merge_fonts.py KeyError: 'cid00001#1'

[benbro]

I've installed master with pip and downloaded all fonts from https://www.google.com/get/noto/
When trying to merge with:

python merge_fonts.py -d individual/unhinted -o NotoSansMerged-Regular.ttf

I first got an error:
KeyError: 'AnatolianHieroglyphs'
I've commented the line "'NotoSansAnatolianHieroglyphs-Regular.ttf'," in merge_fonts.py and tried again and got another error:
KeyError: 'cid00001#1'

[SamJarmanPP]

I saw KeyError: 'cid00001#15' after takign the same step to prevent the first error

[rspilker]

To fix the "AnatolianHieroglyphs" error, a new line 95 should be inserted in merge_noto.py:

    "AnatolianHieroglyphs":"hluw",

So the section would become:

    "OlChiki": "olck",
    "TaiLe": "tale",
    # Following keys are added to satisfy the use case in merge_fonts.py
    # Reference:
    # https://www.google.com/get/noto/#sans-xsux
    # https://www.google.com/get/noto/#sans-cprt
    # https://www.google.com/get/noto/#sans-yiii
    # https://www.microsoft.com/typography/otspec/scripttags.htm
    "AnatolianHieroglyphs":"hluw",
    "Cuneiform": "xsux",
    "Cypriot": "cprt",
    "Yi": "yi  ",

[koumaza]

Is there no choice but to ignore the error now?

[sp96296]

Im getting the same error while trying to merge the CJK fonts with the Regular set.

[sp96296]

Im getting the same error while trying to merge the CJK fonts with the Regular set.

"file" section of merge_fonts.py

files = [  
    # It's recommended to put NotoSans-Regular.ttf as the first element in the  
    # list to maximize the amount of meta data retained in the final merged font.  
    "NotoSerif-Regular.ttf",  
    "NotoSerifCJKjp-Regular.otf",  
    "NotoSerifCJKkr-Regular.otf",  
    "NotoSerifCJKsc-Regular.otf",  
    "NotoSerifCJKtc-Regular.otf",  
    "NotoSerifDisplay-Regular.ttf",  
]  

EDIT:
I tried adding "CJK Ideographic" : "hani", "Hiragana" : "kana", and "Hangul" : "hang" to the SCRIPT_TO_OPENTYPE_SCRIPT_TAG function of merge_noto.py, but no luck.

[koumaza]

This script only merges fonts,
https://github.com/koumaza/Ecchi-fFont/blob/39a47d47ea578abade8b6d8b41203fddf5e8dbef/tool/merge.pe

Does merge_fonts.py here just do the merge? Or is there another function?

[ghost]

In addition to the easy to fix "AnatolianHieroglyphs" problem (thanks @rspilker ), there seem to be other issues making the script merge_fonts.py not working out of the box:

• some fonts are otf
• some fonts have a different UPM

The resulting errors of a dependency seem to be similar to these: fonttools/fonttools#1309
I tried merging with a CJK ttf file I downloaded online, but apparently this exceeds some kind of number range:
struct.error: 'H' format requires 0 <= number <= 65535
For now, I managed to merge 37 fonts (without CJK).

Noto is a great project, but we really need a single font file (for example when generating a pdf with unknown input language).