fix: get content hash for master on local engine branches - #172792

…mpt) (#173169)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards #171790

re-land attempt for #173114 (try 2)
re-land attempt for #172792 (original)

The first commit in this PR is the previously LGTM'd changes for the
above; with tests.

The second commit is the critical change to make this work in post
submits (fixes #173143). It turns out that while LUCI reports the GitHub
private branches; our recipes directly checkout the git sha. This
matters because the content scripts couldn't determine the branch name
and the rev-parse was just HEAD. This lead the scripts down the
merge-base logic, which returns the previous commit.

A test was added specifically for this.

Alternatively to this change, we could have checked for LUCI_CONTEXT
being present in the environment. This is checked by Flutter tools in
some cases, but not by any other scripts in `bin/internal`. The downside
to checking HEAD: if you have a local branch with engine changes and you
move back a revision - `dart`/`flutter` invocations will generate the
hash for your local changes and fail.

…mpt) (flutter#173169)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#173114 (try 2)
re-land attempt for flutter#172792 (original)

The first commit in this PR is the previously LGTM'd changes for the
above; with tests.

The second commit is the critical change to make this work in post
submits (fixes flutter#173143). It turns out that while LUCI reports the GitHub
private branches; our recipes directly checkout the git sha. This
matters because the content scripts couldn't determine the branch name
and the rev-parse was just HEAD. This lead the scripts down the
merge-base logic, which returns the previous commit.

A test was added specifically for this.

Alternatively to this change, we could have checked for LUCI_CONTEXT
being present in the environment. This is checked by Flutter tools in
some cases, but not by any other scripts in `bin/internal`. The downside
to checking HEAD: if you have a local branch with engine changes and you
move back a revision - `dart`/`flutter` invocations will generate the
hash for your local changes and fail.

…3114)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

…lutter#173114)" (flutter#173145)

<!-- start_original_pr_link -->
Reverts: flutter#173114
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: there is still another problem with the merge
queue causing the content hash to be different:

git revision: e13dd53
actual hash: 4b6f7b0f9849efaa59f515c8e95f3f27a6eb2ffb
hash in the queue? 9e5b2eef4ba79b15b4f80dbba812d199d262366f



<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {matanlurey, chingjun}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…mpt) (flutter#173169)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#173114 (try 2)
re-land attempt for flutter#172792 (original)

The first commit in this PR is the previously LGTM'd changes for the
above; with tests.

The second commit is the critical change to make this work in post
submits (fixes flutter#173143). It turns out that while LUCI reports the GitHub
private branches; our recipes directly checkout the git sha. This
matters because the content scripts couldn't determine the branch name
and the rev-parse was just HEAD. This lead the scripts down the
merge-base logic, which returns the previous commit.

A test was added specifically for this.

Alternatively to this change, we could have checked for LUCI_CONTEXT
being present in the environment. This is checked by Flutter tools in
some cases, but not by any other scripts in `bin/internal`. The downside
to checking HEAD: if you have a local branch with engine changes and you
move back a revision - `dart`/`flutter` invocations will generate the
hash for your local changes and fail.

…2792)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790

…lutter#172792)" (flutter#172805)

<!-- start_original_pr_link -->
Reverts: flutter#172792
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: almost, but not quite right. android builders
generate the wrong hash because they are on custom branches.
<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {bdero, matanlurey}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790 

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…3114)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

…lutter#173114)" (flutter#173145)

<!-- start_original_pr_link -->
Reverts: flutter#173114
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: there is still another problem with the merge
queue causing the content hash to be different:

git revision: e13dd53
actual hash: 4b6f7b0f9849efaa59f515c8e95f3f27a6eb2ffb
hash in the queue? 9e5b2eef4ba79b15b4f80dbba812d199d262366f



<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {matanlurey, chingjun}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…mpt) (flutter#173169)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#173114 (try 2)
re-land attempt for flutter#172792 (original)

The first commit in this PR is the previously LGTM'd changes for the
above; with tests.

The second commit is the critical change to make this work in post
submits (fixes flutter#173143). It turns out that while LUCI reports the GitHub
private branches; our recipes directly checkout the git sha. This
matters because the content scripts couldn't determine the branch name
and the rev-parse was just HEAD. This lead the scripts down the
merge-base logic, which returns the previous commit.

A test was added specifically for this.

