tree fcdb532e22827e19e83d878bc53dfdefa4a7c8ab
parent 3c2bfd785f0f0c8a721241a37e1f2c6e6d85fec2
author Andy Phan <andyphan@chromium.org> 1738280141 -0800
committer Pdfium LUCI CQ <pdfium-scoped@luci-project-accounts.iam.gserviceaccount.com> 1738280141 -0800

Fix unnecessary tree traversal in SearchNameNodeByNameInternal()

When SearchNameNodeByNameInternal() is searching the name tree for
`csName`, it encounters nodes with limits, designating a range of what
possible values can be found in its subtree. Logically, when one is
searching for `csName`, and encounters a node where `csName` is greater
than the node's greatest limit, that subtree should be skipped, and the
next sibling node (if it exists) should be processed instead. This was
the behavior prior to https://pdfium-review.googlesource.com/8271, but
that CL introduced a regression where the subtree still gets traversed
anyway. This behavior might make sense when trying to add to the name
tree, where there needs to be some tree balancing, but it is unnecessary
when only searching the tree, leading to slow performance for PDFs with
large name trees.

Fix this behavior by skipping the subtree when only trying to search for
`csName`.

Bug: 372523840
Change-Id: I805e0a32fbd6ffdbece6f831f1be6d3a977587fc
Reviewed-on: https://pdfium-review.googlesource.com/c/pdfium/+/128231
Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Andy Phan <andyphan@chromium.org>