Use gn_helpers to [se]serialize GN lists.

Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we used to take strings and FileArgs, then added support for GN lists using ast.parse which is good enough for 95% of things (we didn't have gn_helpers to do it properly yet). The GN list usage expanded to things that aren't actually GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it to use a GN version instead. The magic "##"->"$" remapping behavior is removed from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is preserved from the previous version even though it's a bit weird. These are are expanded either as a list beginning with '[', or as a string with no quotes. This FileArgs processing is used to do things like extract #define values from build_config files. One of which, java_libraries_list, is a string in the build_config, but expects to be inserted into a Java file with no quotes so the { ... } in the string expand to a Java array. Another of which, pack_relocations, takes a Python dictionary as input. This type of thing makes having more uniform type checking impossible (like by passing all arguments in GN formatting).

The strip_libraries_for_device script is removed, it is never referenced.

BUG=573132, 571022
Committed: https://crrev.com/1783edf919dd8a81d47bc58b7f91af65f79cb175
Cr-Commit-Position: refs/heads/master@{#413498}

[brettw, 2016-01-29 22:51:39]

Description was changed from

==========
USe gn_helpers to derserialize GN lists.

BUG=573132, 571022
==========

to

==========
Use gn_helpers to derserialize GN lists.

BUG=573132, 571022
==========

[brettw, 2016-01-29 23:28:05]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-01-29 23:34:25]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1647353002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1647353002/20001

[commit-bot: I haz the power, 2016-01-29 23:49:34]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-01-29 23:49:35]

Dry run: Try jobs failed on following builders:
  cast_shell_android on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/cast_shell_a...)

[brettw, 2016-01-30 00:14:55]

brettw@chromium.org changed reviewers:
+ dpranke@chromium.org

[brettw, 2016-01-30 00:15:02]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-01-30 00:16:05]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1647353002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1647353002/40001

[Dirk Pranke, 2016-01-30 00:26:48]

mostly lgtm .

https://codereview.chromium.org/1647353002/diff/40001/build/android/gyp/strip...
File build/android/gyp/strip_library_for_device.py (right):

https://codereview.chromium.org/1647353002/diff/40001/build/android/gyp/strip...
build/android/gyp/strip_library_for_device.py:43: libraries =
ast.literal_eval(options.libraries)
if this is in JSON format, why are you using literal_eval rather than
json.loads() ?

[commit-bot: I haz the power, 2016-01-30 00:35:55]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-01-30 00:35:56]

Dry run: Try jobs failed on following builders:
  android_chromium_gn_compile_rel on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_chro...)
  linux_android_rel_ng on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[brettw, 2016-01-30 05:46:47]

Comment and fix

[brettw, 2016-01-30 05:46:48]

The CQ bit was checked by brettw@chromium.org

[brettw, 2016-01-30 05:46:48]

The patchset sent to the CQ was uploaded after l-g-t-m from dpranke@chromium.org
Link to the patchset: https://codereview.chromium.org/1647353002/#ps60001
(title: "Comment and fix")

[commit-bot: I haz the power, 2016-01-30 05:47:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1647353002/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1647353002/60001

[brettw, 2016-01-30 05:48:22]

https://codereview.chromium.org/1647353002/diff/40001/build/android/gyp/strip...
File build/android/gyp/strip_library_for_device.py (right):

https://codereview.chromium.org/1647353002/diff/40001/build/android/gyp/strip...
build/android/gyp/strip_library_for_device.py:43: libraries =
ast.literal_eval(options.libraries)
On 2016/01/30 00:26:48, Dirk Pranke wrote:
> if this is in JSON format, why are you using literal_eval rather than
> json.loads() ?

Clarified to "Python dictionary format"

[commit-bot: I haz the power, 2016-01-30 06:03:32]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-01-30 06:03:33]

Try jobs failed on following builders:
  android_arm64_dbg_recipe on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_arm6...)
  android_compile_dbg on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_comp...)

