My personal build of CMark ✏️
Revert "Remove distinction btw atx and setext header in AST."
This reverts commit a71423f6ee1b77d9f79d42599ea00b4ca99f5da0.
Not quite sure about this change, so reverting for now. Note that we still have a distinction between fenced and indented code blocks in the AST. These two distinctions seem to stand or fall together.
9 files changed, 68 insertions, 37 deletions
| Status | File Name | N° Changes | Insertions | Deletions |
| Modified | api_test/main.c | 12 | 8 | 4 |
| Modified | commonmark.rb | 14 | 11 | 3 |
| Modified | js/lib/blocks.js | 20 | 12 | 8 |
| Modified | js/lib/html-renderer.js | 3 | 2 | 1 |
| Modified | src/blocks.c | 23 | 12 | 11 |
| Modified | src/cmark.h | 6 | 4 | 2 |
| Modified | src/html/html.c | 6 | 4 | 2 |
| Modified | src/node.c | 15 | 10 | 5 |
| Modified | src/print.c | 6 | 5 | 1 |
diff --git a/api_test/main.c b/api_test/main.c @@ -17,7 +17,8 @@ static const cmark_node_type node_types[] = { CMARK_NODE_INDENTED_CODE, CMARK_NODE_HTML, CMARK_NODE_PARAGRAPH, - CMARK_NODE_HEADER, + CMARK_NODE_ATX_HEADER, + CMARK_NODE_SETEXT_HEADER, CMARK_NODE_HRULE, CMARK_NODE_REFERENCE_DEF, CMARK_NODE_STRING, @@ -47,7 +48,8 @@ constructor(test_batch_runner *runner) "get_type %d", type); switch (node->type) { - case CMARK_NODE_HEADER: + case CMARK_NODE_ATX_HEADER: + case CMARK_NODE_SETEXT_HEADER: INT_EQ(runner, cmark_node_get_header_level(node), 1, "default header level is 1"); node->as.header.level = 1; @@ -378,7 +380,8 @@ hierarchy(test_batch_runner *runner) (1 << CMARK_NODE_INDENTED_CODE) | (1 << CMARK_NODE_HTML) | (1 << CMARK_NODE_PARAGRAPH) | - (1 << CMARK_NODE_HEADER) | + (1 << CMARK_NODE_ATX_HEADER) | + (1 << CMARK_NODE_SETEXT_HEADER) | (1 << CMARK_NODE_HRULE) | (1 << CMARK_NODE_REFERENCE_DEF); int all_inlines = @@ -400,7 +403,8 @@ hierarchy(test_batch_runner *runner) test_content(runner, CMARK_NODE_INDENTED_CODE, 0); test_content(runner, CMARK_NODE_HTML, 0); test_content(runner, CMARK_NODE_PARAGRAPH, all_inlines); - test_content(runner, CMARK_NODE_HEADER, all_inlines); + test_content(runner, CMARK_NODE_ATX_HEADER, all_inlines); + test_content(runner, CMARK_NODE_SETEXT_HEADER, all_inlines); test_content(runner, CMARK_NODE_HRULE, 0); test_content(runner, CMARK_NODE_REFERENCE_DEF, 0); test_content(runner, CMARK_NODE_STRING, 0);
diff --git a/commonmark.rb b/commonmark.rb @@ -11,7 +11,7 @@ module CMark typedef :pointer, :node enum :node_type, [:document, :blockquote, :list, :list_item, :fenced_code, :indented_code, :html, :paragraph, - :header, :hrule, :reference_def, + :atx_header, :setext_header, :hrule, :reference_def, :str, :softbreak, :linebreak, :code, :inline_html, :emph, :strong, :link, :image] enum :list_type, [:no_list, :bullet_list, :ordered_list] @@ -55,7 +55,7 @@ class Node b = CMark::cmark_node_next(b) end @string_content = CMark::cmark_node_get_string_content(pointer) - if @type == :header + if @type == :atx_header || @type == :setext_header @header_level = CMark::cmark_node_get_header_level(pointer) end if @type == :list @@ -195,6 +195,14 @@ class Renderer self.code_block(node) end + def setext_header(node) + self.header(node) + end + + def atx_header(node) + self.header(node) + end + def reference_def(node) end @@ -367,7 +375,7 @@ end # Capitalize strings in headers doc.walk do |node| - if node.type == :header + if node.type == :setext_header or node.type == :atx_header node.walk do |subnode| if subnode.type == :str subnode.string_content = subnode.string_content.upcase
diff --git a/js/lib/blocks.js b/js/lib/blocks.js @@ -261,7 +261,8 @@ var incorporateLine = function(ln, line_number) { } break; - case 'Header': + case 'ATXHeader': + case 'SetextHeader': case 'HorizontalRule': // a header can never container > 1 line, so fail to match: all_matched = false; @@ -365,7 +366,7 @@ var incorporateLine = function(ln, line_number) { // ATX header offset = first_nonspace + match[0].length; closeUnmatchedBlocks(this); - container = this.addChild('Header', line_number, first_nonspace); + container = this.addChild('ATXHeader', line_number, first_nonspace); container.level = match[0].trim().length; // number of #s // remove trailing ###s: container.strings = @@ -395,7 +396,7 @@ var incorporateLine = function(ln, line_number) { ((match = ln.slice(first_nonspace).match(/^(?:=+|-+) *$/)))) { // setext header line closeUnmatchedBlocks(this); - container.t = 'Header'; // convert Paragraph to SetextHeader + container.t = 'SetextHeader'; // convert Paragraph to SetextHeader container.level = match[0][0] === '=' ? 1 : 2; offset = ln.length; @@ -468,7 +469,7 @@ var incorporateLine = function(ln, line_number) { // on an empty list item. container.last_line_blank = blank && !(container.t == 'BlockQuote' || - container.t == 'Header' || + container.t == 'SetextHeader' || container.t == 'FencedCode' || (container.t == 'ListItem' && container.children.length === 0 && @@ -499,7 +500,8 @@ var incorporateLine = function(ln, line_number) { } break; - case 'Header': + case 'ATXHeader': + case 'SetextHeader': case 'HorizontalRule': // nothing to do; we already added the contents. break; @@ -510,7 +512,7 @@ var incorporateLine = function(ln, line_number) { } else if (blank) { // do nothing } else if (container.t != 'HorizontalRule' && - container.t != 'Header') { + container.t != 'SetextHeader') { // create paragraph container for line container = this.addChild('Paragraph', line_number, first_nonspace); this.addLine(ln, first_nonspace); @@ -559,7 +561,8 @@ var finalize = function(block, line_number) { } break; - case 'Header': + case 'ATXHeader': + case 'SetextHeader': case 'HtmlBlock': block.string_content = block.strings.join('\n'); break; @@ -629,7 +632,8 @@ var processInlines = function(block) { newblock.inline_content = this.inlineParser.parse(block.string_content.trim(), this.refmap); break; - case 'Header': + case 'SetextHeader': + case 'ATXHeader': newblock.inline_content = this.inlineParser.parse(block.string_content.trim(), this.refmap); newblock.level = block.level;
diff --git a/js/lib/html-renderer.js b/js/lib/html-renderer.js @@ -103,7 +103,8 @@ var renderBlock = function(block, in_tight_list) { return inTags(tag, attr, this.innersep + this.renderBlocks(block.children, block.tight) + this.innersep); - case 'Header': + case 'ATXHeader': + case 'SetextHeader': tag = 'h' + block.level; return inTags(tag, [], this.renderInlines(block.inline_content)); case 'IndentedCode':