Alternatively to this change, we could have checked for LUCI_CONTEXT
being present in the environment. This is checked by Flutter tools in
some cases, but not by any other scripts in `bin/internal`. The downside
to checking HEAD: if you have a local branch with engine changes and you
move back a revision - `dart`/`flutter` invocations will generate the
hash for your local changes and fail.

…2792)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790

…lutter#172792)" (flutter#172805)

<!-- start_original_pr_link -->
Reverts: flutter#172792
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: almost, but not quite right. android builders
generate the wrong hash because they are on custom branches.
<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {bdero, matanlurey}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790 

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…3114)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

…lutter#173114)" (flutter#173145)

<!-- start_original_pr_link -->
Reverts: flutter#173114
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: there is still another problem with the merge
queue causing the content hash to be different:

git revision: e13dd53
actual hash: 4b6f7b0f9849efaa59f515c8e95f3f27a6eb2ffb
hash in the queue? 9e5b2eef4ba79b15b4f80dbba812d199d262366f



<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {matanlurey, chingjun}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…mpt) (flutter#173169)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#173114 (try 2)
re-land attempt for flutter#172792 (original)

The first commit in this PR is the previously LGTM'd changes for the
above; with tests.

The second commit is the critical change to make this work in post
submits (fixes flutter#173143). It turns out that while LUCI reports the GitHub
private branches; our recipes directly checkout the git sha. This
matters because the content scripts couldn't determine the branch name
and the rev-parse was just HEAD. This lead the scripts down the
merge-base logic, which returns the previous commit.

A test was added specifically for this.

Alternatively to this change, we could have checked for LUCI_CONTEXT
being present in the environment. This is checked by Flutter tools in
some cases, but not by any other scripts in `bin/internal`. The downside
to checking HEAD: if you have a local branch with engine changes and you
move back a revision - `dart`/`flutter` invocations will generate the
hash for your local changes and fail.

…2792)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790

…lutter#172792)" (flutter#172805)

<!-- start_original_pr_link -->
Reverts: flutter#172792
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: almost, but not quite right. android builders
generate the wrong hash because they are on custom branches.
<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {bdero, matanlurey}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790 

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…3114)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

…lutter#173114)" (flutter#173145)

<!-- start_original_pr_link -->
Reverts: flutter#173114
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: there is still another problem with the merge
queue causing the content hash to be different:

git revision: e13dd53
actual hash: 4b6f7b0f9849efaa59f515c8e95f3f27a6eb2ffb
hash in the queue? 9e5b2eef4ba79b15b4f80dbba812d199d262366f



<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {matanlurey, chingjun}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…mpt) (flutter#173169)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#173114 (try 2)
re-land attempt for flutter#172792 (original)

The first commit in this PR is the previously LGTM'd changes for the
above; with tests.

The second commit is the critical change to make this work in post
submits (fixes flutter#173143). It turns out that while LUCI reports the GitHub
private branches; our recipes directly checkout the git sha. This
matters because the content scripts couldn't determine the branch name
and the rev-parse was just HEAD. This lead the scripts down the
merge-base logic, which returns the previous commit.

A test was added specifically for this.

Alternatively to this change, we could have checked for LUCI_CONTEXT
being present in the environment. This is checked by Flutter tools in
some cases, but not by any other scripts in `bin/internal`. The downside
to checking HEAD: if you have a local branch with engine changes and you
move back a revision - `dart`/`flutter` invocations will generate the
hash for your local changes and fail.

…utter/flutter#172792)

…2792)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790

…lutter#172792)" (flutter#172805)

<!-- start_original_pr_link -->
Reverts: flutter#172792
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: almost, but not quite right. android builders
generate the wrong hash because they are on custom branches.
<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {bdero, matanlurey}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  flutter#171790 

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…3114)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

…lutter#173114)" (flutter#173145)

<!-- start_original_pr_link -->
Reverts: flutter#173114
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: there is still another problem with the merge
queue causing the content hash to be different:

git revision: e13dd53
actual hash: 4b6f7b0f9849efaa59f515c8e95f3f27a6eb2ffb
hash in the queue? 9e5b2eef4ba79b15b4f80dbba812d199d262366f



<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {matanlurey, chingjun}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#172792  with the following changes:

1. content_aware_hash.(ps1|sh) now consider multiple branches to choose
between HEAD and merge-base.
2. content_aware_hash_test.dart updated for these new requirements 
3. content_aware_hash_test.dart allows for forcing powershell on mac for
testing
4. updated docs/tool/Engine-artifacts.md documentation.

