156902 - Extremely unreadable code

Status: RESOLVED FIXED

(Core :: CSS Parsing and Computation, defect, P4)

Description

[Jochen Suckfuell]

Below is a piece of code from nsRuleNode.h that I stumbled upon and had a hard
time to read: Typecasting "this" to char* and adding an offset for accessing
members is very evil. At the very least a better documentation would be in
order. The best thing is the NOTE: guess what the note says in
nsStyleContext::SetStyle!

nsStyleStruct* CachedStyleData::GetStyleData(const nsStyleStructID& aSID) {
    // NOTE:  nsStyleContext::SetStyle works roughly the same way.
    const StyleStructInfo& info = gInfo[aSID];
    char* resetOrInheritSlot = NS_REINTERPRET_CAST(char*, this) +
info.mCachedStyleDataOffset;
    char* resetOrInherit = NS_REINTERPRET_CAST(char*,
*NS_REINTERPRET_CAST(void**, resetOrInheritSlot));
[...]

Comment 1

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Attachment: patch to add comments explaining the code

Comment 2

[Mike Young]

Do you need review and approval to check in a comment only patch? I was under
the impression that mozilla.org was in desperate need of documentation and there
was blanket permission. (I could be wrong.)

Also, there is a typo at the end where "for" is spelled "fore". Other than that
it looks good to me.

Comment 3

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 90967 [details] [diff] [review]
patch to add comments explaining the code

Ah, yes... this code.. ;)

s/fore/for/  as Mike said, and sr=bzbarsky.

Comment 4

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Fix checked into trunk, 2002-08-06 06:46 PDT.

Comment 5

[Madhur Bhatia]

this does look like a code level verification