karlb / smu Goto Github PK
View Code? Open in Web Editor NEWThis project forked from gottox/smu
Simple MarkUp - markdown/commonmark like syntax
Home Page: http://karlb.github.io/smu/
License: MIT License
This project forked from gottox/smu
Simple MarkUp - markdown/commonmark like syntax
Home Page: http://karlb.github.io/smu/
License: MIT License
smu - a Simple Markup Language ============================== _smu_ is a very simple and minimal markup language. It is designed for use in wiki-like environments. smu makes it very easy to write your documents on the fly and convert them into HTML. smu is capable of parsing very large documents. It scales just great as long as you avoid a huge amount of indents. Syntax ====== smu was started as a rewrite of [markdown](http://daringfireball.net/projects/markdown/) but became something more lightweight and consistent. It differs from [CommonMark](https://commonmark.org/) in the following ways: * No support for _reference style links_ * Stricter indentation rules for lists * Lists don't end paragraphs by themselves (blank line needed) * Horizontal rules (`<hr>`) must use `- - -` as syntax * Code fences have stricter syntax Patches that increase the CommonMark compatibility are welcome as long as they don't increase the code complexity significantly. This project is a fork of the [original smu](https://github.com/gottox/smu) by [Enno Boland (gottox)](https://eboland.de). The main differences to the original smu are: * Support for code fences * Improved [CommonMark](https://commonmark.org/) compatibility. E.g. * Code blocks need four spaces indentation instead of three * Skip empty lines at end of code blocks * Ignore single spaces around code spans * Keep HTML comments in output * Improved spec compliance for lists * Nesting code block in blockquotes works * "Empty" lines in lists behave identically, no matter how much whitespace they contain * No backslash escapes in code blocks * Use first number as start number for ordered lists * Added a simple test suite to check for compliance and avoid regressions Inline patterns --------------- There are several patterns you can use to highlight your text: * Emphasis * Surround your text with `*` or `_` to get *emphasised* text: This *is* cool. This _is_ cool, too. * Surround your text with `**` or `__` to get **strong** text: This **is** cool. This __is__ cool, too. * Surround your text with `***` or `___` to get ***strong and emphasised*** text: This ***is*** cool. This ___is___ cool, too. * But this example won't work as expected: ***Hello** you* This is a wontfix bug because it would make the source too complex. Use this instead: ***Hello*** *you* * inline Code You can produce inline code by surrounding it with backticks. Use `rm -rf /` if you're a N00b. Use ``rm -rf /`` if you're a N00b. Use ```rm -rf /``` if you're a N00b. Double and triple backticks can be used if the code itself contains backticks. Titles ------ Creating titles in smu is very easy. There are two different syntax styles. The first is underlining with at least three characters: Heading ======= Topic ----- This is very intuitive and self explaining. The resulting sourcecode looks like this: <h1>Heading</h1> <h2>Topic</h2> Use the following prefixes if you don't like underlining: # h1 ## h2 ### h3 #### h4 ##### h5 ###### h6 Links ----- The simplest way to define a link is with simple `<>`. <http://s01.de> You can do the same for E-Mail addresses: <[email protected]> If you want to define a label for the url, you have to use a different syntax [smu - simple mark up](http://s01.de/~gottox/index.cgi/proj_smu) The resulting HTML-Code <a href="http://s01.de/~gottox/index.cgi/proj_smu">smu - simple mark up</a></p> Images ------ Images use a syntax similar to the one for links: ![optional alt text](http://example.com/image.png) Lists ----- Defining lists is very straightforward: * Item 1 * Item 2 * Item 3 Result: <ul> <li>Item 