xref: /openbsd-src/gnu/gcc/libstdc++-v3/docs/html/27_io/binary_iostreams_kuehl.txt (revision 404b540a9034ac75a6199ad1a32d1bbc7a0d4210)
1
2From: kuehl@ramsen.informatik.uni-konstanz.de (Dietmar Kuehl)
3Newsgroups: comp.std.c++
4Subject: Re: binary iostreams ?
5Date: Sat,  3 Feb 2001 17:17:49 GMT
6Message-ID: <95hctq$suu$2@news.BelWue.DE>
7
8Hi,
9Plinio Conti (plinio.contiNO@SPAMMINGmclink.it) wrote:
10: Why std c++ library stream classes are only text-oriented?
11
12There is only a text oriented front end to stream buffers because text
13input and output does not vary between platforms. This is very
14different for binary output. For example, binary output has to consider
15
16- word sizes: Is an 'int' two, four, or eight bytes long? The same
17  questions arise for all other built-in types.
18
19- what is the bit pattern of a value? I think that at least implicitly
20  in the standard a binary representation for integer types is required.
21  I don't think that it is required to use two's complement. In any
22  case, the floating point representations do differ, eg. in their
23  number of bytes used.
24
25- what "endianess" is to be used?
26
27Basically it is possible to decide a format for each of those. This,
28however, implies inefficient implementations on platforms where the
29format does not match the internal representation.
30
31What many people asking for binary I/O forget is that binary I/O also
32requires some form of formatting! Assuming that just writing data and
33then reading it in will work is asking for problems, eg. when the
34compiler version changes and they decided to use a 32 bit integer
35rather than a 16 bit integer: It is not even necessary to switch
36platforms to run into problems!
37
38: I mean, if I want to write an int, a float, etc. AS IT IS I can't use
39: streams, because they write and read a human readable text format of
40: numbers.
41
42Which is for most I/O a reasonable approach. If it is not for you, you
43might want to consider a data base: File I/O is not really useful as a
44persistance mechanism. It is fine eg. for user interaction (text I/O),
45logging (text I/O), cross platfrom program interaction (formatted I/O),
46and data exchange (formatted I/O). In all these cases, the I/O is
47formatted, although possible using a binary format. For persistance,
48data bases are used. Depending on your needs, a relational or an object
49oriented one may be better suited.
50
51That said, it is worth to mention that it is easy to create a hierarchy
52similar to IOStreams built on top of stream buffers but doing binary
53formatting. A somewhat aged example is found at
54<ftp://ftp.fmi.uni-konstanz.de/pub/algo/personal/kuehl/binio.tar.gz>.
55This uses XDR formatting of the binary data (well, if I remmeber
56correctly, it is easy to plug in a different binary formatting).
57
58: Does anyone know how to solve the problem?
59
60Use a data base, text formatting, or binary formatting. With the
61details you have given it is impossible to tell which of those is the
62right approach because you haven't told *why* you want a binary format
63and *what* you want to do. That basically means that you came up with
64solution and you want us to confirm that it is the right one without
65telling us what problem is solved! Until I have seen the problem I
66doubt that binary I/O is the right approach...
67
68... and, BTW, using 'std::istream::read()' and 'std::ostream::write()'
69is almost certainly the *wrong* approach! These functions are an
70historical mistake which should have been corrected in the standard:
71It is my understanding that these methods were present in the IOStream
72version predating the rework from Jerry Schwartz and were left in to
73be compatible with the earlier stuff although they were not necessary:
74You could get binary I/O from the stream buffer level. The original
75IOStream library (maybe you remember using <stream.h>) did not have
76stream buffers and thus basic support for binary I/O was also present
77on the streams level.
78
79: What do you think about this choice?
80
81When I wrote the above paragraph about confirming your choice, I haven't
82read this question! As I said above: You told us what solution you have
83choosen without stating what problem is solved. We cannot determine
84whether your choice is the right one. Actually, I'm pretty sure it is
85the wrong one but without seen the details I can't be certain.
86--
87<mailto:dietmar_kuehl@yahoo.com> <http://www.dietmar-kuehl.de/>
88Phaidros eaSE - Easy Software Engineering: <http://www.phaidros.com/>
89
90