Implement task lists. (#50)

Fixes #30.
2019-02-10 22:58:42 +01:00 · 2019-02-10 22:58:42 +01:00 · 8e01a769ea
commit 8e01a769ea
parent 0b95bcc984
8 changed files with 217 additions and 16 deletions
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@ -6,6 +6,14 @@

 New features:

+ * Add extension for GitHub-style task lists:
+
+   ```
+    * [x] foo
+    * [x] bar
+    * [ ] baz
+   ```
+
 * Renamed structure `MD_RENDERER` to `MD_PARSER` and refactorize its contents
   a little bit. Note this is source-level incompatible and initialization code
   in apps may need to be updated.
--- a/README.md
+++ b/README.md
@ -88,6 +88,8 @@ extensions and/or deviations from the specification.

 * With the flag `MD_FLAG_TABLES`, GitHub-style tables are supported.

+ * With the flag `MD_FLAG_TASKLISTS`, GitHub-style task lists are supported.
+
 * With the flag `MD_FLAG_STRIKETHROUGH`, strike-through spans are enabled
   (text enclosed in tilde marks, e.g. `~foo bar~`).

--- a/md2html/md2html.c
+++ b/md2html/md2html.c
@ -208,6 +208,7 @@ static const option cmdline_options[] = {
    { "fcollapse-whitespace",        0,  'W', OPTION_ARG_NONE },
    { "ftables",                     0,  'T', OPTION_ARG_NONE },
    { "fstrikethrough",              0,  'S', OPTION_ARG_NONE },
+    { "ftasklists",                  0,  'X', OPTION_ARG_NONE },
    { 0 }
 };

@ -255,6 +256,7 @@ usage(void)
        "      --fno-html       Same as --fno-html-blocks --fno-html-spans\n"
        "      --ftables        Enable tables\n"
        "      --fstrikethrough Enable strikethrough spans\n"
+        "      --ftasklists     Enable task lists\n"
    );
 }

@ -302,6 +304,7 @@ cmdline_callback(int opt, char const* value, void* data)
        case 'V':   parser_flags |= MD_FLAG_PERMISSIVEAUTOLINKS; break;
        case 'T':   parser_flags |= MD_FLAG_TABLES; break;
        case 'S':   parser_flags |= MD_FLAG_STRIKETHROUGH; break;
+        case 'X':   parser_flags |= MD_FLAG_TASKLISTS; break;

        default:
            fprintf(stderr, "Illegal option: %s\n", value);
--- a/md2html/render_html.c
+++ b/md2html/render_html.c
@ -267,6 +267,20 @@ render_open_ol_block(MD_RENDER_HTML* r, const MD_BLOCK_OL_DETAIL* det)
    RENDER_LITERAL(r, buf);
 }

+static void
+render_open_li_block(MD_RENDER_HTML* r, const MD_BLOCK_LI_DETAIL* det)
+{
+    if(det->is_task) {
+        RENDER_LITERAL(r, "<li class=\"task-list-item\">"
+                          "<input type=\"checkbox\" class=\"task-list-item-checkbox\" disabled");
+        if(det->task_mark == 'x' || det->task_mark == 'X')
+            RENDER_LITERAL(r, " checked");
+        RENDER_LITERAL(r, ">");
+    } else {
+        RENDER_LITERAL(r, "<li>");
+    }
+}
+
 static void
 render_open_code_block(MD_RENDER_HTML* r, const MD_BLOCK_CODE_DETAIL* det)
 {
@ -350,7 +364,7 @@ enter_block_callback(MD_BLOCKTYPE type, void* detail, void* userdata)
        case MD_BLOCK_QUOTE:    RENDER_LITERAL(r, "<blockquote>\n"); break;
        case MD_BLOCK_UL:       RENDER_LITERAL(r, "<ul>\n"); break;
        case MD_BLOCK_OL:       render_open_ol_block(r, (const MD_BLOCK_OL_DETAIL*)detail); break;
-        case MD_BLOCK_LI:       RENDER_LITERAL(r, "<li>"); break;
+        case MD_BLOCK_LI:       render_open_li_block(r, (const MD_BLOCK_LI_DETAIL*)detail); break;
        case MD_BLOCK_HR:       RENDER_LITERAL(r, "<hr>\n"); break;
        case MD_BLOCK_H:        RENDER_LITERAL(r, head[((MD_BLOCK_H_DETAIL*)detail)->level - 1]); break;
        case MD_BLOCK_CODE:     render_open_code_block(r, (const MD_BLOCK_CODE_DETAIL*) detail); break;
--- a/md4c/md4c.c
+++ b/md4c/md4c.c
@ -4285,11 +4285,13 @@ struct MD_BLOCK_tag {

