QRCode Scan Failure - Empty Data Mode 0100 (Byte encoding) #181

semireg · 2020-10-02T00:48:31Z

Hi,

I have an interesting bug reported by a user.

Consider these two QR Codes:

Created with BWIP

Created with TEC-IT

As far as I can tell, these should be the "same" but the BWIP version doesn't scan using a "Socket Mobile S740 barcode scanner." For what it's worth, these both scan using built-in iPhone QR scanner.

Here's where it gets interesting ... Using this online decoder/debugger https://qrlogo.kaarposoft.dk/qrdecode.html we see some interesting differences.

BWIP version:

skew_limit=3.984375
skew=0
left=4 right=253 top=2 bottom=251
size=250
findModuleSize matchVersion version=1 grade=0
matchVersion version=2 finder0=64 finder1=64 finder2=64
matchVersion version=2 timing0=1 timing1=1 alignment=1
matchVersion version=2 format_NW =9 0 format_NESW =9 1 format = 9 ecl = 1 mask = 1
matchVersion version=2 grades(F(V)TAF): 4444->4
findModuleSize matchVersion version=2 grade=4
findModuleSize version=2 grade=4 error_correction_level=1 mask=1
getCodewords mask=1 length=44
getCodewords = 64,2,9,169,126,144,160,78,171,190,119,240,116,235,169,240,144,13,125,41,8,0,236,17,236,17,236,17,236,17,236,17,236,17,171,155,39,228,235,171,69,33,167,7
setBlocks n_blocks_first=1 n_blocks_second=0 n_blocks=1 n_block_words_first=34 n_block_words_second=0 n_block_ec_words=10 total=44
setBlocks block 0 (44): 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43
RS calculateSyndroms: No errors
correctErrors in = 64,2,9,169,126,144,160,78,171,190,119,240,116,235,169,240,144,13,125,41,8,0,236,17,236,17,236,17,236,17,236,17,236,17,171,155,39,228,235,171,69,33,167,7
correctErrors out = 64,2,9,169,126,144,160,78,171,190,119,240,116,235,169,240,144,13,125,41,8,0,236,17,236,17,236,17,236,17,236,17,236,17
error_grade=4
extractData bytes in (34) = 64,2,9,169,126,144,160,78,171,190,119,240,116,235,169,240,144,13,125,41,8,0,236,17,236,17,236,17,236,17,236,17,236,17
extractData mode = 4
extract charcount = 0
extractData mode = 2
extractAlphanum charcount = 19
extractData mode = 4
extract charcount = 3
extractData mode = 0
extractData data(22) = 69,88,45,74,69,66,65,74,76,68,76,69,78,73,65,76,69,45,50,95,74,66

TEC-IT version:

skew_limit=2
skew=0
left=0 right=99 top=0 bottom=99
size=100
matchVersion version=1 finder0=34 finder1=34 finder2=32
findModuleSize matchVersion version=1 grade=0
matchVersion version=2 finder0=64 finder1=64 finder2=64
matchVersion version=2 timing0=1 timing1=1 alignment=1
matchVersion version=2 format_NW =9 0 format_NESW =9 1 format = 9 ecl = 1 mask = 1
matchVersion version=2 grades(F(V)TAF): 4444->4
findModuleSize matchVersion version=2 grade=4
findModuleSize version=2 grade=4 error_correction_level=1 mask=1
getCodewords mask=1 length=44
getCodewords = 32,154,151,233,10,4,234,187,231,127,7,78,186,159,9,0,215,210,144,128,236,17,236,17,236,17,236,17,236,17,236,17,236,17,169,61,21,72,169,146,31,204,44,131
setBlocks n_blocks_first=1 n_blocks_second=0 n_blocks=1 n_block_words_first=34 n_block_words_second=0 n_block_ec_words=10 total=44
setBlocks block 0 (44): 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43
RS calculateSyndroms: No errors
correctErrors in = 32,154,151,233,10,4,234,187,231,127,7,78,186,159,9,0,215,210,144,128,236,17,236,17,236,17,236,17,236,17,236,17,236,17,169,61,21,72,169,146,31,204,44,131
correctErrors out = 32,154,151,233,10,4,234,187,231,127,7,78,186,159,9,0,215,210,144,128,236,17,236,17,236,17,236,17,236,17,236,17,236,17
error_grade=4
extractData bytes in (34) = 32,154,151,233,10,4,234,187,231,127,7,78,186,159,9,0,215,210,144,128,236,17,236,17,236,17,236,17,236,17,236,17,236,17
extractData mode = 2
extractAlphanum charcount = 19
extractData mode = 4
extract charcount = 3
extractData mode = 0
extractData data(22) = 69,88,45,74,69,66,65,74,76,68,76,69,78,73,65,76,69,45,50,95,74,66

Same decoded bytes... but the thing that caught my eye is BWIP's possibly superfluous:

extractData mode = 4
extract charcount = 0

According to Wikipedia QR Encoding:

4 == 0100 == Byte encoding

Thoughts on this?

