# How to write data in binary to a file? **URL:** \ **Category:** ruby-talk\ **Created:** [10 September 2011 00:57 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555 "2011-09-10T00:57:27Z")\ **Posts on this page:** 11\ **Page:** 1
**Author:** ![frank\_hi](https://avatars.discourse-cdn.com/v4/letter/f/77aa72/32.png) [@frank\_hi](https://rubytalk.org/u/frank_hi)\ **Post date:** [10 September 2011 00:57 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/1 "2011-09-10T00:57:27Z")
Hi, I want to write the data into a file in binary mode. The file is opened in binary mode by ofile=File.open("filename",'wb'). However, when I write the data using ofile.write(1), ruby will write the ascii code 31 to the file. I want it to write one byte number withe value 1. How can I do it? I have been googling a lot and could not find the answer. Thanks Frank > **···** > > -- > Posted via [http://www.ruby-forum.com/](http://www.ruby-forum.com/). ---
**Author:** ![Eric\_Wong2](https://avatars.discourse-cdn.com/v4/letter/e/b9e5f3/32.png) [@Eric\_Wong2](https://rubytalk.org/u/Eric_Wong2)\ **Post date:** [10 September 2011 01:32 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/2 "2011-09-10T01:32:59Z")
You need to pack the data into a string before writing. The following will write one unsigned byte:   buffer = [1].pack("C")   file.write(buffer) See `ri Array#pack` for full documentation on things you can do with pack (and String#unpack for the inverse). There are also a nearly-identical `pack'/`unpack' functions in Perl. The "perlpacktut" manpage is a great read for this. > **···** > > frank hi \ wrote: > > > I want to write the data into a file in binary mode. The file is opened > > in binary mode by ofile=File.open("filename",'wb'). However, when I > > write the data using ofile.write(1), ruby will write the ascii code 31 > > to the file. I want it to write one byte number withe value 1. How can I > > do it? I have been googling a lot and could not find the answer. > > -- > Eric Wong ---
**Author:** ![7stud](https://avatars.discourse-cdn.com/v4/letter/7/57b2e6/32.png) [@7stud](https://rubytalk.org/u/7stud)\ **Post date:** [10 September 2011 03:48 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/3 "2011-09-10T03:48:24Z")
frank hi wrote in post #1021069: > Hi, > > I want to write the data into a file in binary mode. The file is opened > in binary mode by ofile=File.open("filename",'wb'). All that does is turn off newline conversions, so if you write \n to a file, \n will be written to the file no matter what operating system the program is running on. > **···** > > -- > Posted via [http://www.ruby-forum.com/\](http://www.ruby-forum.com/%5C). ---
**Author:** ![Dominic\_Rose](https://avatars.discourse-cdn.com/v4/letter/d/a3d4f5/32.png) [@Dominic\_Rose](https://rubytalk.org/u/Dominic_Rose)\ **Post date:** [10 September 2011 17:14 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/4 "2011-09-10T17:14:21Z")
Did you try with File#syswrite? > **···** > > -- > Posted via [http://www.ruby-forum.com/](http://www.ruby-forum.com/). ---
**Author:** ![Brian\_Candler](https://avatars.discourse-cdn.com/v4/letter/b/5f9b8f/32.png) [@Brian\_Candler](https://rubytalk.org/u/Brian_Candler)\ **Post date:** [10 September 2011 16:52 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/5 "2011-09-10T16:52:35Z")
7stud -- wrote in post #1021082: > > I want to write the data into a file in binary mode. The file is opened > > in binary mode by ofile=File.open("filename",'wb'). > > All that does is turn off newline conversions, so if you write \n to a > file, \n will be written to the file no matter what operating system the > program is running on. Well, and this: > > # === ... And sets external encoding to ASCII-8BIT unless explicitly specified. Oh no... yet another inconsistency from ruby 1.9. The documentation you quote is correct in what it says: irb(main):001:0\> f1 = File.open("/tmp/f1","w") =\> #\ irb(main):002:0\> f2 = File.open("/tmp/f2","wb") =\> #\ irb(main):003:0\> f1.external\_encoding =\> nil irb(main):004:0\> f2.external\_encoding =\> #\ However, the behaviour this implies is unexpected: irb(main):005:0\> f2.write("groß") =\> 5 Normally, transcoding a UTF-8 string (which contains non-ASCII characters) to ASCII-8BIT would raise an exception: irb(main):006:0\> "groß".encode("ASCII-8BIT") Encoding::UndefinedConversionError: U+00DF from UTF-8 to ASCII-8BIT But now we have a special case where it \*doesn't\* raise an exception ☹ Regards, Brian. (Tested with "ruby 1.9.2p0 (2010-08-18 revision 29036) [x86\_64-linux]") > **···** > > -- > Posted via [http://www.ruby-forum.com/\](http://www.ruby-forum.com/%5C). ---
**Author:** ![Phillip\_Gawlowski](https://avatars.discourse-cdn.com/v4/letter/p/ecc23a/32.png) [@Phillip\_Gawlowski](https://rubytalk.org/u/Phillip_Gawlowski)\ **Post date:** [10 September 2011 18:05 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/6 "2011-09-10T18:05:07Z")
You obviously aren't aware what Extended ASCII (aka ASCII 8bit) is. Take a look at ISO 8859-1, and check what 11011111 encodes. > **···** > > On Sat, Sep 10, 2011 at 6:52 PM, Brian Candler \ wrote: > > > irb(main):005:0\> f2.write("groß") > > =\> 5 > > > > Normally, transcoding a UTF-8 string (which contains non-ASCII > > characters) to ASCII-8BIT would raise an exception: > > -- > Phillip Gawlowski > > phgaw.posterous.com | twitter.com/phgaw | gplus.to/phgaw > > A method of solution is perfect if we can forsee from the start, > and even prove, that following that method we shall attain our aim. > -- Leibniz ---
**Author:** ![Brian\_Candler](https://avatars.discourse-cdn.com/v4/letter/b/5f9b8f/32.png) [@Brian\_Candler](https://rubytalk.org/u/Brian_Candler)\ **Post date:** [11 September 2011 03:15 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/7 "2011-09-11T03:15:20Z")
Phillip Gawlowski wrote in post #1021166: > > irb(main):005:0\> f2.write("gro") > > =\> 5 > > > > Normally, transcoding a UTF-8 string (which contains non-ASCII > > characters) to ASCII-8BIT would raise an exception: > > You obviously aren't aware what Extended ASCII (aka ASCII 8bit) is. I'm not? Then please explain it to me. What I know so far about ruby 1.9 encoding I have documented at > There are some 200+ behaviours there, reverse-engineered with tests. However, I'm quite happy to have gaps in my knowledge filled out. > Take a look at ISO 8859-1, and check what 11011111 encodes. What has ISO-8859-1 got to do with this? \*If\* the file had an external encoding of ISO-8859-1, then the character "ß" (two bytes in UTF-8) would have been translated to the single byte 0xDF as it was written out. But in this example, the external encoding is ASCII-8BIT, which is ruby's encoding for "ASCII characters in the low 128 values, and unknown binary in the high 128 values" Ruby does not let you transcode UTF-8 to ASCII-8BIT, or back again, if there are any high-value characters in it. You can confirm this easily: > > s1 = "groß" =\> "groß" > > s1.encode("ASCII-8BIT") Encoding::UndefinedConversionError: U+00DF from UTF-8 to ASCII-8BIT   from (irb):2:in `encode'   from (irb):2   from /usr/local/bin/irb192:12:in `\' > > s2 = "gro\xDF".force\_encoding("ASCII-8BIT") =\> "gro\xDF" > > s2.encode("UTF-8") Encoding::UndefinedConversionError: "\xDF" from ASCII-8BIT to UTF-8   from (irb):4:in `encode'   from (irb):4   from /usr/local/bin/irb192:12:in `\' Now, if you open a file for write and set an external encoding (the default is nil) it means "transcode to this encoding". But for some reason, setting external encoding to ASCII-8BIT bypasses this rule. > **···** > > > On Sat, Sep 10, 2011 at 6:52 PM, Brian Candler \ \> wrote: > > -- > Posted via [http://www.ruby-forum.com/\](http://www.ruby-forum.com/%5C). ---
**Author:** ![Robert\_K1](https://yyz1.discourse-cdn.com/flex029/user_avatar/rubytalk.org/robert_k1/32/1830_2.png) [@Robert\_K1](https://rubytalk.org/u/Robert_K1)\ **Post date:** [12 September 2011 11:20 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/8 "2011-09-12T11:20:56Z")
> Phillip Gawlowski wrote in post #1021166: > > > > irb(main):005:0\> f2.write("gro") > > > =\> 5 > > > > > > Normally, transcoding a UTF-8 string (which contains non-ASCII > > > characters) to ASCII-8BIT would raise an exception: > > > > You obviously aren't aware what Extended ASCII (aka ASCII 8bit) is. > > I'm not? Then please explain it to me. What I know so far about ruby 1.9 > encoding I have documented at > [string19/string19.rb at master · candlerb/string19 · GitHub](https://github.com/candlerb/string19/blob/master/string19.rb) irb(main):001:0\> u=Encoding::BINARY =\> #\ irb(main):002:0\> u.names =\> ["ASCII-8BIT", "BINARY"] In other words: binary and ASCII-8BIT are the same. I find the behavior totally consistent: when writing a String with \*any\* encoding to BINARY then simply bytes are dumped as is and \_no conversion\_ is done. So there cannot be an encoding error. irb(main):003:0\> s="groß" =\> "groß" irb(main):004:0\> s.size =\> 4 irb(main):005:0\> s.bytesize =\> 5 irb(main):006:0\> s.encoding =\> #\ irb(main):007:0\> File.open("x","wb"){|io| p io.external\_encoding, io.internal\_encoding ; io.write(s)} #\ nil =\> 5 irb(main):008:0\> File.stat("x").size =\> 5 Now I can read the file with encoding UTF-8 properly: irb(main):009:0\> t = File.open("x","r:UTF-8") {|io| io.read} =\> "groß" irb(main):010:0\> t.size =\> 4 irb(main):011:0\> t.bytesize =\> 5 irb(main):012:0\> t.encoding =\> #\ irb(main):013:0\> s == t =\> true > Now, if you open a file for write and set an external encoding (the > default is nil) it means "transcode to this encoding". But for some > reason, setting external encoding to ASCII-8BIT bypasses this rule. See above. The name ASCII-8BIT is probably not the best one to choose for this but if you think about it as BINARY then everything fits nicely together. Kind regards robert > **···** > > On Sun, Sep 11, 2011 at 5:15 AM, Brian Candler \ wrote: > > > > On Sat, Sep 10, 2011 at 6:52 PM, Brian Candler \ \>\> wrote: > > -- > remember.guy do |as, often| as.you\_can - without end > [http://blog.rubybestpractices.com/](http://blog.rubybestpractices.com/) ---
**Author:** ![Brian\_Candler](https://avatars.discourse-cdn.com/v4/letter/b/5f9b8f/32.png) [@Brian\_Candler](https://rubytalk.org/u/Brian_Candler)\ **Post date:** [12 September 2011 17:45 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/9 "2011-09-12T17:45:48Z")
Robert Klemme wrote in post #1021429: > irb(main):002:0\> u.names > =\> ["ASCII-8BIT", "BINARY"] > > In other words: binary and ASCII-8BIT are the same. Indeed. > I find the behavior totally consistent Consistent with what? If you set external\_encoding to something other than nil, you are telling ruby to transcode strings on output to the given encoding - \*unless\* the given encoding is "ASCII-8BIT"/"BINARY". So it seems that "ASCII-8BIT" is handled as a special case. In fact, looking at io.c, that encoding is handled as a special case all over the place: $ egrep -n "= rb\_ascii8bit" io.c 232:#define NEED\_WRITECONV(fptr) ((fptr-\>encs.enc != NULL && fptr-\>encs.enc != rb\_ascii8bit\_encoding()) || NEED\_NEWLINE\_DECORATOR\_ON\_WRITE(fptr) || (fptr-\>encs.ecflags & (ECONV\_DECORATOR\_MASK|ECONV\_STATEFUL\_DECORATOR\_MASK))) 774: if (!fptr-\>encs.enc || (fptr-\>encs.enc == rb\_ascii8bit\_encoding() && !fptr-\>encs.enc2)) { 924: else if (fptr-\>encs.enc != rb\_ascii8bit\_encoding()) 3948: fptr-\>encs.enc = rb\_ascii8bit\_encoding(); 4203: if (intern == NULL && ext != rb\_ascii8bit\_encoding()) > **···** > > -- > Posted via [http://www.ruby-forum.com/\](http://www.ruby-forum.com/%5C). ---
**Author:** ![Gary\_Wright](https://avatars.discourse-cdn.com/v4/letter/g/58956e/32.png) [@Gary\_Wright](https://rubytalk.org/u/Gary_Wright)\ **Post date:** [12 September 2011 18:54 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/10 "2011-09-12T18:54:13Z")
> > Now, if you open a file for write and set an external encoding (the > > default is nil) it means "transcode to this encoding". But for some > > reason, setting external encoding to ASCII-8BIT bypasses this rule. > > See above. The name ASCII-8BIT is probably not the best one to choose > for this but if you think about it as BINARY then everything fits > nicely together. I think the anomaly that Brian pointed out was that the transcoding during IO behaved differently than explicit transcoding: > Normally, transcoding a UTF-8 string (which contains non-ASCII > characters) to ASCII-8BIT would raise an exception: > > irb(main):006:0\> "groß".encode("ASCII-8BIT") > Encoding::UndefinedConversionError: U+00DF from UTF-8 to ASCII-8BIT > > But now we have a special case where it \*doesn't\* raise an exception ☹ It makes sense to me that transcoding to 'binary' (or 'ASCII-8BIT') would just copy bytes but it doesn't make sense that implicit transcoding via IO and explicit transcoding via #encode behave differently. Gary Wright > **···** > > On Sep 12, 2011, at 7:20 AM, Robert Klemme wrote: ---
**Author:** ![Robert\_K1](https://yyz1.discourse-cdn.com/flex029/user_avatar/rubytalk.org/robert_k1/32/1830_2.png) [@Robert\_K1](https://rubytalk.org/u/Robert_K1)\ **Post date:** [12 September 2011 21:58 UTC](https://rubytalk.org/t/how-to-write-data-in-binary-to-a-file/63555/11 "2011-09-12T21:58:30Z")
The only special thing is that it is the "encoding" of a String's raw data, I'd say. Cheers robert > **···** > > On Mon, Sep 12, 2011 at 7:45 PM, Brian Candler \ wrote: > > > Robert Klemme wrote in post #1021429: > > > > > irb(main):002:0\> u.names > > > =\> ["ASCII-8BIT", "BINARY"] > > > > > > In other words: binary and ASCII-8BIT are the same. > > > > Indeed. > > > > > I find the behavior totally consistent > > > > Consistent with what? If you set external\_encoding to something other > > than nil, you are telling ruby to transcode strings on output to the > > given encoding - \*unless\* the given encoding is "ASCII-8BIT"/"BINARY". > > > > So it seems that "ASCII-8BIT" is handled as a special case. In fact, > > looking at io.c, that encoding is handled as a special case all over the > > place: > > -- > remember.guy do |as, often| as.you\_can - without end > [http://blog.rubybestpractices.com/](http://blog.rubybestpractices.com/)