[brettw, 2016-01-30 06:33:45]

The CQ bit was checked by brettw@chromium.org

[brettw, 2016-01-30 06:33:45]

The patchset sent to the CQ was uploaded after l-g-t-m from dpranke@chromium.org
Link to the patchset: https://codereview.chromium.org/1647353002/#ps80001
(title: " ")

[brettw, 2016-01-30 06:33:45]

The CQ bit was unchecked by brettw@chromium.org

[brettw, 2016-01-30 06:37:46]

Description was changed from

==========
Use gn_helpers to derserialize GN lists.

BUG=573132, 571022
==========

to

==========
Use gn_helpers to deserialize GN lists.

Some Android build files were using build_utils to parse Python lists. But
build_utils treats [] as a GN dictionary instead (the handling of which this
patch is trying to fix). So switching to GN parsing of the cases explicitly
passing Python lists here breaks. The cases where this happens have been updated
to call ast directly so the format of the list is well-defined (it doesn't need
the magic GN-vs-GYP detection that build_utils uses).

BUG=573132, 571022
==========

[brettw, 2016-01-30 06:37:52]

The CQ bit was checked by brettw@chromium.org

[commit-bot: I haz the power, 2016-01-30 06:38:05]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1647353002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1647353002/80001

[commit-bot: I haz the power, 2016-01-30 06:56:54]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-01-30 06:56:55]

Try jobs failed on following builders:
  android_compile_dbg on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_comp...)

[brettw, 2016-01-30 07:19:49]

I'm giving up on this as I explained in
https://code.google.com/p/chromium/issues/detail?id=573132

[brettw, 2016-08-17 21:16:42]

rebase

[brettw, 2016-08-17 21:16:43]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 21:17:22]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 21:35:26]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 21:35:27]

Dry run: Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[brettw, 2016-08-17 22:01:01]

replace all

[brettw, 2016-08-17 22:01:02]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 22:01:57]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[brettw, 2016-08-17 22:20:14]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 22:20:18]

Dry run: Try jobs failed on following builders:
  cast_shell_android on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/cast_shell_a...)

[commit-bot: I haz the power, 2016-08-17 22:20:39]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 22:42:26]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 22:42:27]

Dry run: Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[brettw, 2016-08-17 23:35:42]

fixes

[brettw, 2016-08-17 23:35:43]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 23:36:09]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-18 00:12:22]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-18 00:12:23]

Dry run: Try jobs failed on following builders:
  android_clang_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_clan...)

[brettw, 2016-08-18 18:13:25]

merge

[brettw, 2016-08-18 18:13:26]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-18 18:13:50]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-18 18:45:30]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-18 18:45:31]

Dry run: Try jobs failed on following builders:
  android_arm64_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_arm6...)

[brettw, 2016-08-18 20:00:17]

native library fix

[brettw, 2016-08-18 20:00:17]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-18 20:00:59]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[brettw, 2016-08-18 20:09:34]

Description was changed from

==========
Use gn_helpers to deserialize GN lists.

Some Android build files were using build_utils to parse Python lists. But
build_utils treats [] as a GN dictionary instead (the handling of which this
patch is trying to fix). So switching to GN parsing of the cases explicitly
passing Python lists here breaks. The cases where this happens have been updated
to call ast directly so the format of the list is well-defined (it doesn't need
the magic GN-vs-GYP detection that build_utils uses).

BUG=573132, 571022
==========

to

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. This type of thing makes having
more uniform type checking impossible (like by passing all arguments in GN
formatting).

BUG=573132, 571022
==========

[brettw, 2016-08-18 20:19:19]

Description was changed from

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. This type of thing makes having
more uniform type checking impossible (like by passing all arguments in GN
formatting).

BUG=573132, 571022
==========

to

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. Another of which,
pack_relocations, takes a Python dictionary as input. This type of thing makes
having more uniform type checking impossible (like by passing all arguments in
GN formatting).

