WDYT?
A specific alternative might be to define `FollowRedirectParams` that includes `new_url` + `HttpRequestHeadersUpdateParams`, while I don't have any specific reason to (or not to) do so.
(bot failures are due to recent addition of `FollowRedirect` overrides; I'll rebase/rerun later).
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
network::HttpRequestHeadersUpdateParams headers_update_params,Now we pass by value instead of reference. Does that mean we add extra copies?
bool StructTraits<network::mojom::HttpRequestHeadersUpdateParamsDataView,Maybe nice to add unitteest.
struct HttpRequestHeadersUpdateParams {```
// Typemapped to `network::HttpRequestHeadersUpdateParams`.
```
Also, consider using `IFFT` so that we don't forget update them each other?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
network::HttpRequestHeadersUpdateParams headers_update_params,Now we pass by value instead of reference. Does that mean we add extra copies?
No, `HttpRequestHeadersUpdateParams` is passed by move, not copy.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
network::HttpRequestHeadersUpdateParams headers_update_params,Hiroshige HayashizakiNow we pass by value instead of reference. Does that mean we add extra copies?
No, `HttpRequestHeadersUpdateParams` is passed by move, not copy.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
bool StructTraits<network::mojom::HttpRequestHeadersUpdateParamsDataView,Maybe nice to add unitteest.
Created: https://crrev.com/c/7855487
```
// Typemapped to `network::HttpRequestHeadersUpdateParams`.
```Also, consider using `IFFT` so that we don't forget update them each other?
```
// Typemapped to `network::HttpRequestHeadersUpdateParams`.
```
Done.
Also, consider using `IFFT` so that we don't forget update them each other?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
bool StructTraits<network::mojom::HttpRequestHeadersUpdateParamsDataView,Hiroshige HayashizakiMaybe nice to add unitteest.
Created: https://crrev.com/c/7855487
It would be nice to submit the test together with the implementation, ideally factoring out in two CLs: 1. impl + test, 2. other stuff in this CL.
I don't meant to block this CL though.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
bool StructTraits<network::mojom::HttpRequestHeadersUpdateParamsDataView,Hiroshige HayashizakiMaybe nice to add unitteest.
Kenichi IshibashiCreated: https://crrev.com/c/7855487
It would be nice to submit the test together with the implementation, ideally factoring out in two CLs: 1. impl + test, 2. other stuff in this CL.
I don't meant to block this CL though.
Merged https://crrev.com/c/7855487.
What do you mean by "other stuff in this CL"?
Everything related to the `network::mojom::URLLoader::FollowRedirect` change should be committed at once and can't be split.
The other parts of
- Defines `network.mojom.HttpRequestHeadersUpdateParams` mojo struct.
- `services/network/public/mojom/http_request_headers.mojom`
- Maps it to `network::HttpRequestHeadersUpdateParams`.
- `services/network/public/mojom/BUILD.gn`
- `services/network/public/cpp/http_request_headers_mojom_traits.*`
can be split into a preparation CL, but is this what you mean?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
bool StructTraits<network::mojom::HttpRequestHeadersUpdateParamsDataView,Hiroshige HayashizakiMaybe nice to add unitteest.
Kenichi IshibashiCreated: https://crrev.com/c/7855487
Hiroshige HayashizakiIt would be nice to submit the test together with the implementation, ideally factoring out in two CLs: 1. impl + test, 2. other stuff in this CL.
I don't meant to block this CL though.
Merged https://crrev.com/c/7855487.
What do you mean by "other stuff in this CL"?
Everything related to the `network::mojom::URLLoader::FollowRedirect` change should be committed at once and can't be split.
The other parts of
- Defines `network.mojom.HttpRequestHeadersUpdateParams` mojo struct.
- `services/network/public/mojom/http_request_headers.mojom`
- Maps it to `network::HttpRequestHeadersUpdateParams`.
- `services/network/public/mojom/BUILD.gn`
- `services/network/public/cpp/http_request_headers_mojom_traits.*`can be split into a preparation CL, but is this what you mean?
