ECI: ECI 899 binary in `UNICODE_MODE` now converted from UTF-8,
manual: use modified "haddock.theme" for nicer syntax highlighting;
general: suppress clang-tidy-21/22 warnings;
Hi Ulrich, thanks for pointing this out. Eventually got this working by reverting to setting the ENVIRONMENT directly and backslashing semicolons in $ENV{PATH}, rather than using ENVIRONMENT_MODIFICATION and path_list_prepend. It seems to work now in the CI and locally, regards, Martin
CI: windows: try env_path backslashing semicolons hack
CI: windows: try adding backend_qt to PATH
test_qzint: try setting CMAKE_CURRENT_SOURCE_DIR
test suite: Windows: use old-fashioned setting of ENVIROMENT PATH
gs1: Use new `gs1_encoders_init_ex()` API;
win32: add "zint_dll_vc6" sub-directory with VC6 workspace for
Good to hear! Yes, it was built from the 2.15.0 source. I'll add the workspace to git in a while... thanks for the feedback, it's very useful to me to know how people are using zint, regards, Martin
Oh I see. I've attached a Zint 2.15.0 "zint.dll" (and its "zint.lib") which was built with VC6, so should work on XP. If it checks out ok for you I'll add the VC6 workspace that created them to git.
That DLL was for Tcl 8, so "zint2150t.dll" would be the equivalent. May I ask why do you need a "zint.dll"? What issue are you having? Note that none of the Tcl DLLs depend on Qt, and the "zint.exe" and "qtZint.exe" shipped are self-contained and do not require "zint.dll" or Qt DLLs. (If you're building your own app, see the first part "of "win32\README", "Visual Studio 2022", for how to build "zint.dll" (Win32) - only requires installing Visual Studio 2022 Community Edition.)
Hi, do you mean the "zint.dll" that used to be in the "tcl" sub-directory? This is now in the "tcl\zint2.15.0" sub-directory in 2 versions: "tcl9zint2150.dll" for Tcl 9, and "zint2150t.dll" for Tcl 8.
CI: macOS: use homebrew Qt PATH
CI: macOS: add Qt PATH
CI: Windows: try GS1 Syntax Engine with fixed CMakeLists.txt; try Qt
CI: Windows: leave GS1 Syntax Engine out
CI: Windows: fix GS1SE_ROOT; free-bsd: try offscreen for qt test
CI: free-bsd: remove -C from ctest
CI: macos: try qt; free-bsd: try explicit --config Release
CI: macos: try homebrew; free-bsd: fix build, try ctest
CI: macos: try LD_LIBRARY_PATH; free-bsd: try building
CI: Windows: shell missing?; macos: shared zlib/lpng install
CI: Windows: try with updated CMakeLists.txt; macos: try with PNG
CI: Windows: try with BUILD_TYPE
CI: Windows: try no BUILD_TYPE
CI: Windows: GS1 Syntax Engine cmake
CI: Windows: fix ${} syntax
CI: Windows: try GS1 Syntax Engine install again
CI: Windows: try non-relative XXX_ROOTs
CI: Windows: don'try GS1 Syntax Engine install
CI: Windows: try GS1 Syntax Engine install
CI: Windows: specify WIN32 for zlib/libpng
CI: Windows: try using libpng
CI: Windows: try slashes
CI: Windows: try installing libpng
CI: Windows: wot's in build?
CI: Windows: leave out zconf.h mv
CI: Windows: separate cmds
CI: Windows: try zlib again
CI: Windows: leave for now
CI: Windows: try installing zlib
CLI: `--gs1XXX` args now imply `--gs1
GS1 Syntax Engine: update Windows README and various project files
Suppress gcc-15 warning -Wunterminated-string-initialization
github: macos: forget it
github ci: macos: try hack around ldconfig not found
github ci: macos: try installing GS1 Syntax Engine
github ci: ubuntu: try installing GS1 Syntax Engine
And in tests cmake
cmake: fix GS1SE NOTFOUND test
Integrate GS1 Syntax Engine
general: fix some typos using https://github.com/crate-ci/typos
RAW_TEXT: change `source` to be unconverted, i.e. UTF-8 (unless
library: use "filemem" for TXT `dump_plot()` -> `txt_hex_plot()`
(Your posts got marked as requiring approval for some reason, hence the delay in them appearing.) Yes, it would mean the code would have to be written from scratch. As far as I understand things, it being LGPL and not GPL doesn't change things - it still has a viral effect at source level, so can't be integrated without the license applying.
