replace std::map with <ext/hash_map>

[g]

hello!

I wanna replace an std::map<std::s tring,Services* >
with hash_map.How I will do this?

any link with examples?

transactions.in sert(std::pair< std::string,Ser vices*>("Aservi ce",new
xxxx));
transactions.in sert(std::pair< std::string,Ser vices*>("Vservi ce",new
fffff));
transactions.in sert(std::pair< std::string,Ser vices*>("Rservi ce",new
ttt));
...............

how I should replace this with hash_map?

thanks for any help.

[Victor Bazarov]

g wrote:

I wanna replace an std::map<std::s tring,Services* >
with hash_map.How I will do this?
What's "hash_map"? There is no such standard construct (yet, anyway).
any link with examples?
Have you tried Google?
transactions.in sert(std::pair< std::string,Ser vices*>("Aservi ce",new
xxxx));
transactions.in sert(std::pair< std::string,Ser vices*>("Vservi ce",new
fffff));
transactions.in sert(std::pair< std::string,Ser vices*>("Rservi ce",new
ttt));
...............

how I should replace this with hash_map?

It depends entirely what 'hash_map' exposes and what interface it has.
Do you think that knowing what "hash_map" is a prerequisite here? It's
not. Does it comply with the requirements for an associative container?
If it does, you probably need to use 'hash_map::valu e_type' instead of
'std::pair'. Maybe it's the same thing. Maybe not. Have you actually
tried compiling to see if it fails? If yes, where does it fail and with
what error message? If not, why not?

V
--
Please remove capital 'A's when replying by e-mail
I do not respond to top-posted replies, please don't ask

[Howard Hinnant]

In article <e2**********@n ews.datemas.de> ,
"Victor Bazarov" <v.********@com Acast.net> wrote:

g wrote:

I wanna replace an std::map<std::s tring,Services* >
with hash_map.How I will do this?

What's "hash_map"? There is no such standard construct (yet, anyway).

Fwiw, when (not if) it becomes standard, the likely name is
std::unordered_ map.

-Howard

[g]

hi !
I am googling but I haven't found something......

I wanna use the gnu/sgi extension

this code doesn't compile,it's from my std::map
transactions.in sert(std::pair< std::string,Ser vices*>("Aservi ce",new
xxxx));

I want to convert it to hash_map.....

hash_map<std::s tring,Services* ,hash<std::stri ng>> transactions;

this is the correct declaration?

[Howard Hinnant]

In article <11************ **********@t31g 2000cwb.googleg roups.com>,
"g" <ge*******@yaho o.com> wrote:

hash_map<std::s tring,Services* ,hash<std::stri ng>> transactions;

Try:

hash_map<std::s tring,Services* ,hash<std::stri ng> > transactions;
^
space

Fwiw, that will be fixed in C++0X too... :-)

-Howard

[Alex Buell]

On Thu, 20 Apr 2006 21:01:31 GMT Howard Hinnant
<ho************ @gmail.com> waved a wand and this message magically
appeared:

hash_map<std::s tring,Services* ,hash<std::stri ng> > transactions;
^
space

Fwiw, that will be fixed in C++0X too... :-)

What will it look like in C++0X? I do find that '<some type<different
type> >' syntax a bit ugly.

--
http://www.munted.org.uk

Take a nap, it saves lives.

[Howard Hinnant]

In article <20************ *************** *****@munted.or g.uk>,
Alex Buell <al********@mun ted.org.uk> wrote:

On Thu, 20 Apr 2006 21:01:31 GMT Howard Hinnant
<ho************ @gmail.com> waved a wand and this message magically
appeared:

hash_map<std::s tring,Services* ,hash<std::stri ng> > transactions;
^
space

Fwiw, that will be fixed in C++0X too... :-)

What will it look like in C++0X? I do find that '<some type<different
type> >' syntax a bit ugly.

Sorry for the abbreviated message. My impression is that:

some_template<s ome_other_templ ate<some_type> > my_object;

will be equivalent to:

some_template<s ome_other_templ ate<some_type>> my_object;

But that's just my personal (and rushed) interpretation.

My latest factual information is that:

http://www.open-std.org/jtc1/sc22/wg...005/n1757.html

has been adopted into the C++0X working draft (which should be publicly
available in a few more days).

-Howard

[g]

thanks for your comments.

I still have a problem:

_gnu_cxx::hash_ map<std::string ,Services*, __gnu_cxx::hash <std::string>

transactions

transactions["CreateUser "]=new CreateUser;

the compiler give me this :

error: expected unqualified-id before '[' token

how to deal with?

thanks

[Jeff Flinn]

g wrote:

thanks for your comments.

I still have a problem:

_gnu_cxx::hash_ map<std::string ,Services*, __gnu_cxx::hash <std::string>

transactions

Does your code include a ';' here? or was this a typo?

transactions["CreateUser "]=new CreateUser;

the compiler give me this :

error: expected unqualified-id before '[' token

how to deal with?

See above. Also I assume you have the appropriate includes.

Jeff Flinn

[g]

>Does your code include a ';' here? or was this a typo?

yes it does include ;.

I use <ext/hash_map>