add a method on the Selector class, to export back the selector to css

[arthurdarcet]

My use case is the following:

• parse a selector group
• filter out some rules I don't want
• export the remaining selectors back to CSS

I have added a css method on the Selector class (and on each "sub"-selectors), to be able to export them back to CSS

[codecov-io]

Codecov Report

Merging #75 into master will increase coverage by 0.03%.
The diff coverage is 91.3%.

@@            Coverage Diff             @@
##           master      #75      +/-   ##
==========================================
+ Coverage   95.02%   95.05%   +0.03%     
==========================================
  Files           3        3              
  Lines         703      748      +45     
  Branches      114      124      +10     
==========================================
+ Hits          668      711      +43     
- Misses         20       21       +1     
- Partials       15       16       +1

Impacted Files	Coverage Δ
cssselect/xpath.py	94.86% <60%> (+0.03%)	✅
cssselect/parser.py	95.12% <95.12%> (+0.02%)	✅

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3036cba...08b0e47. Read the comment docs.

[redapple]

Hi @arthurdarcet , sorry for the delayed feedback.
I quite like the idea, makes sense to me.
I wonder about the .css() name though. I was thinking about .canonical() perhaps. No strong opinion though.

[redapple]

Is this fixing a bug? I don't quite get this changing to selector.value.value.

[arthurdarcet]

it's the symmetrical change for this: https://github.com/scrapy/cssselect/pull/75/files/08b0e472175197d4887612cd90d74815ab10d90a#diff-adc3ae8f2cf8b1931771d84ca7af6275R596

needed because i had to distinguish between different types of values here: https://github.com/scrapy/cssselect/pull/75/files/08b0e472175197d4887612cd90d74815ab10d90a#diff-adc3ae8f2cf8b1931771d84ca7af6275R651 (strings have to be serialised with quotes, number have to skip them, …)

I'm not sure I remember why I needed

[arthurdarcet]

Your call for the method name, you're right canonical is closer to what it does but it feels a bit too abstract to me. canonical_text or canonical_css?

[codecov]

Codecov Report

Merging #75 into master will increase coverage by 0.32%.
The diff coverage is 96.15%.

@@            Coverage Diff             @@
##           master      #75      +/-   ##
==========================================
+ Coverage   95.25%   95.57%   +0.32%     
==========================================
  Files           2        2              
  Lines         695      746      +51     
  Branches      113      125      +12     
==========================================
+ Hits          662      713      +51     
  Misses         19       19              
  Partials       14       14

Impacted Files	Coverage Δ
cssselect/parser.py	96.03% <100%> (+0.47%)	⬆️
cssselect/xpath.py	94.86% <60%> (+0.03%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update cb7a7e2...8d0ff3e. Read the comment docs.

[dangra]

@Gallaecio I think this could be useful and almost ready. Could you take a look? thanks

[Gallaecio]

It looks good to me. However, I wonder if, now that #87 has been merged, we need to take it into account here as well.

[Gallaecio]

Nevermind, :scope > * works just fine.

[kmike]

Thanks @arthurdarcet! And thanks @redapple, @dangra and @Gallaecio for reviews.