diff --git a/src/dom/test/sgml-parser.c b/src/dom/test/sgml-parser.c index 9ac6817a6..23b2ef680 100644 --- a/src/dom/test/sgml-parser.c +++ b/src/dom/test/sgml-parser.c @@ -13,6 +13,7 @@ #include "dom/configuration.h" #include "dom/node.h" +#include "dom/sgml/dump.h" #include "dom/sgml/parser.h" #include "dom/stack.h" @@ -265,6 +266,7 @@ main(int argc, char *argv[]) enum sgml_parser_code code = 0; enum dom_config_flag normalize_flags = 0; int normalize = 0; + int dump = 0; int complete = 1; size_t read_stdin = 0; struct dom_string uri = STATIC_DOM_STRING("dom://test"); @@ -338,6 +340,10 @@ main(int argc, char *argv[]) flags |= SGML_PARSER_INCREMENTAL; complete = 0; + } else if (!strcmp(arg, "dump")) { + type = SGML_PARSER_TREE; + dump = 1; + } else if (!strcmp(arg, "error")) { flags |= SGML_PARSER_DETECT_ERRORS; @@ -355,7 +361,7 @@ main(int argc, char *argv[]) parser->error_func = sgml_error_function; if (normalize) add_dom_config_normalizer(&parser->stack, normalize_flags); - else + else if (!dump) add_dom_stack_context(&parser->stack, NULL, &sgml_parser_test_context_info); if (read_stdin > 0) { @@ -402,7 +408,7 @@ main(int argc, char *argv[]) pop_dom_node(&parser->stack); } - if (normalize) { + if (normalize || dump) { struct dom_stack stack; /* Note, that we cannot free nodes when walking the DOM @@ -412,7 +418,10 @@ main(int argc, char *argv[]) /* XXX: This context needs to be added first because it * assumes the parser can be accessed via * stack->contexts[0].data. */ - add_dom_stack_context(&stack, parser, &sgml_parser_test_context_info); + if (normalize) + add_dom_stack_context(&stack, parser, &sgml_parser_test_context_info); + else if (dump) + add_sgml_file_dumper(&stack, stdout); walk_dom_nodes(&stack, parser->root); done_dom_stack(&stack); done_dom_node(parser->root); diff --git a/src/dom/test/test-sgml-dump-basic b/src/dom/test/test-sgml-dump-basic new file mode 100755 index 000000000..2e4ee6fe9 --- /dev/null +++ b/src/dom/test/test-sgml-dump-basic @@ -0,0 +1,173 @@ +#!/bin/sh +# +# Copyright (c) 2005 Jonas Fonseca +# + +test_description='Test dumping of SGML documents. + +Test that DOM documents are consistently dumped to SGML. +Note, this also test whether attribute nodes are sorted +correctly. +' + +. "$TEST_LIB" + +test_sgml_dump_exact () { + desc="$1"; shift + src="$1"; shift + + sgml-parser --dump --src "$src" > output + echo -n "$src" > expected + + test_expect_success "$desc" 'cmp output expected' +} + +test_sgml_dump_equals () { + desc="$1"; shift + src="$1"; shift + out="$1"; shift + + sgml-parser --dump --src "$src" > output + echo -n "$out" > expected + + test_expect_success "$desc" 'cmp output expected' +} + + +################################################################ +# Parse various SGML node types. + +test_sgml_dump_exact \ +'Parse a small document.' \ +'

Hello World!

' + +test_sgml_dump_exact \ +'Parse an enclosed comment.' \ +'' + +test_sgml_dump_exact \ +'Parse comment combinations. (I)' \ +'' + +test_sgml_dump_exact \ +'Parse comment combinations. (II).' \ +'s-->tu' + +test_sgml_dump_exact \ +'Parse empty comment.' \ +'s' + +test_sgml_dump_exact \ +'Parse an enclosed CDATA section.' \ +'...]]>' + +test_sgml_dump_exact \ +'Parse non-enclosed CDATA section.' \ +'' + +test_sgml_dump_exact \ +'Parse attributes.' \ +'' + +test_sgml_dump_exact \ +'Parse XML stylesheet processing instructions.' \ +'' + +test_sgml_dump_exact \ +'Parse entity references.' \ +'&-*' + +############################################################################# +# Test tidy up dumping + +test_sgml_dump_equals \ +'Parse elements.' \ +'a' \ +'a' + +test_sgml_dump_equals \ +'Parse attributes with garbage.' \ +"" \ +"" + +test_sgml_dump_equals \ +'Parse bad comment. (II)' \ +'bad comment' \ +'bad comment' + +test_sgml_dump_equals \ +'Parse empty notation.' \ +'s' \ +'s' + +test_sgml_dump_equals \ +'Parse a bad CDATA section.' \ +'' + +test_sgml_dump_equals \ +'Parse tag soup elements. (I)' \ +'a' \ +'a' + +test_sgml_dump_equals \ +'Parse tag soup elements. (II)' \ +'< a >< b < c / >< / >' \ +'' + +test_sgml_dump_equals \ +'Parse attribute with non-quoted values.' \ +'...' \ +'...' + +# Just how these should be gracefully handled is not clear to me. +test_sgml_dump_equals \ +'Parse badly formatted entity references.' \ +'& m33p;-&.:-copy;-&;-&#;-&#xx;' \ +'& m33p;-&.:-copy;-&;-&#;-&#xx;' + +test_sgml_dump_equals \ +'Parse processing instructions.' \ +' +... +' \ +' +... +' + +test_sgml_dump_equals \ +'Parse XML processing instructions.' \ +'?>-' \ +'?>-?>' + +test_sgml_dump_equals \ +'Parse exotic processing instructions.' \ +'+?>-?>---' \ +'+?>-?>---' + +test_sgml_dump_equals \ +'Parse incorrect processing instructions. (II)' \ +'------' \ + +test_sgml_dump_equals \ +'Parse incorrect processing instructions (II).' \ +'' + +test_sgml_dump_equals \ +'Skip spaces not inside text.' \ +'< +root +ns:attr += +"value" +>< / root >' \ +'' + +test_done