## Pre-launch Checklist

- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [ ] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>

…mpt) (flutter#173169)

The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards flutter#171790

re-land attempt for flutter#173114 (try 2)
re-land attempt for flutter#172792 (original)

The first commit in this PR is the previously LGTM'd changes for the
above; with tests.

The second commit is the critical change to make this work in post
submits (fixes flutter#173143). It turns out that while LUCI reports the GitHub
private branches; our recipes directly checkout the git sha. This
matters because the content scripts couldn't determine the branch name
and the rev-parse was just HEAD. This lead the scripts down the
merge-base logic, which returns the previous commit.

A test was added specifically for this.

Alternatively to this change, we could have checked for LUCI_CONTEXT
being present in the environment. This is checked by Flutter tools in
some cases, but not by any other scripts in `bin/internal`. The downside
to checking HEAD: if you have a local branch with engine changes and you
move back a revision - `dart`/`flutter` invocations will generate the
hash for your local changes and fail.

Roll Flutter from 20788c0 to cfbaa18 (43 revisions)

flutter/flutter@20788c0...cfbaa18

2025-07-28 [email protected] Remove shortner (flutter/flutter#172844)
2025-07-28 [email protected] Update `flutter pub get` to use `flutter.version.json` (instead of `version`) (flutter/flutter#172798)
2025-07-27 [email protected] Roll Fuchsia Linux SDK from i7RS-4iBtdndwa3Jy... to tK_PAaLeo0pUxi8hv... (flutter/flutter#172827)
2025-07-26 [email protected] Roll Fuchsia Linux SDK from 4IewCwqcEm7r4QsXd... to i7RS-4iBtdndwa3Jy... (flutter/flutter#172808)
2025-07-26 98614782+auto-submit[bot]@users.noreply.github.com Reverts "fix: get content hash for master on local engine branches (#172792)" (flutter/flutter#172805)
2025-07-26 [email protected] fix: build size_exp artifacts in post (flutter/flutter#172802)
2025-07-26 [email protected] fix: get content hash for master on local engine branches (flutter/flutter#172792)
2025-07-25 [email protected] Update `main`/`master` repoExceptions analysis set (flutter/flutter#172796)
2025-07-25 [email protected] Remove deprecated `--[no-]-disable-dds` (flutter/flutter#172791)
2025-07-25 [email protected] `last_engine_commit.ps1`: Use `$flutterRoot` instead of `$gitTopLevel` (flutter/flutter#172786)
2025-07-25 [email protected] Check file contents, remove tests that are now invalid (`: 'none'`) (flutter/flutter#172788)
2025-07-25 [email protected] windows: allow pwsh.bat wrapper (flutter/flutter#171778)
2025-07-25 [email protected] [Impeller] Log glCheckFrameBufferStatus output when it fails in blit pass. (flutter/flutter#172592)
2025-07-25 [email protected] Remove deprecated `AssetManifest.json` file (flutter/flutter#172594)
2025-07-25 [email protected] Refactor `verifyVersion` into typed `Version` object, use it (flutter/flutter#172780)
