Add an encoder/decoder for quoted-printable.

Adds a way to encode/decode quoted-printable strings.
This is going to be used for MHTML support.

BUG=30863
TEST=Run the unit-tests.

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=76273

[Avi (use Gerrit), 2011-02-22 23:09:51]

If you're writing code to implement the bug, can you switch the bug from
unconfirmed? :)

[Avi (use Gerrit), 2011-02-22 23:10:46]

If you're writing code to implement the bug, can you switch the bug from
unconfirmed? :)

[brettw, 2011-02-25 03:05:25]

This doesn't really sound like something that should be in base. Does it need to
be there? Can it be somewhere like chrome/browser/net instead?

[Jay Civelli, 2011-02-28 06:23:27]

On 2011/02/25 03:05:25, brettw wrote:
> This doesn't really sound like something that should be in base. Does it need
to
> be there? Can it be somewhere like chrome/browser/net instead?
May be, not sure. I put it in there because that is where the base64 code is 
and this is somehow similar. If you feel strongly about it, I can certainly move
it to chrome/browser/net.

[brettw, 2011-02-28 06:32:41]

On Sun, Feb 27, 2011 at 10:23 PM,  <jcivelli@chromium.org> wrote:
> On 2011/02/25 03:05:25, brettw wrote:
>>
>> This doesn't really sound like something that should be in base. Does it
>> need
>
> to
>>
>> be there? Can it be somewhere like chrome/browser/net instead?
>
> May be, not sure. I put it in there because that is where the base64 code is
> and this is somehow similar. If you feel strongly about it, I can certainly
> move
> it to chrome/browser/net.

Yeah, if it doesn't need to be in base, it shouldn't be there. Base is
so big, it's not clear there will be another user of this code, and we
can always move it later if there is.

Brett

[Jay Civelli, 2011-02-28 17:20:37]

On 2011/02/28 06:32:41, brettw wrote:
> On Sun, Feb 27, 2011 at 10:23 PM,  <mailto:jcivelli@chromium.org> wrote:
> > On 2011/02/25 03:05:25, brettw wrote:
> >>
> >> This doesn't really sound like something that should be in base. Does it
> >> need
> >
> > to
> >>
> >> be there? Can it be somewhere like chrome/browser/net instead?
> >
> > May be, not sure. I put it in there because that is where the base64 code is
> > and this is somehow similar. If you feel strongly about it, I can certainly
> > move
> > it to chrome/browser/net.
> 
> Yeah, if it doesn't need to be in base, it shouldn't be there. Base is
> so big, it's not clear there will be another user of this code, and we
> can always move it later if there is.
> 
> Brett
OK, moved to chrome/browser/net.

[brettw, 2011-02-28 17:33:14]

LGTM

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable.cc (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:20: int HexASCIICharToInt(char c) {
You should be able to use HexDigitToInt in string_util instead of writing this.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:41: iter!= input.end(); ++iter) {
Space before the !

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:125: char r = static_cast<char>((i1 << 4
& 0xF0) | (i2 &0x0F));
Can you use an extra set of parens around the first part, I'm not sure of the
order pf operations for << and &.

Space after the second &

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable.h (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.h:14: namespace chrome_browser_net {
We're trying to standardize on using nested namespaces. Although this doesn't
match the existing code, I'd prefer doin the new way, so this would be
namespace chrome {
namespace browser {
namespace net {

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable_unittest.cc (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable_unittest.cc:76:
"=0E=0F=10=11=12=13=14=15=16=17=18=19=1A=1B=1C=1D=1E=1F !\"#$%&'()*+,-./01234="
Looks like these are too long.

[tfarina, 2011-02-28 18:32:44]

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable.cc (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:47: if (c >= 33 && c <= 126 && c != 61) {
Is there a reason for not using the letters instead? I think they will make the
code more easy to read. Same thing for the other occurrences below.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable.h (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.h:27: int IsEOL(const
std::string::const_iterator& iter, const std::string& input);
Is this going to be used by the consumers (i.e is this part of the public API as
the two functions above)? If not, can you move this to the source file?

[tfarina, 2011-02-28 18:38:20]

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable.cc (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:14: const char kHexTable[] = {
You can simplify this as:

const char kHexTable[] = "0123456789ABCDEF";

[Jay Civelli, 2011-02-28 19:25:53]

All comments addressed.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable.cc (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:14: const char kHexTable[] = {
On 2011/02/28 18:38:21, tfarina wrote:
> You can simplify this as:
> 
> const char kHexTable[] = "0123456789ABCDEF";

Done.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:20: int HexASCIICharToInt(char c) {
On 2011/02/28 17:33:14, brettw wrote:
> You should be able to use HexDigitToInt in string_util instead of writing
this.

Done.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:41: iter!= input.end(); ++iter) {
On 2011/02/28 17:33:14, brettw wrote:
> Space before the !

Done.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:47: if (c >= 33 && c <= 126 && c != 61) {
On 2011/02/28 18:32:44, tfarina wrote:
> Is there a reason for not using the letters instead? I think they will make
the
> code more easy to read. Same thing for the other occurrences below.

Done.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.cc:125: char r = static_cast<char>((i1 << 4
& 0xF0) | (i2 &0x0F));
On 2011/02/28 17:33:14, brettw wrote:
> Can you use an extra set of parens around the first part, I'm not sure of the
> order pf operations for << and &.
> 
> Space after the second &

Done.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable.h (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.h:14: namespace chrome_browser_net {
On 2011/02/28 17:33:14, brettw wrote:
> We're trying to standardize on using nested namespaces. Although this doesn't
> match the existing code, I'd prefer doin the new way, so this would be
> namespace chrome {
> namespace browser {
> namespace net {

Done.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable.h:27: int IsEOL(const
std::string::const_iterator& iter, const std::string& input);
On 2011/02/28 18:32:44, tfarina wrote:
> Is this going to be used by the consumers (i.e is this part of the public API
as
> the two functions above)? If not, can you move this to the source file? 
As the comment mentions, this is also used by the unit-tests, this is why it is
exposed.

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
File chrome/browser/net/quoted_printable_unittest.cc (right):

http://codereview.chromium.org/6542069/diff/15001/chrome/browser/net/quoted_p...
chrome/browser/net/quoted_printable_unittest.cc:76:
"=0E=0F=10=11=12=13=14=15=16=17=18=19=1A=1B=1C=1D=1E=1F !\"#$%&'()*+,-./01234="
On 2011/02/28 17:33:14, brettw wrote:
> Looks like these are too long.

Done.