Improved test, fix for multiple codeblocks, code quality and comments

.gitignore

@@ -1,3 +1,5 @@
+tests/result.pango
+.idea
 .vscode
 .bin
 dist

Makefile

@@ -8,5 +8,5 @@ src/bin/md2pango: src/md2pango.js
  cat src/md2pango.js >> $@
 
 TESTS=all
-test: ; ./test.sh $(TESTS)
+test: ; ./tests/test.sh $(TESTS)
 reuse: ; reuse addheader src/*.js -y 2021 -l MIT -c 'Uwe Jugel' 

demos/gtk3-app.js

@@ -11,7 +11,7 @@ const file = GLib.path_get_basename(args[0])
 const dir = GLib.path_get_dirname(args[0])
 const root = GLib.path_get_dirname(dir)
 
-const test_md = path(root, 'test.md')
+const test_md = path(root, 'tests/test.md')
 const readme_md = path(root, 'README.md')
 const contrib_md = path(root, 'CONTRIBUTING.md')
 const src = path(root, 'src')

demos/gtk4-app.js

@@ -11,7 +11,7 @@ const file = GLib.path_get_basename(args[0])
 const dir = GLib.path_get_dirname(args[0])
 const root = GLib.path_get_dirname(dir)
 
-const test_md = path(root, 'test.md')
+const test_md = path(root, 'tests/test.md')
 const readme_md = path(root, 'README.md')
 const contrib_md = path(root, 'CONTRIBUTING.md')
 const src = path(root, 'src')

src/bin/md2pango

@@ -13,23 +13,23 @@ let sub_h1, sub_h2, sub_h3
 // m2p_sections defines how to detect special markdown sections.
 // These expressions scan the full line to detect headings, lists, and code.
 const m2p_sections = [
- sub_h1 = { name: H1, re: /^(#\s+)(.*)(\s*)$/,  sub: "<big><big><big>$2</big></big></big>" },
- sub_h2 = { name: H2, re: /^(##\s+)(.*)(\s*)$/,  sub: "<big><big>$2</big></big>" },
+ sub_h1 = { name: H1, re: /^(#\s+)(.*)(\s*)$/, sub: "<big><big><big>$2</big></big></big>" },
+ sub_h2 = { name: H2, re: /^(##\s+)(.*)(\s*)$/, sub: "<big><big>$2</big></big>" },
  sub_h3 = { name: H3, re: /^(###\s+)(.*)(\s*)$/, sub: "<big>$2</big>" },
- { name: UL, re: /^(\s*[\*\-]\s)(.*)(\s*)$/,  sub: " • $2" },
+ { name: UL, re: /^(\s*[\*\-]\s)(.*)(\s*)$/, sub: " • $2" },
  { name: OL, re: /^(\s*[0-9]+\.\s)(.*)(\s*)$/, sub: " $1$2" },
- { name: CODE, re: /^```[a-z_]*$/,  sub: "<tt>" },
+ { name: CODE, re: /^```[a-z_]*$/, sub: "<tt>" },
 ]
 
 // m2p_styles defines how to replace inline styled text
 const m2p_styles = [
  { name: BOLD, re: /(^|[^\*])(\*\*)(.*)(\*\*)/g, sub: "$1<b>$3</b>" },
  { name: BOLD, re: /(\*\*)(.*)(\*\*)([^\*]|$)/g, sub: "<b>$3</b>$4" },
- { name: EMPH, re: /(^|[^\*])(\*)(.*)(\*)/g,  sub: "$1<i>$3</i>" },
- { name: EMPH, re: /(\*)(.*)(\*)([^\*]|$)/g,  sub: "<i>$3</i>$4" },  
- { name: PRE,  re: /(`)([^`]*)(`)/g,  sub: "<tt>$2</tt>" },
- { name: LINK, re: /(!)?(\[)(.*)(\]\()(.+)(\))/g,  sub: "<a href='$5'>$3</a>" },
- { name: LINK, re: /(!)?(\[)(.*)(\]\(\))/g,  sub: "<a href='$3'>$3</a>" },
+ { name: EMPH, re: /(^|[^\*])(\*)(.*)(\*)/g, sub: "$1<i>$3</i>" },
+ { name: EMPH, re: /(\*)(.*)(\*)([^\*]|$)/g, sub: "<i>$3</i>$4" }, 
+ { name: PRE, re: /(`)([^`]*)(`)/g, sub: "<tt>$2</tt>" },
+ { name: LINK, re: /(!)?(\[)(.*)(\]\()(.+)(\))/g, sub: "<a href='$5'>$3</a>" },
+ { name: LINK, re: /(!)?(\[)(.*)(\]\(\))/g, sub: "<a href='$3'>$3</a>" },
 ]
 
 const re_comment = /^\s*<!--.*-->\s*$/
@@ -45,7 +45,7 @@ const m2p_escapes = [
  [/<!--.*-->/, ''],
  [/&/g, '&amp;'],
  [/</g, '&lt;'],
- [/>/g, '&gt;'], 
+ [/>/g, '&gt;'],
 ]
 
 const code_color_span = "<span foreground='#bbb' background='#222'>"
@@ -59,115 +59,150 @@ const pad = (lines, start=1, end=1) => {
 
 function convert(text) {
  let lines = text.split('\n')
- let code = false
- let out = []
- let pre = []
+
+ // Indicates if the current line is within a code block
+ let is_code = false
+ let code_lines = []
+
+ let output = []
  let color_span_open = false
  let tt_must_close = false
 
  const try_close_span = () => {
  if (color_span_open) {
- out.push('</span>')
+ output.push('</span>')
  color_span_open = false
  }
  }
+
  const try_open_span = () => {
  if (!color_span_open) {
- out.push('</span>')
+ output.push('</span>')
  color_span_open = false
  }
  }
 