2025-07-25 [email protected] Bump the customer tests to pick up an update to Zulip's tests, for th… (flutter/flutter#172733)
2025-07-25 [email protected] Rename `AppRunLogger`, stop writing status messages that break JSON (flutter/flutter#172591)
2025-07-25 [email protected] Add `--dart-define`, `-D` to `assemble`, deprecate `--define`, `-d`. (flutter/flutter#172510)
2025-07-25 [email protected] Fix previous text input connection interrupts enter/space (flutter/flutter#171973)
2025-07-25 [email protected] Update CHANGELOG.md for 3.32.8 (flutter/flutter#172779)
2025-07-25 [email protected] Fix an `omit-type` lint that is breaking the tree (flutter/flutter#172776)
2025-07-25 [email protected] feat(tool): Respect user-data-dir flag from web-browser-flag (flutter/flutter#169445)
2025-07-25 [email protected] [web] Fix empty first frame in multiview mode (flutter/flutter#172493)
2025-07-25 [email protected] Migrate to null aware elements - Part 3 (flutter/flutter#172307)
2025-07-25 [email protected] Update few links in docs (flutter/flutter#172166)
2025-07-25 [email protected] Roll Skia from 768ace3fd144 to 409e1c7ba09b (1 revision) (flutter/flutter#172745)
2025-07-25 [email protected] Roll Fuchsia Linux SDK from a_55b5hqpRnigzuNC... to 4IewCwqcEm7r4QsXd... (flutter/flutter#172742)
2025-07-25 [email protected] Roll Skia from 094ac350125f to 768ace3fd144 (18 revisions) (flutter/flutter#172735)
2025-07-25 [email protected] Remove 2023 deprecated `'platforms'` key from daemon output (flutter/flutter#172593)
2025-07-24 [email protected] Manual roll of Dart from 26dece893751 to 5ede85aff798 (flutter/flutter#172731)
2025-07-24 [email protected] 169108 ci debug (flutter/flutter#172589)
2025-07-24 [email protected] Improve assertion message in `_MixedBorderRadius.resolve()` (flutter/flutter#172100)
2025-07-24 [email protected] Make sure that a BottomAppBar doesn't crash in 0x0 environment (flutter/flutter#172097)
2025-07-24 [email protected] Handle#6537 bottom navigation bar (flutter/flutter#172404)
2025-07-24 [email protected] Make sure that a MaterialBanner doesn't crash in 0x0 environment (flutter/flutter#172074)
2025-07-24 [email protected] Remove support for `pluginClass: none` (on `master`; `> 3.35.0`) (flutter/flutter#172579)
2025-07-24 [email protected] Make sure that a Card doesn't crash in 0x0 environment (flutter/flutter#172373)
2025-07-24 [email protected] Make sure that an AboutDialog doesn't crash in 0x0 environment (flutter/flutter#172444)
2025-07-24 [email protected] Make sure that an ActionChip doesn't crash in 0x0 environment (flutter/flutter#172684)
2025-07-24 [email protected] Delete (unused) `dev/checks_tool` directory (flutter/flutter#172715)
2025-07-24 [email protected] [Android] Add missing ci configuration for `Linux android_java17_tool_integration_tests` (flutter/flutter#172702)
2025-07-24 [email protected] Flutter-Cherrypick-Process.md - Spell cherry-pick consistently (flutter/flutter#172697)
2025-07-24 [email protected] Bringup `linux_web_engine_tests`; failing 4+ times in a row (flutter/flutter#172714)

