Add a utility class to retrieve the components of an MSI product

Add a utility class to retrieve the components of an MSI product

This will be used to map dlls to a particular program.

BUG=717696
Review-Url: https://codereview.chromium.org/2854143005
Cr-Commit-Position: refs/heads/master@{#471193}
Committed: https://chromium.googlesource.com/chromium/src/+/50cb83d908964cb4f1c14c78d7943ae2fcf5880f

[Patrick Monette, 2017-05-03 19:29:51]

pmonette@chromium.org changed reviewers:
+ chrisha@chromium.org

[Patrick Monette, 2017-05-03 19:29:52]

PTAL. This CL addressed the comment raised in the bigger CL.

[chrisha1, 2017-05-05 20:01:40]

chrisha@google.com changed reviewers:
+ chrisha@google.com

[chrisha1, 2017-05-05 20:01:41]

Nice! A handful of nits.

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
File chrome/browser/conflicts/msi_util_win.cc (right):

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:22: if
(::MsiGetProductInfo(product_guid.c_str(), INSTALLPROPERTY_LOCALPACKAGE,
Document that you're querying first for the required buffer length.

OR

If you know this is typically less than say, MAX_PATH, then you could start with
a string of this size and only call out a second time if the result doesn't fit.

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:45: buffer_size++;
Ditto for this read. If you have reason to believe there's a reasonable starting
buffer size, start with that to minimize round trips.

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:89: if
(::MsiGetComponentPath(product_guid.c_str(), component_guid.c_str(), L"",
Ditto.

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:109: // So this function retrieves the
path to the installer, extract the component
extracts*

[Patrick Monette, 2017-05-06 00:54:37]

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
File chrome/browser/conflicts/msi_util_win.cc (right):

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:22: if
(::MsiGetProductInfo(product_guid.c_str(), INSTALLPROPERTY_LOCALPACKAGE,
On 2017/05/05 20:01:40, chrisha1 wrote:
> Document that you're querying first for the required buffer length.
> 
> OR
> 
> If you know this is typically less than say, MAX_PATH, then you could start
with
> a string of this size and only call out a second time if the result doesn't
fit.

Done. Let me know if you have any idea to make the code looks nicer.

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:45: buffer_size++;
On 2017/05/05 20:01:40, chrisha1 wrote:
> Ditto for this read. If you have reason to believe there's a reasonable
starting
> buffer size, start with that to minimize round trips.

Done.

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:89: if
(::MsiGetComponentPath(product_guid.c_str(), component_guid.c_str(), L"",
On 2017/05/05 20:01:40, chrisha1 wrote:
> Ditto.

Done.

https://codereview.chromium.org/2854143005/diff/20001/chrome/browser/conflict...
chrome/browser/conflicts/msi_util_win.cc:109: // So this function retrieves the
path to the installer, extract the component
On 2017/05/05 20:01:40, chrisha1 wrote:
> extracts*

Done.

[chrisha, 2017-05-09 18:57:21]

No brilliant ideas on how to make that code prettier, but that's a very common
pattern we see with MS APIs. So, lgtm!

[Patrick Monette, 2017-05-12 00:25:03]

Thanks!

[Patrick Monette, 2017-05-12 00:25:11]

The CQ bit was checked by pmonette@chromium.org

[Patrick Monette, 2017-05-12 00:25:11]

The patchset sent to the CQ was uploaded after l-g-t-m from chrisha@chromium.org
Link to the patchset: https://codereview.chromium.org/2854143005/#ps80001
(title: "Rebase")

[commit-bot: I haz the power, 2017-05-12 00:26:23]

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, 2017-05-12 02:43:51]

CQ is committing da patch.

Bot data: {"patchset_id": 80001, "attempt_start_ts": 1494548711213560,
"parent_rev": "6f29a51e720b94230809fec020c5098bd9aaf16b", "commit_rev":
"50cb83d908964cb4f1c14c78d7943ae2fcf5880f"}

[commit-bot: I haz the power, 2017-05-12 02:44:03]

Description was changed from

==========
Add a utility class to retrieve the components of an MSI product

This will be used to map dlls to a particular program.

BUG=717696
==========

to

==========
Add a utility class to retrieve the components of an MSI product

This will be used to map dlls to a particular program.

BUG=717696

Review-Url: https://codereview.chromium.org/2854143005
Cr-Commit-Position: refs/heads/master@{#471193}
Committed:
https://chromium.googlesource.com/chromium/src/+/50cb83d908964cb4f1c14c78d794...
==========

[commit-bot: I haz the power, 2017-05-12 02:44:04]

Committed patchset #5 (id:80001) as
https://chromium.googlesource.com/chromium/src/+/50cb83d908964cb4f1c14c78d794...