X-Git-Url: http://nitlanguage.org diff --git a/lib/markdown/markdown.nit b/lib/markdown/markdown.nit index 0e7edbd..b29372f 100644 --- a/lib/markdown/markdown.nit +++ b/lib/markdown/markdown.nit @@ -30,9 +30,6 @@ import template # SEE: `String::md_to_html` for a shortcut. class MarkdownProcessor - # `MarkdownEmitter` used for ouput. - var emitter: MarkdownEmitter is noinit - # Work in extended mode (default). # # Behavior changes when using extended mode: @@ -41,33 +38,39 @@ class MarkdownProcessor # # In normal markdown the following: # - # This is a paragraph - # * and this is not a list + # ~~~md + # This is a paragraph + # * and this is not a list + # ~~~ # # Will produce: # - #
This is a paragraph - # * and this is not a list
+ # ~~~html + #This is a paragraph + # * and this is not a list
+ # ~~~ # - # When using extended mode this changes to: + # When using extended mode this changes to: # - #This is a paragraph
- #This is a paragraph
+ #import markdown
+ # ~~~html
+ # import markdown
#
- # print "Hello World!".md_to_html
- #
+ # print "Hello World!".md_to_html
+ #
+ # ~~~
#
# * Underscores (Emphasis)
#
# Underscores in the middle of a word like:
#
- # Con_cat_this
+ # ~~~md
+ # Con_cat_this
+ # ~~~
#
- # normally produces this:
+ # normally produces this:
#
- # Concatthis
+ # ~~~html + #Concatthis
+ # ~~~ # # With extended mode they don't result in emphasis. # - #Con_cat_this
+ # ~~~html + #Con_cat_this
+ # ~~~ # # * Strikethrough # # Like in [GFM](https://help.github.com/articles/github-flavored-markdown), # strikethrought span is marked with `~~`. # - # ~~Mistaken text.~~ + # ~~~md + # ~~Mistaken text.~~ + # ~~~ # # becomes # - #"
+ var meta = block.meta
+ if meta != null then
+ v.add ""
else
v.add ""
end
@@ -847,9 +917,36 @@ class HTMLDecorator
private var allowed_id_chars: Array[Char] = ['-', '_', ':', '.']
end
+# Location in a Markdown input.
+class MDLocation
+
+ # Starting line number (starting from 1).
+ var line_start: Int
+
+ # Starting column number (starting from 1).
+ var column_start: Int
+
+ # Stopping line number (starting from 1).
+ var line_end: Int
+
+ # Stopping column number (starting from 1).
+ var column_end: Int
+
+ redef fun to_s do return "{line_start},{column_start}--{line_end},{column_end}"
+
+ # Return a copy of `self`.
+ fun copy: MDLocation do
+ return new MDLocation(line_start, column_start, line_end, column_end)
+ end
+end
+
# A block of markdown lines.
# A `MDBlock` can contains lines and/or sub-blocks.
class MDBlock
+
+ # Position of `self` in the input.
+ var location: MDLocation
+
# Kind of block.
# See `Block`.
var kind: Block = new BlockNone(self) is writable
@@ -902,7 +999,14 @@ class MDBlock
# Split `self` creating a new sub-block having `line` has `last_line`.
fun split(line: MDLine): MDBlock do
- var block = new MDBlock
+ # location for new block
+ var new_loc = new MDLocation(
+ first_line.as(not null).location.line_start,
+ first_line.as(not null).location.column_start,
+ line.location.line_end,
+ line.location.column_end)
+ # create block
+ var block = new MDBlock(new_loc)
block.first_line = first_line
block.last_line = line
first_line = line.next
@@ -910,13 +1014,16 @@ class MDBlock
if first_line == null then
last_line = null
else
- first_line.prev = null
+ first_line.as(not null).prev = null
+ # update current block loc
+ location.line_start = first_line.as(not null).location.line_start
+ location.column_start = first_line.as(not null).location.column_start
end
if first_block == null then
first_block = block
last_block = block
else
- last_block.next = block
+ last_block.as(not null).next = block
last_block = block
end
return block
@@ -928,10 +1035,10 @@ class MDBlock
first_line = line
last_line = line
else
- last_line.next_empty = line.is_empty
- line.prev_empty = last_line.is_empty
+ last_line.as(not null).next_empty = line.is_empty
+ line.prev_empty = last_line.as(not null).is_empty
line.prev = last_line
- last_line.next = line
+ last_line.as(not null).next = line
last_line = line
end
end
@@ -941,12 +1048,12 @@ class MDBlock
if line.prev == null then
first_line = line.next
else
- line.prev.next = line.next
+ line.prev.as(not null).next = line.next
end
if line.next == null then
last_line = line.prev
else
- line.next.prev = line.prev
+ line.next.as(not null).prev = line.prev
end
line.prev = null
line.next = null
@@ -1013,6 +1120,12 @@ class MDBlock
text.append "\n"
line = line.next
end
+ var block = first_block
+ while block != null do
+ text.append block.text
+ text.append "\n"
+ block = block.next
+ end
return text.write_to_string
end
end
@@ -1025,10 +1138,10 @@ abstract class Block
var block: MDBlock
# Output `self` using `v.decorator`.
- fun emit(v: MarkdownEmitter) do v.emit_in(self)
+ fun emit(v: MarkdownProcessor) do v.emit_in(self)
# Emit the containts of `self`, lines or blocks.
- fun emit_in(v: MarkdownEmitter) do
+ fun emit_in(v: MarkdownProcessor) do
block.remove_surrounding_empty_lines
if block.has_lines then
emit_lines(v)
@@ -1038,7 +1151,7 @@ abstract class Block
end
# Emit lines contained in `block`.
- fun emit_lines(v: MarkdownEmitter) do
+ fun emit_lines(v: MarkdownProcessor) do
var tpl = v.push_buffer
var line = block.first_line
while line != null do
@@ -1056,13 +1169,35 @@ abstract class Block
end
# Emit sub-blocks contained in `block`.
- fun emit_blocks(v: MarkdownEmitter) do
+ fun emit_blocks(v: MarkdownProcessor) do
var block = self.block.first_block
while block != null do
+ v.push_loc(block.location)
block.kind.emit(v)
+ v.pop_loc
block = block.next
end
end
+
+ # The raw content of the block as a multi-line string.
+ fun raw_content: String do
+ var infence = self isa BlockFence
+ var text = new FlatBuffer
+ var line = self.block.first_line
+ while line != null do
+ if not line.is_empty then
+ var str = line.value
+ if not infence and str.has_prefix(" ") then
+ text.append str.substring(4, str.length - line.trailing)
+ else
+ text.append str
+ end
+ end
+ text.append "\n"
+ line = line.next
+ end
+ return text.write_to_string
+ end
end
# A block without any markdown specificities.
@@ -1103,6 +1238,9 @@ end
class BlockCode
super Block
+ # Any string found after fence token.
+ var meta: nullable Text
+
# Number of char to skip at the beginning of the line.
#
# Block code lines start at 4 spaces.
@@ -1129,9 +1267,6 @@ end
class BlockFence
super BlockCode
- # Any string found after fence token.
- var meta: nullable Text
-
# Fence code lines start at 0 spaces.
redef var line_start = 0
end
@@ -1140,7 +1275,15 @@ end
class BlockHeadline
super Block
- redef fun emit(v) do v.decorator.add_headline(v, self)
+ redef fun emit(v) do
+ var loc = block.location.copy
+ loc.column_start += start
+ v.push_loc(loc)
+ v.decorator.add_headline(v, self)
