iherman / respec2epub Goto Github PK

View Code? Open in Web Editor NEW

4.0 4.0 1.0 2.99 MB

Python based service to convert a respec document into EPUB3

License: Other

Python 86.26% HTML 11.62% CSS 2.12%

respec2epub's People

Contributors

Stargazers

Watchers

Forkers

gitter-badger

respec2epub's Issues

Solving the file handling

The usage of os.path join and split may be a good approach to solve the current issue when producing a folder

HTTP CORS?

Hello Ivan,
first of all: cool project :)

I was playing with the demo site:
https://labs.w3.org/epub-generator/
...using this Respec URL:
http://w3c.github.io/dpub-accessibility/
(works great)

The HTTP GET URL works fine too:
https://labs.w3.org/epub-generator/cgi-bin/epub-generator.py?type=html&url=http%3A%2F%2Fw3c.github.io%2Fdpub-accessibility
...but unfortunately there are no CORS headers included in the HTTP response, so I cannot fetch the application/epub+zip payload from a web browser (that is to say, from XmlHttpRequest).

Here's the list of HTTP headers returned by a curl HEAD request:

$ curl --head "https://labs.w3.org/epub-generator/cgi-bin/epub-generator.py?type=html&url=http%3A%2F%2Fw3c.github.io%2Fdpub-accessibility"

HTTP/1.1 200 OK
Date: Thu, 10 Mar 2016 16:05:00 GMT
Server: Apache/2.2.22 (Debian)
Expires: Thu, 10 Mar 2016 16:05:00 GMT
Content-Disposition: attachment; filename=index.epub
Last-Modified: Thu, 10 Mar 2016 16:05:00 GMT
Content-Length: 145203
Content-Type: application/epub+zip
Strict-Transport-Security: max-age=15552015; preload
Public-Key-Pins: pin-sha256="cN0QSpPIkuwpT6iP2YjEo1bEwGpH/yiUn6yhdy+HNto="; pin-sha256="WGJkyYjx1QMdMe0UqlyOKXtydPDVrk7sl2fV+nNm1r4="; pin-sha256="LrKdTxZLRTvyHM4/atX2nquX9BeHRZMCxg3cf4rhc2I="; max-age=864000

So, I tried using a CORS proxy, unfortunately to no avail:
https://crossorigin.me/https://labs.w3.org/epub-generator/cgi-bin/epub-generator.py?type=html&url=http%3A%2F%2Fw3c.github.io%2Fdpub-accessibility

$ curl --head "https://crossorigin.me/https://labs.w3.org/epub-generator/cgi-bin/epub-generator.py?type=html&url=http%3A%2F%2Fw3c.github.io%2Fdpub-accessibility"

HTTP/2.0 500
server:cloudflare-nginx
date:Thu, 10 Mar 2016 16:07:56 GMT
content-type:text/html; charset=iso-8859-1
set-cookie:__cfduid=d2232775d992c99ef835d9fcfc48f8dd61457626075; expires=Fri, 10-Mar-17 16:07:55 GMT; path=/; domain=.crossorigin.me; HttpOnly
access-control-allow-origin:*
access-control-allow-credentials:false
access-control-allow-headers:Content-Type
expires:Fri, 11 Mar 2016 16:07:55 GMT
cf-ray:281805be638334fa-LHR

There are alternative CORS proxies, such as:

http://cors.io/?u=https%3A%2F%2Flabs.w3.org%2Fepub-generator%2Fcgi-bin%2Fepub-generator.py%3Ftype%3Dhtml%26url%3Dhttp%253A%252F%252Fw3c.github.io%252Fdpub-accessibility
...but unfortunately in this case the response content-type is incorrect.

So ideally, could https://labs.w3.org/epub-generator/cgi-bin/epub-generator.py itself be updated to emit CORS headers?

PS: I am just toying with the idea of using Readium to read ReSpec EPUBs directly from your HTTP URLs :)

Thanks!

Cover page and base.css

The cover page refers to base.css in Assets. That has to change to be adapted to the new set up with css files

Cannot generate EPUB for a ReSpec document

I was trying to export an EPUB version of https://w3c.github.io/jlreq/ , but got the following error:

Traceback (most recent call last):
  File "/home/ivan/htdocs/cgi-bin/epub-generator.py", line 189, in 
    Generator().process()
  File "/home/ivan/htdocs/cgi-bin/epub-generator.py", line 175, in process
    wrapper = self.generate_ebook()
  File "/home/ivan/htdocs/cgi-bin/epub-generator.py", line 163, in generate_ebook
    logger=logger
  File "/home/ivan/lib/python/rp2epub/doc2epub.py", line 84, in __init__
    session = HttpSession(CONVERTER + url if is_respec else url, raise_exception=True, is_respec=is_respec)
  File "/home/ivan/lib/python/rp2epub/utils.py", line 514, in __init__
    handle_exception("There seems to be a problem with the spec generator service ('%s' should be tested separately)" % url)
  File "/home/ivan/lib/python/rp2epub/utils.py", line 502, in handle_exception
    raise R2EError(message)
R2EError: "There seems to be a problem with the spec generator service ('https://labs.w3.org/spec-generator/?type=respec&url=https://w3c.github.io/jlreq/' should be tested separately)"

See https://labs.w3.org/epub-generator/cgi-bin/epub-generator.py?type=respec&url=https%3A%2F%2Fw3c.github.io%2Fjlreq%2F for the error.

I'm not sure it is an issue of spec generator or epub generator, so filing an issue here first.

The w3c_home logo is still generated

And relies on the exitence in Asset (whilst it is not there any more). This is in the header

Margin is still necessary...

The margin has disappeared with the new setup, but that may not work; it seems that iBook still needs it (like in the old days:-(

W3C manifest spec throws

Trying to gen from http://w3c.github.io/manifest/ I get:

Epub Generator Exception

Traceback (most recent call last):
  File "/home/ivan/htdocs/cgi-bin/epub-generator.py", line 176, in 
    Generator().process()
  File "/home/ivan/htdocs/cgi-bin/epub-generator.py", line 162, in process
    wrapper = self.generate_ebook()
  File "/home/ivan/htdocs/cgi-bin/epub-generator.py", line 150, in generate_ebook
    logger=logger
  File "/home/ivan/lib/python/rp2epub/doc2epub.py", line 79, in __init__
    session = HttpSession(CONVERTER + url if is_respec else url, raise_exception=True)
  File "/home/ivan/lib/python/rp2epub/utils.py", line 386, in __init__
    handle_exception("%s cannot be reached" % url)
  File "/home/ivan/lib/python/rp2epub/utils.py", line 375, in handle_exception
    raise R2EError(message)
R2EError: https://labs.w3.org/spec-generator/?type=respec&url=http://w3c.github.io/manifest/ cannot be reached

Drastically reduce the config files

For the new setup with CSS the config files should be drastically reduced

Provide option to remove document status like "Recommendation"

My Tolino ebook reader seems to be having a problem with the status of the document on the left side like "Recommendation", it either distorts the text or overlays part of it so it cannot be read anymore. It would be nice if there is an option to remove this in the epub output.

<item href="annotation-protocol/ https://w3c.github.io/test-results/annotation-protocol/all.xhtml" id="annotation-protocol- https:--w3c-github-io-test-results-annotation-protocol-all-xhtml" media-type="application/xhtml+xml"/>