Yeah, that's what I meant:
But I understand that splitting this into two CLs is bothersome, so lgtm.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Commit-Queue | +1 |
+mea...@chromium.org as IPC reviewer for `mojom` file (manually assigned to the owner that randomly assigned in https://crrev.com/c/7869458).
Sounds reasonable. Created: https://crrev.com/c/7869458
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
+blundell@ PTAL for chrome and components OWNER review.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
peconn@ PTAL for android_webview OWNER review.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Commit-Queue | +1 |
+arthursonzogni@ PTAL for content OWNER review.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Commit-Queue | +1 |
+tjudkins@ PTAL for extensions OWNER review.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Commit-Queue | +1 |
+ddorwin@ PTAL for fuchsia OWNER review.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
horo@ PTAL for storage/browser/blob/blob_url_loader.* OWNER review.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
horo@ PTAL for storage/browser/blob/blob_url_loader.* OWNER review.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Thanks, Hiroshige! This CL has quite a wide spread over //chrome and //components. I don't mind doing the reviews for all of this, but I'd prefer to do them in smaller CLs to be confident that I don't end up missing anything, if I'm understanding correctly that the changes here could be split across CLs without any problem?
Thanks, Hiroshige! This CL has quite a wide spread over //chrome and //components. I don't mind doing the reviews for all of this, but I'd prefer to do them in smaller CLs to be confident that I don't end up missing anything, if I'm understanding correctly that the changes here could be split across CLs without any problem?
Thanks, Hiroshige! This CL has quite a wide spread over //chrome and //components. I don't mind doing the reviews for all of this, but I'd prefer to do them in smaller CLs to be confident that I don't end up missing anything, if I'm understanding correctly that the changes here could be split across CLs without any problem?
As discussed in the other CL https://crrev.com/c/7835809/comments/a0032e36_83efc0a0?tab=comments, this CL also can't be split, and I'll add subdir OWNERs.
PTAL,
mthiesse@ for components/navigation_interception
igorruvinov@ for components/enterprise/platform_auth
chrome-signin-reviews@ for chrome/browser/signin/
curranmax@ for chrome/browser/offline_pages
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Shadowed: er...@google.com
Reviewer source(s):
er...@google.com, msa...@chromium.org is from context(googleclient/chrome/chromium_gwsq/components/signin/config.gwsq)
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
PTAL
jinsukkim@ for components/embedder_support/android/util
nhiroki@ for chrome/browser/preloading/prefetch/search_prefetch
PTAL kouhei@ for content/browser/web_package
| Code-Review | +1 |
| Code-Review | +1 |
| Code-Review | +1 |
extensions/* LGTM! Thanks!
| Code-Review | +1 |
`chrome/browser/signin/chrome_signin_proxying_url_loader_factory.cc` LGTM.
PTAL +greengrape@ for components/webapps/isolated_web_apps/url_loading/url_loader.cc
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
PTAL +majewski@ for chrome/browser/ash/fileapi/
PTAL +yaoxia@ for content/browser/browsing_topics/
PTAL behamilton@ for content/browser/interest_group
+caseq@ PTAL for content/browser/devtools
| Code-Review | +1 |
Hiroshige HayashizakiThanks, Hiroshige! This CL has quite a wide spread over //chrome and //components. I don't mind doing the reviews for all of this, but I'd prefer to do them in smaller CLs to be confident that I don't end up missing anything, if I'm understanding correctly that the changes here could be split across CLs without any problem?
In the future for CLs like this I would recommend finding somebody with owners-override powers as this is a largely mechanical change, so you only need 1-2 reviewers.
| Code-Review | +1 |
| Code-Review | +1 |
stamp-ish lgtm
url_loader_->FollowRedirect({}, std::nullopt);nit: `/*headers_update_params=*/`
Ditto below.
| Code-Review | +1 |
Hiroshige HayashizakiThanks, Hiroshige! This CL has quite a wide spread over //chrome and //components. I don't mind doing the reviews for all of this, but I'd prefer to do them in smaller CLs to be confident that I don't end up missing anything, if I'm understanding correctly that the changes here could be split across CLs without any problem?