BUG=573132, 571022
==========

[brettw, 2016-08-18 20:20:02]

Description was changed from

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. Another of which,
pack_relocations, takes a Python dictionary as input. This type of thing makes
having more uniform type checking impossible (like by passing all arguments in
GN formatting).

BUG=573132, 571022
==========

to

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. Another of which,
pack_relocations, takes a Python dictionary as input. This type of thing makes
having more uniform type checking impossible (like by passing all arguments in
GN formatting).

The strip_libraries_for_device script is removed, it is never referenced.

BUG=573132, 571022
==========

[brettw, 2016-08-18 20:21:06]

more

[brettw, 2016-08-18 20:21:08]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-18 20:21:39]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-18 22:28:00]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-18 22:28:01]

Dry run: Try jobs failed on following builders:
  win_chromium_x64_rel_ng on master.tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_x64_...)

[Dirk Pranke, 2016-08-20 00:07:59]

dpranke@chromium.org changed reviewers:
+ jbudorick@chromium.org

[Dirk Pranke, 2016-08-20 00:08:00]

lgtm, though jbudorick should probably stamp it as well ...

[jbudorick, 2016-08-22 02:40:11]

lgtm

[brettw, 2016-08-22 17:30:00]

The CQ bit was checked by brettw@chromium.org

[commit-bot: I haz the power, 2016-08-22 17:30:30]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-22 19:05:27]

Description was changed from

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. Another of which,
pack_relocations, takes a Python dictionary as input. This type of thing makes
having more uniform type checking impossible (like by passing all arguments in
GN formatting).

The strip_libraries_for_device script is removed, it is never referenced.

BUG=573132, 571022
==========

to

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. Another of which,
pack_relocations, takes a Python dictionary as input. This type of thing makes
having more uniform type checking impossible (like by passing all arguments in
GN formatting).

The strip_libraries_for_device script is removed, it is never referenced.

BUG=573132, 571022
==========

[commit-bot: I haz the power, 2016-08-22 19:05:29]

Committed patchset #13 (id:220001)

[commit-bot: I haz the power, 2016-08-22 19:09:48]

Description was changed from

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. Another of which,
pack_relocations, takes a Python dictionary as input. This type of thing makes
having more uniform type checking impossible (like by passing all arguments in
GN formatting).

The strip_libraries_for_device script is removed, it is never referenced.

BUG=573132, 571022
==========

to

==========
Use gn_helpers to [se]serialize GN lists.

Removes GYP/GN command-line parsing logic. My read of the history is that we
used to take strings and FileArgs, then added support for GN lists using
ast.parse which is good enough for 95% of things (we didn't have gn_helpers to
do it properly yet). The GN list usage expanded to things that aren't actually
GN lists which made it impossible to do proper GN list parsing.

Now that we don't need to support GYP this removes the GYP variant and forces it
to use a GN version instead. The magic "##"->"$" remapping behavior is removed
from the scripts as well as the BUILD files (this was for GYP compat).

This also changes the FileArg expansion to write GN lists. The logic is
preserved from the previous version even though it's a bit weird. These are are
expanded either as a list beginning with '[', or as a string with no quotes.
This FileArgs processing is used to do things like extract #define values from
build_config files. One of which, java_libraries_list, is a string in the
build_config, but expects to be inserted into a Java file with no quotes so the
{ ... } in the string expand to a Java array. Another of which,
pack_relocations, takes a Python dictionary as input. This type of thing makes
having more uniform type checking impossible (like by passing all arguments in
GN formatting).

The strip_libraries_for_device script is removed, it is never referenced.

BUG=573132, 571022

Committed: https://crrev.com/1783edf919dd8a81d47bc58b7f91af65f79cb175
Cr-Commit-Position: refs/heads/master@{#413498}
==========

[commit-bot: I haz the power, 2016-08-22 19:09:50]

Patchset 13 (id:??) landed as
https://crrev.com/1783edf919dd8a81d47bc58b7f91af65f79cb175
Cr-Commit-Position: refs/heads/master@{#413498}