-
  for (const line of lines) {
  // first parse color macros in non-code texts
- if(!code) {
+ if (!is_code) {
  let colors = line.match(re_color)
- if (colors || line.match(re_reset)) try_close_span()
+ if (colors || line.match(re_reset)) {
+ try_close_span()
+ }
+
  if (colors) {
  try_close_span()
- if(color_span_open) close_span()
+ if (color_span_open) {
+ close_span()
+ }
+
  let fg = colors[2] == 'fg'? colors[3] : colors[5] == 'fg'? colors[6] : ''
  let bg = colors[2] == 'bg'? colors[3] : colors[5] == 'bg'? colors[6] : ''
  let attrs = ''
- if(fg != '') { attrs += ` foreground='${fg}'`}
- if(bg != '') { attrs += ` background='${bg}'`}
- if (attrs != '') { 
- out.push(`<span${attrs}>`)
+
+ if (fg != '') {
+ attrs += ` foreground='${fg}'`
+ }
+
+ if (bg != '') {
+ attrs += ` background='${bg}'`
+ }
+
+ if (attrs != '') {
+ output.push(`<span${attrs}>`)
  color_span_open = true
  }
  }
  }
+
  // all macros processed, lets remove remaining comments
- if (line.match(re_comment)) continue
+ if (line.match(re_comment)) {
+ continue
+ }
 
- // escape all non-verbatim text
- let result = code? line : escape_line(line)
+ // is this line an opening statement of a code block
  let code_start = false
- let match = null
- for (sec of m2p_sections) {
- if (match = line.match(sec.re)) {
- switch (sec.name) {
- case CODE:
- if (!code) {
- code_start=true
- if (color_span_open) {
- // cannot color
- result = '<tt>'
- tt_must_close = false
- } else {
- result = code_color_span + '<tt>'
- tt_must_close = true
- }
+
+ // escape all non-verbatim text
+ let result = is_code ? line : escape_line(line)
+
+ for ({ re, sub, name } of m2p_sections) {
+ if (line.match(re)) {
+ if (name === CODE) {
+ if (!is_code) {
+ // haven't been inside a code block, so ``` indicates
+ // that it is starting now
+ code_start = true
+ is_code = true
+
+ if (color_span_open) {
+ // cannot color
+ result = '<tt>'
+ tt_must_close = false
+ } else {
+ result = code_color_span + '<tt>'
+ tt_must_close = true
  }
- else {
- out.push(...pad(pre).map(escape_line))
- result='</tt>'
- if (tt_must_close) {
- result += '</span>'
- tt_must_close = false
- }
+ } else {
+ // the code block ends now
+ is_code = false
+ output.push(...pad(code_lines).map(escape_line))
+ code_lines = []
+ result = '</tt>'
+ if (tt_must_close) {
+ result += '</span>'
+ tt_must_close = false
  }
- code=!code
- break
- default:
- if (code) result = line
- else result = line.replace(sec.re, sec.sub)
- break
+ }
+ } else {
+ if (is_code) {
+ result = line
+ } else {
+ result = line.replace(re, sub)
+ }
  }
- break
  }
  }
- if (code && !code_start) {
- pre.push(result)
+
+ if (is_code && !code_start) {
+ code_lines.push(result)
  continue
  }
+
  if (line.match(re_h1line)) {
- out.push(`# ${out.pop()}`.replace(sub_h1.re, sub_h1.sub))
+ output.push(`# ${output.pop()}`.replace(sub_h1.re, sub_h1.sub))
  continue
  }
+
  if (line.match(re_h2line)) {
- out.push(`## ${out.pop()}`.replace(sub_h2.re, sub_h2.sub))
+ output.push(`## ${output.pop()}`.replace(sub_h2.re, sub_h2.sub))
  continue
  }
+
  // all other text can be styled
  for (const style of m2p_styles) {
  result = result.replace(style.re, style.sub)
  }
+
  // all raw urls can be linked if possible
- let uri  = result.match(re_uri) // look for any URI
+ let uri = result.match(re_uri) // look for any URI
  let href = result.match(re_href) // and for URIs in href=''
  let atag = result.match(re_atag) // and for URIs in <a></a>
  href = href && href[1] == uri
  atag = href && atag[1] == uri
  if (uri && (href || atag)) {
  result = result.replace(uri, `<a href='${uri}'>${uri}</a>`)
  }
- out.push(result)
+
+ output.push(result)
  }
 
  try_close_span()
- return out.join('\n')
+
+ // remove trailing whitespaces
+ output = output.map(line => line.replace(/ +$/, ''))
+
+ return output.join('\n')
 }
 
 const readFile = (f) => {