[Confluence] Fix pagination for get_all_* methods and unify _get_paged across Cloud/Server

[Zircoz]

Summary

Fixes #1598

get_all_pages_from_space and related get_all_* methods only returned the first page of results because they called self.get() directly instead of self._get_paged(). Additionally, pagination broke when the Confluence API returned _links.next as a plain URL string rather than a {"href": "..."} dict.

---

Changes

Pagination fix (core — #1598)

• Switch get_all_* methods to use _get_paged on Server so they return fully-paginated generators instead of single-page dicts
• Add get_all_pages_from_space and get_all_blog_posts_from_space to Cloud — these were missing from the Cloud implementation entirely
• Unify _get_paged into ConfluenceBase — removed duplicate implementations from ConfluenceCloudBase and ConfluenceServerBase, replacing them with a single method that handles both str and dict formats for _links.next
• Fix relative pagination URLs — when the API returns a relative path (e.g. /rest/api/content?start=25), the scheme and host are extracted from self.url via urlparse and correctly prepended

Routing improvements (Confluence wrapper)

• Fix URL substring sanitization (CodeQL) — replaced naive "atlassian.net" in url checks with urlparse(url).hostname to prevent false matches on paths like evil.com/fake/atlassian.net/
• Recognize api.atlassian.com — the Confluence wrapper now correctly routes OAuth2 API gateway URLs to ConfluenceCloud
• Support explicit cloud= kwarg — allows callers to force Cloud or Server routing regardless of URL heuristics

Out of scope (intentionally not in this PR)

The Cloud api_root and api_version defaults (wiki/api/v2, "2") are unchanged. An earlier iteration of this branch changed these to wiki/rest/api / "latest" (v1 REST API), but that was reverted — it was unrelated to #1598 and would break existing Cloud users who have already set up v2 OAuth scopes. Users who need v1 REST API access can pass api_root="wiki/rest/api", api_version="latest" explicitly to the constructor.

---

Breaking change

The following methods now return generators instead of dicts:

get_all_pages_from_space, get_all_blog_posts_from_space, get_all_pages_by_label, get_all_blog_posts_by_label, get_all_draft_pages_from_space, get_all_draft_blog_posts_from_space, get_trash_content, get_all_pages_from_space_trash, get_all_blog_posts_from_space_trash

Before (broken — only first page):

result = confluence.get_all_pages_from_space("MYSPACE")
pages = result["results"]  # KeyError on Cloud; only 25 results on Server

After (correct — all pages):

for page in confluence.get_all_pages_from_space("MYSPACE"):
    process(page)

# or collect all at once
pages = list(confluence.get_all_pages_from_space("MYSPACE"))

---

🤖 Generated with Claude Code

[gonchik]

So let's check how AI will work, fortunately it works to me

[gonchik]

@Zircoz looks like a lot of extra code

[Zircoz]

Ill tell claude to review and rethink. :P

More seriously tho, I'll take a closer look myself with assistance from claude and come back to ya on this.

[Zircoz]

Migration note for maintainers

Hi — a note for when this gets merged, in case it's useful for release notes or a changelog entry:

What changed

Several get_all_* methods that previously returned a single-page dict now return a generator that automatically paginates through all results. This is the fix for #1598 — the old behaviour silently truncated results at the first page.

Affected methods:

• get_all_pages_from_space
• get_all_blog_posts_from_space
• get_all_pages_by_label
• get_all_blog_posts_by_label
• get_all_draft_pages_from_space
• get_all_draft_blog_posts_from_space
• get_trash_content
• get_all_pages_from_space_trash
• get_all_blog_posts_from_space_trash

Upgrading

# Before — broken (only first page returned, dict access required)
result = confluence.get_all_pages_from_space("MYSPACE")
pages = result["results"]

# After — iterate the generator directly
for page in confluence.get_all_pages_from_space("MYSPACE"):
    process(page)

# or collect everything at once
pages = list(confluence.get_all_pages_from_space("MYSPACE"))

Cloud users on v1 REST API

If you use api.atlassian.com with a v1 API token, pass these explicitly to the constructor — the Cloud defaults remain wiki/api/v2:

confluence = Confluence(
    url="https://api.atlassian.com/ex/confluence/<tenant-id>",
    username=email,
    password=api_token,
    cloud=True,
    api_root="wiki/rest/api",
    api_version="latest",
)

---

A couple of requests:

• Could you add something along the above lines to the changelog / release notes when merging? Happy to submit a separate PR for that if you'd prefer.
• Could you trigger the CI checks once more? There were some automated review comments from GitHub Advanced Security that I'd like to make sure are resolved.
• Any feedback on the approach or anything you'd like changed before merging is very welcome.

Thanks!

🤖 Generated with Claude Code