FX_CMapDwordToDword considered harmful.

FX_CMapDwordToDword considered harmful.

Lookups are log(n), but random insertions could result in n^2
behaviour.  Replace with maps and sets.

R=thestig@chromium.org

Committed: https://pdfium.googlesource.com/pdfium/+/85d5c4af4a9546970b34dd413c473d10fef8534b

[Tom Sepez, 2015-08-17 23:00:33]

tsepez@chromium.org changed reviewers:
+ thestig@chromium.org

[Tom Sepez, 2015-08-17 23:00:33]

Lei, for review.

[Lei Zhang, 2015-08-17 23:45:13]

lgtm

Be sure to do: git gs FX_CMapDwordToDword and make sure there's not any more
laying around.

https://codereview.chromium.org/1289703003/diff/1/core/src/fpdfapi/fpdf_font/...
File core/src/fpdfapi/fpdf_font/ttgsubtable.cpp (right):

https://codereview.chromium.org/1289703003/diff/1/core/src/fpdfapi/fpdf_font/...
core/src/fpdfapi/fpdf_font/ttgsubtable.cpp:99: if (m_featureMap.find(i) ==
m_featureMap.end()) {
Assuming we started with an empty map on line 95 and the for-loop as is, this
will eval to true every time.

[Tom Sepez, 2015-08-18 16:18:16]

On 2015/08/17 23:45:13, Lei Zhang wrote:
> lgtm
> 
> Be sure to do: git gs FX_CMapDwordToDword and make sure there's not any more
> laying around.
Done.

[Tom Sepez, 2015-08-18 16:20:33]

Committed patchset #2 (id:20001) manually as
85d5c4af4a9546970b34dd413c473d10fef8534b (presubmit successful).

[Tom Sepez, 2015-08-18 16:21:42]

https://codereview.chromium.org/1289703003/diff/1/core/src/fpdfapi/fpdf_font/...
File core/src/fpdfapi/fpdf_font/ttgsubtable.cpp (right):

https://codereview.chromium.org/1289703003/diff/1/core/src/fpdfapi/fpdf_font/...
core/src/fpdfapi/fpdf_font/ttgsubtable.cpp:99: if (m_featureMap.find(i) ==
m_featureMap.end()) {
On 2015/08/17 23:45:13, Lei Zhang wrote:
> Assuming we started with an empty map on line 95 and the for-loop as is, this
> will eval to true every time.
Indeed it will.