1</li> <li>Item 2</li> <li>Item 3</li> </ul> Defining ordered lists is also very easy: 1. Item 1 2. Item 2 3. Item 3 Only the first number in a list is meaningful. All following list items are continously counted. If you want a list starting at 2, you could write: 2. Item 1 2. Item 2 2. Item 3 and get the following HTML which will render with the numbers 2, 3, 4: <ol start="2"> <li>Item 1</li> <li>Item 2</li> <li>Item 3</li> </ol> Code & Blockquote ----------------- Use the `> ` as a line prefix for defining blockquotes. Blockquotes are interpreted as well. This makes it possible to embed links, headings and even other quotes into a quote: > Hello > This is a quote with a [link](http://s01.de/~gottox) Result: <blockquote><p> Hello This is a quote with a <a href="http://s01.de/~gottox">link</a></p> </blockquote> You can define a code block with a leading Tab or with __4__ leading spaces this.is(code) this.is(code, too) Result: <pre><code>this.is(code)</code></pre> <pre><code>this.is(code, too) </code></pre> Please note that you can't use HTML or smu syntax in a code block. Another way to write code blocks is to use code fences: ```json {"some": "code"} ``` This has two advantages: * The optional language identifier will be turned into a `language-` class name * You can keep the original indentation which helps when doing copy & paste Tables ------ Tables can be generated with the following syntax: | Heading1 | Heading2 | | -------- | -------- | | Cell 1 | Cell2 | Aligning the columns make the input nicer to read, but is not necessary to get correct table output. You could just write | Heading1 | Heading2 | | --- | --- | | Cell 1 | Cell2 | To align the content of table cells, use `|:--|` for left, `|--:|` for right and `|:--:|` for centered alignment in the row which separates the header from the table body. | Heading1 | Heading2 | Heading3 | | :------- | :------: | -------: | | Left | Center | Right | Other interesting stuff ----------------------- * to insert a horizontal rule simple add `- - -` into an empty line: Hello - - - Hello2 Result: <p> Hello <hr /> Hello2</p> * Any ASCII punctuation character may escaped by precedeing them with a backslash to avoid them being interpreted: !"#$%&'()*+,-./:;<=>?@[]^_`{|}~\ * To force a linebreak simple add two spaces to the end of the line: No linebreak here. But here is one. embed HTML ---------- You can include arbitrary HTML code in your documents. The HTML will be passed through to the resulting document without modification. This is a good way to work around features that are missing in smu. If you don't want this behaviour, use the `-n` flag when executing smu to stricly escape the HTML tags.
This is because of the optional "title" attribute for links.
The spec says "If both link destination and link title are present, they must be separated by spaces, tabs, and up to one line ending." (https://spec.commonmark.org/0.30/#links), so it should be possible to preserve single quotes that come before a whitespace character followed by a quote inside the link parenthesis.
A workaround is to use URL percent-encoding, but this is inconvenient when copying and pasting links.
https://www.w3schools.com/html/html_entities.asp
it will be interpreted as ·
-> &middot;
For the following file:
* a
* b
a
b
smu
renders:
<ul>
<li><p>a
<ul>
<li>b
<pre><code>a
</code></pre>
</li>
</ul>
</p>
<pre><code> b
</code></pre>
</li>
</ul>
and I would rather have it render something like:
<ul>
<li>a
<ul>
<li>b
<pre><code>a
b
</code></pre>
</li>
</ul>
</li>
</ul>
Can I have it so smu
respect blank lines in code block within lists? Any recommended way of solving this?
Hi,
I couldn't find the very much needed table support, so I've added a Parser for it. It is using a minimalistic state-machine in the spirit of simplicity. Known issues:
<p></p>
Supports normal coloumns, |---|
, right aligned |--:|
and centered |:-:|
. Licensed under MIT, use it as you wish.
