diff options
| -rw-r--r-- | CMakeLists.txt | 2 | ||||
| -rwxr-xr-x | runtests.py | 178 | ||||
| -rw-r--r-- | src/html/html.c | 2 |
3 files changed, 104 insertions, 78 deletions
diff --git a/CMakeLists.txt b/CMakeLists.txt index 637099d..afa3d22 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -24,7 +24,7 @@ enable_testing() # To get verbose output: cmake --build build --target "test" -- ARGS='-V' add_test(spectest_library python "${CMAKE_SOURCE_DIR}/runtests.py" "--spec" - "${CMAKE_SOURCE_DIR}/spec.txt" "--library_dir" "${CMAKE_BINARY_DIR}/src" + "${CMAKE_SOURCE_DIR}/spec.txt" "--library-dir" "${CMAKE_BINARY_DIR}/src" ) add_test(spectest_executable python "${CMAKE_SOURCE_DIR}/runtests.py" "--spec" "${CMAKE_SOURCE_DIR}/spec.txt" "--program" "${CMAKE_BINARY_DIR}/src/cmark" diff --git a/runtests.py b/runtests.py index 906a573..b3c8d98 100755 --- a/runtests.py +++ b/runtests.py @@ -10,6 +10,7 @@ import argparse from HTMLParser import HTMLParser from htmlentitydefs import name2codepoint import re +import cgi if __name__ == "__main__": parser = argparse.ArgumentParser(description='Run cmark tests.') @@ -19,8 +20,13 @@ if __name__ == "__main__": help='path to spec') parser.add_argument('--pattern', dest='pattern', nargs='?', default=None, help='limit to sections matching regex pattern') - parser.add_argument('--library_dir', dest='library_dir', nargs='?', + parser.add_argument('--library-dir', dest='library_dir', nargs='?', default=None, help='directory containing dynamic library') + parser.add_argument('--no-normalize', dest='normalize', + action='store_const', const=False, default=True) + parser.add_argument('--debug-normalization', dest='debug_normalization', + action='store_const', const=True, + default=False, help='filter stdin through normalizer for testing') args = parser.parse_args(sys.argv[1:]) if not args.program: @@ -60,106 +66,119 @@ class MyHTMLParser(HTMLParser): self.last = "starttag" self.in_pre = False self.output = u"" + self.last_tag = "" def handle_data(self, data): - if self.in_pre: - self.output += data - else: - data = normalize_whitespace_re.sub(' ', data) - data_strip = data.strip() - if (self.last == "ref") and data_strip and data[0] == " ": - self.output += " " - self.data_end_in_space_not_empty = (data[-1] == ' ' and data_strip) - self.output += data_strip - self.last = "data" + after_tag = self.last == "endtag" or self.last == "starttag" + after_block_tag = after_tag and self.is_block_tag(self.last_tag) + if after_tag and self.last_tag == "br": + data = data.lstrip('\n') + data = normalize_whitespace_re.sub(' ', data) + if after_block_tag and not self.in_pre: + if self.last == "starttag": + data = data.lstrip() + elif self.last == "endtag": + data = data.strip() + self.output += data + self.last = "data" def handle_endtag(self, tag): if tag == "pre": self.in_pre = False + if self.is_block_tag(tag): + self.output = self.output.rstrip() self.output += "</" + tag + ">" + self.last_tag = tag self.last = "endtag" def handle_starttag(self, tag, attrs): if tag == "pre": self.in_pre = True self.output += "<" + tag - attrs = filter(lambda attr: attr[0] in significant_attrs, attrs) + # For now we don't strip out 'extra' attributes, because of + # raw HTML test cases. + # attrs = filter(lambda attr: attr[0] in significant_attrs, attrs) if attrs: attrs.sort() - for attr in attrs: - self.output += " " + attr[0] + "=" + '"' + attr[1] + '"' + for (k,v) in attrs: + self.output += " " + k + if v != None: + self.output += ("=" + '"' + cgi.escape(v,quote=True) + '"') self.output += ">" + self.last_tag = tag self.last = "starttag" def handle_startendtag(self, tag, attrs): - """Ignore closing tag for self-closing void elements.""" + """Ignore closing tag for self-closing """ self.handle_starttag(tag, attrs) + self.last_tag = tag + self.last = "endtag" + def handle_comment(self, data): + self.output += '<!--' + data + '-->' + self.last = "comment" + def handle_decl(self, data): + self.output += '<!' + data + '>' + self.last = "decl" + def handle_unknown_decl(self, data): + self.output += '<!' + data + '>' + self.last = "decl" + def handle_pi(self,data): + self.output += '<?' + data + '>' + self.last = "pi" def handle_entityref(self, name): - self.add_space_from_last_data() try: - self.output += unichr(name2codepoint[name]) + c = unichr(name2codepoint[name]) except KeyError: - self.output += name + c = None + self.output_char(c, '&' + name + ';') self.last = "ref" def handle_charref(self, name): - self.add_space_from_last_data() try: if name.startswith("x"): c = unichr(int(name[1:], 16)) else: c = unichr(int(name)) - self.output += c except ValueError: - self.output += name + c = None + self.output_char(c, '&' + name + ';') self.last = "ref" # Helpers. - def add_space_from_last_data(self): - """Maintain the space at: `a <span>b</span>`""" - if self.last == 'data' and self.data_end_in_space_not_empty: - self.output += ' ' + def output_char(self, c, fallback): + if c == u'<': + self.output += "<" + elif c == u'>': + self.output += ">" + elif c == u'&': + self.output += "&" + elif c == u'"': + self.output += """ + elif c == None: + self.output += fallback + else: + self.output += c -def normalize(html): + def is_block_tag(self,tag): + return (tag in ['article', 'header', 'aside', 'hgroup', 'blockquote', + 'hr', 'iframe', 'body', 'li', 'map', 'button', 'object', 'canvas', + 'ol', 'caption', 'output', 'col', 'p', 'colgroup', 'pre', 'dd', + 'progress', 'div', 'section', 'dl', 'table', 'td', 'dt', + 'tbody', 'embed', 'textarea', 'fieldset', 'tfoot', 'figcaption', + 'th', 'figure', 'thead', 'footer', 'tr', 'form', 'ul', + 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'video', 'script', 'style']) + +def normalize_html(html): r""" - Return normalized form of HTML which igores insignificant output differences. - Multiple inner whitespaces to a single space - >>> normalize("<p>a \t\nb</p>") - u'<p>a b</p>' - Surrounding whitespaces are removed: - >>> normalize("<p> a</p>") - u'<p>a</p>' - >>> normalize("<p>a </p>") - u'<p>a</p>' - TODO: how to deal with the following cases without a full list of the void tags? - >>> normalize("<p>a <b>b</b></p>") - u'<p>a<b>b</b></p>' - >>> normalize("<p><b>b</b> c</p>") - u'<p><b>b</b>c</p>' - >>> normalize("<p>a <br></p>") - u'<p>a<br></p>' - `pre` elements preserve whitespace: - >>> normalize("<pre>a \t\nb</pre>") - u'<pre>a \t\nb</pre>' - Self-closing tags: - >>> normalize("<p><br /></p>") - u'<p><br></p>' - References are converted to Unicode: - >>> normalize("<p><</p>") - u'<p><</p>' - >>> normalize("<p><</p>") - u'<p><</p>' - >>> normalize("<p><</p>") - u'<p><</p>' - >>> normalize("<p>中</p>") - u'<p>\u4e2d</p>' - Spaces around entities are kept: - >>> normalize("<p>a < b</p>") - u'<p>a < b</p>' - >>> normalize("<p>a<b</p>") - u'<p>a<b</p>' - Most attributes are ignored: - >>> normalize('<p id="a"></p>') - u'<p></p>' - Critical attributes are considered and sorted alphabetically: - >>> normalize('<a href="a"></a>') - u'<a href="a"></a>' - >>> normalize('<img src="a" alt="a">') - u'<img alt="a" src="a">' + Return normalized form of HTML which ignores insignificant output + differences: + + * Multiple inner whitespaces are collapsed to a single space (except + in pre tags). + * Outer whitespace (outside block-level tags) is removed. + * Self-closing tags are converted to open tags. + * Attributes are sorted and lowercased. + * References are converted to unicode, except that '<', '>', '&', and + '&' are rendered using entities. + + Known limitations: + + * HTMLParser just swallows CDATA. + * HTMLParser seems to treat unknown declarations as comments. """ parser = MyHTMLParser() parser.feed(html.decode(encoding='UTF-8')) @@ -170,13 +189,17 @@ def print_test_header(headertext, example_number, start_line, end_line): print "Example %d (lines %d-%d) %s" % (example_number,start_line,end_line,headertext) def do_test(markdown_lines, expected_html_lines, headertext, - example_number, start_line, end_line, prog=None): + example_number, start_line, end_line, prog, normalize): real_markdown_text = ''.join(markdown_lines).replace('→','\t') [retcode, actual_html, err] = md2html(real_markdown_text, prog) if retcode == 0: actual_html_lines = actual_html.splitlines(True) expected_html = ''.join(expected_html_lines) - if normalize(actual_html) == normalize(expected_html): + if normalize: + passed = normalize_html(actual_html) == normalize_html(expected_html) + else: + passed = actual_html == expected_html + if passed: return 'pass' else: print_test_header(headertext, example_number,start_line,end_line) @@ -192,7 +215,7 @@ def do_test(markdown_lines, expected_html_lines, headertext, print(err) return 'error' -def do_tests(specfile, prog, pattern): +def do_tests(specfile, prog, pattern, normalize): line_number = 0 start_line = 0 end_line = 0 @@ -208,7 +231,7 @@ def do_tests(specfile, prog, pattern): header_re = re.compile('#+ ') if pattern: - pattern_re = re.compile(pattern) + pattern_re = re.compile(pattern, re.IGNORECASE) with open(specfile, 'r') as specf: for line in specf: @@ -228,7 +251,8 @@ def do_tests(specfile, prog, pattern): if active: result = do_test(markdown_lines, html_lines, headertext, example_number, - start_line, end_line, prog) + start_line, end_line, prog, + normalize) if result == 'pass': passed = passed + 1 elif result == 'fail': @@ -248,7 +272,9 @@ def do_tests(specfile, prog, pattern): return (failed == 0 and errored == 0) if __name__ == "__main__": - if do_tests(args.spec, args.program, args.pattern): + if args.debug_normalization: + print normalize_html(sys.stdin.read()) + elif do_tests(args.spec, args.program, args.pattern, args.normalize): exit(0) else: exit(1) diff --git a/src/html/html.c b/src/html/html.c index 8110f87..0c6c399 100644 --- a/src/html/html.c +++ b/src/html/html.c @@ -210,7 +210,7 @@ static void inlines_to_html(strbuf *html, cmark_node* ils) escape_html(html, ils->as.link.title, -1); } - strbuf_puts(html, "\"/>"); + strbuf_puts(html, "\" />"); break; case NODE_STRONG: |