FPDF 'latin-1' encoding error

[falkecarlsen]

It seems FPDF assumes Windows specific, non-standard 'latin-1' encoding according to: https://stackoverflow.com/questions/3942888/unicodeencodeerror-latin-1-codec-cant-encode-character. This makes FPDF error out with the following message: UnicodeEncodeError: 'latin-1' codec can't encode character '\U0001f438' in position 2953: ordinal not in range(256).

Hotfix that worked for me: replace 'latin1' with 'utf-8' for encoding in fpdf.py at line 1170 for fpdf v1.7.2:
p = self.pages[n].encode("utf-8") if PY3K else self.pages[n]

Note this is a copy of issue 16 in original repo