+ v.pop_loc
+ end
+
+ private var start = 0
# Depth of the headline used to determine the headline level.
var depth = 0
@@ -1150,6 +1293,7 @@ class BlockHeadline
if depth > 0 then return
var level = 0
var line = block.first_line
+ if line == null then return
if line.is_empty then return
var start = line.leading
while start < line.value.length and line.value[start] == '#' do
@@ -1169,6 +1313,7 @@ class BlockHeadline
line.leading = 0
line.trailing = 0
end
+ self.start = start
depth = level.min(6)
end
end
@@ -1188,6 +1333,7 @@ abstract class BlockList
# Split list block into list items sub-blocks.
private fun init_block(v: MarkdownProcessor) do
var line = block.first_line
+ if line == null then return
line = line.next
while line != null do
var t = v.line_kind(line)
@@ -1283,6 +1429,9 @@ end
# A markdown line.
class MDLine
+ # Location of `self` in the original input.
+ var location: MDLocation
+
# Text contained in this line.
var value: String is writable
@@ -1317,8 +1466,8 @@ class MDLine
leading = 0
trailing = 0
is_empty = true
- if prev != null then prev.next_empty = true
- if next != null then next.prev_empty = true
+ if prev != null then prev.as(not null).next_empty = true
+ if next != null then next.as(not null).prev_empty = true
end
# Number or leading spaces on this line.
@@ -1452,8 +1601,8 @@ class MDLine
# Used by `check_html`.
private fun read_xml_comment(first_line: MDLine, start: Int): Int do
var line: nullable MDLine = first_line
- if start + 3 < line.value.length then
- if line.value[2] == '-' and line.value[3] == '-' then
+ if start + 3 < line.as(not null).value.length then
+ if line.as(not null).value[2] == '-' and line.as(not null).value[3] == '-' then
var pos = start + 4
while line != null do
while pos < line.value.length and line.value[pos] != '-' do
@@ -1494,7 +1643,7 @@ class LineEmpty
super Line
redef fun process(v) do
- v.current_line = v.current_line.next
+ v.current_line = v.current_line.as(not null).next
end
end
@@ -1506,7 +1655,7 @@ class LineOther
redef fun process(v) do
var line = v.current_line
# go to block end
- var was_empty = line.prev_empty
+ var was_empty = line.as(not null).prev_empty
while line != null and not line.is_empty do
var t = v.line_kind(line)
if (v.in_list or v.ext_mode) and t isa LineList then
@@ -1522,29 +1671,30 @@ class LineOther
line = line.next
end
# build block
+ var current_block = v.current_block.as(not null)
if line != null and not line.is_empty then
- var block = v.current_block.split(line.prev.as(not null))
+ var block = current_block.split(line.prev.as(not null))
if v.in_list and not was_empty then
block.kind = new BlockNone(block)
else
block.kind = new BlockParagraph(block)
end
- v.current_block.remove_leading_empty_lines
+ current_block.remove_leading_empty_lines
else
var block: MDBlock
if line != null then
- block = v.current_block.split(line)
+ block = current_block.split(line)
else
- block = v.current_block.split(v.current_block.last_line.as(not null))
+ block = current_block.split(current_block.last_line.as(not null))
end
if v.in_list and (line == null or not line.is_empty) and not was_empty then
block.kind = new BlockNone(block)
else
block.kind = new BlockParagraph(block)
end
- v.current_block.remove_leading_empty_lines
+ current_block.remove_leading_empty_lines
end
- v.current_line = v.current_block.first_line
+ v.current_line = current_block.first_line
end
end
@@ -1559,15 +1709,16 @@ class LineCode
line = line.next
end
# split at block end line
+ var current_block = v.current_block.as(not null)
var block: MDBlock
if line != null then
- block = v.current_block.split(line.prev.as(not null))
+ block = current_block.split(line.prev.as(not null))
else
- block = v.current_block.split(v.current_block.last_line.as(not null))
+ block = current_block.split(current_block.last_line.as(not null))
end
block.kind = new BlockCode(block)
block.remove_surrounding_empty_lines
