Closed kjk closed 13 years ago
Hey Krzysztof, thanks for your concern!
There is full compatibility with the official Markdown Test Suite (1.0.3) -- the link to the test suite from peg-markdown looks like the correct version. If your tests are not failing, make sure that:
Your upskirt.c
file is rendering with no extensions enabled:
upshtml_renderer(&renderer, 0);
ups_markdown(ob, ib, &renderer, 0);
upshtml_free_renderer(&renderer);
You are running the test suite from the official MarkdownTest.pl
script, like this:
./MarkdownTest.pl --script ~/src/upskirt/upskirt --tidy
Note the importance of the --tidy
flag: the generated HTML must be tidied up before comparing because e.g. Markdown.pl may generate
<p>This is HTML</p>
And Upskirt may generate:
<p>
This is HTML
</p>
Both are exactly the same piece of HTML, but a naive comparison with diff
will consider them different.
I hope this fixes your issue -- if you happen to find any failing tests, please report back.
Cheers!
Thanks for the info. I'm not using tidy and Markdown.pl since I'm on windows and it's a bit hard to set up so I'm just comparing what upskirt generates with the reference .html file (after normalizing whitespace).
It's plausible that after tidy > and > become the same thing (> is in the reference html and upskirt generates >)
This maybe is a question rather than a bug report.
Since the tests are not part of the sources, I've modified upscript.c in my repo (https://github.com/kjk/upskirt/blob/master/examples/upskirt.c) to run the test against 1.0.3 markdown test files from https://github.com/jgm/peg-markdown/tree/master/MarkdownTest_1.0.3/Tests
The test reads the .text file, renders it to html with ups_markdown() and compares to .html reference result (after trimming whitespace at the beginning of the line and collapsing empty lines).
10 tests fail.
Am I using the wrong test files? (those are the only ones I managed to track down).
The tests do look a bit fishy, e.g. in "Amps and angle encoding.text" < in "4 < 5." is supposed to be converted to < while > in "6 > 5." stays as > (and the code, apparently correctly, converts it to >).
So: