Revert

2026-06-05 23:20:19 +00:00 · 2016-02-13 10:28:20 -05:00
parent e643f20d7c
commit e240f4dcf2
6 changed files with 37 additions and 138 deletions
@@ -3,28 +3,18 @@
 This is a very simple web service that will take a given URL, and return
 a Markdown representation of that page.

-Powered by [Readability](http://readability.com/),
-[Requests](http://python-guide.org/),
-[html2text](http://www.aaronsw.com/2002/html2text/),
-[markdown](http://pythonhosted.org/Markdown/),
-and [Flask](http://flask.pocoo.org/).
+Powered by [Readability](http://readability.com/), [Requests](http://python-guide.org/), [html2text](http://www.aaronsw.com/2002/html2text/), and [Flask](http://flask.pocoo.org/).

 ## Usage


    $ curl http://url2markdown.herokuapp.com/?url=http://kennethreitz.org
-
+    
    # Hi, there.
-
+    
    My name is Kenneth Reitz.
    ...

-Or, if you understand code:
-
-    $ mkvirtualenv url2markdown
-    $ pip install -r requirements.txt
-    $ READABILITY_TOKEN="XXX" python service.py
-
 Enjoy!

 ## Configuration
@@ -38,4 +28,3 @@ You can use [autoenv](https://github.com/kennethreitz/autoenv) to do this easily
 ## License

 Unfortunately, this code is released under [GPLv3](http://www.gnu.org/copyleft/gpl.html).
-
@@ -1,44 +1,36 @@
 # -*- coding: utf-8 -*-

+
+
+
 import os

 import requests
 from html2text import html2text

+READABILITY_URL = 'https://www.readability.com/api/content/v1/parser'

-_READABILITY_URL = 'https://www.readability.com/api/content/v1/parser'
-
-
-def _get_readability_html_and_title(url):
+def readability(url):
    token = os.environ.get('READABILITY_TOKEN')
    params = {'url': url, 'token': token}

-    r = requests.get(_READABILITY_URL, params=params)
-    decoded_content = (
-        r.json()['content'],
-        r.json()['title'],
-    )
-    return decoded_content
+    r = requests.get(READABILITY_URL, params=params)
+    return r.json()['content'], r.json()['title']

-
-def _convert_html_to_markdown(html, title=None):
+def convert(html, title=None):
    if title:
        title = '# {}'.format(title)
        html = '\n\n'.join([title, html])

-    text_from_html = html2text(html)
-    return text_from_html
+    return html2text(html)

-
-def get_readable_content_from_url(url):
+def meh(url):
    try:
-        content, title = _get_readability_html_and_title(url)
-        markdown = _convert_html_to_markdown(content, title=title)
-        return markdown
+        content, title = readability(url)
+        return convert(content, title=title)
    except KeyError:
        return None


 if __name__ == '__main__':
-    print get_readable_content_from_url('http://kennethreitz.org/')
-
+    print meh('http://kennethreitz.org/')
@@ -8,4 +8,3 @@ html2text==3.200.3
 itsdangerous==0.23
 requests==2.0.0
 wsgiref==0.1.2
-markdown==2.3.1
@@ -1,40 +1,19 @@
 # -*- coding: utf-8 -*-

-from flask import Flask, request, render_template, Markup
-from converter import get_readable_content_from_url
-from markdown import markdown
-
+from flask import Flask, request, redirect, url_for, render_template
+from converter import meh

 app = Flask(__name__)

-
-def _markdown_to_html(text):
-    return Markup(markdown(text))
-
-
@app.route('/')
 def fuck_gpl3():
    url = request.args.get('url')
-    type = request.args.get('type', 'markdown')
-
-    content = get_readable_content_from_url(url)

    if url:
-        if not content:
-            return '404 Not Found', 404
-
-        if type == 'html':
-            markdown_url_contents = _markdown_to_html(content)
-            return render_template(
-                'index.html',
-                converted_url_contents=markdown_url_contents,
-                page_url=url,
-            )
-        else:
+        content = meh(url)
+        if content:
            return content, 200, {'Content-Type': 'text/x-markdown; charset=UTF-8'}
+        else:
+            return '404 Not Found', 404
    else:
        return render_template('index.html')
-
-
-if __name__ == '__main__':
-    app.run()
@@ -1,72 +1,21 @@
 <!DOCTYPE html>
 <html>
-    <head>
-        <title>Markdown Please!</title>
-        <link
-            href="{{ url_for('static', filename='css/bootstrap.min.css' ) }}"
-            rel="stylesheet"
-        />
-    </head>
-    <body>
-        <nav class="navbar navbar-default navbar-fixed-top">
-            <div class="container">
-                <ul class="nav">
-                <a class="navbar-brand" href="/">Markdown Please!</a>
+<head>
+    <title>url2markdown</title>
+</head>
+<body>
+<h1>url2markdown</h1>
+<p>This is a very simple web service that will take a given URL, and return
+a Markdown representation of that page.</p>

-                <form action="/" method="get" class="navbar-form navbar-left">
-                    <fieldset>
-                        <input
-                            type="text"
-                            name="url"
-                            class="form-control"
-                            style="width: 300px;"
-                            placeholder="http://en.wikipedia.org/wiki/Markdown"
-                            {% if page_url %}
-                                value="{{ page_url }}"
-                            {% endif %}
-                        />
-                        <input type="hidden" name="type" value="html" />
-                        <button type="submit" class="btn btn-default">
-                            Go
-                        </button>
-                    </fieldset>
-                </form>
+<form action="/" method="get">
+    URL: <input type="text" name="url">
+    <button type='submit'>Submit</button>
+</form>

-                {% if page_url %}
-                    <a
-                        href="/?url={{ page_url }}"
-                        class="btn btn-default navbar-btn navbar-right"
-                    >
-                        Get the Markdown
-                    </a>
-                {% endif %}
-                </ul>
-            </div>
-        </nav>
+<p>A <a href='http://kennethreitz.org/projects/'>Kenneth Reitz</a> project.</p>

-        <div class="container">
-            <div style="padding-top:40px;padding-bottom:80px;">
-                {% if page_url %}
-                    <div>
-                        {{ converted_url_contents }}
-                    </div>
-                {% endif %}
+<a href="https://github.com/kennethreitz/url2markdown"><img style="position: absolute; top: 0; right: 0; border: 0;" src="https://s3.amazonaws.com/github/ribbons/forkme_right_darkblue_121621.png" alt="Fork me on GitHub"></a>

-                <a href="https://github.com/kennethreitz/url2markdown">
-                    <img
-                        style="position: absolute; top: 0; right: 0; border: 0; z-index: 10000;"
-                        src="https://s3.amazonaws.com/github/ribbons/forkme_right_darkblue_121621.png"
-                        alt="Fork me on GitHub"
-                    />
-                </a>
-            </div>
-
-        </div>
-        <footer class="navbar navbar-fixed-bottom panel-footer">
-            <p class="container">
-                A <a href="http://kennethreitz.org/projects/">Kenneth Reitz</a>/
-                <a href="http://lumbercoder.com/">Gil Goncalves</a> project.
-            </p>
-        </footer>
-    </body>
-</html>
+</body>
+</html>