Michael Thiessen
In the future for CLs like this I would recommend finding somebody with owners-override powers as this is a largely mechanical change, so you only need 1-2 reviewers.
This is on me - I had asked Hiroshige to find closer OWNERS for //chrome and //components as I didn't feel comfortable reviewing all of that spread. It definitely introduced a fair bit of process overhead, for which I apologize.
Thanks, Hiroshige! I appreciate your effort here. Remaining //chrome and //components files LGTM.
| Code-Review | +1 |
chrome/browser/signin/chrome_signin_proxying_url_loader_factory.cc,LGTM, thanks!
| Code-Review | +1 |
+ntfschr@ PTAL as android_webview/browser/network_service OWNER
(Removing peconn@ instead due to OOO)
url_loader_->FollowRedirect({}, std::nullopt);Hiroshige Hayashizakinit: `/*headers_update_params=*/`
Ditto below.
Done
| Code-Review | +1 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
Hiroshige HayashizakiWDYT?
A specific alternative might be to define `FollowRedirectParams` that includes `new_url` + `HttpRequestHeadersUpdateParams`, while I don't have any specific reason to (or not to) do so.
(bot failures are due to recent addition of `FollowRedirect` overrides; I'll rebase/rerun later).
Proceeding to the current approach for now.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
40 is the latest approved patch-set.
No files were changed between the latest approved patch-set and the submitted one.
Use HttpRequestHeadersUpdateParams in mojom::URLLoader::FollowRedirect
Follow-up of https://crrev.com/c/7835601 that introduced
`network::HttpRequestHeadersUpdateParams`.
This CL uses `network.mojom.HttpRequestHeadersUpdateParams` (defined
in https://crrev.com/c/7869458) in
`network::mojom::URLLoader::FollowRedirect` in
`services/network/public/mojom/url_loader.mojom`, and adjust its
overrides accordingly.
### Purely mechanical changes (See Patch Set 35):
Replaces
```
FollowRedirect\(\s*
const std::vector<std::string>&\s*\w*,\s*
const (::)?net::HttpRequestHeaders&\s*\w*,\s*
const (::)?net::HttpRequestHeaders&\s*\w*,
```
with
```
FollowRedirect(
network::HttpRequestHeadersUpdateParams headers_update_params,
```
Replaces the followings (ignoring whitespaces):
```
FollowRedirect({}, {}, {}, std::nullopt)
FollowRedirect({}, {}, {}, {})
FollowRedirect(/*removed_headers=*/{},
/*modified_headers=*/{},
/*modified_cors_exempt_headers=*/{},
/*new_url=*/std::nullopt)
FollowRedirect(/*removed_headers=*/{},
/*modified_headers=*/{},
/*modified_cors_exempt_headers=*/{},
std::nullopt)
FollowRedirect(
std::vector<std::string>() /* sremoved_headers */,
net::HttpRequestHeaders() /* modified_headers */,
net::HttpRequestHeaders() /* smodified_cors_exempt_headers*/,
std::nullopt)
```
with
```
FollowRedirect(/*headers_update_params=*/{}, /*new_url=*/std::nullopt)
```
### Manual changes
Still mostly straightforward, with notable manual changes:
- `PrefetchURLLoader` has been intentionally dropped some of
`HttpRequestHeadersUpdateParams` since before. This CL has
additional lines to preserve the existing behavior:
- `content/browser/loader/prefetch_url_loader.cc`
- Before this CL, we had some copies just because `FollowRedirect()`
has been passing `const&` arguments. This CL removes the copies and
modifies the passed `headers_update_params` directly.
- `android_webview/browser/network_service/aw_proxying_url_loader_factory.cc`
- `chrome/browser/signin/chrome_signin_proxying_url_loader_factory.cc`
- `content/browser/loader/subresource_proxying_url_loader.cc`
- `services/network/cors/cors_url_loader.cc`
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |