mmmd 0.1.1 → 0.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f64700645c2b97717469cd6da2ecd780138d0722f43f1503375e367f9574b718
-  data.tar.gz: 153b9712bdb5ed0fe66681f64105152360bd50abf9aa907e97e5ef2f33ca3f2b
+  metadata.gz: 1a8790d6e9e6e5a9e3164e1f6f3ecf019cc73876d231676c562eb33925cea1a1
+  data.tar.gz: 738942fca7cea39dbd6fd4fcb15b46d643646e5c5fbb90593f7bf330f967c8ae
 SHA512:
-  metadata.gz: dfe1a17ba166b8eb2076ef3097552d40fba6f1ea040b85d741ab650fe0794320ec20528b9887ac83c9b7620d5b6844805d0b12c0c7a8c547fd2cad889af1b046
-  data.tar.gz: cd476ec80f1022b99115b3e6a0213ceca078d736fe3c3dce47aa3acd7404f15766d41e1324d2d90f84781b82a7f3a959df2c479584a02d28a3bf73b9f5ea5c81
+  metadata.gz: cc7e8c75e20ced1af9c502ebfc5453489c427a61f2c8b7e9e0fa26d50dcaf56df3f522c5e560fca0b27f4d06603ca8e6bb9a720f9dc00929c9c847f408301709
+  data.tar.gz: 03e2bcfff6462f28dc3439afdf02667a3944b6053e2b42da86fa2e1bfd4fffb8faa1bbfe2bcd4a0eaef006e6f22c9120a6cab365d475e294ccc3780e16dbba41

data/lib/mmmd/blankshell.rb

@@ -367,7 +367,7 @@ module PointBlank
 
       # (see ::PointBlank::Parsing::NullParser#consume)
       def consume(line, parent = nil, lazy: false)
-        @lazy_triggered = lazy || @lazy_triggered
+        @lazy_triggered = lazy || @lazy_triggered unless line.strip.empty?
         return [nil, nil] if line.match?(/\A {0,3}\Z/)
         return [nil, nil] if @closed
         return [nil, nil] if check_candidates(line, parent)
@@ -957,9 +957,9 @@ module PointBlank
         parts = tokens
         @valid_parsers.each do |parser|
           newparts = []
-          parts.each do |x|
+          parts.each_with_index do |x, i|
             if x.is_a? String
-              newparts.append(*parser.tokenize(x))
+              newparts.append(*parser.tokenize(x, newparts.last, parts[i + 1]))
             else
               newparts.append(x)
             end
@@ -1013,8 +1013,10 @@ module PointBlank
 
       # Tokenize a string
       # @param string [String]
+      # @param before [String, ::PointBlank::DOM::DOMObject]
+      # @param after [String, ::PointBlank::DOM::DOMObject]
       # @return [Array<Array(String, Class, Symbol), String>]
-      def self.tokenize(string)
+      def self.tokenize(string, _before, _after)
         [string]
       end
 
@@ -1142,18 +1144,11 @@ module PointBlank
     # Code inline parser
     class CodeInline < NullInline
       # (see ::PointBlank::Parsing::NullInline#tokenize)
-      def self.tokenize(string)
-        open = {}
+      def self.tokenize(string, *_lookaround)
         iterate_tokens(string, "`") do |_before, current_text, matched|
           if matched
             match = current_text.match(/^`+/)[0]
-            if open[match]
-              open[match] = nil
-              [match, self, :close]
-            else
-              open[match] = true
-              [match, self, :open]
-            end
+            [match, self, :open]
           else
             current_text[0]
           end
@@ -1170,9 +1165,9 @@ module PointBlank
           text = (part.is_a?(Array) ? part.first : part)
           buffer += text
           next unless part.is_a? Array
+          next if idx.zero?
 
-          break (cutoff = idx) if part.first == opening &&
-                                  part.last == :close
+          break (cutoff = idx) if part.first == opening
         end
         buffer = construct_literal(buffer[opening.length..(-1 - opening.length)])
         [cutoff.positive? ? build([buffer]) : opening, parts[(cutoff + 1)..]]
@@ -1182,7 +1177,7 @@ module PointBlank
     # Autolink inline parser
     class AutolinkInline < NullInline
       # (see ::PointBlank::Parsing::NullInline#tokenize)
-      def self.tokenize(string)
+      def self.tokenize(string, *_lookaround)
         iterate_tokens(string, /[<>]/) do |_before, current_text, matched|
           if matched
             if current_text.start_with?("<")
@@ -1238,11 +1233,10 @@ module PointBlank
         linkinfo = capture[-1][2]
         obj = build(capture[1..-2])
         if linkinfo[:label]
-          if (props = doc.root.properties[:linkdefs][linkinfo[:label]])
-            linkinfo = props
-          else
+          unless (props = doc.root.properties[:linkdefs][linkinfo[:label]])
             return nil
           end
+          linkinfo = props
         end
         obj.properties = linkinfo
         obj
@@ -1277,7 +1271,7 @@ module PointBlank
       end
 
       # (see ::PointBlank::Parsing::NullInline#tokenize)
-      def self.tokenize(string)
+      def self.tokenize(string, *_lookaround)
         iterate_tokens(string, /(?:!\[|\]\()/) do |_before, text, matched|
           next text[0] unless matched
           next ["![", self, :open] if text.start_with? "!["
@@ -1296,7 +1290,7 @@ module PointBlank
       end
 
       # (see ::PointBlank::Parsing::NullInline#tokenize)
-      def self.tokenize(string)
+      def self.tokenize(string, *_lookaround)
         iterate_tokens(string, /(?:\[|\][(\[])/) do |_before, text, matched|
           next text[0] unless matched
           next ["[", self, :open] if text.start_with? "["
@@ -1308,20 +1302,61 @@ module PointBlank
       end
     end
 
+    # TODO: this seems way too complicated for something that's supposed
+    # to be a goddamn emphasis markup parser. i'd blame it on commonmark's
+    # convoluted specs.
+    # (P.S: it could be possible to make this easier for implementers by
+    # making a claims system with pointers that do not modify the string
+    # while it's being parsed. however that would just move complexity from
+    # the parser into the scanner instead. and it does not resolve the
+    # problem of overlapping claims as efficiently as simply splitting text
+    # into tokens and remaining string bits.)
+
     # Emphasis and strong emphasis inline parser
     class EmphInline < NullInline
       INFIX_TOKENS = /^[^\p{S}\p{P}\p{Zs}_]_++[^\p{S}\p{P}\p{Zs}_]$/
       # (see ::PointBlank::Parsing::NullInline#tokenize)
-      def self.tokenize(string)
+      def self.tokenize(string, before, after)
+        bfrb = extract_left(before)
+        afra = extract_right(after)
         iterate_tokens(string, /(?:_++|\*++)/) do |bfr, text, matched|
           token, afr = text.match(/^(_++|\*++)(.?)/)[1..2]
-          left = left_token?(bfr[-1] || "", token, afr)
-          right = right_token?(bfr[-1] || "", token, afr)
+          bfr = bfr[-1] || bfrb || ""
+          afr = afr.empty? ? afra || "" : afr
+          left = left_token?(bfr, token, afr)
+          right = right_token?(bfr, token, afr)
           break_into_elements(token, [bfr[-1] || "", token, afr].join(''),
                               left, right, matched)
         end
       end
 
+      # Extract left-flanking token from before the tokenized string
+      # @param bfr [String, ::PointBlank::DOM::DOMObject, Array(String, Class, Symbol)]
+      # @return [String]
+      def self.extract_left(bfr)
+        case bfr
+        when String
+          bfr[-1]
+        when ::PointBlank::DOM::DOMObject
+          "."
+        when Array
+          bfr.first[-1]
+        end
+      end
+
+      # Extract right-flanking token from after the tokenized string
+      # @param afr [String, ::PointBlank::DOM::DOMObject, Array(String, Class, Symbol)]
+      # @return [String]
+      def self.extract_right(afr)
+        case afr
+        when String
+          afr[0]
+        when ::PointBlank::DOM::DOMObject
+          "."
+        when Array
+          afr.first[0]
+        end
+      end
       # Is this token, given these surrounding characters, left-flanking?
       # @param bfr [String]
       # @param token [String]
@@ -1386,8 +1421,9 @@ module PointBlank
                                ((blk.first.length % 3).zero? &&
                                 (closer.first.length % 3).zero?))
             (open ? capture : before).prepend(blk)
-            next unless blk.is_a?(Array)
-            return backlog unless blk[1].check_contents(capture)
+            next unless blk.is_a?(Array) && !open
+
+            return backlog unless closer[1].check_contents(capture)
           end
           return backlog if open
 
@@ -1431,7 +1467,7 @@ module PointBlank
     # Hard break
     class HardBreakInline < NullInline
       # (see ::PointBlank::Parsing::NullInline#tokenize)
-      def self.tokenize(string)
+      def self.tokenize(string, *_lookaround)
         iterate_tokens(string, /(?:  \n|\\\n)/) do |_before, token, matched|
           next ["\n", self, :close] if token.start_with?("  \n")
           next ["\n", self, :close] if matched

data/test.md

@@ -1,15 +1 @@
-Jabronicle Mitch's wild ride
-============================
-
-crusty cocks
-
--------------
-
-- DIIIIICKS
-- - let
-  - me
-  - smell
-  - your
-  - dick
-- [anime feet](https://google.com)
-- [goog](https://goog.com 'goog')
+tokens blahblah **[test](/pee)** more tokens

data/test2.md

@@ -0,0 +1,12 @@
+> **Quote block**
+>
+> a**.not even emphasis.**b
+>
+> a**[neither is this emphasis](/uri 'but should it be?')**x
+>
+> **[emphasized image link ![image](/uri)](/link)**
+> 
+> titlet
+> ======
+>
+> amongus

data/test3.md

@@ -0,0 +1 @@
+**test<http://autolink`>[stuff`[dat](#lonk)[`...](#stuff)**

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: mmmd
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors:
 - Yessiest
@@ -9,7 +9,7 @@ autorequire:
 bindir:
 - bin
 cert_chain: []
-date: 2025-03-09 00:00:00.000000000 Z
+date: 2012-01-01 00:00:00.000000000 Z
 dependencies: []
 description: |
   MMMD (short for Mark My Manuscript Down) is a Markdown processor
@@ -24,6 +24,8 @@ extra_rdoc_files:
 - architecture.md
 - security.md
 - test.md
+- test2.md
+- test3.md
 files:
 - README.md
 - architecture.md
@@ -37,6 +39,8 @@ files:
 - lib/mmmd/util.rb
 - security.md
 - test.md
+- test2.md
+- test3.md
 homepage: https://adastra7.net/git/Yessiest/rubymark
 licenses:
 - AGPL-3.0-or-later