From: Artem <ne...@gm...> - 2008-06-23 20:23:42
|
Hi, I extended test suite with some markdown2 tests, failed tests listed below. I didn't include some tests from tm-cases, because they use option files. I committed version with extended test suite here: http://gitorious.org/projects/python-markdown/repos/gsoc2008 tm-cases: ====== **blockquote** ">" transforms to ">" [Trent wrote] > no way [Jeff wrote] > way other implementations consider ">" in this test as blockquote **escapes** problem with hash characters escaping at the end of header **lists2** ul followed by ol transforms to one ul **mising_link_defn** works fine in markdown 1.7, but don't work in markdown from repository. **parens_in_url_4** this input: [Inline link 4 with non-escaped parens](</url(test)> "title"). gives invalid output: <p><a href="</url(test">Inline link 4 with non-escaped parens</a>> "title"). </p> **raw_html** HTML comments escaping test <!-- *blah* --> **sublist_para** test fails, sublist with 2-spaces indentation **tricky_anchors** fails on: [only open [bracket(/in/) text](/url/ 'a title') but all the other markdown implementations pass this test. php-markdown-cases: ============== **Blackslash escapes** Seems right but differ from other markdown implementations **Empasis** inline patterns problem **Images (Untitled)** Behavior differs from others implementations on:  expected: <p><img src="/url/" alt="alt text"/> </p> but gives that: <p><img src="</url/>" alt="alt text"/> </p> **Inline HTML (Simple)** skips markdown paragraph **Inline HTML (Span)** don't recognize html blocks **Links, inline style** **Nesting** Inline patterns problem **Parens in URL** **PHP-Specific Bugs** **Tight blocks** blockquote problem tm-cases excluded list: =============== basic_safe_mode basic_safe_mode_escape auto_link_safe_mode code_safe_emphasis emacs_head_vars emacs_tail_vars footnotes footnotes_letters footnotes_markup footnotes_safe_mode_escape nested_list_safe_mode issue2_safe_mode_borks_markup issue3_bad_code_color_hack link_defn_spaces_in_url link_patterns link_patterns_double_hit link_patterns_edge_cases mismatched_footnotes nested_lists_safe_mode pyshell syntax_color |
From: Yuri T. <qar...@gm...> - 2008-06-23 21:43:01
|
Glad to see you are starting to make commits! > http://gitorious.org/projects/python-markdown/repos/gsoc2008 Hmm, something is seriously not right about this repository. For some reason, your "first commit" shows all files as newly created. Did you merge those changes in from somewhere? Can you try to re-do this? Basically, we need a commit that shows a clear diff from the previous revision, both for the sake of maintaining history and also to allow me to cherry-pick the commits. Consider the difference between: http://gitorious.org/projects/python-markdown/repos/gsoc2008/commits/356cd58e06e38ace4a4ea65ac0cbbdbe910e5dfc http://gitorious.org/projects/python-markdown/repos/gsoc2008/commits/448c3397c0dc53f4aa8207d8a61f300768647524 In the first case we know exactly what changed. In the second case, it looks like everything is new. I would suggest doing the following: copy your files somewhere else, start with a brand new clone from mainline, then copy files into it in groups, making incremental commits. The smaller commits, the better. E.g., if you fixed a bug in markdown.py, then commit that change by itself, or together with the test cases that specifically relate to this change. Something like this: cp ../artems-modified-files/markdown.py . git add markdown.py git commit # enter a message explaining the change cp ../artems-modified-files/tests/ cp -r ../artems-modified-files/tests/foo tests/ git add ... # add specific files that need to be added git commit # enter a message - yuri > > tm-cases: > ====== > > **blockquote** > ">" transforms to ">" > > [Trent wrote] > > no way > > [Jeff wrote] > > way > > other implementations consider ">" in this test as blockquote > > > **escapes** > problem with hash characters escaping at the end of header > > > **lists2** > ul followed by ol transforms to one ul > > > **mising_link_defn** > works fine in markdown 1.7, but don't work in markdown from repository. > > > **parens_in_url_4** > this input: > > [Inline link 4 with non-escaped parens](</url(test)> "title"). > > gives invalid output: > > <p><a href="</url(test">Inline link 4 with non-escaped > parens</a>> "title"). > </p> > > > **raw_html** > HTML comments escaping > > test <!