Added command line interface.

[rmax]

$ python -m parsel --help
usage: python -m parsel [-h] [--xpath] [--re PATTERN] [--encoding ENCODING]
                        [--repr] [-v]
                        EXPRESSION [FILE]

Parsel command line interface.

positional arguments:
  EXPRESSION           A CSSexpression, or a XPath expression if --xpath is
                       given.
  FILE                 If missing, it reads the HTML content from the standard
                       input.

optional arguments:
  -h, --help           show this help message and exit
  --xpath              Given expression is a XPath expression.
  --re PATTERN         Apply given regular expression.
  --encoding ENCODING  Input encoding. Default: utf-8.
  --repr               Output result object representation instead of as text.
  -v, --version        show program's version number and exit

$ curl http://scrapy.org/ | python -m parsel --xpath '//a/@href'  | sort | uniq -c
   1 http://scrapy.org
   1 ../community/
   2 ../companies/
   1 ../doc/
   1 ../download/
   1 ../support/
   1 http://codecov.io/github/scrapy/scrapy?branch=master
   1 http://doc.scrapy.org/en/1.1/intro/overview.html
   1 http://doc.scrapy.org/en/latest/faq.html
   1 http://doc.scrapy.org/en/latest/topics/ubuntu.html
   1 http://github.com/scrapy/scrapy/graphs/contributors
   1 http://pypi.python.org/pypi/Scrapy
   1 http://scrapinghub.com
   1 http://scrapinghub.com/scrapy-cloud/
   1 http://stackoverflow.com/tags/scrapy/info
   1 https://anaconda.org/scrapinghub/scrapy
   2 https://github.com/scrapy/scrapy
   1 https://github.com/scrapy/scrapy/archive/1.1.zip
   1 https://github.com/scrapy/scrapy/wiki/Python-3-Porting
   1 https://github.com/scrapy/scrapyd
   1 https://groups.google.com/forum/?fromgroups#!aboutgroup/scrapy-users
   3 https://pypi.python.org/pypi/Scrapy
   2 https://twitter.com/ScrapyProject

[codecov-io]

Current coverage is 98.29%

No coverage report found for master at fde9087.

Powered by Codecov. Last updated by fde9087...9043836

[rmax]

If #41 gets merged, I have a patch to add a --base-url option to return absolute links. This could lead to a nicer CLI interaction:

$ URL=https://sites.google.com/site/automl2016/accepted-papers; curl -q -s $URL | python -m parsel --base-url $URL 'a[href*=pdf]::attr(href)' | xargs -n1 wget
...
2016-06-23 15:23:39 (300 KB/s) - 'AutoML challenge_ system description of Lisheng Sun.pdf?attredirects=0&d=1' saved [212855/212855]
...
2016-06-23 15:23:41 (527 KB/s) - 'Bayesian optimization for automated model selection.pdf?attredirects=0&d=1' saved [454152/454152]
...

[eliasdorneles]

Wouldn't this be better in a separate project?

It would be easier to maintain and could have more dependencies (click, tabulate, colorama, etc).

[rmax]

@eliasdorneles indeed, I aimed for the simpler cli but having a separate project would lead to more useful features. Is parsel-cli a good name?

[eliasdorneles]

Yea, looks pretty cool, I thought it deserves its own project. :)

parsel-cli is a decent name, +1
(nothing you can't change later, also)

[rmax]

Moving to its own project 🎉