Auto-Submit | +1 |
@tha...@chromium.org - here we go last batch. after only a few mini_chromium/crashpad/tools... will be left (mini_chromium is already prepared/WIP -> https://crrev.com/c/5523573)
```
git grep -l base::StringPiece | awk -F "/" '{r[$1] +=1} END{for(x in r) { printf("%05d\t %s\n",r[x], x) }}'|sort
00001 chrome
00001 PRESUBMIT.py
00001 styleguide
00001 ui
00002 net
00004 docs
00005 base
00017 tools
00033 third_party
```
close to closing bug 🥳
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
ping
Code-Review | +1 |
LG. First comment is something you might want to address before hitting cq+2. The other two are for the future.
// This header is deprecated. `std::string_view` is now `std::string_view`.
🤨
Probably want to keep the LHS here and then remove this file in a follow-up instead.
[[maybe_unused]] std::string_view piece = f(); // expected-error {{object backing the pointer will be destroyed at the end of the full-expression}}
Not sure if having a .nc test for a libc++ type is useful. Maybe we should toss this too now that this is done?
[[maybe_unused]] std::string_view b(s1);
Similar to the .nc comment: Do we still need this unittest file? (If so, we should probably rename it? Not in this CL though.)
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
thanks for your feedback, removed the not-needed anymore tests.
not sure if i fully understand the first comment, about the string_piece.h.
ready for submit?
// This header is deprecated. `std::string_view` is now `std::string_view`.
🤨
Probably want to keep the LHS here and then remove this file in a follow-up instead.
changed the comment, will remove the file in a follow up - once mini_chromium and tools are done
[[maybe_unused]] std::string_view piece = f(); // expected-error {{object backing the pointer will be destroyed at the end of the full-expression}}
Not sure if having a .nc test for a libc++ type is useful. Maybe we should toss this too now that this is done?
TIL about no-compile tests, thank you, removed the case.
[[maybe_unused]] std::string_view b(s1);
Similar to the .nc comment: Do we still need this unittest file? (If so, we should probably rename it? Not in this CL though.)
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Code-Review | +1 |
Commit-Queue | +2 |
// This header is deprecated. use `std::string_view` and <string_view> instead.
very very nit: upper case after a period
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Auto-Submit | +1 |
@tha...@chromium.org - this seems to fail on some non-public code.
on slack it was mentioned that it most likely be of a transitive include (in non-public chromeos).
as i cannot see any build-messages, could you be so kind and take a look? i am happy and ready to change whatever is needed.
@pkas...@chromium.org could you help me unblock this? seems to be some non-public code transitively relying on string_piece.h include. i cannot see the code nor the error message.
happy to address what ever is needed, this is the last big CL - for stringpiece removal, after this, only a few minors and a lot of followups (like method names; GetStringPiece) 🙌
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
I'll help take a look. Note that there are still some base::StringPiece usage in the rest of the Chromium code base.
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
I'll help take a look. Note that there are still some base::StringPiece usage in the rest of the Chromium code base.
I sent out a CL in a Google-internal repo that I have never touched. We'll see how that goes.
You may be able to make progress by splitting the base/strings/string_util* part into a separate CL and landing that first?
Lei ZhangI'll help take a look. Note that there are still some base::StringPiece usage in the rest of the Chromium code base.
I sent out a CL in a Google-internal repo that I have never touched. We'll see how that goes.
You may be able to make progress by splitting the base/strings/string_util* part into a separate CL and landing that first?
Landed the Google-internal CL. It may need a DEPS roll. I also have https://crrev.com/c/5653995 in flight.
Lei ZhangI'll help take a look. Note that there are still some base::StringPiece usage in the rest of the Chromium code base.
Lei ZhangI sent out a CL in a Google-internal repo that I have never touched. We'll see how that goes.
You may be able to make progress by splitting the base/strings/string_util* part into a separate CL and landing that first?
Landed the Google-internal CL. It may need a DEPS roll. I also have https://crrev.com/c/5653995 in flight.
Based on where the build error was, yeah we'll need a DEPS roll.
If we want to get as much of this in as quickly as possible, I might leave the string_piece.h #include in string_split.h, with a TODO to remove it?
Auto-Submit | +1 |
Lei ZhangI'll help take a look. Note that there are still some base::StringPiece usage in the rest of the Chromium code base.
Lei ZhangI sent out a CL in a Google-internal repo that I have never touched. We'll see how that goes.
You may be able to make progress by splitting the base/strings/string_util* part into a separate CL and landing that first?
Peter KastingLanded the Google-internal CL. It may need a DEPS roll. I also have https://crrev.com/c/5653995 in flight.
Based on where the build error was, yeah we'll need a DEPS roll.
If we want to get as much of this in as quickly as possible, I might leave the string_piece.h #include in string_split.h, with a TODO to remove it?
thank you both! uploaded a PS - with the revert of #include
@the...@chromium.org - i know some places still not done. but most of them are due to the fact that mini_chromium needs to change.
```
git grep -l base::StringPiece | awk -F "/" '{r[$1"/"$2] +=1} END{for(x in r) { printf("%05d\t %s\n",r[x], x) }}'|sort
00001 base/json
00001 chrome/updater
00001 docs/debugging_with_crash_keys.md
00001 docs/security
00001 docs/speed
00001 docs/threading_and_tasks_testing.md
00001 PRESUBMIT.py/
00001 services/screen_ai
00001 styleguide/c++
00001 third_party/blink
00001 third_party/leveldatabase
00001 third_party/liburlpattern
00001 third_party/rust
00001 tools/ipc_fuzzer
00001 tools/metrics
00001 tools/win
00001 ui/ozone
00002 net/device_bound_sessions
00002 tools/mac
00003 third_party/hunspell
00003 third_party/webrtc_overrides
00003 third_party/zxcvbn-cpp
00003 tools/clang
00009 tools/json_schema_compiler
00020 third_party/crashpad
````
already started the work on mini_chromium: https://crrev.com/c/5523573
(was mentioned to first get //base done in chromium itself)
once min_chromium is done, the crashpad's matches can be tackled.
with a few "false-positiv" matches of the `grep` - i think after that we are almost there.
what a ride! 🤣
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Auto-Submit | +1 |
not 100% sure about DEPS - if the above CL is landed, i should be able to rebase this one - and it should pass? would'nt it make sense to wait for this? or does it take longer for the DEPS cl to land?
rebasing is pretty straight forward - not sure what is better
//TODO vs "just wait a bit"
(sorry if that might be a stupid question)
You can chain this CL to my CL and then run try jobs. I don't know if there are any other issues in the Chrome OS build.
Or just sync your checkout, as that CL landed.
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
(For future reference: if you chain a CL atop another, then run tryjobs, the tryjobs will include all CLs in the chain up to the CL in question. So the currently-running tryjobs should include Lei's change.)
Convert base::StringPiece to std::string_view final //base
The changes of this CL are made using the following script.
Script: https://issues.chromium.org/issues/40506050#comment343
Bug: 40506050
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |