Re: [Zint-barcode] Possible MaxiCode encoding error

[Ed D. <ede...@gm...>]

Hi Herald,

Yes. This change looks good to me.

Thank you for taking it, and thank you for taking on the task of
maintaining the Zint C/C++ repository.

Ed

On Mon, Aug 17, 2015 at 5:52 AM, Harald Oehlmann
<har...@el...> wrote:
> Ed,
>
> thank you providing the patch.
>
> Is this ok for you:
> http://sourceforge.net/u/oehhar/zint/
> ?
>
> Thank you,
> Harald
>
> Am 12.08.2015 um 05:01 schrieb Ed Deinstadt:
>>
>> I have a code change which fixes the issue below. The changes are in
>> maxicode.c.
>>
>> There are actually two issues showing up in this test. First this code:
>>      /* Number compression */
>>      char substring[11];
>>      int value;
>>
>>      for(j = 0; j < 10; j++) {
>>          substring[j] = character[i + j];
>>      }
>>      substring[10] = '\0';
>>      value = atoi(substring);
>>
>> It extracts 10 characters from character, puts them in substring and
>> null terminates the new string. It should however only extract 9
>> characters as the maxicode algorithm only allows for the compression
>> of 9 numeric characters in a single run. With the current code, and a
>> run of 10 or more digits available, "value" will be too large (10
>> digits worth). The fix is straightforward:
>>
>>      /* Number compression */
>>      char substring[10];
>>      int value;
>>
>>      for(j = 0; j < 9; j++) {
>>          substring[j] = character[i + j];
>>      }
>>      substring[9] = '\0';
>>      value = atoi(substring);
>>
>> After making the above change, the data I supplied is now decoded
>> properly, except for the last zero. The following loop:
>>
>>      /* Add shift and latch characters */
>>      current_set = 1;
>>      i = 0;
>>      do {
>>
>>         if(set[i] != current_set ) {
>>              switch(set[i]) {
>>              case 1:
>>                   if(set[i + 1] == 1) {
>>                       if(set[i + 2] == 1) {
>>                           if(set[i + 3] == 1) {
>>                                /* Latch A */
>>                                       .
>>                                       .
>>                                       .
>>
>>               case 6:
>>                     /* Number Compressed */
>>                     /* Do nothing */
>>                     break;
>>               }
>>               i++;
>>           }
>>           i++;
>>      } while(i < 145);
>>
>> skips over the character after each set[i] value of 6 when setting the
>> shift and latch characters (i is incremented twice even though no new
>> Shift/Latch control characters have been inserted). This change fixes
>> this issue:
>>
>>      /* Add shift and latch characters */
>>      current_set = 1;
>>      i = 0;
>>      do {
>>
>>         if(set[i] != current_set && set[i] != 6) {
>>                                       .
>>                                       .
>>                                       .
>> Please consider putting these changes into the Zint library.
>>
>> Thanks
>> Ed Deinstadt
>>
>>
>>
>>
>>
>> On Mon, Aug 10, 2015 at 12:48 PM, Ed Deinstadt <ede...@gm...>
>> wrote:
>>>
>>> Hi,
>>>
>>> I am a c/C++ developer, I am new to Zint. I have been using the Zint
>>> backend library (version 2.4.3 cloned from
>>> git://git.code.sf.net/p/zint/code zint-code). I am producing MaxiCode
>>> symbols, but I have come across what seems to be an error when
>>> encoding some specific data in a MaxiCode symbol. I can reproduce the
>>> symptom in Zint BarCode studio 2.4. I am encoding the following data
>>> in Mode 4.
>>>
>>> When I encode the data: "MaxiCode
>>> ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz1234567890"
>>> (without the quotes), and scan the symbol my scanner, a Honeywell
>>> 4600g, I get "MaxiCode
>>> ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz160826066,".
>>>
>>> The data returned by the scanner matches the input except for the last
>>> 9 bytes. I have observed that if I shorten the data by removing the
>>> last character(the zero), the data returned matches the input.
>>>
>>> I was able to successfully encode, and scan the data above with the
>>> ByteScout encoder available from:
>>> http://bytescout.com/files/BarCodeGenerator.exe
>>>
>>> I have attached .png file (ZintMaxiCode.png) generated by Zint BarCode
>>> Studio (which does not scan properly), and a .png file generated by
>>> ByteScout's BarCodeGenerator which I have scanned properly.
>>>
>>> Has anyone seen this issue before? Is there a fix available? Perhaps I
>>> have exceeded some limit. I haven't been able to find a post, but like
>>> I said I am new to Zint...
>>>
>>> Thanks for any help!
>>>
>>> Ed Deinstadt