pyyaml: pyyaml does not support literals in unicode over codepoint 0xffff

See https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=806826

the yaml spec says that

“The allowed character range explicitly excludes the surrogate block #xD800-#xDFFF, DEL #x7F, the C0 control block #x0-#x1F (except for #x9, #xA, and #xD), the C1 control block #x80-#x9F, #xFFFE, and #xFFFF.”

however pyyaml has chosen to negate that check and apply it to only plane 0. This means that any yaml document that contains unicode literals in higher planes will fail to parse (and, on output, use the rather unfriendly \Uxxxxxxxx format).

The attached patch fixes this in a minimally intrusive way, by extending the checks to cover the additional codepoints where appropriate. A better fix would be to use the check as the spec specifies it, but that would be a bigger change.

Index: pyyaml-3.11/lib/yaml/emitter.py

— pyyaml-3.11.orig/lib/yaml/emitter.py +++ pyyaml-3.11/lib/yaml/emitter.py @@ -8,9 +8,13 @@

all = [‘Emitter’, ‘EmitterError’]

+import sys + from error import YAMLError from events import *

+has_ucs4 = sys.maxunicode > 0xffff + class EmitterError(YAMLError): pass

@@ -701,7 +705,8 @@ class Emitter(object): line_breaks = True if not (ch == u’\n’ or u’\x20’ <= ch <= u’\x7E’): if (ch == u’\x85’ or u’\xA0’ <= ch <= u’\uD7FF’

  •                    or u'\uE000' <= ch <= u'\uFFFD') and ch != u'\uFEFF':
    
  •                    or u'\uE000' <= ch <= u'\uFFFD'
    
  •                    or ((not has_ucs4) or (u'\U00010000' <= ch < u'\U0010ffff'))) and ch != u'\uFEFF':
                 unicode_characters = True
                 if not self.allow_unicode:
                     special_characters = True
    

    Index: pyyaml-3.11/lib/yaml/reader.py

    — pyyaml-3.11.orig/lib/yaml/reader.py +++ pyyaml-3.11/lib/yaml/reader.py @@ -19,7 +19,9 @@ all = [‘Reader’, ‘ReaderError’]

    from error import YAMLError, Mark

-import codecs, re +import codecs, re, sys + +has_ucs4 = sys.maxunicode > 0xffff

class ReaderError(YAMLError):

@@ -134,7 +136,10 @@ class Reader(object): self.encoding = ‘utf-8’ self.update(1)

  • NON_PRINTABLE = re.compile(u’[^\x09\x0A\x0D\x20-\x7E\x85\xA0-\uD7FF\uE000-\uFFFD]')

  • if has_ucs4:

  •    NON_PRINTABLE = re.compile(u'[^\x09\x0A\x0D\x20-\x7E\x85\xA0-\uD7FF\uE000-\uFFFD\U00010000-\U0010ffff]')
    
  • else:

  •    NON_PRINTABLE = re.compile(u'[^\x09\x0A\x0D\x20-\x7E\x85\xA0-\uD7FF\uE000-\uFFFD]')
    

    def check_printable(self, data): match = self.NON_PRINTABLE.search(data) if match: Index: pyyaml-3.11/lib3/yaml/emitter.py

    — pyyaml-3.11.orig/lib3/yaml/emitter.py +++ pyyaml-3.11/lib3/yaml/emitter.py @@ -698,7 +698,8 @@ class Emitter: line_breaks = True if not (ch == ‘\n’ or ‘\x20’ <= ch <= ‘\x7E’): if (ch == ‘\x85’ or ‘\xA0’ <= ch <= ‘\uD7FF’

  •                    or '\uE000' <= ch <= '\uFFFD') and ch != '\uFEFF':
    
  •                    or '\uE000' <= ch <= '\uFFFD'
    
  •                    or '\U00010000' <= ch < '\U0010ffff') and ch != '\uFEFF':
                 unicode_characters = True
                 if not self.allow_unicode:
                     special_characters = True
    

    Index: pyyaml-3.11/lib3/yaml/reader.py

    — pyyaml-3.11.orig/lib3/yaml/reader.py +++ pyyaml-3.11/lib3/yaml/reader.py @@ -134,7 +134,7 @@ class Reader(object): self.encoding = ‘utf-8’ self.update(1)

  • NON_PRINTABLE = re.compile(‘[^\x09\x0A\x0D\x20-\x7E\x85\xA0-\uD7FF\uE000-\uFFFD]’)

  • NON_PRINTABLE = re.compile(‘[^\x09\x0A\x0D\x20-\x7E\x85\xA0-\uD7FF\uE000-\uFFFD\U00010000-\U0010ffff]’) def check_printable(self, data): match = self.NON_PRINTABLE.search(data) if match:

About this issue

  • Original URL
  • State: closed
  • Created 8 years ago
  • Reactions: 3
  • Comments: 20 (8 by maintainers)

Commits related to this issue

Most upvoted comments

Just thought I’d share, in case others are in the same boat: After patching/working around this for a while, we realized https://pypi.python.org/pypi/ruamel.yaml handles higher Unicodes just fine, and it’s worked out well for us.

Fixed by #63

this was merged in #63 but still hasn’t been released 😦