- v.current_line = v.current_block.first_line
+ v.current_line = current_block.first_line
end
end
@@ -1577,12 +1728,14 @@ class LineXML
redef fun process(v) do
var line = v.current_line
+ if line == null then return
+ var current_block = v.current_block.as(not null)
var prev = line.prev
- if prev != null then v.current_block.split(prev)
- var block = v.current_block.split(line.xml_end_line.as(not null))
+ if prev != null then current_block.split(prev)
+ var block = current_block.split(line.xml_end_line.as(not null))
block.kind = new BlockXML(block)
- v.current_block.remove_leading_empty_lines
- v.current_line = v.current_block.first_line
+ current_block.remove_leading_empty_lines
+ v.current_line = current_block.first_line
end
end
@@ -1592,6 +1745,7 @@ class LineBlockquote
redef fun process(v) do
var line = v.current_line
+ var current_block = v.current_block.as(not null)
# go to bquote end
while line != null do
if not line.is_empty and (line.prev_empty and
@@ -1602,9 +1756,9 @@ class LineBlockquote
# build sub block
var block: MDBlock
if line != null then
- block = v.current_block.split(line.prev.as(not null))
+ block = current_block.split(line.prev.as(not null))
else
- block = v.current_block.split(v.current_block.last_line.as(not null))
+ block = current_block.split(current_block.last_line.as(not null))
end
var kind = new BlockQuote(block)
block.kind = kind
@@ -1612,7 +1766,7 @@ class LineBlockquote
kind.remove_block_quote_prefix(block)
v.current_line = line
v.recurse(block, false)
- v.current_line = v.current_block.first_line
+ v.current_line = current_block.first_line
end
end
@@ -1622,11 +1776,13 @@ class LineHR
redef fun process(v) do
var line = v.current_line
- if line.prev != null then v.current_block.split(line.prev.as(not null))
- var block = v.current_block.split(line.as(not null))
+ if line == null then return
+ var current_block = v.current_block.as(not null)
+ if line.prev != null then current_block.split(line.prev.as(not null))
+ var block = current_block.split(line)
block.kind = new BlockRuler(block)
- v.current_block.remove_leading_empty_lines
- v.current_line = v.current_block.first_line
+ current_block.remove_leading_empty_lines
+ v.current_line = current_block.first_line
end
end
@@ -1636,7 +1792,8 @@ class LineFence
redef fun process(v) do
# go to fence end
- var line = v.current_line.next
+ var line = v.current_line.as(not null).next
+ var current_block = v.current_block.as(not null)
while line != null do
if v.line_kind(line) isa LineFence then break
line = line.next
@@ -1647,16 +1804,17 @@ class LineFence
# build fence block
var block: MDBlock
if line != null then
- block = v.current_block.split(line.prev.as(not null))
+ block = current_block.split(line.prev.as(not null))
else
- block = v.current_block.split(v.current_block.last_line.as(not null))
+ block = current_block.split(current_block.last_line.as(not null))
end
- var meta = block.first_line.value.meta_from_fence
+ block.remove_surrounding_empty_lines
+ var meta = block.first_line.as(not null).value.meta_from_fence
block.kind = new BlockFence(block, meta)
- block.first_line.clear
+ block.first_line.as(not null).clear
var last = block.last_line
if last != null and v.line_kind(last) isa LineFence then
- block.last_line.clear
+ block.last_line.as(not null).clear
end
block.remove_surrounding_empty_lines
v.current_line = line
@@ -1669,14 +1827,16 @@ class LineHeadline
redef fun process(v) do
var line = v.current_line
+ if line == null then return
+ var current_block = v.current_block.as(not null)
var lprev = line.prev
- if lprev != null then v.current_block.split(lprev)
- var block = v.current_block.split(line.as(not null))
+ if lprev != null then current_block.split(lprev)
+ var block = current_block.split(line)
var kind = new BlockHeadline(block)
block.kind = kind
kind.transform_headline(block)
- v.current_block.remove_leading_empty_lines
