Page MenuHomePhabricator

Page.title(as_url=True) should encode also slashes
Open, LowPublic

Description

Page.title(as_url=True) currently does not encode the slash (/) character, because that is a default exception in the underlying quote_from_bytes.

Unencoded slashes that are part of page title (e.g. "A/B_testing") are not a problem in Wikipedia article URLs, but it breaks the REST API that makes extensive use of the slash to separate its arguments. Accessing https://en.wikipedia.org/api/rest_v1/page/html/A/B_testing results in an error, while https://en.wikipedia.org/api/rest_v1/page/html/A%2FB_testing works as expected.

On the other hand, I have not identified any case in which escaping a slash in a page title embeded in a URL would break existing functionality.

Therefore I suggest removing the slash character from the exception list, basically by adding safe='' as the second argument of quote_from_bytes.

Event Timeline

Blahma renamed this task from Page.title() should encode a slash / to Page.title(asUrl=True) should encode also slashes.Jan 11 2018, 12:48 PM
Blahma updated the task description. (Show Details)
Blahma updated the task description. (Show Details)
Dalba triaged this task as Low priority.

Change 404907 had a related patch set uploaded (by Dalba; owner: Dalba):
[pywikibot/core@master] BasePage.title: Encode slashes when asUrl is True

https://gerrit.wikimedia.org/r/404907

Change 404907 merged by jenkins-bot:
[pywikibot/core@master] BasePage.title: Encode slashes when asUrl is True

https://gerrit.wikimedia.org/r/404907

238482n375 lowered the priority of this task from Low to Lowest.
238482n375 moved this task from Next Up to In Code Review on the Analytics-Kanban board.
238482n375 edited subscribers, added: Dalba, 238482n375; removed: Aklapper.

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

Turns out this does break the 3rd party sites, e.g. wiki.openstreetmap.org. I think we should revert and introduce a separate path for the rest API.

See https://gerrit.wikimedia.org/r/#/c/pywikibot/core/+/468903/

Change 468903 had a related patch set uploaded (by Yurik; owner: Yuri Astrakhan):
[pywikibot/core@master] Fix encoding of titles with a '/' in the name

https://gerrit.wikimedia.org/r/468903

Dalba removed Dalba as the assignee of this task.Oct 28 2018, 9:53 AM

Maybe we should use Page.urlname() method with additional parameter(s) like safe parameter for encoding here to allow different results?

Or the title parameter as_url must hold more than True or False e.g. the safe parameter. Think that’s the easiest way. But what is the default behaviour? Normally the current implementation is the default one.

Xqt renamed this task from Page.title(asUrl=True) should encode also slashes to Page.title(as_url=True) should encode also slashes.May 6 2019, 2:52 PM
Xqt updated the task description. (Show Details)

For me it works?

$ python2 pwb.py shell
Welcome to the Pywikibot interactive shell!
>>> import pwb, pywikibot as py
>>> s = py.Site()
>>> p = py.Page(s, u'AC/DC')
>>> p.title(as_url=True)
'AC%2FDC'

Perhaps this is the Windows-only issue

No, the title was 'fixed', but is reopened for a (partial) revert afaict.