    /* MD_BLOCK_H:      Header level (1 - 6)
     * MD_BLOCK_CODE:   Non-zero if fenced, zero if indented.
-     * MD_BLOCK_TABLE:  Column count (as determined by the table underline)
+     * MD_BLOCK_LI:     Task mark character (0 if not task list item, 'x', 'X' or ' ').
+     * MD_BLOCK_TABLE:  Column count (as determined by the table underline).
     */
    unsigned data      : 16;

    /* Leaf blocks:     Count of lines (MD_LINE or MD_VERBATIMLINE) on the block.
+     * MD_BLOCK_LI:     Task mark offset in the input doc.
     * MD_BLOCK_OL:     Start item number.
     */
    unsigned n_lines;
@ -4298,10 +4300,12 @@ struct MD_BLOCK_tag {
 struct MD_CONTAINER_tag {
    CHAR ch;
    unsigned is_loose    : 8;
+    unsigned is_task     : 8;
    unsigned start;
    unsigned mark_indent;
    unsigned contents_indent;
    OFF block_byte_off;
+    OFF task_mark_off;
 };


@ -4515,6 +4519,7 @@ md_process_all_blocks(MD_CTX* ctx)
        union {
            MD_BLOCK_UL_DETAIL ul;
            MD_BLOCK_OL_DETAIL ol;
+            MD_BLOCK_LI_DETAIL li;
        } det;

        switch(block->type) {
@ -4529,6 +4534,12 @@ md_process_all_blocks(MD_CTX* ctx)
                det.ol.mark_delimiter = (CHAR) block->data;
                break;

+            case MD_BLOCK_LI:
+                det.li.is_task = (block->data != 0);
+                det.li.task_mark = (CHAR) block->data;
+                det.li.task_mark_offset = (OFF) block->n_lines;
+                break;
+
            default:
                /* noop */
                break;
@ -5240,11 +5251,13 @@ md_enter_child_containers(MD_CTX* ctx, int n_children, unsigned data)
                MD_CHECK(md_push_container_bytes(ctx,
                                (is_ordered_list ? MD_BLOCK_OL : MD_BLOCK_UL),
                                c->start, data, MD_BLOCK_CONTAINER_OPENER));
-                MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI, 0, data, MD_BLOCK_CONTAINER_OPENER));
+                MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI,
+                                c->task_mark_off, (c->is_task ? CH(c->task_mark_off) : 0),
+                                MD_BLOCK_CONTAINER_OPENER));
                break;

            case _T('>'):
-                MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_QUOTE, 0, data, MD_BLOCK_CONTAINER_OPENER));
+                MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_QUOTE, 0, 0, MD_BLOCK_CONTAINER_OPENER));
                break;

