2010/11/1 Rao, Anant :

Hi, I have a ‘char *’. I need to be able to traverse thru the string in both fwd and reverse directions. Say, I have a string ‘world’. I will start with a fwd iter pointing at ‘w’ and start traversing. When I reach the letter ‘r’, I might need to start traversing in the backward direction i.e., the next char should be ‘o’ and I might again need to traverse fwd.

How will You be traversing? Won't operators ++ and -- on a plain iterator/char* suffice? Regards Kris

Thanks Krzysztof, Patrick! While it’s true that a char* seems to be equivalent to an iterator (in this case), I prefer the iterator/reverse_iterator much better. For example, if I use an iterator and it reaches end(), I can do --it to get to the last char. In a char * model, once I reach the NULL ptr (at the end of the string), I can’t back-track. Similarly, rbegin()/reverse_iter. Maintaining a sentinel value for the start of a string is more difficult (other than indices). IMHO, those are the solid advantages I see with an iterator. I re-read the article published by David Abrahams et al on ‘reverse_iterator’ (http://www.boost.org/doc/libs/1_44_0/libs/iterator/doc/reverse_iterator.html...) and felt it’s adequate for my purposes. Basically, I will use that reverse_iterator for both fwd and reverse traversals by using ++ and --. A couple of simple questions I have: - Why is it not possible to define a boost::iterator as easily as a boost::reverse_iterator on a char * string as in the above web link? (Again, it doesn’t block me – as I mentioned, I’ll just use a reverse_iter and use ++ and – on it to satisfy my needs. More of curiosity). - Looks like boost::reverse_iterator can’t be assigned to; can only be created. If I need to define it as a class member that needs to be made to point to diff strings as the program progresses, it doesn’t seem to be possible. So, I’m thinking of using a ptr to a boost::reverse_iterator so that I can re-seat the reverse_iterator to diff strings. Is this a convoluted approach or correct approach (given my needs)? If former, what’d be the correct approach? Thanks again, Anant From: boost-users-bounces@lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Patrick Horgan Sent: Monday, November 01, 2010 4:34 PM To: boost-users@lists.boost.org Subject: Re: [Boost-users] [Iterators] iterators over char * On 11/01/2010 01:16 PM, Rao, Anant wrote: Hi, I have a ‘char *’. I need to be able to traverse thru the string in both fwd and reverse directions. A char* is a reversible iterator already. Reversible iterators are modeled to look like pointers. Patrick The information contained in this email message and its attachments is intended only for the private and confidential use of the recipient(s) named above, unless the sender expressly agrees otherwise. Transmission of email over the Internet is not a secure communications medium. If you are requesting or have requested the transmittal of personal data, as defined in applicable privacy laws by means of email or in an attachment to email, you must select a more secure alternate means of transmittal that supports your obligations to protect such personal data. If the reader of this message is not the intended recipient and/or you have received this email in error, you must take no action based on the information in this email and you are hereby notified that any dissemination, misuse or copying or disclosure of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by email and delete the original message.

2010/11/2 Rao, Anant :

While it’s true that a char* seems to be equivalent to an iterator (in this case), I prefer the iterator/reverse_iterator much better. For example, if I use an iterator and it reaches end(), I can do --it to  get to the last char. In a char * model, once I reach the NULL ptr (at the end of the string), I can’t back-track.

I think You are confusing NULL pointer with null character.

Similarly, rbegin()/reverse_iter. Maintaining a sentinel value for the start of a string is more difficult (other than indices). IMHO, those are the solid advantages I see with an iterator.

I re-read the article published by David Abrahams et al on ‘reverse_iterator’ (http://www.boost.org/doc/libs/1_44_0/libs/iterator/doc/reverse_iterator.html...)

and felt it’s adequate for my purposes. Basically, I will use that reverse_iterator for both fwd and reverse traversals by using ++ and --.

I think You are confusing bidirectional iterator with reverse iterator

A  couple  of simple questions I have:

-          Why is it not possible to define a boost::iterator as easily as a boost::reverse_iterator on a char * string  as in the above web link? (Again, it doesn’t block me – as I mentioned, I’ll just use a reverse_iter and use ++ and – on it to satisfy my needs. More of curiosity).

Because char* is an iterator itself.

-          Looks like boost::reverse_iterator can’t be assigned to; can only be created. If I need to define it as a class member that needs to be made to point to diff strings as the program progresses, it doesn’t seem to be possible. So, I’m thinking of using a ptr to a boost::reverse_iterator so that I can re-seat the reverse_iterator to diff strings. Is this a convoluted approach or correct approach (given my needs)? If former, what’d be the correct approach?

I think it can be assigned to, but that is irrelevant, because You don't seem to need it anyway. Regards, Kris.

