Merge pull request #33 from increments/merge-upstream-v0.23.6

Merge upstream v0.23.6

Author: kato647

.rubocop.yml

@@ -1,3 +1,5 @@
+inherit_from: .rubocop_todo.yml
+
 inherit_gem:
   rubocop-standard:
     - config/default.yml

.rubocop_todo.yml

@@ -0,0 +1,25 @@
+# This configuration was generated by
+# `rubocop --auto-gen-config --exclude-limit 999999`
+# on 2022-10-03 03:50:46 UTC using RuboCop version 1.36.0.
+# The point is for the user to remove these configuration records
+# one by one as the offenses are removed from the code base.
+# Note that changes in the inspected code, or installation of new
+# versions of RuboCop, may require this file to be generated again.
+
+# Offense count: 3
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyle.
+# SupportedStyles: def_self, self_class
+Style/ClassMethodsDefinitions:
+  Exclude:
+    - 'lib/qiita_marker.rb'
+    - 'lib/qiita_marker/config.rb'
+
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyleForMultiline.
+# SupportedStylesForMultiline: comma, consistent_comma, no_comma
+Style/TrailingCommaInArguments:
+  Exclude:
+    - 'lib/qiita_marker/renderer.rb'
+    - 'test/test_commonmark.rb'

CHANGELOG.md

@@ -6,6 +6,8 @@
 
 ### Changed
 
+- [#33](https://github.com/increments/qiita_marker/pull/33): Update base CommonMarker version to `v0.23.6`.
+
 ## 0.23.5.1 - 2022-06-21
 
 ### Changed

ext/qiita_marker/autolink.c

@@ -286,6 +286,22 @@ static cmark_node *match(cmark_syntax_extension *ext, cmark_parser *parser,
   // inline was finished in inlines.c.
 }
 
+static bool validate_protocol(char protocol[], uint8_t *data, int rewind) {
+  size_t len = strlen(protocol);
+
+  // Check that the protocol matches
+  for (int i = 1; i <= len; i++) {
+    if (data[-rewind - i] != protocol[len - i]) {
+      return false;
+    }
+  }
+
+  char prev_char = data[-rewind - len - 1];
+
+  // Make sure the character before the protocol is non-alphanumeric
+  return !cmark_isalnum(prev_char);
+}
+
 static void postprocess_text(cmark_parser *parser, cmark_node *text, int offset,
                              int depth, cmark_syntax_extension *ext) {
   // postprocess_text can recurse very deeply if there is a very long line of
@@ -296,6 +312,8 @@ static void postprocess_text(cmark_parser *parser, cmark_node *text, int offset,
   uint8_t *data = text->as.literal.data,
     *at;
   size_t size = text->as.literal.len;
+  bool auto_mailto = true;
+  bool is_xmpp = false;
   int rewind, max_rewind,
       nb = 0, np = 0, ns = 0;
 
@@ -322,8 +340,18 @@ static void postprocess_text(cmark_parser *parser, cmark_node *text, int offset,
     if (strchr(".+-_", c) != NULL)
       continue;
 
-    if (c == '/')
-      ns++;
+    if (strchr(":", c) != NULL) {
+      if (validate_protocol("mailto:", data, rewind)) {
+        auto_mailto = false;
+        continue;
+      }
+
+      if (validate_protocol("xmpp:", data, rewind)) {
+        auto_mailto = false;
+        is_xmpp = true;
+        continue;
+      }
+    }
 
     break;
   }
@@ -343,6 +371,8 @@ static void postprocess_text(cmark_parser *parser, cmark_node *text, int offset,
       nb++;
     else if (c == '.' && link_end < size - 1 && cmark_isalnum(data[link_end + 1]))
       np++;
+    else if (c == '/' && is_xmpp)
+      continue;
     else if (c != '-' && c != '_')
       break;
   }
@@ -368,7 +398,8 @@ static void postprocess_text(cmark_parser *parser, cmark_node *text, int offset,
   }
   cmark_strbuf buf;
   cmark_strbuf_init(parser->mem, &buf, 10);
-  cmark_strbuf_puts(&buf, "mailto:");
+  if (auto_mailto)
+    cmark_strbuf_puts(&buf, "mailto:");
   cmark_strbuf_put(&buf, data - rewind, (bufsize_t)(link_end + rewind));
   link_node->as.link.url = cmark_chunk_buf_detach(&buf);
 

ext/qiita_marker/cmark-gfm_version.h

@@ -1,7 +1,7 @@
 #ifndef CMARK_GFM_VERSION_H
 #define CMARK_GFM_VERSION_H
 
-#define CMARK_GFM_VERSION ((0 << 24) | (29 << 16) | (0 << 8) | 4)
-#define CMARK_GFM_VERSION_STRING "0.29.0.gfm.4"
+#define CMARK_GFM_VERSION ((0 << 24) | (29 << 16) | (0 << 8) | 6)
+#define CMARK_GFM_VERSION_STRING "0.29.0.gfm.6"
 
 #endif

ext/qiita_marker/inlines.c

@@ -43,6 +43,8 @@ typedef struct bracket {
   bool image;
   bool active;
   bool bracket_after;
+  bool in_bracket_image0;
+  bool in_bracket_image1;
 } bracket;
 
 typedef struct subject{
@@ -526,6 +528,8 @@ static void push_bracket(subject *subj, bool image, cmark_node *inl_text) {
   bracket *b = (bracket *)subj->mem->calloc(1, sizeof(bracket));
   if (subj->last_bracket != NULL) {
     subj->last_bracket->bracket_after = true;
+    b->in_bracket_image0 = subj->last_bracket->in_bracket_image0;
+    b->in_bracket_image1 = subj->last_bracket->in_bracket_image1;
   }
   b->image = image;
   b->active = true;
@@ -534,6 +538,11 @@ static void push_bracket(subject *subj, bool image, cmark_node *inl_text) {
   b->previous_delimiter = subj->last_delim;
   b->position = subj->pos;
   b->bracket_after = false;
+  if (image) {
+    b->in_bracket_image1 = true;
+  } else {
+    b->in_bracket_image0 = true;
+  }
   subj->last_bracket = b;
 }
 
@@ -1265,6 +1274,17 @@ static cmark_node *handle_close_bracket(cmark_parser *parser, subject *subj) {
       }
       opener = opener->previous;
     }
+    bool in_bracket_image1 = false;
+    if (opener) {
+      in_bracket_image1 = opener->in_bracket_image1;
+    }
+    bracket *opener2 = subj->last_bracket;
+    while (opener2 != opener) {
+      if (opener2->image) {
+        opener2->in_bracket_image1 = in_bracket_image1;
+      }
+      opener2 = opener2->previous;
+    }
   }
 
   return NULL;
@@ -1674,10 +1694,15 @@ cmark_chunk *cmark_inline_parser_get_chunk(cmark_inline_parser *parser) {
 }
 
 int cmark_inline_parser_in_bracket(cmark_inline_parser *parser, int image) {
-  for (bracket *b = parser->last_bracket; b; b = b->previous)
-    if (b->active && b->image == (image != 0))
-      return 1;
-  return 0;
+  bracket *b = parser->last_bracket;
+  if (!b) {
+    return 0;
+  }
+  if (image != 0) {
+    return b->in_bracket_image1;
+  } else {
+    return b->in_bracket_image0;
+  }
 }
 
 void cmark_node_unput(cmark_node *node, int n) {