as t with its values clipped to clip_value_min and clip_value_max.Any values less than clip_value_min are set to clip_value_min. Any valuesgreater than clip_value_max are set to clip_value_max.

demo2

<strong>Note:</strong><span> <code translate="no" dir="ltr">clip_value_min</code> needs to be smaller or equal to <code translate="no" dir="ltr">clip_value_max</code> forcorrect results.</span>

---

Note: clip_value_min needs to be smaller or equal to clip_value_max forcorrect results.

demo3

<h4 id="for_example" is-upgraded="">For example:</h4>

---

For example:

demo4：todo ，换行的字符不是特别好

k = '<pre class="prettyprint lang-python" translate="no" dir="ltr" is-upgraded=""><code translate="no" dir="ltr">A = tf.constant([[1, 20, 13], [3, 21, 13]])B = tf.clip_by_value(A, clip_value_min=0, clip_value_max=3) # [[1, 3, 3],[3, 3, 3]]C = tf.clip_by_value(A, clip_value_min=0., clip_value_max=3.) # throws `TypeError`as input and clip_values are of different dtype</code></pre>'

---

A = tf.constant([[1, 20, 13], [3, 21, 13]])B = tf.clip_by_value(A, clip_value_min=0, clip_value_max=3) # [[1, 3, 3],[3, 3, 3]]C = tf.clip_by_value(A, clip_value_min=0., clip_value_max=3.) # throws `TypeError`as input and clip_values are of different dtype

demo5:

<li><b><code translate="no" dir="ltr">t</code></b>: A <code translate="no" dir="ltr">Tensor</code> or <code translate="no" dir="ltr">IndexedSlices</code>.</li><li><b><code translate="no" dir="ltr">clip_value_min</code></b>: A 0-D (scalar) <code translate="no" dir="ltr">Tensor</code>, or a <code translate="no" dir="ltr">Tensor</code> with the same shapeas <code translate="no" dir="ltr">t</code>. The minimum value to clip by.</li><li><b><code translate="no" dir="ltr">clip_value_max</code></b>: A 0-D (scalar) <code translate="no" dir="ltr">Tensor</code>, or a <code translate="no" dir="ltr">Tensor</code> with the same shapeas <code translate="no" dir="ltr">t</code>. The maximum value to clip by.</li><li><b><code translate="no" dir="ltr">name</code></b>: A name for the operation (optional).</li>

---

• t: A Tensor or IndexedSlices.
• clip_value_min: A 0-D (scalar) Tensor, or a Tensor with the same shapeas t. The minimum value to clip by.
• clip_value_max: A 0-D (scalar) Tensor, or a Tensor with the same shapeas t. The maximum value to clip by.
• name: A name for the operation (optional).

demo6:

<h4 id="raises" is-upgraded="">
    Raises:
    <button role="button" class="devsite-heading-link button-flat material-icons" title="Copy link to this section">
    </button>
</h4> 

---

Raises:

demo7:

<li>
	<b> <code translate="no" dir="ltr">ValueError</code> </b> : If the clip tensors would trigger array broadcastingthat would make the returned tensor larger than the input.
</li>

<li>
	<b><code translate="no" dir="ltr">TypeError</code></b>: If dtype of the input is <code translate="no" dir="ltr">int32</code> and dtype of the <code translate="no" dir="ltr">clip_value_min or</code> clip_value_max <code translate="no" dir="ltr">is</code> float32
</li> 

---

• ValueError: If the clip tensors would trigger array broadcastingthat would make the returned tensor larger than the input.
• TypeError: If dtype of the input is int32 and dtype ofthe clip_value_min or clip_value_max is float32

demo8:

<a href="/api_docs/python/tf/clip_by_value"><code>tf.compat.v2.clip_by_value</code></a>

---

• tf.compat.v2.clip_by_value

demo9:

<img src="https://www.baidu.com/img/bd_logo1.png">
<img src="https://www.baidu.com/img/bd_logo1.png" alt="百度logo">

---

吐槽

这正则会引发： maximum recursion depth exceeded in comparison 什么毛病，其他就没有！！！！ 问题：由于一个正则替换引发的无限递归

def remove_attrs(block):
  content=block
  # remove_h1 = re.sub(r'<h1(.*?)">', '<h1>', content)
  # remove_h2 = re.sub(r'<h2(.*?)">', '<h2>', remove_h1)
  # remove_h3 = re.sub(r'<h3(.*?)">', '<h3>', remove_h2)
  # remove_h4 = re.sub(r'<h4(.*?)">', '<h4>', remove_h3)
  # remove_h5 = re.sub(r'<h5(.*?)">', '<h5>', remove_h4)
  # remove_h6 = re.sub(r'<h6(.*?)">', '<h6>', remove_h5)
  # remove_code = re.sub(r'<code(.*?)">', '<code>', remove_h6)
  # remove_span = re.sub(r'<span(.*?)">', '<span>', remove_code)
  remove_b = re.sub(r'<b(.*?)">', '<b>', content)
  # remove_button = re.sub(r'<button(.*?)">', '<button>', content) # 这个就没报错，很奇怪
  # remove_div = re.sub(r'<div(.*?)">', '<div>', content)
  # remove_a = re.sub(r'<a(.*?)">', '<a>', remove_div)
  return remove_b

# todo table
import re
from pyhtmd.core import Pyhtmd
array=[
'<h3 id="aliases" is-upgraded="">Aliases:<button role="button" class="devsite-heading-link button-flat material-icons" title="Copy link to this section"></button></h3>'
]
for item in array:
  mk=Pyhtmd(item).markdown()
  print('===========================')
  print(mk)
  print('===========================')