- v.current_line = v.current_block.first_line
+ current_block.remove_leading_empty_lines
+ v.current_line = current_block.first_line
end
end
@@ -1686,16 +1846,18 @@ class LineHeadline1
redef fun process(v) do
var line = v.current_line
+ if line == null then return
+ var current_block = v.current_block.as(not null)
var lprev = line.prev
- if lprev != null then v.current_block.split(lprev)
- line.next.clear
- var block = v.current_block.split(line.as(not null))
+ if lprev != null then current_block.split(lprev)
+ line.next.as(not null).clear
+ var block = current_block.split(line)
var kind = new BlockHeadline(block)
kind.depth = 1
kind.transform_headline(block)
block.kind = kind
- v.current_block.remove_leading_empty_lines
- v.current_line = v.current_block.first_line
+ current_block.remove_leading_empty_lines
+ v.current_line = current_block.first_line
end
end
@@ -1705,22 +1867,24 @@ class LineHeadline2
redef fun process(v) do
var line = v.current_line
+ if line == null then return
+ var current_block = v.current_block.as(not null)
var lprev = line.prev
- if lprev != null then v.current_block.split(lprev)
- line.next.clear
- var block = v.current_block.split(line.as(not null))
+ if lprev != null then current_block.split(lprev)
+ line.next.as(not null).clear
+ var block = current_block.split(line)
var kind = new BlockHeadline(block)
kind.depth = 2
kind.transform_headline(block)
block.kind = kind
- v.current_block.remove_leading_empty_lines
- v.current_line = v.current_block.first_line
+ current_block.remove_leading_empty_lines
+ v.current_line = current_block.first_line
end
end
# A markdown list line.
# Mainly used to factorize code between ordered and unordered lists.
-class LineList
+abstract class LineList
super Line
redef fun process(v) do
@@ -1733,19 +1897,20 @@ class LineList
line = line.next
end
# build list block
+ var current_block = v.current_block.as(not null)
var list: MDBlock
if line != null then
- list = v.current_block.split(line.prev.as(not null))
+ list = current_block.split(line.prev.as(not null))
else
- list = v.current_block.split(v.current_block.last_line.as(not null))
+ list = current_block.split(current_block.last_line.as(not null))
end
var kind = block_kind(list)
list.kind = kind
- list.first_line.prev_empty = false
- list.last_line.next_empty = false
+ list.first_line.as(not null).prev_empty = false
+ list.last_line.as(not null).next_empty = false
list.remove_surrounding_empty_lines
- list.first_line.prev_empty = false
- list.last_line.next_empty = false
+ list.first_line.as(not null).prev_empty = false
+ list.last_line.as(not null).next_empty = false
kind.init_block(v)
var block = list.first_block
while block != null do
@@ -1790,14 +1955,17 @@ end
# Some tokens have a specific markup behaviour that is handled here.
abstract class Token
- # Position of `self` in markdown input.
+ # Location of `self` in the original input.
+ var location: nullable MDLocation
+
+ # Position of `self` in input independant from lines.
var pos: Int
# Character found at `pos` in the markdown input.
var char: Char
# Output that token using `MarkdownEmitter::decorator`.
- fun emit(v: MarkdownEmitter) do v.addc char
+ fun emit(v: MarkdownProcessor) do v.decorator.add_char(v, char)
end
# A token without a specific meaning.
@@ -1865,14 +2033,15 @@ abstract class TokenCode
super Token
redef fun emit(v) do
+ var current_text = v.current_text.as(not null)
var a = pos + next_pos + 1
- var b = v.processor.find_token(v.current_text.as(not null), a, self)
+ var b = v.find_token(current_text, a, self)
if b > 0 then
v.current_pos = b + next_pos
- while a < b and v.current_text[a] == ' ' do a += 1
+ while a < b and current_text[a] == ' ' do a += 1
if a < b then
- while v.current_text[b - 1] == ' ' do b -= 1
- v.decorator.add_span_code(v, v.current_text.as(not null), a, b)
+ while current_text[b - 1] == ' ' do b -= 1
+ v.decorator.add_span_code(v, current_text, a, b)