diff --git a/src/blocks.c b/src/blocks.c @@ -97,9 +97,9 @@ static inline bool can_contain(cmark_node_type parent_type, cmark_node_type chil static inline bool accepts_lines(cmark_node_type block_type) { return (block_type == NODE_PARAGRAPH || - block_type == NODE_HEADER || - block_type == NODE_INDENTED_CODE || - block_type == NODE_FENCED_CODE); + block_type == NODE_ATX_HEADER || + block_type == NODE_INDENTED_CODE || + block_type == NODE_FENCED_CODE); } static void add_line(cmark_node* cmark_node, chunk *ch, int offset) @@ -293,7 +293,8 @@ static void process_inlines(cmark_node* cur, reference_map *refmap) while (cur != NULL) { switch (cur->type) { case NODE_PARAGRAPH: - case NODE_HEADER: + case NODE_ATX_HEADER: + case NODE_SETEXT_HEADER: parse_inlines(cur, refmap); break; @@ -547,7 +548,8 @@ void cmark_process_line(cmark_doc_parser *parser, const char *buffer, all_matched = false; } - } else if (container->type == NODE_HEADER) { + } else if (container->type == NODE_ATX_HEADER || + container->type == NODE_SETEXT_HEADER) { // a header can never contain more than one line all_matched = false; @@ -623,7 +625,7 @@ void cmark_process_line(cmark_doc_parser *parser, const char *buffer, } else if ((matched = scan_atx_header_start(&input, first_nonspace))) { offset = first_nonspace + matched; - container = add_child(parser, container, NODE_HEADER, parser->line_number, offset + 1); + container = add_child(parser, container, NODE_ATX_HEADER, parser->line_number, offset + 1); int hashpos = chunk_strchr(&input, '#', first_nonspace); int level = 0; @@ -653,7 +655,7 @@ void cmark_process_line(cmark_doc_parser *parser, const char *buffer, strbuf_strrchr(&container->string_content, '\n', strbuf_len(&container->string_content) - 2) < 0) { - container->type = NODE_HEADER; + container->type = NODE_SETEXT_HEADER; container->as.header.level = lev; offset = input.len - 1; @@ -730,7 +732,7 @@ void cmark_process_line(cmark_doc_parser *parser, const char *buffer, // on an empty list item. container->last_line_blank = (blank && container->type != NODE_BLOCK_QUOTE && - container->type != NODE_HEADER && + container->type != NODE_SETEXT_HEADER && container->type != NODE_FENCED_CODE && !(container->type == NODE_LIST_ITEM && container->first_child == NULL && @@ -789,7 +791,7 @@ void cmark_process_line(cmark_doc_parser *parser, const char *buffer, // ??? do nothing - } else if (container->type == NODE_HEADER) { + } else if (container->type == NODE_ATX_HEADER) { chop_trailing_hashtags(&input); add_line(container, &input, first_nonspace); @@ -800,8 +802,7 @@ void cmark_process_line(cmark_doc_parser *parser, const char *buffer, add_line(container, &input, first_nonspace); - } else if (container->type != NODE_HRULE && - container->type != NODE_HEADER) { + } else if (container->type != NODE_HRULE && container->type != NODE_SETEXT_HEADER) { // create paragraph container for line container = add_child(parser, container, NODE_PARAGRAPH, parser->line_number, first_nonspace + 1);