/* smu table parser, Copyright(C) bzt 2022 MIT */
static char intable = 0, inrow, incell; /* table state */
static long int calign;
int
dotable(const char *begin, const char *end, int newblock) {
const char *p;
int i, l = (int)sizeof(calign) * 4;
if(*begin != '|')
return 0;
if(inrow && (begin + 1 >= end || begin[1] == '\n')) { /* close cell and row and if ends, table too */
fprintf(stdout, "</t%c></tr>", inrow == -1 ? 'h' : 'd');
inrow = 0;
if(begin + 2 >= end || begin[2] == '\n') {
intable = 0;
fputs("\n</table>", stdout);
return 2;
}
return 1;
}
if(begin < end && (begin[1] == '-' || begin[1] == ':')) { /* only load cell aligns from 2nd line */
for(i = -1, p = begin; p < end && *p != '\n'; p++)
if(*p == '|') {
i++;
if(i < l && p[1] == ':') {
calign |= 1 << (i * 2); p++;
}
} else
if(i < l && *p == ':')
calign |= 1 << (i * 2 + 1);
return p - begin + 1;
}
if(!intable) { /* open table */
intable = 1; inrow = -1; incell = 0; calign = 0;
fputs("<table>\n<tr>", stdout);
}
if(!inrow) { /* open row */
inrow = 1; incell = 0;
fputs("<tr>", stdout);
}
if(incell) /* close cell */
fprintf(stdout, "</t%c>", inrow == -1 ? 'h' : 'd');
l = incell < l ? (calign >> (incell * 2)) & 3 : 0; /* open cell */
fprintf(stdout, "<t%c%s>", inrow == -1 ? 'h' : 'd',
l == 2 ? " class=\"right\"" : (l == 3 ? " class=\"center\"" : ""));
incell++;
for(p = begin + 1; p < end && *p == ' '; p++);
return p - begin;
}
Cheers,
bzt
Hi,
This code does not work if there are UTF-8 characters in the headings. The reason for this is, it expects at least as many =
or -
characters in the next line as the heading's length in bytes. Now this is only true for English headings, encoded as ASCII. With UTF-8, you can't use the number of bytes, you have to use the number of characters instead.
For example:
abc (l=3)
=== (j=3, so correctly detected as a heading)
ábc (l=4)
=== (j=3, so incorrectly NOT detected as a heading)
This patch fixes this, while keeping backward compatibility.
diff --git a/smu.c b/smu.c
index fea4fbd..9272855 100644
--- a/smu.c
+++ b/smu.c
@@ -589,19 +589,20 @@ dosurround(const char *begin, const char *end, int newblock) {
int
dounderline(const char *begin, const char *end, int newblock) {
- unsigned int i, j, l;
+ unsigned int i, j, l, k;
const char *p;
if (!newblock)
return 0;
p = begin;
- for (l = 0; p + l != end && p[l] != '\n'; l++);
+ for(l = k = 0; p + l != end && p[l] != '\n'; l++)
+ if(p[l] > 0 || ((unsigned char)p[l] & 0xC0) == 0xC0) k++;
p += l + 1;
if (l == 0)
return 0;
for (i = 0; i < LENGTH(underline); i++) {
for (j = 0; p + j < end && p[j] != '\n' && p[j] == underline[i].search[0]; j++);
- if (j >= l) {
+ if (j >= k) {
fputs(underline[i].before, stdout);
if (underline[i].process)
process(begin, begin + l, 0);
This code uses l
for the length (number of bytes, just as the original), but it also counts number of multi-byte characters in k
, and then compares j
with that latter.
Cheers,
bzt
If I run the following command:
echo 'Hello<br><br>world' | smu
I expect to see:
<p>Hello<br><br>world</p>
but I get this:
<p>Hello<br><br>world</p>
If I use only one <br>
then smu
works as expected, but two <br>
s mangles the second one. Markdown.pl
works in both cases. Do you have any suggestions on how to get around this?
Note: I know there are other ways to embed newlines in the rendered HTML, but I need the Markdown source to be on one line because I need to pass it through sort
support tables?
HTML does not allow lists in paragraphs, so
foo:
- bar
should result in
<p>foo:</p>
<ul>
<li>bar</li>
</ul>
not
<p>foo:
<ul>
<li>bar</li>
</ul>
</p>
as it does now.
The readme says: "You can include arbitrary HTML code in your documents. The HTML will be
passed through to the resulting document without modification."
But when you embed top-level HTML tags like "div", they get wrapped in "p" tags. I think the ideal behavior would as the readme suggests, which is to not add the "p" tags.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.