#!/bin/sh
#
# Copyright (c) 2005 Jonas Fonseca
#
test_description='Test the very basic parsing of SGML documents.
This test runs very basic features, like checking that nodes are placed
correctly in the DOM tree.
'
. ./libtest
test_output_equals () {
desc="$1"; shift
src="$1"; shift
out="$1"; shift
URI="test:$(echo "$desc" | sed '
s/^[ \t]*\[[^]]*\][ \t]*//;
s/[:., \t][:., \t]*/-/g;
s/_/-/g;
# *cough*
y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/;
s/[^a-zA-Z0-9-]//g;')"
sgml-parser --uri "$URI" --src "$src" $@ | sed 's/^ //' > output
echo "#document: $URI" > expected
echo "$out" | sed -n '2,$p' >> expected
test_expect_success "$desc" 'cmp -b output expected'
}
################################################################
# Parse various SGML node types.
test_output_equals \
'Parse a small document.' \
'Hello World!
' \
'
element: html
element: body
element: p
#text: Hello World!'
test_output_equals \
'Parse elements.' \
'a' \
'
element: root
element: child
attribute: attr -> value
element: child2
element: child3
#text: a'
test_output_equals \
'Check tag soup elements.' \
'a' \
'
element: parent
attribute: attr -> value
element: child:1
element: child:2
#text: a'
test_output_equals \
'Parse an enclosed comment.' \
'' \
'
element: root
#comment: Hello World! '
test_output_equals \
'Parse comment combinations.' \
'' \
'
element: root
#comment: s' \
'
#comment: ->s'
test_output_equals \
'Parse an enclosed CDATA section.' \
'...]]>' \
'
element: root
#cdata-section: ...] ]>...'
test_output_equals \
'Parse non-enclosed CDATA section.' \
'' \
'
#cdata-section: ...'
test_output_equals \
'Parse a bad CDATA section.' \
'' \
'
element: root
attribute: lang -> fr
attribute: attr ->
attribute: name -> value with &foo; " \
'
element: root
attribute: a -> b
attribute: c -> d
attribute: g -> h
attribute: i -> j
attribute: k -> '
test_output_equals \
'Parse attribute with non-quoted values.' \
'...' \
'
element: root
attribute: color -> #abc
attribute: path -> /to/%61-&\one";files
#text: ...'
test_output_equals \
'Parse entity references.' \
'&-*' \
'
entity-reference: amp
#text: -
entity-reference: #42'
# Just how these should be gracefully handled is not clear to me.
test_output_equals \
'Parse badly formatted entity references.' \
'& m33p;-&.:-copy;-&;-&#;-&#xx;' \
'
#text: & m33p;
#text: -
entity-reference: .:-copy
#text: -
#text: &;
#text: -
entity-reference: #
#text: -
entity-reference: #xx'
test_output_equals \
'Parse processing instructions.' \
'
...
' \
'
proc-instruction: xml -> encoding="UTF8"
attribute: encoding -> UTF8
#text: \n...\n
proc-instruction: ecmascript -> var val=2;\n'
test_output_equals \
'Parse XML processing instructions.' \
'?>-' \
'
proc-instruction: xml -> version="1.0" />
attribute: version -> 1.0
proc-instruction: xml -> /'
test_output_equals \
'Parse XML stylesheet processing instructions.' \
'' \
'
proc-instruction: xml-stylesheet -> type="text/xsl" href="url"
attribute: type -> text/xsl
attribute: href -> url'
test_output_equals \
'Parse exotic processing instructions.' \
'+?>-?>---' \
'
proc-instruction: xml -> ?+>+
#text: -?>-
proc-instruction: js ->
#text: -
proc-instruction: ->
#text: -'
test_output_equals \
'Parse incorrect processing instructions.' \
'--- <
#text: -
proc-instruction: -> <=";&
#text: -'
test_output_equals \
'Parse incorrect processing instructions (II).' \
' >< / root >' \
'
element: root
attribute: ns:attr -> value
proc-instruction: target -> data'
test_output_equals \
'Check line numbers. (I)' \
' number ' \
'
1' \
--print-lines
test_output_equals \
'Check line numbers. (II)' \
'<
line:2
line:3
=
"line:5"
>' \
'
10' \
--print-lines
test_output_equals \
'Check line numbers. (III)' \
'1
2
3
4
5
6
7
8' \
'
8' \
--print-lines
test_done