end
else
v.addc char
@@ -1925,11 +2094,12 @@ abstract class TokenLinkOrImage
end
# Emit the hyperlink as link or image.
- private fun emit_hyper(v: MarkdownEmitter) is abstract
+ private fun emit_hyper(v: MarkdownProcessor) is abstract
# Check if the link is a valid link.
- private fun check_link(v: MarkdownEmitter, out: FlatBuffer, start: Int, token: Token): Int do
+ private fun check_link(v: MarkdownProcessor, out: FlatBuffer, start: Int, token: Token): Int do
var md = v.current_text
+ if md == null then return -1
var pos
if token isa TokenLink then
pos = start + 1
@@ -1944,9 +2114,9 @@ abstract class TokenLinkOrImage
pos += 1
pos = md.skip_spaces(pos)
if pos < start then
- var tid = name.write_to_string.to_lower
- if v.processor.link_refs.has_key(tid) then
- var lr = v.processor.link_refs[tid]
+ var tid = name.as(not null).write_to_string.to_lower
+ if v.link_refs.has_key(tid) then
+ var lr = v.link_refs[tid]
is_abbrev = lr.is_abbrev
link = lr.link
comment = lr.title
@@ -1981,6 +2151,7 @@ abstract class TokenLinkOrImage
if pos == -1 then return -1
end
end
+ if pos < start then return -1
if md[pos] != ')' then return -1
else if md[pos] == '[' then
pos += 1
@@ -1993,16 +2164,16 @@ abstract class TokenLinkOrImage
else
id = name
end
- var tid = id.write_to_string.to_lower
- if v.processor.link_refs.has_key(tid) then
- var lr = v.processor.link_refs[tid]
+ var tid = id.as(not null).write_to_string.to_lower
+ if v.link_refs.has_key(tid) then
+ var lr = v.link_refs[tid]
link = lr.link
comment = lr.title
end
else
- var tid = name.write_to_string.replace("\n", " ").to_lower
- if v.processor.link_refs.has_key(tid) then
- var lr = v.processor.link_refs[tid]
+ var tid = name.as(not null).write_to_string.replace("\n", " ").to_lower
+ if v.link_refs.has_key(tid) then
+ var lr = v.link_refs[tid]
link = lr.link
comment = lr.title
pos = old_pos
@@ -2054,7 +2225,7 @@ class TokenHTML
# Is the HTML valid?
# Also take care of link and mailto shortcuts.
- private fun check_html(v: MarkdownEmitter, out: FlatBuffer, md: Text, start: Int): Int do
+ private fun check_html(v: MarkdownProcessor, out: FlatBuffer, md: Text, start: Int): Int do
# check for auto links
var tmp = new FlatBuffer
var pos = md.read_until(tmp, start + 1, ':', ' ', '>', '\n')
@@ -2135,7 +2306,7 @@ class TokenEscape
redef fun emit(v) do
v.current_pos += 1
- v.addc v.current_text[v.current_pos]
+ v.addc v.current_text.as(not null)[v.current_pos]
end
end
@@ -2179,18 +2350,11 @@ redef class Text
if c == '\\' and pos + 1 < length then
pos = escape(out, self[pos + 1], pos)
else
- var end_reached = false
- for n in nend do
- if c == n then
- end_reached = true
- break
- end
- end
- if end_reached then break
+ for n in nend do if c == n then break label
out.add c
end
pos += 1
- end
+ end label
if pos == length then return -1
return pos
end
@@ -2442,7 +2606,7 @@ redef class String
# var md = "**Hello World!**"
# var html = md.md_to_html
# assert html == "Hello World!
\n"
- fun md_to_html: Streamable do
+ fun md_to_html: Writable do
var processor = new MarkdownProcessor
return processor.process(self)
end