Summary: | Web Inspector: REGRESSION(r249831): content view is empty if a breakpoint is hit in the main resource | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Devin Rousso <hi> | ||||||
Component: | Web Inspector | Assignee: | Devin Rousso <hi> | ||||||
Status: | RESOLVED FIXED | ||||||||
Severity: | Normal | CC: | commit-queue, hi, inspector-bugzilla-changes, joepeck, simon.fraser, webkit-bug-importer | ||||||
Priority: | P2 | Keywords: | InRadar | ||||||
Version: | WebKit Nightly Build | ||||||||
Hardware: | All | ||||||||
OS: | All | ||||||||
See Also: | https://bugs.webkit.org/show_bug.cgi?id=203748 | ||||||||
Bug Depends on: | 201535 | ||||||||
Bug Blocks: | |||||||||
Attachments: |
|
Description
Devin Rousso
2019-10-31 21:26:12 PDT
Created attachment 382549 [details]
Patch
Created attachment 382565 [details] Patch This gets part of the way there, in that the content is now visible. But there appears to be another issue, which is that the initial content is now just the inline script content, rather than the entire HTML document. This seems to have been caused by r249831. My guess as to why that's happening is that the timing of showing the pause location within the inline script is now not working correctly since we have to wait for the formatted HTML content, which is asynchronous. I think we should either not auto-format HTML resources if we're showing a pause location the first time the resource is shown or try to delay the highlighting code until after the format is complete. Comment on attachment 382565 [details]
Patch
rs=me
(In reply to Devin Rousso from comment #3) > Created attachment 382565 [details] > Patch > > This gets part of the way there, in that the content is now visible. > > But there appears to be another issue, which is that the initial content is > now just the inline script content, I think that might have been expected behavior. If I recall correctly, the frontend would have paused in a Script but not necessarily the main resource content... as backwards as that sounds. We've had situations in the past where we would show blank content and just a <script>...</script> if you pause in script during page load. That said, if we have the full content, we should address it! (In reply to Joseph Pecoraro from comment #5) > (In reply to Devin Rousso from comment #3) > > Created attachment 382565 [details] > > Patch > > > > This gets part of the way there, in that the content is now visible. > > > > But there appears to be another issue, which is that the initial content is now just the inline script content, > > I think that might have been expected behavior. > > If I recall correctly, the frontend would have paused in a Script but not necessarily the main resource content... as backwards as that sounds. We've had situations in the past where we would show blank content and just a <script>...</script> if you pause in script during page load. > > That said, if we have the full content, we should address it! Before r249831 though, we did show the *full* content. This is 100% a regression. I'm pretty sure we also have the full content by this point, so we should definitely show it. Comment on attachment 382565 [details] Patch I'm going to land this as is since it's already a positive change. I've created a followup for the "missing the rest of the content" issue: - <https://webkit.org/b/203748> Web Inspector: REGRESSION(r249831): hitting a breakpoint in an inline <script> doesn't show the surrounding HTML content Comment on attachment 382565 [details] Patch Clearing flags on attachment: 382565 Committed r251932: <https://trac.webkit.org/changeset/251932> All reviewed patches have been landed. Closing bug. |