"""XML Parser based (loosely) on the XML Spec's EBNF
This is a hand-coded parser based on the W3C's XML specification,
there was a lot of busy-work rewriting to make the syntax agree,
but also a number of signficant structural changes required by
the limitations of the SimpleParse engine, and the completely
procedural definition of References in the XML spec (the References
don't occur in most places they can occur, and they are seen as
altering the buffer directly as soon as they are encountered, this
isn't something that fits readily into the mx.TextTools engine.
http://www.w3.org/TR/REC-xml#sec-references
Major Deviations from Spec:
No support for the unicode-style character classes
No support for UTF-16 (or Unicode at all, for that matter)
No support for References that alter the production
being parsed, so you can't have a Reference to an
item "</this>and<this>" or similar non-structure-
respecting References. References have
particular locations they can occur, and they are
just ignored elsewhere
No support for parsing the contents of References within
the primary parsing pass
No support for excluded start/end tags
Comments allowed in both tags and declarations (but not
inside content-specifiers).
Allows end tags of the form </>
"""
declaration = """
# Simple (changable) literals
# These should be chosen based on the encoding
# of the file, which is actually embedded in the
# file :(
<S> := [\x20\x09\x0D\x0A]+
<letter> := [a-zA-Z]
<namestart> := letter/[_:]
<namechar> := letter/digit/[-._:]
# don't change for XML, but would change for SGML or HTML
<Eq> := '='
<REFO> := '&'
<PREFO> := '%'
<REFC> := ';'
<PIO> := '<?'
<PIC> := '?>'
<STagO> := '<'
<STagC> := '>'
<ETagO> := '</'
<ETagC> := '>'
<EmptyElemTagC> := '/>'
# an XML-comment, note that this follows
# SGML semantics, so that you can embed comment_sets
# in the middle of the various declarations...
>Comment< := "<!", comment_set,(S?,comment_set)*,S?,">"
>comment_set< := '--', xml_comment,'--'
xml_comment := -'--'*
# whitespace in tag (including possible comment)
>TS< := (Comment/S)+
# general structures
AttValue := ('"', (Reference/ -[&"] )*, '"') / ( "'", (Reference / -[&'])*, "'")
# Names
Name := namestart, namechar*
Names := Name, (S,Name)*
Nmtoken := namechar+
Nmtokens := Nmtoken, (S,Nmtoken)*
# processing instructions
PI := PIO, PITarget, S?, PIContent, PIC
PIContent := -PIC*
PITarget := ?-( [Xx],[Mm],[Ll]), Name
## references
# character reference
CharRef := REFO,'#',('x',hex)/(int),REFC
# entity reference
EntityRef := REFO, Name, REFC
# parsed entity ref
PEReference := PREFO, Name, REFC
Reference := EntityRef / CharRef
Misc := Comment/S
### PROLOG definitions...
prolog := XMLDecl?, Misc*, (doctypedecl, Misc*)?
XMLDecl := '<?xml', VersionInfo, EncodingDecl?, SDDecl?, TS?, '?>'
VersionInfo := TS?, 'version', TS?, Eq, TS?, (('"',VersionNum,'"')/("'",VersionNum,"'"))
VersionNum := [a-zA-Z0-9_.:-]+
### Document-type declarations (DTDs)
doctypedecl := '<!DOCTYPE', TS, Name, (TS, ExternalID)?, TS?,('[', (markupdecl / DeclSep)*, ']', TS?)?, '>'
DeclSep := PEReference / S
markupdecl := elementdecl / AttlistDecl / EntityDecl / NotationDecl / PI / Comment
EncodingDecl := TS, 'encoding', Eq, (('"', EncName, '"') / ("'", EncName, "'") )
EncName := [A-Za-z],[A-Za-z0-9._-]*
SDDecl := TS, 'standalone', Eq, (("'", ('yes' / 'no'), "'") / ('"', ('yes' / 'no'), '"'))
ExternalID := ('SYSTEM', TS?, SystemLiteral) / ('PUBLIC', TS?, PubidLiteral, TS?, SystemLiteral ) / PEReference
NDataDecl := (TS, 'NDATA', TS, Name)/ (TS,PEReference,TS,(Name/ PEReference)?)