diff --git a/src/cmark.h b/src/cmark.h @@ -20,7 +20,8 @@ typedef enum { CMARK_NODE_INDENTED_CODE, CMARK_NODE_HTML, CMARK_NODE_PARAGRAPH, - CMARK_NODE_HEADER, + CMARK_NODE_ATX_HEADER, + CMARK_NODE_SETEXT_HEADER, CMARK_NODE_HRULE, CMARK_NODE_REFERENCE_DEF, @@ -203,7 +204,8 @@ char *cmark_markdown_to_html(const char *text, int len); #define NODE_INDENTED_CODE CMARK_NODE_INDENTED_CODE #define NODE_HTML CMARK_NODE_HTML #define NODE_PARAGRAPH CMARK_NODE_PARAGRAPH - #define NODE_HEADER CMARK_NODE_HEADER + #define NODE_ATX_HEADER CMARK_NODE_ATX_HEADER + #define NODE_SETEXT_HEADER CMARK_NODE_SETEXT_HEADER #define NODE_HRULE CMARK_NODE_HRULE #define NODE_REFERENCE_DEF CMARK_NODE_REFERENCE_DEF #define NODE_STRING CMARK_NODE_STRING
diff --git a/src/html/html.c b/src/html/html.c @@ -149,7 +149,8 @@ static void node_to_html(strbuf *html, cmark_node *node) break; } - case NODE_HEADER: + case NODE_ATX_HEADER: + case NODE_SETEXT_HEADER: cr(html); start_header[2] = '0' + cur->as.header.level; strbuf_puts(html, start_header); @@ -315,7 +316,8 @@ finish_node(strbuf *html, cmark_node *node, bool tight) break; } - case NODE_HEADER: + case NODE_ATX_HEADER: + case NODE_SETEXT_HEADER: end_header[3] = '0' + node->as.header.level; strbuf_puts(html, end_header); break;
diff --git a/src/node.c b/src/node.c @@ -13,7 +13,8 @@ cmark_node_new(cmark_node_type type) { node->type = type; switch (node->type) { - case CMARK_NODE_HEADER: + case CMARK_NODE_ATX_HEADER: + case CMARK_NODE_SETEXT_HEADER: node->as.header.level = 1; break; @@ -57,7 +58,8 @@ S_type_string(cmark_node *node) case CMARK_NODE_INDENTED_CODE: return "INDENTED_CODE"; case CMARK_NODE_HTML: return "HTML"; case CMARK_NODE_PARAGRAPH: return "PARAGRAPH"; - case CMARK_NODE_HEADER: return "HEADER"; + case CMARK_NODE_ATX_HEADER: return "ATX_HEADER"; + case CMARK_NODE_SETEXT_HEADER: return "SETEXT_HEADER"; case CMARK_NODE_HRULE: return "HRULE"; case CMARK_NODE_REFERENCE_DEF: return "REFERENCE_DEF"; case CMARK_NODE_STRING: return "STRING"; @@ -157,7 +159,8 @@ cmark_node_set_string_content(cmark_node *node, const char *content) { int cmark_node_get_header_level(cmark_node *node) { switch (node->type) { - case CMARK_NODE_HEADER: + case CMARK_NODE_ATX_HEADER: + case CMARK_NODE_SETEXT_HEADER: return node->as.header.level; default: @@ -174,7 +177,8 @@ cmark_node_set_header_level(cmark_node *node, int level) { } switch (node->type) { - case CMARK_NODE_HEADER: + case CMARK_NODE_ATX_HEADER: + case CMARK_NODE_SETEXT_HEADER: node->as.header.level = level; return 1; @@ -389,7 +393,8 @@ S_can_contain(cmark_node *node, cmark_node *child) return child->type == CMARK_NODE_LIST_ITEM; case CMARK_NODE_PARAGRAPH: - case CMARK_NODE_HEADER: + case CMARK_NODE_ATX_HEADER: + case CMARK_NODE_SETEXT_HEADER: case CMARK_NODE_EMPH: case CMARK_NODE_STRONG: case CMARK_NODE_LINK:
diff --git a/src/print.c b/src/print.c @@ -132,7 +132,11 @@ static void print_blocks(cmark_node* b, int indent) } print_blocks(b->first_child, indent + 2); break; - case NODE_HEADER: + case NODE_ATX_HEADER: + printf("atx_header (level=%d)\n", b->as.header.level); + print_inlines(b->first_child, indent + 2); + break; + case NODE_SETEXT_HEADER: printf("setext_header (level=%d)\n", b->as.header.level); print_inlines(b->first_child, indent + 2); break;