Hi, yes it would be nice to support JAB Code and it is on the agenda at some stage. It having an open source encoder and decoder available already is huge, but the main problem is that the code is GPL licensed so we'd have to implement are own version to comply with our BSD-3-Clause licence.
Thanks Ulrich, done with commit [39380d]. It makes some of the source a bit "chunky" (e.g. bmp/emf/gif etc) but overall it clarifies it quite a bit I think. Went with just z_ prefix for "common.h" functions. Best regards, Martin
Tcl: run "autoconfig" and "./configure" after committing PR #9
backend_tcl: add tcl 9.0 compatibility define (CONST)
backend_tcl: configure tk
general: prefix all `INTERNAL` funcs/tables with `zint_`, except
Oh I see, I didn't know that. That's pretty crap from a library point of view - and it's quite surprising this issue hasn't been run into before. Will follow your advice, and prefix all INTERNAL functions with zint_, except probably the ones in "common.h" - will prefix them with just z_ - hope this will be enough. Thanks
Hi Ulrich, channel is marked INTERNAL, which is defined in "backend/common.h" as hidden, as long as ZINT_TEST and __MINGW32__ aren't defined, so shouldn't be visible outside the library. Could you do e.g. readelf -s ../../libs/libzint.a | grep channel on your zint library? It should list channel as GLOBAL HIDDEN. Thanks, Martin
gs1: update to latest gs1-syntax-dictionary (key -> gcppos1,
gs1: update to latest gs1-syntax-dictionary (keyoff1 for 8006/8026)
AZTEC/PDF417: stripf ecc feedback (was getting 32-bit difference)
iso4217: add 396 (XAD) ISO 4217 AMENDMENT NUMBER 179
DBAR_OMN/LTD: replace unnecessary `large_uint` with `uint64_t`
GS1PARENS_MODE: allow parentheses in AI data if backslashed and
Re commit [789e04] and [86363f] allow prefixes without check digit
- Add new symbologies `BARCODE_EAN8`, `BARCODE_EAN_2ADDON`,
Fix installation of cmake files
test suite: testcommon.c: NOLINT clang-tidy-20.1.2 false positive
EAN-8: specify in check digit error
RAW_TEXT: expand UPC-A, UPC-E to EAN-13, and EAN-8 if has add-on,
Thanks Milton, hopefully commit [182c84] has addressed this. There's a later document available at https://www.royalmail.com/sites/royalmail.com/files/2024-12/Mailmark-Barcode-definition-document-181224-v1_5-final.pdf, but it doesn't change the DPS/RTS stuff, all the best, Martin
MAILMARK_2D: fix postcode validation: no limited alphanumerics,
ITF-14/EAN-14/NVE-18: allow check digit to be given and then
test suite: test_upcean: fix `test_hrt` not actually testing
Re last commit, keep line length <= 118.
dxfilmedge: make raw_segs consistent with zxing-cpp
Addressed by commit [789e04].
DBAR_OMN/DBAR_LTD/DBAR_STK/DBAR_OMNSTK: prefix `raw_segs` with
Yes, can do that.
Revert "[[fallthrough]]" -> "fall through" to avoid gcc warning
ZINT_CAP_STACKABLE: missed CODABLOCKF/HIBC_BLOCKF previously - not
- CODABLOCKF: fix misencodation of extended ASCII 0xB0-0xB9 when
test_common: rt_cpy etc.: zeroize symbol
github: macOS: ctest -VV
general: change BARCODE_RAW_TEXT to write to new `zint_symbol`
Thanks Harald, fixed with commit [d1bf02].
gs1: csumalpha: improve warning, report both checksum chars
Thanks Milton, fixed with commit [de8a62].
UPU_S10: fix Service Indicator warning re "H" (ticket #331, props
VIN Code prefix
You're welcome Milton, all the best, Martin - closing.
DBAR_STK/DBAR_OMNSTK/DBAR_EXPSTK (including composites): output HRT
- composite/MAXICODE: set `option_1/2` to values used in encodation