view contrib/xml.rnc @ 25783:1f6878c87c25

templater: introduce one-pass parsing of nested template strings

Instead of re-parsing quoted strings as templates, the tokenizer can delegate
the parsing of nested template strings to the parser. It has two benefits:

 1. syntax errors can be reported with absolute positions
 2. nested template can use quotes just like shell: "{"{rev}"}"

It doesn't sound nice that the tokenizer recurses into the parser. We could
instead make the tokenize itself recursive, but it would be much more
complicated because we would have to adjust binding strengths carefully and
put dummy infix operators to concatenate template fragments.

Now "string" token without r"" never appears. It will be removed by the next
patch.

# RelaxNG schema for "xml" log style
# Inspired by Subversion's XML log format.

start = log
node.type = xsd:string  {minLength = "40" maxLength = "40"}

log = element log { logentry+ }
logentry = element logentry {
    logentry.attlist,
    branch*, tag*, hgparent*,
    author, date,
    msg, paths?, copies?, extra*
}
logentry.attlist =
    attribute revision {xsd:nonNegativeInteger}
  & attribute node {node.type}
branch = element branch { text }
tag = element tag { text }
hgparent = element parent {hgparent.attlist, text}
hgparent.attlist =
    attribute revision {xsd:integer {minInclusive = "-1"} }
  & attribute node {node.type}
author = element author { author.attlist, text }
author.attlist =
    attribute email {text}
date = element date {xsd:dateTime}
msg = element msg {msg.attlist, text}
msg.attlist =
    attribute xml:space {"preserve"}
paths = element paths { path* }
path = element path { path.attlist, text }
path.attlist =
    # Action: (A)dd, (M)odify, (R)emove
    attribute action {"A"|"M"|"R"}
copies = element copies { copy+ }
copy = element copy { copy.attlist, text }
copy.attlist =
    attribute source {text}
extra = element extra {extra.attlist, text}
extra.attlist =
    attribute key {text}