nerdery
The goal of polished is to show all of the meticulous tweaks that go into a website. My resume is a good example, dozens of hours of work and tweaking to come up with this pretty basic final product. Showing that blood, sweat and hilarious tears in between should be pretty entertaining. Watch pages undulate, stretch, break, grow, and shrink into place.
Once you've installed polished, it works like this:
@polish out the kinks so it's a nice smooth videoResume page video without polishing
Polished resume page video
And I had to "polish" these videos to get them just right, fix bad links for some commits:
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from polished.backends import PelicanBackend
from polished.decorators import polish
class EricPelicanBackend(PelicanBackend):
def _patch_image_srcs(self):
wait = WebDriverWait(self.DRIVER, 10)
element = wait.until(EC.visibility_of_element_located((By.TAG_NAME, 'img')))
self.DRIVER.execute_script("""
var img_array = document.getElementsByTagName('img');
for(var i=0; i<img_array.length; i++) {
var href_replaced = img_array[i].getAttribute('src').replace(/^\/images/, "../images");
img_array[i].setAttribute("src", href_replaced);
}
""")
@polish(urls=["output/pages/about.html"], commit_indexes=range(112, 135))
def fix_image_links_on_about_me_page(self):
self._patch_image_srcs()
@polish(urls=["output/pages/resume.html"], commit_indexes=range(68,134))
def fix_resume_page_broken_images(self):
self._patch_image_srcs()
Requires:
Then the usual
> pip install polished
For more detailed instructions please check out the repo readme.
> polished
The default behavior is to capture "index.html" each commit
> polished output/index.html
Local file
> polished http://localhost:8000/
Local server
By default the files are saved to polished_output/<commit count>.<sha>.polished.png and polished_output/output.mp4
I've been looking make a little python package to launch on pypi for quite some time. With all of my recent Pelican blog tweaking, I am having about 50 urges per second to improve the development a little bit.
One of the things that keeps irking me while writing posts was constantly having to double check that I put in the correct value for each link in markdown
So there I was, at [Some Restaurant I can't Remember]() downtown and I totally saw Becky macking on Jonathon!
Right as I am getting into the juicy gossip I normally would blog about, now I have worries:
My first idea was to make some kind of hook in the Pelican system that re-wrote [text](url) links where the url was
empty with the first google search result of the text description.
That idea sucks for so many reasons:
After playing around with a few different ideas I finally thought: screw this. I was thinking too hard, all of the solutions were too convoluted. When I stepped away from the computer and thought for a while I realized the problem I was really trying to fix: broken links!
After some searching I didn't find a nice simple python module that ran through static html files, tried links, and spit out which ones were bad.
I want this to run quickly or I'll never use it, so my plan is to use requests asychronously after scanning all of the
files for broken links. However, after trying my damndest I couldn't get grequests to work right... I was having weird
errors in the background I couldn't debug easily.
Then I saw this nice little example, here's my version:
def async_check_url(url, file_name, line_number):
try:
urllib2.urlopen(url)
except urllib2.URLError:
BROKEN_URLS.append((url, file_name, line_number))
def check_urls(urls):
'''
expected format of urls is list of tuples (url, file name, source line) i.e. ("google.com", "index.html", 32)
'''
threads = list()
for u in urls:
t = Thread(target=async_check_url, args=(u[0], u[1], u[2]))
t.start()
threads.append(t)
for thread in threads:
thread.join()
.pypirc in home directory with login info[distutils]
index-servers = pypi
[pypi]
username:ckcollab
password:hunter2
import os
from setuptools import setup
try:
with open('README.md') as readme:
long_description = readme.read()
except IOError, ImportError:
long_description = ''
setup(
install_requires = [
"lxml>=3.3.4",
"cssselect>=0.9.1"
],
name="existence",
py_modules=["existence"],
version="0.0.8",
author="Eric Carmichael",
author_email="[email protected]",
description="Checks static .html files for bad links",
long_description=long_description,
license="MIT",
keywords="link checker",
url="https://github.com/ckcollab/existence",
classifiers=[
'Intended Audience :: Developers',
'Natural Language :: English',
'License :: OSI Approved :: MIT License',
'Operating System :: OS Independent',
'Programming Language :: Python',
'Programming Language :: Python :: 2',
'Programming Language :: Python :: 2.6',
'Programming Language :: Python :: 2.7',
'Topic :: Software Development :: Libraries :: Python Modules',
],
)
> python setup.py register
> python setup.py sdist upload
import os
from existence import get_bad_urls
from fabric.api import *
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
def clean():
if os.path.isdir(DEPLOY_PATH):
local('rm -rf {deploy_path}'.format(**env))
local('mkdir {deploy_path}'.format(**env))
def deploy():
clean()
local("make html")
print "Checking URLs"
bad_urls = get_bad_urls(DEPLOY_PATH)
if not bad_urls:
print "URL's are looking good"
local("git push")
local("git push heroku master")
else:
for url in bad_urls:
print "Broken link found in file %s on line %s linking to %s" % (url[1], url[2], url[0])
Now I just run
> fab deploy
and I will never ever, ever have to worry about broken links in my blog! Thanks existence!