SystemLiteral := ('"', -["]*, '"') / ("'", -[']*, "'") / PEReference
PubidLiteral := ('"', [\x20\x0D\x0Aa-zA-Z0-9'()+,./:=?;!*#@$_%-]*, '"') / ("'", [\x20\x0D\x0Aa-zA-Z0-9()+,./:=?;!*#@$_%-]*, "'") / PEReference
PublicID := ('PUBLIC', TS, PubidLiteral) / PEReference
### Element-type declarations
# hack to try and get PEReference parsing for the "normal case"
# where the PEReference doesn't change the production level, which
# seems to be suggested by the spec...
elementdecl := '<!ELEMENT', (
(TS, Name, TS, contentspec)/
elementdecl_pe
), TS?,'>'
>elementdecl_pe< := (TS, PEReference, TS?, contentspec?)
contentspec := 'EMPTY' / 'ANY' / Mixed / children
Mixed := ('(', S?, '#PCDATA', (S?, '|', S?, (Name/PEReference))*, S?, ')*' ) /('(', S?, '#PCDATA', S?, ')')
repetition_specifier := ('?' / '*' / '+')?
children := (choice / seq/ PEReference), repetition_specifier
cp := (choice / seq / Name/ PEReference ), repetition_specifier
choice := '(', S?, cp, ( S?, '|', S?, cp )+, S?, ')'
seq := '(', S?, cp, ( S?, ',', S?, cp )*, S?, ')'
### Attribute list declarations...
AttlistDecl := '<!ATTLIST', TS, ((Name, AttDef*, TS?)/(PEReference, AttDef*, TS?)), '>'
AttDef := TS, ((Name, TS, AttType, TS, DefaultDecl)/(PEReference, TS?, AttType?, TS?, DefaultDecl?))
AttType := StringType / TokenizedType / EnumeratedType/ PEReference
StringType := 'CDATA'
TokenizedType := 'ID' / 'IDREF' / 'IDREFS' / 'ENTITY' / 'ENTITIES' / 'NMTOKEN' / 'NMTOKENS'
EnumeratedType := NotationType / Enumeration
NotationType := 'NOTATION', TS, ('(', NameOrList, ')')/PEReference
Enumeration := '(', (NmTokenOrList/PEReference), ')'
>NameOrList< := S?, (Name/PEReference), (S?, '|', S?, (Name/PEReference))*, S?
>NmTokenOrList< := S?, (Nmtoken/PEReference), (S?, '|', S?, (Nmtoken/PEReference))*, S?
DefaultDecl := '#REQUIRED' / '#IMPLIED' / ((('#FIXED', TS)/PEReference)?, (AttValue/PEReference)) / PEReference
### Entity declarations
EntityDecl := GEDecl / PEDecl
GEDecl := '<!ENTITY', TS, ((Name, TS, EntityDef)/(PEReference,TS?,EntityDef?)), TS?, '>'
PEDecl := '<!ENTITY', TS, '%', TS, ((Name, TS, PEDef)/(PEReference,TS?,PEDef?)), TS?, '>'
EntityDef := EntityValue / (ExternalID, NDataDecl?) / PEReference
PEDef := EntityValue / ExternalID / PEReference
EntityValue := ('"', (PEReference / Reference / -[%&"])*, '"') / ("'", (PEReference / Reference / -[%&'])*, "'")
NotationDecl := '<!NOTATION', TS, Name, TS, (ExternalID / PublicID), TS?, '>'
### elements (nodes/tags/you-know :) )
# limitations in the SimpleParse engine mean that this
# particular structure will be basically useless...
element := EmptyElemTag / (STag, content, ETag)
EmptyElemTag := STagO, Name, (TS, Attribute)*, TS?, EmptyElemTagC
STag := STagO, Name, (TS, Attribute)*, TS?, STagC
ETag := ETagO, Name?, TS?, ETagC
content := (element / Reference / CDSect / PI / Comment / CharData)*
Attribute := (Name, Eq, (AttValue/Reference))/(Reference,(Eq,(AttValue/Reference))?)
# general content of an element
CharData := ( -[<&]+ / -(STag / EmptyElemTag / ETag / Reference / CDSect / PI / Comment) )+
# special non-parsed character data sections
CDSect := CDStart, CData, CDEnd
<CDStart> := '<![CDATA['
CData := -CDEnd*
<CDEnd> := ']]>'
document := prolog, element, Misc*
"""
from simpleparse.common import numbers,strings,chartypes
|