            default:
@ -5275,8 +5288,9 @@ md_leave_child_containers(MD_CTX* ctx, int n_keep)
            case _T('-'):
            case _T('+'):
            case _T('*'):
-                MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI, 0,
-                                0, MD_BLOCK_CONTAINER_CLOSER));
+                MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI,
+                                c->task_mark_off, (c->is_task ? CH(c->task_mark_off) : 0),
+                                MD_BLOCK_CONTAINER_CLOSER));
                MD_CHECK(md_push_container_bytes(ctx,
                                (is_ordered_list ? MD_BLOCK_OL : MD_BLOCK_UL), 0,
                                c->ch, MD_BLOCK_CONTAINER_CLOSER));
@ -5310,6 +5324,7 @@ md_is_container_mark(MD_CTX* ctx, unsigned indent, OFF beg, OFF* p_end, MD_CONTA
        off++;
        p_container->ch = _T('>');
        p_container->is_loose = FALSE;
+        p_container->is_task = FALSE;
        p_container->mark_indent = indent;
        p_container->contents_indent = indent + 1;
        *p_end = off;
@ -5320,9 +5335,10 @@ md_is_container_mark(MD_CTX* ctx, unsigned indent, OFF beg, OFF* p_end, MD_CONTA
    if(off+1 < ctx->size  &&  ISANYOF(off, _T("-+*"))  &&  (ISBLANK(off+1) || ISNEWLINE(off+1))) {
        p_container->ch = CH(off);
        p_container->is_loose = FALSE;
+        p_container->is_task = FALSE;
        p_container->mark_indent = indent;
        p_container->contents_indent = indent + 1;
-        *p_end = off+1;
+        *p_end = off + 1;
        return TRUE;
    }

@ -5338,9 +5354,10 @@ md_is_container_mark(MD_CTX* ctx, unsigned indent, OFF beg, OFF* p_end, MD_CONTA
    if(off+1 < ctx->size  &&  (CH(off) == _T('.') || CH(off) == _T(')'))   &&  (ISBLANK(off+1) || ISNEWLINE(off+1))) {
        p_container->ch = CH(off);
        p_container->is_loose = FALSE;
+        p_container->is_task = FALSE;
        p_container->mark_indent = indent;
        p_container->contents_indent = indent + off - beg + 1;
-        *p_end = off+1;
+        *p_end = off + 1;
        return TRUE;
    }

@ -5726,6 +5743,28 @@ redo:
        n_parents = ctx->n_containers;
    }

+    /* Check for task mark. */
+    if((ctx->parser.flags & MD_FLAG_TASKLISTS)  &&  n_brothers + n_children > 0  &&
+       ISANYOF_(ctx->containers[ctx->n_containers-1].ch, _T("-+*.)")))
+    {
+        OFF tmp = off;
+
+        while(tmp < ctx->size  &&  tmp < off + 3  &&  ISBLANK(tmp))
+            tmp++;
+        if(tmp + 2 < ctx->size  &&  CH(tmp) == _T('[')  &&
+           ISANYOF(tmp+1, _T("xX "))  &&  CH(tmp+2) == _T(']')  &&
+           (tmp + 3 == ctx->size  ||  ISBLANK(tmp+3)  ||  ISNEWLINE(tmp+3)))
+        {
+            MD_CONTAINER* task_container = (n_children > 0 ? &ctx->containers[ctx->n_containers-1] : &container);
+            task_container->is_task = TRUE;
+            task_container->task_mark_off = tmp + 1;
+            off = tmp + 3;
+            while(ISWHITESPACE(off))
+                off++;
+            line->beg = off;
+        }
+    }
+
 done:
    /* Scan for end of the line.
     *
@ -5787,8 +5826,16 @@ done_on_eol:
    /* Enter any container we found a mark for. */
    if(n_brothers > 0) {
        MD_ASSERT(n_brothers == 1);
-        MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI, 0, 0,
-                MD_BLOCK_CONTAINER_CLOSER | MD_BLOCK_CONTAINER_OPENER));
+        MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI,
+                    ctx->containers[n_parents].task_mark_off,
+                    (ctx->containers[n_parents].is_task ? CH(ctx->containers[n_parents].task_mark_off) : 0),
+                    MD_BLOCK_CONTAINER_CLOSER));
+        MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI,
+                    container.task_mark_off,
+                    (container.is_task ? CH(container.task_mark_off) : 0),
+                    MD_BLOCK_CONTAINER_OPENER));
+        ctx->containers[n_parents].is_task = container.is_task;
+        ctx->containers[n_parents].task_mark_off = container.task_mark_off;
    }

    if(n_children > 0)
--- a/md4c/md4c.h
+++ b/md4c/md4c.h
@ -47,7 +47,8 @@ typedef unsigned MD_OFFSET;


 /* Block represents a part of document hierarchy structure like a paragraph
- * or list item. */
+ * or list item.
+ */
 typedef enum MD_BLOCKTYPE {
    /* <body>...</body> */
    MD_BLOCK_DOC = 0,
@ -63,7 +64,8 @@ typedef enum MD_BLOCKTYPE {
     * Detail: Structure MD_BLOCK_OL_DETAIL. */
    MD_BLOCK_OL,

-    /* <li>...</li> */
+    /* <li>...</li>
+     * Detail: Structure MD_BLOCK_LI_DETAIL. */
    MD_BLOCK_LI,

    /* <hr> */
@ -186,7 +188,7 @@ typedef enum MD_ALIGN {
 * So, for example, lets consider an image has a title attribute string
 * set to "foo &quot; bar". (Note the string size is 14.)
 *
- * Then:
+ * Then the attribute MD_SPAN_IMG_DETAIL::title shall provide the following:
 *  -- [0]: "foo "   (substr_types[0] == MD_TEXT_NORMAL; substr_offsets[0] == 0)
 *  -- [1]: "&quot;" (substr_types[1] == MD_TEXT_ENTITY; substr_offsets[1] == 4)
 *  -- [2]: " bar"   (substr_types[2] == MD_TEXT_NORMAL; substr_offsets[2] == 10)
@ -207,17 +209,24 @@ typedef struct MD_ATTRIBUTE {

 /* Detailed info for MD_BLOCK_UL. */
 typedef struct MD_BLOCK_UL_DETAIL {
-    int is_tight;           /* Non-zero if tight list, zero of loose. */
+    int is_tight;           /* Non-zero if tight list, zero if loose. */
    MD_CHAR mark;           /* Item bullet character in MarkDown source of the list, e.g. '-', '+', '*'. */
 } MD_BLOCK_UL_DETAIL;

 /* Detailed info for MD_BLOCK_OL. */
 typedef struct MD_BLOCK_OL_DETAIL {
    unsigned start;         /* Start index of the ordered list. */
-    int is_tight;           /* Non-zero if tight list, zero of loose. */
+    int is_tight;           /* Non-zero if tight list, zero if loose. */
    MD_CHAR mark_delimiter; /* Character delimiting the item marks in MarkDown source, e.g. '.' or ')' */
 } MD_BLOCK_OL_DETAIL;

+/* Detailed info for MD_BLOCK_LI. */
+typedef struct MD_BLOCK_LI_DETAIL {
+    int is_task;            /* Can be non-zero only with MD_FLAG_TASKLISTS */
+    MD_CHAR task_mark;      /* If is_task, then one of 'x', 'X' or ' '. Undefined otherwise. */
+    MD_OFFSET task_mark_offset;  /* If is_task, then offset in the input of the char between '[' and ']'. */
+} MD_BLOCK_LI_DETAIL;
+
 /* Detailed info for MD_BLOCK_H. */
 typedef struct MD_BLOCK_H_DETAIL {
    unsigned level;         /* Header level (1 - 6) */
@ -262,6 +271,7 @@ typedef struct MD_SPAN_IMG_DETAIL {
 #define MD_FLAG_TABLES                      0x0100  /* Enable tables extension. */
 #define MD_FLAG_STRIKETHROUGH               0x0200  /* Enable strikethrough extension. */
 #define MD_FLAG_PERMISSIVEWWWAUTOLINKS      0x0400  /* Enable WWW autolinks (even without any scheme prefix, if they begin with 'www.') */
+#define MD_FLAG_TASKLISTS                   0x0800  /* Enable task list extension. */

 #define MD_FLAG_PERMISSIVEAUTOLINKS         (MD_FLAG_PERMISSIVEEMAILAUTOLINKS | MD_FLAG_PERMISSIVEURLAUTOLINKS | MD_FLAG_PERMISSIVEWWWAUTOLINKS)
 #define MD_FLAG_NOHTML                      (MD_FLAG_NOHTMLBLOCKS | MD_FLAG_NOHTMLSPANS)
@ -276,7 +286,7 @@ typedef struct MD_SPAN_IMG_DETAIL {
 * extensions, bringing the dialect closer to the original, are implemented.
 */
 #define MD_DIALECT_COMMONMARK               0
-#define MD_DIALECT_GITHUB                   (MD_FLAG_PERMISSIVEAUTOLINKS | MD_FLAG_TABLES | MD_FLAG_STRIKETHROUGH)
+#define MD_DIALECT_GITHUB                   (MD_FLAG_PERMISSIVEAUTOLINKS | MD_FLAG_TABLES | MD_FLAG_STRIKETHROUGH | MD_FLAG_TASKLISTS)

 /* Renderer structure.
 */
--- a/scripts/run-tests.sh
+++ b/scripts/run-tests.sh
@ -54,6 +54,10 @@ echo
 echo "Strikethrough extension:"
 $PYTHON "$TEST_DIR/spec_tests.py" -s "$TEST_DIR/strikethrough.txt" -p "$PROGRAM --fstrikethrough"

+echo
+echo "Task lists extension:"
+$PYTHON "$TEST_DIR/spec_tests.py" -s "$TEST_DIR/tasklists.txt" -p "$PROGRAM --ftasklists"
+
 echo
 echo "Pathological input:"
 $PYTHON "$TEST_DIR/pathological_tests.py" -p "$PROGRAM"
--- a/test/tasklists.txt
+++ b/test/tasklists.txt
@ -0,0 +1,113 @@
+
+# Tasklists
+
+With the flag `MD_FLAG_TASKLISTS`, MD4C enables extension for recognition of
+task lists.
+
+Basic task list may look as follows:
+
+```````````````````````````````` example
+ * [x] foo
+ * [X] bar
+ * [ ] baz
+.
+<ul>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>foo</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>bar</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>baz</li>
+</ul>
+````````````````````````````````
+
+Task lists can also be in ordered lists:
+```````````````````````````````` example
+ 1. [x] foo
+ 2. [X] bar
+ 3. [ ] baz
+.
+<ol>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>foo</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>bar</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>baz</li>
+</ol>
+````````````````````````````````
+
+Task lists can also be nested in ordinary lists:
+```````````````````````````````` example
+ * xxx:
+   * [x] foo
+   * [x] bar
+   * [ ] baz
+ * yyy:
+   * [ ] qux
+   * [x] quux
+   * [ ] quuz
+.
+<ul>
+<li>xxx:
+<ul>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>foo</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>bar</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>baz</li>
+</ul></li>
+<li>yyy:
+<ul>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>qux</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>quux</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>quuz</li>
+</ul></li>
+</ul>
+````````````````````````````````
+
+Or in a parent task list:
+```````````````````````````````` example
+ 1. [x] xxx:
+    * [x] foo
+    * [x] bar
+    * [ ] baz
+ 2. [ ] yyy:
+    * [ ] qux
+    * [x] quux
+    * [ ] quuz
+.
+<ol>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>xxx:
+<ul>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>foo</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>bar</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>baz</li>
+</ul></li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>yyy:
+<ul>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>qux</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>quux</li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>quuz</li>
+</ul></li>
+</ol>
+````````````````````````````````
+
+Also, ordinary lists can be nested in the task lists.
+```````````````````````````````` example
+ * [x] xxx:
+   * foo
+   * bar
+   * baz
+ * [ ] yyy:
+   * qux
+   * quux
+   * quuz
+.
+<ul>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled checked>xxx:
+<ul>
+<li>foo</li>
+<li>bar</li>
+<li>baz</li>
+</ul></li>
+<li class="task-list-item"><input type="checkbox" class="task-list-item-checkbox" disabled>yyy:
+<ul>
+<li>qux</li>
+<li>quux</li>
+<li>quuz</li>
+</ul></li>
+</ul>
+````````````````````````````````