If this roll has caused a breakage, revert this CL and stop the roller
using the controls here:
...

Roll Flutter from 20788c0 to cfbaa18 (43 revisions)

flutter/flutter@20788c0...cfbaa18

2025-07-28 [email protected] Remove shortner (flutter/flutter#172844)
2025-07-28 [email protected] Update `flutter pub get` to use `flutter.version.json` (instead of `version`) (flutter/flutter#172798)
2025-07-27 [email protected] Roll Fuchsia Linux SDK from i7RS-4iBtdndwa3Jy... to tK_PAaLeo0pUxi8hv... (flutter/flutter#172827)
2025-07-26 [email protected] Roll Fuchsia Linux SDK from 4IewCwqcEm7r4QsXd... to i7RS-4iBtdndwa3Jy... (flutter/flutter#172808)
2025-07-26 98614782+auto-submit[bot]@users.noreply.github.com Reverts "fix: get content hash for master on local engine branches (#172792)" (flutter/flutter#172805)
2025-07-26 [email protected] fix: build size_exp artifacts in post (flutter/flutter#172802)
2025-07-26 [email protected] fix: get content hash for master on local engine branches (flutter/flutter#172792)
2025-07-25 [email protected] Update `main`/`master` repoExceptions analysis set (flutter/flutter#172796)
2025-07-25 [email protected] Remove deprecated `--[no-]-disable-dds` (flutter/flutter#172791)
2025-07-25 [email protected] `last_engine_commit.ps1`: Use `$flutterRoot` instead of `$gitTopLevel` (flutter/flutter#172786)
2025-07-25 [email protected] Check file contents, remove tests that are now invalid (`: 'none'`) (flutter/flutter#172788)
2025-07-25 [email protected] windows: allow pwsh.bat wrapper (flutter/flutter#171778)
2025-07-25 [email protected] [Impeller] Log glCheckFrameBufferStatus output when it fails in blit pass. (flutter/flutter#172592)
2025-07-25 [email protected] Remove deprecated `AssetManifest.json` file (flutter/flutter#172594)
2025-07-25 [email protected] Refactor `verifyVersion` into typed `Version` object, use it (flutter/flutter#172780)
2025-07-25 [email protected] Bump the customer tests to pick up an update to Zulip's tests, for th… (flutter/flutter#172733)
2025-07-25 [email protected] Rename `AppRunLogger`, stop writing status messages that break JSON (flutter/flutter#172591)
2025-07-25 [email protected] Add `--dart-define`, `-D` to `assemble`, deprecate `--define`, `-d`. (flutter/flutter#172510)
2025-07-25 [email protected] Fix previous text input connection interrupts enter/space (flutter/flutter#171973)
2025-07-25 [email protected] Update CHANGELOG.md for 3.32.8 (flutter/flutter#172779)
2025-07-25 [email protected] Fix an `omit-type` lint that is breaking the tree (flutter/flutter#172776)
2025-07-25 [email protected] feat(tool): Respect user-data-dir flag from web-browser-flag (flutter/flutter#169445)
2025-07-25 [email protected] [web] Fix empty first frame in multiview mode (flutter/flutter#172493)
2025-07-25 [email protected] Migrate to null aware elements - Part 3 (flutter/flutter#172307)
2025-07-25 [email protected] Update few links in docs (flutter/flutter#172166)
2025-07-25 [email protected] Roll Skia from 768ace3fd144 to 409e1c7ba09b (1 revision) (flutter/flutter#172745)
2025-07-25 [email protected] Roll Fuchsia Linux SDK from a_55b5hqpRnigzuNC... to 4IewCwqcEm7r4QsXd... (flutter/flutter#172742)
2025-07-25 [email protected] Roll Skia from 094ac350125f to 768ace3fd144 (18 revisions) (flutter/flutter#172735)
2025-07-25 [email protected] Remove 2023 deprecated `'platforms'` key from daemon output (flutter/flutter#172593)
2025-07-24 [email protected] Manual roll of Dart from 26dece893751 to 5ede85aff798 (flutter/flutter#172731)
2025-07-24 [email protected] 169108 ci debug (flutter/flutter#172589)
2025-07-24 [email protected] Improve assertion message in `_MixedBorderRadius.resolve()` (flutter/flutter#172100)
2025-07-24 [email protected] Make sure that a BottomAppBar doesn't crash in 0x0 environment (flutter/flutter#172097)
2025-07-24 [email protected] Handle#6537 bottom navigation bar (flutter/flutter#172404)
2025-07-24 [email protected] Make sure that a MaterialBanner doesn't crash in 0x0 environment (flutter/flutter#172074)
2025-07-24 [email protected] Remove support for `pluginClass: none` (on `master`; `> 3.35.0`) (flutter/flutter#172579)
2025-07-24 [email protected] Make sure that a Card doesn't crash in 0x0 environment (flutter/flutter#172373)
2025-07-24 [email protected] Make sure that an AboutDialog doesn't crash in 0x0 environment (flutter/flutter#172444)
2025-07-24 [email protected] Make sure that an ActionChip doesn't crash in 0x0 environment (flutter/flutter#172684)
2025-07-24 [email protected] Delete (unused) `dev/checks_tool` directory (flutter/flutter#172715)
2025-07-24 [email protected] [Android] Add missing ci configuration for `Linux android_java17_tool_integration_tests` (flutter/flutter#172702)
2025-07-24 [email protected] Flutter-Cherrypick-Process.md - Spell cherry-pick consistently (flutter/flutter#172697)
2025-07-24 [email protected] Bringup `linux_web_engine_tests`; failing 4+ times in a row (flutter/flutter#172714)

If this roll has caused a breakage, revert this CL and stop the roller
using the controls here:
...

…172792)" (#172805)

<!-- start_original_pr_link -->
Reverts: flutter/flutter#172792
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: almost, but not quite right. android builders
generate the wrong hash because they are on custom branches.
<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: jtmcdole
<!-- end_original_pr_author -->

<!-- start_reviewers -->
Reviewed By: {bdero, matanlurey}
<!-- end_reviewers -->

<!-- start_revert_body -->
This change reverts the following previous change:
The content hash doesn't exist for local engine changes, except for on
CI. If we detect we're on a branch with committed or uncommitted changes
to engine files; use "master".

towards  #171790 

<!-- end_revert_body -->

Co-authored-by: auto-submit[bot] <[email protected]>