The text was updated successfully, but these errors were encountered:

metafloor · 2020-10-02T01:41:06Z

Please show the options/values used in bwip-js to generate the barcode.

semireg · 2020-10-02T01:42:07Z

Hi @metafloor, they are linked in the URL above.

https://bwipjs-api.metafloor.com/?bcid=qrcode&text=EX-JEBAJLDLENIALE-2_JB&eclevel=L&scale=2

metafloor · 2020-10-02T01:47:06Z

Please verify the following barcode with your scanner:

semireg · 2020-10-02T01:51:25Z

I've submitted to my customer. I'll report back when I hear from them. 👍 Thank you!

semireg · 2020-10-02T14:35:05Z

@metafloor, the customer reports this barcode is not scannable with their scanner. Hmm. 🤔

metafloor · 2020-10-02T15:17:01Z

@terryburton: Hope you are following this. Comparing the debug output provided for the BWIPP/bwip-js generated symbol:

extractData bytes in (34) = 64,2,9,169,126,144,160,78,171,190,119,240,116,235,169,240,144,13,125,41,8,0,236,17,236,17,236,17,236,17,236,17,236,17
extractData mode = 4
extract charcount = 0
extractData mode = 2
extractAlphanum charcount = 19
extractData mode = 4
extract charcount = 3
extractData mode = 0

Against the TEC-IT symbol:

extractData bytes in (34) = 32,154,151,233,10,4,234,187,231,127,7,78,186,159,9,0,215,210,144,128,236,17,236,17,236,17,236,17,236,17,236,17,236,17
extractData mode = 2
extractAlphanum charcount = 19
extractData mode = 4
extract charcount = 3
extractData mode = 0

Note how the BWIPP symbol starts in mode 4 and encodes zero bytes, then switches to mode 2. I am wondering if that bit of oddness is what the scanner is glitching on. That last barcode that I had the OP check was produced by BWIPP's online generator. It has identical encoding as the original bwip-js symbol. (bwip-js doesn't have all the new ECI updates you've recently added.)

terryburton · 2020-10-02T15:23:21Z

@metafloor:I suspect your analysis is correct. I'm working on the optimisation code now...

terryburton · 2020-10-02T15:37:42Z

@semireg Please could you ask the customer to test the following which has been generated using bwipp/postscriptbarcode@cad3e87 which fixes the erroneous byte-mode selection for the start:

semireg · 2020-10-02T15:38:22Z

@terryburton Will do!

semireg · 2020-10-02T17:48:15Z

The user has confirmed they can scan the latest barcode in this thread. 😄

terryburton · 2020-10-02T18:25:19Z

The user has confirmed they can scan the latest barcode in this thread.

Good. BWIPP 2020-10-02 released.

metafloor · 2020-10-02T19:32:22Z

Looks like the PS to JS trace compiler is having issues with some of the new code constructs. Will need to do some sleuthing this weekend to track it down...

metafloor · 2020-10-03T16:21:12Z

Working through the PS/JS translation issue is going to take more time than I have right now. So I have patched the current version of BWIPP that is being used by bwip-js with the additional optimization check that was added to the latest release. This is the barcode image produced by the patched version of bwip-js:

It is different from the one Terry uploaded as it is missing the ECI mode (not needed for ASCII data).

@semireg: Please have the customer verify against their scanner.

@terryburton: Thank you for the quick response!

terryburton · 2020-10-03T17:48:49Z

It is different from the one Terry uploaded as it is missing the ECI mode (not needed for ASCII data).

@metafloor LGTM. The only difference it that your symbol is generated with eclevel=L whereas mine is eclevel=M.

semireg · 2020-10-03T21:22:31Z

@metafloor, if you push to a feature branch I'll build a beta for my user to test on Monday morning. Thank you so much.

metafloor · 2020-10-04T16:01:29Z

Not sure why github auto-closed this issue. The patch passes all tests so pushing to the master branch. Please give it a try.

metafloor closed this as completed in d6a5c1e Oct 4, 2020

metafloor reopened this Oct 4, 2020

jabibi pushed a commit to nodoo/bwip-js that referenced this issue Dec 26, 2022

Patch to resolve metafloor#181

aa3ea42

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

QRCode Scan Failure - Empty Data Mode 0100 (Byte encoding) #181

QRCode Scan Failure - Empty Data Mode 0100 (Byte encoding) #181

semireg commented Oct 2, 2020 •

edited

Loading

metafloor commented Oct 2, 2020

semireg commented Oct 2, 2020

metafloor commented Oct 2, 2020

semireg commented Oct 2, 2020

semireg commented Oct 2, 2020

metafloor commented Oct 2, 2020

terryburton commented Oct 2, 2020

terryburton commented Oct 2, 2020

semireg commented Oct 2, 2020

semireg commented Oct 2, 2020

terryburton commented Oct 2, 2020

metafloor commented Oct 2, 2020

metafloor commented Oct 3, 2020

terryburton commented Oct 3, 2020 •

edited

Loading

semireg commented Oct 3, 2020

metafloor commented Oct 4, 2020 •

edited

Loading

QRCode Scan Failure - Empty Data Mode 0100 (Byte encoding) #181

QRCode Scan Failure - Empty Data Mode 0100 (Byte encoding) #181

Comments

semireg commented Oct 2, 2020 • edited Loading

metafloor commented Oct 2, 2020

semireg commented Oct 2, 2020

metafloor commented Oct 2, 2020

semireg commented Oct 2, 2020

semireg commented Oct 2, 2020

metafloor commented Oct 2, 2020

terryburton commented Oct 2, 2020

terryburton commented Oct 2, 2020

semireg commented Oct 2, 2020

semireg commented Oct 2, 2020

terryburton commented Oct 2, 2020

metafloor commented Oct 2, 2020

metafloor commented Oct 3, 2020

terryburton commented Oct 3, 2020 • edited Loading

semireg commented Oct 3, 2020

metafloor commented Oct 4, 2020 • edited Loading

semireg commented Oct 2, 2020 •

edited

Loading

terryburton commented Oct 3, 2020 •

edited

Loading

metafloor commented Oct 4, 2020 •

edited

Loading