-- *blah* --> > > > **sublist_para** > test fails, sublist with 2-spaces indentation > > > **tricky_anchors** > fails on: > > [only open [bracket(/in/) text](/url/ 'a title') > > but all the other markdown implementations pass this test. > > > php-markdown-cases: > ============== > > > **Blackslash escapes** > Seems right but differ from other markdown implementations > > > **Empasis** > inline patterns problem > > > **Images (Untitled)** > Behavior differs from others implementations on: > >  > > expected: > > <p><img src="/url/" alt="alt text"/> > </p> > > but gives that: > > <p><img src="</url/>" alt="alt text"/> > </p> > > > **Inline HTML (Simple)** > skips markdown paragraph > > > **Inline HTML (Span)** > don't recognize html blocks > > > **Links, inline style** > > > **Nesting** > Inline patterns problem > > > **Parens in URL** > > > **PHP-Specific Bugs** > > > **Tight blocks** > blockquote problem > > > tm-cases excluded list: > =============== > > basic_safe_mode > basic_safe_mode_escape > auto_link_safe_mode > code_safe_emphasis > emacs_head_vars > emacs_tail_vars > footnotes > footnotes_letters > footnotes_markup > footnotes_safe_mode_escape > nested_list_safe_mode > issue2_safe_mode_borks_markup > issue3_bad_code_color_hack > link_defn_spaces_in_url > link_patterns > link_patterns_double_hit > link_patterns_edge_cases > mismatched_footnotes > nested_lists_safe_mode > pyshell > syntax_color > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Python-markdown-discuss mailing list > Pyt...@li... > https://lists.sourceforge.net/lists/listinfo/python-markdown-discuss > -- http://sputnik.freewisdom.org/ |
From: Artem <ne...@gm...> - 2008-06-24 17:45:22
|
Yuri Takhteyev wrote: > Hmm, something is seriously not right about this repository. For some > reason, your "first commit" shows all files as newly created. Did you > merge those changes in from somewhere? Can you try to re-do this? > Yes, first time I merged clone with my previous repository. I re-created repository. Everything seems right now. |
From: Yuri T. <qar...@gm...> - 2008-06-24 18:38:24
|
> Yes, first time I merged clone with my previous repository. > I re-created repository. Everything seems right now. Yes, this looks much better! Can you explain what all the .tags files are? Also, are those all just Trent's tests "as is" or did you have to adjust them in anyway? Anyway, the thing to do with those now is to decide which of them we actually want to get working and which we don't. Also, can you add a file somewhere saying that those are Trent's tests and are under a different license from python markdown itself? - yuri -- http://sputnik.freewisdom.org/ |
From: Artem <ne...@gm...> - 2008-06-24 19:03:38
|
Yuri Takhteyev wrote: > Yes, this looks much better! > Good :) > Can you explain what all the .tags files are? I don't know exactly, suppose it's for some additional information, it's useless for us, I'll remove them in the next commit. > Also, are those all > just Trent's tests "as is" or did you have to adjust them in anyway? > Yes, I wrote a script that reformatted Trent's tests, because there are some differences in line breaks between markdown and markdown2 output. It's in tests2/formater.py > Anyway, the thing to do with those now is to decide which of them we > actually want to get working and which we don't. Also, can you add a > file somewhere saying that those are Trent's tests and are under a > different license from python markdown itself? > Ok. |
From: Trent M. <tr...@gm...> - 2008-06-24 20:12:36
|
>> Can you explain what all the .tags files are? > I don't know exactly, suppose it's for some additional information, it's > useless for us, I'll remove them in the next commit. They are used by the test harness I use in python-markdown2. Yes, they are useless for you. Trent -- Trent Mick tr...@gm... |