Hi, I need to iterate over a char string and I wrote the code as follows: char *s = "abcdefg"; iterator_range ir( &s[0], &s[ strlen(s) ] ); char *cp; for (cp = (char*)ir.begin(); cp != (char*)ir.end(); ++cp) { cout << "cp in ri [" << *cp << "]\n"; } boost::range_reverse_iterator< char*>::type irb = boost::rbegin( ir ); boost::range_reverse_iterator< char*>::type ire = boost::rend( ir ); for (; irb != ire; ++irb) { cout << "reverse cp [" << *irb << "]\n"; } This works like a charm – both in fwd and reverse traversal. However, what I want is – in reverse traversal also, I want to use the same pointer (“cp” in this case). In other words, I want to use the same pointer to traverse in both directions. Here’s what I want to be able to write, but get a compile error with both attempts and I was wondering if you could help. Option 1: for (cp = boost::rbegin(ir); cp != boost::rend(ir); ++cp) { cout << "cp reverse in ri [" << *cp << "]\n"; } Option 2: while( cp != ire ) { //Do something cp++ } Thanks so much in advance, Anant The information contained in this email message and its attachments is intended only for the private and confidential use of the recipient(s) named above, unless the sender expressly agrees otherwise. Transmission of email over the Internet is not a secure communications medium. If you are requesting or have requested the transmittal of personal data, as defined in applicable privacy laws by means of email or in an attachment to email, you must select a more secure alternate means of transmittal that supports your obligations to protect such personal data. If the reader of this message is not the intended recipient and/or you have received this email in error, you must take no action based on the information in this email and you are hereby notified that any dissemination, misuse or copying or disclosure of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by email and delete the original message.

Hi Kris, Thanks for a quick response! This is part of a larger issue that I'm addressing. So, there's absolutely no way to use the same ptr to traverse backwards? I tried the options based on my limited knowledge, but if there're other options, please let me know (using the same ptr). It's ironic a simple pointer could go fwd, but not back. Anant From: boost-users-bounces@lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Krzysztof Czainski Sent: Friday, December 10, 2010 9:00 AM To: boost-users@lists.boost.org Subject: Re: [Boost-users] Range_iterator - char* - reverse 2010/12/10 Rao, Anant mailto:Anant.Rao@ironmountain.com> char *s = "abcdefg"; iterator_range ir( &s[0], &s[ strlen(s) ] ); char *cp; boost::range_reverse_iterator< char*>::type irb = boost::rbegin( ir ); boost::range_reverse_iterator< char*>::type ire = boost::rend( ir ); for (; irb != ire; ++irb) { cout << "reverse cp [" << *irb << "]\n"; } This works like a charm - both in fwd and reverse traversal. However, what I want is - in reverse traversal also, I want to use the same pointer ("cp" in this case). In other words, I want to use the same pointer to traverse in both directions. Here's what I want to be able to write, but get a compile error with both attempts and I was wondering if you could help. Option 1: for (cp = boost::rbegin(ir); cp != boost::rend(ir); ++cp) { cout << "cp reverse in ri [" << *cp << "]\n"; } Option 2: while( cp != ire ) { //Do something cp++ } I think You can't do neither. Why do You want to do that? Regards, Kris The information contained in this email message and its attachments is intended only for the private and confidential use of the recipient(s) named above, unless the sender expressly agrees otherwise. Transmission of email over the Internet is not a secure communications medium. If you are requesting or have requested the transmittal of personal data, as defined in applicable privacy laws by means of email or in an attachment to email, you must select a more secure alternate means of transmittal that supports your obligations to protect such personal data. If the reader of this message is not the intended recipient and/or you have received this email in error, you must take no action based on the information in this email and you are hereby notified that any dissemination, misuse or copying or disclosure of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by email and delete the original message.

Hi Eric, Yes, it was a typo on my part to use ++ when I shd have had a --. But, I was talking about the compile error. That is, even if I had a '--', it would still complain. The offending line is: while( cp != ire ) { //Do something cp--; } Please bear with me as I (hopefully) make my requirement statement more precise: Fwd: iterator_range ir( &s[0], &s[ strlen(s) ] ); for (cp = (char*)ir.begin(); cp != (char*)ir.end(); ++cp) { cout << "cp in ri [" << *cp << "]\n"; } I want similar code (using 'cp') to traverse backwards using 'ir' or another container-in-reverse mode. Your help much appreciated. Thanks a lot again, Anant -----Original Message----- From: boost-users-bounces@lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Eric MALENFANT Sent: Friday, December 10, 2010 11:23 AM To: boost-users@lists.boost.org Subject: Re: [Boost-users] Range_iterator - char* - reverse De : Rao, Anant

So, there's absolutely no way to use the same ptr to traverse backwards? I tried the options based on my limited knowledge, but if there're other options, please let me know (using the same ptr).

It's ironic a simple pointer could go fwd, but not back.

Well, it can go forward or back, but you have to ask it. "++" to go forward, "--" to go back.

char *cp; boost::range_reverse_iterator< char*>::type irb = boost::rbegin( ir ); boost::range_reverse_iterator< char*>::type ire = boost::rend( ir ); for (; irb != ire; ++irb) { cout << "reverse cp [" << *irb << "]\n"; }

This works like a charm - both in fwd and reverse traversal.

This works because irb is a reverse iterator, with operator++ defined to go "backwards". That behavior will not change because the initial value happens to come from a reverse_iterator. A char* is a char*, regardless of the source of its initial value.

Option 1: for (cp = boost::rbegin(ir); cp != boost::rend(ir); ++cp) { cout << "cp reverse in ri [" << *cp << "]\n"; }

cp is a char* here, and "++" on a char* is defined to increment it, moving it "forward". To move it "backward", use "--" _______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users The information contained in this email message and its attachments is intended only for the private and confidential use of the recipient(s) named above, unless the sender expressly agrees otherwise. Transmission of email over the Internet is not a secure communications medium. If you are requesting or have requested the transmittal of personal data, as defined in applicable privacy laws by means of email or in an attachment to email, you must select a more secure alternate means of transmittal that supports your obligations to protect such personal data. If the reader of this message is not the intended recipient and/or you have received this email in error, you must take no action based on the information in this email and you are hereby notified that any dissemination, misuse or copying or disclosure of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by email and delete the original message.